Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download MARKETING RESEARCH EXPLAINED
Document related concepts
no text concepts found
Transcript
PREFACE Marketing research is generally considered as a complex subject. I have tried to take a macro picture of the subject in the book with a focus on the practical example from the professional life. This book is for everyone who wants to understand what exactly marketing research is all about. This is not intended to be a book which will make all you managers experts in marketing research buy yes this book will help you in getting to know what is marketing research? How it can be used? What are its uses? How you can apply it in your practical professional life? Also it is for the people around the world who want to understand the value of market research- and its role in improving the way company make decisions, based on objective, reliable data and insights. You may be working in any field, from government to a service industry, in manufacturing, HR, Finance or sales or may be just studying one of these areas. Even though your job may not include the word marketing but still every company has customers to which it caters and once customers role come into picture, marketing research automatically gain significance.. Thus your main interest is in the role market research plays in understanding those customers, how market research works and some of its impact on business and wider society. I hope this book will help you discover the enormous possibilities that market research provides in 21st century, not only for decision makers but also for the society. Remember to keep in mind that market researcher have the privilege of working side by side with those who decide the fate of products, services, brands and government policies. All with the aim of helping to create a better life for people. That will be always a challenge for a market researcher. ACKNOWLEDGEMENT I am really grateful to Ajay sir in giving me an opportunity to put my ideas into perspective regarding marketing research. I am also humbled by the confidence shown by sir in allowing me to write this book which for me is a first time experience. I. I am also grateful to my clients who gave me much opportunity to broaden my approach towards marketing research as it is currently practiced. Also a word of thanks towards my parent company AC Nielsen working in which I have able to get so many insightful experience which I have tried to put into this book. Last but not the least wants to thank Sapna my wife who helped me with her patience during all those days and night I have used up in completing this book. 1 I hope I had done full justice with what is expected of me. PREFACE ..................................................................................................................................................... 1 ACKNOWLEDGEMENT ........................................................................................................................... 1 CHAPTER ONE ........................................................................................................................................... 5 1.0 INTRODUCTION ..................................................................................................................................... 5 1.1 SCOPE OF MARKET RESEARCH-MYTHS ................................................................................................ 6 1.2 ROLE OF MARKETING RESEARCH IN MANAGERIAL DECISION MAKING? ................................................ 7 1.2.1Environment Analysis ................................................................................................................... 7 1.2.2 Figuring out the future strategy ................................................................................................... 7 1.2.3 Marketing program Development ................................................................................................ 7 1.2.4 Actual Implementation ................................................................................................................. 8 1.3 PRACTICAL CASE STUDY EXAMPLE ....................................................................................................... 9 CHAPTER TWO .........................................................................................................................................10 2.0 MARKETING RESEARCH PROCESS........................................................................................................11 2.1 IDENTIFYING THE DECISION PROBLEM ................................................................................................11 2.2 RESEARCH OBJECTIVE/GAP/OPPORTUNITY ANALYSIS/DECISION ALTERNATIVES ...............................12 2.3 RESEARCH QUESTION ..........................................................................................................................12 2.4 DEVELOPMENT OF HYPOTHESIS ...........................................................................................................13 2.5 MARKETING RESEARCH DESIGN ..........................................................................................................13 2.6 ERRORS IN SAMPLING ..........................................................................................................................14 2.7 MARKETING RESEARCH DATA COLLECTION .......................................................................................16 2.8 SURVEY DATA ANALYSIS ....................................................................................................................16 2.9 REPORTING AND PRESENTATION ..........................................................................................................17 CHAPTER THREE.....................................................................................................................................18 3. RESEARCH DISCUSSION AND DATA COLLECTION METHODS .................................................................19 3.1 DESCRIPTIVE RESEARCH: ....................................................................................................................19 3.2 EXPLORATORY RESEARCH...................................................................................................................19 3.3 CAUSAL RESEARCH .............................................................................................................................19 3.4 DATA COLLECTION METHODS..............................................................................................................19 3.5 PRIMARY DATA RESEARCH .................................................................................................................20 3.5.1 Diaries ........................................................................................................................................20 2 3.5.2 Interviews ....................................................................................................................................20 3.5.3 Telephonic interviews .................................................................................................................20 3.5.4 Face to Face Interviews ..............................................................................................................21 3.5.5 Internet ........................................................................................................................................21 3.5.6 Mail Survey: ................................................................................................................................22 3.5.7 Focus Groups ..............................................................................................................................22 3.5.8 Mystery Shopping........................................................................................................................23 3.5.9 Projective Techniques .................................................................................................................23 3.5.10 Product Tests ............................................................................................................................24 3.5.11 Omnibus Studies........................................................................................................................24 3.6 SECONDARY DATA RESEARCH .............................................................................................................24 3.6.1 Internal Sources ..........................................................................................................................25 3.6.2 External Sources .........................................................................................................................25 3.7 DATA COLLECTION THROUGH THE LENS OF OUR CASE STUDY? ...........................................................27 3.7.1 Use of Primary data research .....................................................................................................27 3.7.2 Use of secondary Data research .................................................................................................27 CHAPTER FOUR .......................................................................................................................................28 4.0 DESIGNING THE QUESTIONNAIRE.........................................................................................................29 4.1 SO WHAT IS A QUESTIONNAIRE? ...........................................................................................................29 4.2 HOW THE ANSWERS BE FRAMED IN A QUESTIONNAIRE .........................................................................30 4.2.1 Scaled ..........................................................................................................................................30 4.2.2 Open-Ended ................................................................................................................................30 4.2.3 Fixed Choice/Pre-coded/Closed .................................................................................................30 4.3 QUALITIES OF A GOOD QUESTIONNAIRE ..............................................................................................30 4.4 QUESTIONNAIRE OF OUR CASE STUDY..................................................................................................33 4.5 PRETESTING ....................................................................................................................................36 CHAPTER FIVE .........................................................................................................................................37 5.0 SAMPLING............................................................................................................................................38 5.1 PROBABILITY SAMPLING .....................................................................................................................39 5.1.1 Random sampling........................................................................................................................39 5.1.2 Systematic sampling ....................................................................................................................39 5.1.2 Stratified sampling: .....................................................................................................................39 5.2 NON PROBABILITY SAMPLING .............................................................................................................39 5.2.1 Convenience sampling ................................................................................................................40 5.2.1 Judgment sampling .....................................................................................................................40 5.2.2 Quota sampling ...........................................................................................................................40 5.2.3 Snowball sampling ......................................................................................................................40 5.3 SAMPLING PROCESS FLOW CHART ............................................................................................40 5.4 SAMPLING ERRORS ..............................................................................................................................42 5.5 NON SAMPLING ERRORS ......................................................................................................................42 5.5.1 Design Error: ..............................................................................................................................42 5.5.2 Administrative Errors .................................................................................................................42 5.5.3 Response Error ...........................................................................................................................43 5.5.4 Non Response Errors ..................................................................................................................43 5.6 SAMPLING PLAN FOR OUR CASE STUDY ................................................................................................43 5.6.1 Household Survey .......................................................................................................................43 5.6.2 Household /Retailer survey .........................................................................................................44 CHAPTER SIX ............................................................................................................................................46 6.0 PREPARATION OF DATA FOR DATA ANALYSIS .....................................................................................46 6.1 DATA EDITING ................................................................................................................................47 6.2 CODING .............................................................................................................................................47 6.3 ADJUSTMENT OF DATA STATISTICALLY ...............................................................................................47 6.3.1 Scale transformation ...................................................................................................................48 3 6.3.2 Dummy Variables........................................................................................................................48 6.3.3 Weighting ....................................................................................................................................48 6.3.4 Variable Respecification .............................................................................................................48 6.4 TABULATION ...................................................................................................................................48 6.4.1 Frequency Distribution ...............................................................................................................48 6.5 DESCRIPTIVE STATISTICS.....................................................................................................................49 6.6 CROSS –TABULATIONS .................................................................................................................51 6.7 SELECTION CRITERION FOR STATISTICAL TECHNIQUES.....................................................52 6.7.1 Research Design: ........................................................................................................................52 6.7.2 Type of Data................................................................................................................................53 CHAPTER SEVEN .....................................................................................................................................54 7.0 DATA ANALYSIS TECHNIQUES.............................................................................................................55 7.1 UNIVARIATE STATISTICAL TECHNIQUE ................................................................................................55 7.2 MULTIVARIATE STATISTICAL TECHNIQUE ...........................................................................................55 7.2.1 Conjoint Analysis ........................................................................................................................55 7.2.2 Multidimensional scaling ............................................................................................................56 7.2.3 Factor Analysis ...........................................................................................................................56 7.2.4 Cluster Analysis ..........................................................................................................................56 7.2.5 Discriminant Analysis .................................................................................................................56 7.3 CONCEPT OF HYPOTHESIS TESTING ..........................................................................................56 7.4 SIGNIFICANCE LEVEL ...........................................................................................................................57 7.5 STATISTICAL TESTS .............................................................................................................................58 7.6 EXPERIMENTAL ERRORS ......................................................................................................................58 7.6.1 Type I Error ................................................................................................................................58 7.6.2 Type II Error ...............................................................................................................................59 7.6.3 Type III Errors ............................................................................................................................59 7.8 DEGREES OF FREEDOM.........................................................................................................................59 7.9 ONE OR TWO TAIL TEST ......................................................................................................................60 7.10 HYPOTHESIS TESTING OF MEAN, PROPORTION AND ANOVA ...............................................................60 7.10.1 Hypothesis testing about single mean .......................................................................................60 7.10.2 Hypothesis testing of proportions .............................................................................................64 7.11 EFFECT OF SAMPLE SIZE AND TEST RESULTS .....................................................................................64 7.12 ANALYSIS OF VARIANCE (ANOVA) ..........................................................................................65 7.12.1 Use of ANOVA in our case study ..............................................................................................66 CHAPTER EIGHT .....................................................................................................................................69 8.0 PRESENTATION, APPLICATION AND ROLE OF MARKETING RESEARCH ..................................................69 8.1 APPLICATION OF MARKETING RESEARCH ............................................................................................69 8.1.1 New product Research ................................................................................................................69 8.1.2 Test marketing .............................................................................................................................70 8.1.3 Pricing Research .........................................................................................................................70 8.1.4 Distribution research: .................................................................................................................70 8.1.5 Promotional Research.................................................................................................................70 8.1.6 Assessing Competitive advantage ...............................................................................................71 8.1.7 Measuring Brand Equity .............................................................................................................71 8.2 MARKETING RESEARCH IN 21ST CENTURY ...........................................................................................72 CHAPTER NINE ........................................................................................................................................74 9.0 MARKETING RESEARCH AND RURAL MARKETING ..............................................................................74 9.1 SALIENT FEATURES MAKING RURAL MARKET AS AN ATTRACTIVE PIT STOP. .......................................74 9.2 ROLE OF MARKETING RESEARCHER IN RURAL MARKET .....................................................................74 9.2.1 Ways in which marketing research can help rural manager ......................................................75 9.3 SOME RURAL SUCCESS STORIES ..........................................................................................................76 REFERENCES ............................................................................................................................................78 4 CHAPTER ONE 1.0 Introduction “Entry of ITC in rural sector through E-Chaupal Model” “Pillsbury Presses flour Power in India” “Success of Dell in Latin American Countries” “Launch of Diet Coke and Pepsi Zero” “Success of Boursin Cheese as a Brand in France” …………and Many more. Reading the above examples which is the common binding point among them? Marketing Research is the common point which made all the above things happen. So Marketing research is the common tool which helps the businesses to grow and work out of the box. First we’ll start with the definition of Marketing Research. Marketing Research, which includes social and opinion research is the systematic gathering and interpretation of information about individuals or organizations using the statistical and analytical methods and techniques of the applied social sciences to gain insight or support decision making. The identity of the respondents will not be revealed to the user of the information without explicit consent and no sale s approach will be made to them as a direct result of their having provided information. Now let us go through the official Definition of marketing research Marketing research is the function that links the consumer, customer, and public to the marketer through information-information used to identify and define marketing opportunities and problems; generate, refine and evaluate marketing actions; monitor 5 marketing performance; and improve understanding of marketing as a process. Marketing research specifies the information required to address these issues, design the method for collecting information, manages and implements the data collection process, analyses and communicates the finding and their implication. American Marketing Association In short we can say that marketing research is about listening to people, analyzing the information to help organization make better decisions and reducing the risk. This foresight aspect of marketing research makes it a vital business discipline as it can play a key role in : Helping a company to plan expansion in new retail sector or market Helping a company to understand better customer and employee satisfaction Helping a company to understand the views of citizen regarding future government policy Helping a company to develop or test a strategy for a new or improved product or service ……….What Marketing research is not A Mystery Market research is a discipline which requires specific knowledge and skills but there is nothing mysterious about it. It encompasses a mix of discipline including sociology, Psychology, anthropology, semiotics, mathematics, statistics, economics and management services. A Sales Technique Professional marketing research is many steps removed from selling. There are strict guidelines and codes of practice on the use of market research and is regulated by international and national bodies. A Seer Marketing research provides objective insights based on in depth understandings and can anticipate likely scenarios in the future. Independent, systematic and rigorous research, supported by experience and imagination, can help to identify trends in markets, technology and demographics - and define opportunities, understand choices and reduce risk. 1.1 Scope of Market Research-Myths MYTH ONE Market research is unintelligible. False. Like any other discipline, it may appear complex to the non specialist, but is logical, useful and understandable. It does however; require a certain willingness to accept structural rules of discipline MYTH TWO 6 Advertising is more efficient than Marketing research. False. Advertising and Marketing research are two different and complimentary disciplines. Market research helps organization understand consumers and markets better in order to improve strategy. Without good marketing research, the risk of making strategic mistake increases and without good strategy the chance of advertising reaching its goal are very much reduce. MYTH THREE Marketing research is very expensive and only for big Organizations. False. Many market research methods are simple and accessible. Market research covers a wide range of costs- from less costly desk research and omnibus survey to broader, costly, global, tailored surveys. MYTH FOUR Marketing research is only used in Mass consumption Industries. False. Today marketing research is a common tool for service companies, advertising agencies, media communications, agriculture companies, management consultancies and business to business (B2B).Business to Consumer (B2C) was only the starting of marketing research. 1.2 Role of marketing research in Managerial decision making? Well the above question can be answered in 4 stage sequential process. These can be also termed as market planning process 1.2.1Environment Analysis Analyzing the present condition in which the market is operating. To look for the opportunities in the market Vis a Vis competitor poisoning. To focus on political, regulatory, economic, social and technological trends. 1.2.2 Figuring out the future strategy Defining the Segments to target and the business scope of organization. How to compete in the market better than the competitors. Goals of the Business and the ultimate differentiating factor of the Product or the service being offered. 1.2.3 Marketing program Development Product and distribution decision Pricing decision Customer satisfaction decision Advertising and Promotion decision 7 This is the step in which major bulk of ongoing marketing research is directed. Action program usually focuses on a single objective in support of one element of the overall business strategy. 1.2.4 Actual Implementation This is the decision step which measures the action program in relation to committed objectives, timelines and budget constraints. Specific objective are set like Sales goal by sales force etc and role of marketing research is to provide measures against these objectives and to analyze the results obtained thereof. At this point before going in depth to understand Marketing research technically, let me present to you some example to show the importance of marketing research in today’s world. “When Unilever wanted to take the No 1 French cheese brand Boursin, across to UK, effective research enabled the company to evaluate opportunity. Research revealed that consumers in UK like the French connection of Boursin cheese the most. Thus the strength of Boursin’s French association became the focus for the new advertising strategy. Subsequent research revealed extraordinary rich and strong brand imagery for Boursin. The advertising campaign was responsible for the brand’s success as an authentic product from rural France in UK.” “Philips is working successfully in key markets of Greater China such as Domestic Appliances, Consumer electronics, personal care and Healthcare division since 1920.The reason for their success lie in marketing research. Philips design teams have regularly embarked on research in China that aimed to meet the needs to the specific Chinese context and its layer of complexity. Lessons learnt through these ongoing research programs are invaluable in deepening Philips understanding of China.” “Tesco one of the world’s largest retailers was facing the issue of families with babies not purchasing the baby product from Tesco. When Tesco carried out market research with these group of customers it found out there was a trust issue. Mothers felt that they should go to specialized retailers for their baby needs, despite the fact that they were paying more. Tesco on the basis of inferences obtained from marketing research set up a number of initiatives. This led to Tesco gaining rapid share of this market and now Tesco sells as much as baby products as each of UK’s specialized baby retailers.” Above given example illustrates the importance of marketing research in today’s cut throat competition. It is not wrong to say that companies which give proper importance to the marketing research have been able to withstand competition and grow success fully sometimes even faster than their segment. To give you all a better understanding of the whole process of marketing research, I will try to frame each and every step in the easiest and simplest way forward. 8 1.3 Practical case study example To give you all Agri business manager’s better example of practical use of Marketing research I will be simultaneously using an example of practical case study of marketing research which was done by me with my partner for an organization working in Agri business division. This case study was done in semi urban setting of Gujarat. The objective of the case study were the following (1) To access the consumer preferences for three daily use food products (A), (B) and (C) under the brand name “MILKY” owned by owner “SAGAR”. (2) To Design a Marketing Strategy to Improve the Sales of the above three in the region. Scope: Again the scope of the study is the area in which the study would be carried out and for this given case it was a semi urban area of North Gujarat with its adjoining rural areas. Thus I’ll be following this above case study example in course of this book. I will try to connect with the topic covered in the book by drawing an inference with the above practical case study. 9 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 10 CHAPTER TWO 2.0 Marketing Research Process The marketing research process includes the systematic identification, collection, analysis and distribution of information for the purpose of knowledge development and decision making. The reasons and times at which company or organization might consider performing marketing research varies, but the general purpose of gaining intelligence for decision making remains constant throughout. Customers occupy the central role in the marketing research process. As a company or organization, the overwhelming majority of research which a manger can consider likely revolves around the customers which can be: Current customers Prospective customers Lost customers Members Community Employees (internal customers) Shareholders (internal customers) When a manger is creating a new marketing research program or perhaps revising an existing marketing research program, following major steps generally constitute the whole process. Step One: Identifying and defining the research problem Step Two: Estimating the Research Objective /Gap/Opportunity Analysis/Decision Alternatives Step Three: Establishing the research design and strategy Step Four: Collecting the data Step five: Performing the data analysis Step Six: Final reporting and presentation 2.1 Identifying the Decision Problem This step is always the first of the marketing research steps. At this point, the problem will have been recognized by at least one level of management, and internal discussions will have taken place. Here at the outset of the marketing research steps, the most common tools are internal and external secondary research. Secondary research intelligence consists of information that was collected for another purpose, but can be useful for other purposes. Examples of internal secondary research are sales revenues, sales forecasts, customer demographics, purchase patterns, and other information that has been collected about the customer. Often referred to as data mining, this information can be critical in diagnosing the problem for further exploration and should be leveraged when available and 11 appropriate. The amount of internal secondary information that can be applied is typically limited. External secondary research is typically far more available, especially since the Internet age. Most external secondary information is produced via research conducted for other purposes, financial performance data, expert opinions and analysis, corporate executive interviews, legal proceedings, competitive intelligence firms, etc. Now taking our case study perspective our decision problem can be framed directly as “What marketing strategy should be followed so as to increase the sales of products “A”, “B” and “C”?” Thus we can see from our case study example that decision problem is the problem which has been directly recognized by the management of the organization and hence have taken the services of research mangers to find a solution. 2.2 Research Objective/Gap/Opportunity Analysis/Decision alternatives Once the problem is better defined, manager can move onto developing marketing research approach, which will generally be around a defined set of objectives. Clear objectives developed will lend themselves to better marketing research approach development. This approach consist of honestly assessing the manger’s and its team’s market research skills, establishing a budget, understanding the environment and its influencing factors, developing an analysis model, and formulating hypotheses. Research objective have three components: The first is the research question. It specifies the information the decision maker needs. The second step consists of Hypothesis development that are basically alternatives answers to the research questions. The final step determines the Scope or boundary of the research. Now taking our case study perspective our Research Objectives can be framed directly as to assess preferences of consumers and customers for fresh milk and fresh milk products: Study the perception of MILKY Brand among customers Study the strategies adopted by the competitors in the region 2.3 Research Question The research question searches for all the information which is required to achieve the research purpose. The information provided by research question should help in decision making. It is also possible based on specific cases to have multiple research questions for a single research purpose. For example If the purpose is to determine if a new crop seed will be used by farmers, following questions can be posed: Will the seed be noticed in Agri market by adequate farmers? Will the farmer be ready for new experimentation? 12 Will it influence the using pattern of farmers across the region? Now taking our case study perspective our research question can be framed directly as To identify the consumer segments which MILKY brand Owner should target, positioning of its products and promotional strategies, to be followed to increase the sales. Thus we can clearly see from the case study example that research question is framed in the way that the information provided by it would help the researcher in solving the decision problem. As we all know marketing strategy would depend upon the Promotion strategy, positioning of product, targeted consumer segment, hence first of we as an researcher need to get the answer of these question so as to reach to the solution of our decision problem 2.4 Development of hypothesis This is basically alternative answers to the research questions. The research determines which of the alternative is correct. Normally, there will be several competing hypothesis, either specified or implied. Thus, one objective of the research is to choose among the possible hypothesis. The answer to the way in which a manager can develop a hypothesis is multifold. Whatever information is available with the manger is generally used to speculate on which answers to research questions are possible and which are likely. Firstly the manager takes help from the previous research efforts. A second source of Hypothesis is theory from various disciplines like geography, marketing, economics, psychology etc. like for example geography could help in indicating about the rainfall pattern suited to different type of seed germination. A third but the most important source is the experience of the manager with the related problem coupled with the knowledge of the problem situation and use of judgment. 2.5 Marketing Research Design Based upon a well-defined approach from Step 2, a framework for the designing the marketing research design is the next step which follows. Marketing research design is the most encompassing of all steps in the marketing research process, requiring the greatest amount of thought, time and expertise — and is the point at which those less experienced with market research will obtain assistance from an internal market research expert or perhaps partner with an external marketing research provider. Since the intelligence eventually gained from the research is so closely related to the selected marketing research design, this is the single most import step in the research process and the step most vulnerable to common marketing research errors Marketing research design includes secondary information analysis, qualitative research, methodology selection, question measurement & scale selection, questionnaire design, sample design & size and determining data analysis to be used. We now try to see this step in the form of a Flow chart 1, clearing enumerating in flow pattern the steps included while designing the research design. Each of these steps are 13 important in their own self as the final execution of the research will be dependent on the market research project conceived and planned. Research Design process is very critical step as a lot of research objective and result will depend upon the way in which the research is designed. While preparing research design invariably error crops up. There are majorly 2 types of error Sampling Error and Non Sampling error. 2.6 Errors in Sampling Although researchers tries to apply as much as foolproof method as possible but invariably some errors always creeps up while doing the sampling. Generally two types of error are seen while following the procedure of sampling, these are sampling error and non sampling error. If the difference in value between the population variable and the sample statistic is only because of sampling, then this error is termed as sampling error. But if simply while surveying a population error creeps up then that error is termed as non sampling error. 14 Flow chart-1 Research Objectives Research question Hypothesis development Research boundaries Estimate Value of Research Information Research Approach Exploratory Descriptive Figure: 1 Choice of data collection method Secondary and standardized data Qualitative methods Surveys Experiments Role of research supplier Project design Raw data collection Data Collection and Analysis Data collection Field work Data processing Data analysis Statistical analysis Interpretation 15 Causal Sampling error in turn can be minimized by increasing the sample size of the study but this may in turn reduce the quality of the research study and also it may lead to increase in non-sampling error. For example increasing the sample size may results in higher proportion of non response error from the sample size. Thus we need to have a perfect sample size so as to draw a balance between the sampling and non sampling error. Non Sampling error can be observed both in census as well as in a sample. Some of the common source of non sampling error includes measurement error, data-recording error, data analysis error and non response error. Since non sampling error can occur from varying sources, it is difficult to identify and control them. Therefore researcher tries to reduce them as much as possible. 2.7 Marketing Research Data Collection Marketing research data collection (often called survey fielding) is the point at which the finalized questionnaire (survey instrument) is used in gathering information among the chosen sample segments. There are a variety of data collection methodologies to consider. Selecting which is the most appropriate marketing research data collection methodology for a particular research project takes place during Steps 2 & 3 of the marketing research process. Marketing research data collection typically begins with field testing the final questionnaire with a small portion of the respondent sample to make sure it is gathering information correctly. Then data collection can be fairly automatic throughout the remainder of the marketing research data collection process. When quota groups and/or sample subgroups are being screened for, data collection will require more oversight, maintenance time and cost. Regardless of the data collection methodology chosen, the data collection process often takes from 25 percent to 50 percent of the total time needed to complete a research project. Market research data collection methods: Computer Assisted Telephone Interviewing (CATI) Internet survey Interactive Voice Response (IVR) Mail survey Mall intercepts Traditional telephone interviewing Internet panel Mail panel In-home panel 2.8 Survey Data Analysis Any survey data analysis will depend on how the survey questionnaire was constructed. Less complex survey data analysis can be handled with any of a number of office suite tools, while more complex questionnaire data analysis requires dedicated market research analysis programs. 16 Types of statistical survey data analysis that might be performed are simple frequency distributions, crosstab analysis, multiple regression (driver analysis), cluster analysis, factor analysis, perceptual mapping (multidimensional scaling), structural equation modeling and data mining. The more complex the needed level of statistical data analysis is, the more time and cost it will take to execute. Advanced Statistical Analyses ANALYSIS Multiple Regression (Driver Analysis) Cluster Analysis Factor Analysis DESCRIPTION Describes the relationship of each variable in a set (and the set of variables as a whole) to a single variable. Identifies homogeneous sub-groups within a much larger group of respondents. Reduces a complicated data matrix into its more basic structural essentials. Perceptual Mapping (Multidimensional Scaling) Extracts multiple dimensions from a variable set and positions concepts within that space. Structural Equation Tests how well observed Modeling data confirm an entire theoretical model. Data Mining Detects useful and sometimes unexpected patterns among variables in a data set. EXAMPLE APPLICATION Determine key "drivers" of overall customer satisfaction with your service. Identify customer profiles or market segments, groups of customers or potential customers who make similar decisions and perceive products and services similarly. Uncover basic dimensions employees use to evaluate how satisfied they are working for your organization. Visualize how customers mentally organize competitors in your product or service category and your brand's position relative to your competitors. Describe the process by which customer loyalty is built for your particular product or service category. Increase revenues by cross-selling your products. 2.9 Reporting and presentation The final step consists of reporting the results and the presentation. In this step the researcher shares his/her research work with the intended audience in a way they should understand. After this step researcher should be able to solve the questions related to research which have cropped up at the start of the process. 17 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 18 CHAPTER THREE 3. Research Discussion and Data Collection Methods All the research is classified into broadly three main categories: Descriptive, exploratory and causal. 3.1 Descriptive Research: - This research is used in the cases wherein an accurate result is to be obtained of a specific part. The purpose of this research is to provide an accurate snapshot of some particular aspect of the market environment. In this type of research Hypothesis often will exist, but they may be tentative and speculative. Here the development of hypothesis provides guidance to researcher by introducing more details to the research question. Descriptive research helps us in showing that the variables on which researcher would be working as related or associated. For example for a Tractor company it will be important to know ,the proportion of all possible outlets that are carrying, displaying or merchandising the product parts of tractor. 3.2 Exploratory Research:-This research is used in the cases where the research hypothesis is either vaguely defined or do not exist at all. This is the case where the researcher is usually seeking info about the general nature of the problem, the possible decision alternatives and the relevant variables which will be taken into consideration. Here in exploratory research the research method are highly flexible, unstructured and qualitative. This type of research is also used for knowing the preferences among the research questions and to go for the creative and unorthodox solutions. For example a dairy company can ask which product appeal will be effective in advertising. 3.3 Causal Research: - This type of research is done in the cases where there is cause-effect relationship. In these cases one variable causes or determines the values of other variable. Here in causal research, the research question and relevant hypothesis are very much specific. For example suppose a seed company has evidence that territories with good sales coverage had higher per capita sales of Seeds. Are their sufficient grounds to increase the sales coverage in the area where there is fewer sales? So here the causal analysis would be of great help for a researcher. 3.4 Data Collection methods There are 2 main sources of data- primary and secondary. Primary data collection is done from scratch. It is original and collected to solve the problem in hand. 19 Secondary data collection is also called as secondary research; here the data already exists since it has been collected previously for other purpose. 3.5 Primary Data Research This is called as primary research data because the data collection is done for the first time and for specific purpose. It is original and collected to solve a specific problem with specific objective. It’s expensive, and more time consuming, but is more focused than secondary data collection. There are many ways of doing this data collection which are enumerated below. 1. Diaries 2. Interviews 3. Focus Groups 4. Mystery shopping 5. Projective techniques 6. Product Tests 7. Omnibus Studies 3.5.1 Diaries: - Diaries are used by a number of specially recruited consumers. They are asked to complete a diary that lists and records their purchasing behavior of a period of time (Weeks, Months or years). It demands a substantial commitment on the part of the respondent. However, by collecting a series of diaries with a number of entries, the researcher has a reasonable picture of purchasing behavior of intended consumer segment. 3.5.2 Interviews: - This is the most common and most practiced technique associated with the marketing researchers. Interviews can be telephone, face to face or over the internet. 3.5.3 Telephonic interviews: - Telephonic interview is a very common way of getting data in developed countries. It’s ideal for collecting data from geographically dispersed sample. The interviews tend to be very structured and tend to lack depth. Telephone interviews are cheaper to conduct than face to face interviews. Advantages of telephonic interview : Cheaper than face to face interviews Random Samples can be collected Can be geographically spread Can be Set up fastly and conducted relatively cheaply Disadvantages of Telephonic interviews: Respondents sometimes are not interested in being interviewed Interview tends to be a lot shorter Visual aids cannot be used 20 Researcher cannot study or draw inference from Behavior or body language In short the summary of Telephonic interview can be summarized as : WHAT Telephonic interviews, based on set of questions, often computer aided WHEN Short questionnaires, god for B2B WHY USE IT Inexpensive, fast, accurate WHY NOT Only suitable for short, structured questionnaires, difficult to accommodate open ended responses, no non verbal signals Table: 1 3.5.4 Face to Face Interviews: - These types of interview are conducted between a market researcher and a respondent. Data is collected on a survey. Some surveys are very rigid and structured and uses closed questions. Data is easily compared. Other face to face interviews are more “in depth” and depend upon more open form of questioning. The research probes and develops various points of interest for researcher. Advantages of Face to Face interview Body language can be studied They allow more “in depth” data to be collected Physical prompts such as product and pictures can be used Respondents can be observed at the same time Disadvantages of Face to Face Interview Interviews can be expensive It usually takes a long period of time to arrange and conduct the interview Some respondents give biased responses when face to face with the researcher. In short the Face to face interview process can be summarized as WHAT Face to face interviews, based on standard set of questions WHEN For specific target groups, measuring specific views using visual images or products WHY USE IT Flexible, explanation possible, can measure non-verbal responses, accuracy, product placement possible WHY NOT Costly, time consuming and problems of getting people to co-operate Table: 2 3.5.5 Internet: - The internet can be used in a number of ways to collect primary data. Visitors to sites can be asked to complete electronic questionnaires. However responses will increase if an incentive is offered such as a free membership, free alerts etc. Other important data is collected when visitor sign up for membership. 21 Advantages of Internet Relatively Inexpensive Uses graphics and visual aids Random samples can be selected Visitors tend to be loyal to particular sites and are willing to give up time to complete the information. Data released by respondents are unbiased as their identity may be unknown to researcher. Disadvantages of Internet Survey can be done only of Current, not potential customers It needs knowledge of software to set up questionnaires Methods of processing data may deter visitor from visiting the sites. In short the data collection through Internet can be summarized as WHAT Internet based surveys, either by panels or a broader non-targeted group WHEN Short questionnaires, easy and fast to answer questionnaires, for measuring fast changing events WHY USE IT Inexpensive, very fast, large samples , broad based WHY NOT Only short and easy questionnaires, can’t reach people with no internal access( e.g. old people, or specific geographies), lack of personal contact Table: 3 3.5.6 Mail Survey: - In many countries, the mail survey is the most appropriate way to get primary data. Lists are collated, or purchased, and a predestined questionnaire is mailed to a sample of respondents. Mail surveys do not tend to generate more than 5-10 percent response rate. However, a second mailing to prompt or remind respondents tends to improve response rates. Mail surveys are less popular with the advent of technologies such as internet and telephones. WHAT WHEN WHY USE IT WHY NOT Mailed, self completion questionnaire For existing customer base, or where target audience is geographically dispersed Low cost, more complex lengthy questionnaires possible, no interview bias Low response rate, long read time Table: 4 3.5.7 Focus Groups:-Focus groups are made up from a number of selected respondents based together in the same room. Highly experienced researchers work with the focus 22 groups to gather in depth qualitative feedback. Groups tend to be made up of 10 to 18 participants. Discussions, opinion, and beliefs are encouraged and the research will probe into specific areas that are of interest to the company commissioning the research. Advantages of Focus Groups All participants and the research intact Areas of specific interest can be covered in greater depth Visual aids and tangible products can be circulated and opinions taken Commissioning marketers often observe the group from behind a one way screen Disadvantages of Focus Groups Highly expensive researchers are needed. Complex to organize because of large number of respondents. Can be very expensive in comparison to other methods Again in Short the Focus group can be summarized as WHAT Group of 6 to 10 people with an experienced moderator WHEN Broad Market studies, creative development, ideas needing unconscious thought WHY USE IT Fast, interactive, flexible, inexpensive and specific WHY NOT Group represents a small sample, requires good recruitment, cannot forecast precise business results Table: 5 3.5.8 Mystery Shopping: - Companies set up mystery shopping campaigns on an organization behalf. Often used in retailing, banking, travel, cafes and restaurants and many other customer focused organizations, mystery shopper will enter, posing as real customers. They collect data on customer service and customer experience. Findings are reported back to the commissioning agencies. There are many issues surrounding the ethics of such an approach to research. 3.5.9 Projective Techniques: - Projective techniques are borrowed from the field of psychology. They will generate highly subjective qualitative data. There are many examples of such approach including: Inkblots tests- look for images in a series of inkblots cartoons- complete the bubble on a cartoon series sentence or a story completion word psychodrama- imagine that you are a product and describe what it is like to be operated, warn ,or used. WHAT WHEN Psychological technique based on projecting as social image To understand attitudes and opinions which may be difficult to express, or for socially 23 WHY USE IT WHY NOT sensitive subjects Goof for difficult issues Participants can unconsciously affect results by not lowering their barriers. Results may be difficult to interpret Table: 6 3.5.10 Product Tests: - Product tests are often completed as part of the test marketing process. Products are displayed in a mall of shopping center. Potential customers are asked to visit the store and their purchase behavior is observed. Observers will contemplate how the product is handled, how the packing is read, how much time the consumer spends with the product and so on. 3.5.11 Omnibus Studies: - An Omnibus study is where an organization purchases a single or a few questions on a “hybrid” interview (either face to face or by Telephone). The organization will be one of many that simply want to straightforward answer to simple question. An Omnibus survey could include questions from companies in sectors as diverse as health care and tobacco. The research is far cheaper and commits less time and effort than conducting your own research. In short for Omnibus survey WHAT WHEN WHY USE IT WHY NOT General study with specific questions for large number of categories, products and brands For obtaining basic information, first approach to new market, measuring impact of advertising Cost effective, easily accessible Not very flexible, only basic information Table: 7 3.6 Secondary Data research Secondary data are data that were collected by persons or agencies for the purpose other than the problem or objective at hand. It already exists in one form or the other. This is taken from 2 major sources External and Internal. Secondary data research is a basic but important step that researcher takes initially before committing to the expensive task of generating primary data. This type of data research yields good result but has shortcomings. The information can be out of date, may not quite fit the definition or may be too basic for the project. It is relatively cheap, and can be conducted quickly. Also it may be untargeted, and difficult to use to make comparisons (e.g. financial data gather on Indian people below poverty line to data on American population below poverty line). There are a number of sources available to marketer and are explained below. Contd. 24 3.6.1 Internal Sources These are the in house sources providing the most basic data: a) Internal Records: - These are the record which provides the most basics data on marketing inputs and the resulting outcomes. The principal benefits of these data are availability, reasonable accessibility on a continuing basis, and relevance to the organization’s situation. Their can be various forms of reports which help in getting the required info. Budgets, Schedules of expenditures Billing records Sales report Sales person’s call reports b) Customer Feedback:-Now a day’s companies are not only relying on internal records but are also collecting data in various forms from customer so as to augment their data base. Complaint letters are also being used as source of data on product quality and service. This helps in providing insights to the company. But there is also flip side to it as Manufacturers would be completely cut off from the actual reason of dissatisfaction of customers, because most complaints are only voice to the retailers and there is little systematic feedback from retailers to manufacturers. Also people who usually give feedback are not typical client or customers but highly educated ultra modern customers. In case of Agri business fold, customer feedback can be a very good tool for data analysis as the field in which agro business companies are working is more related to base root level but the problem of mostly unaware and less educated customers poses a big challenge in getting customer feedbacks. c) Customer Database: - Customer database is another important way of collecting data as it contains raw information which can be sorted and enhanced to produce useful information. Records of customers and their buying habit is recorded in the database and this helps in finding the difference/similarity between the buying habits of various set of customers. Now a days these customer databases are used by research manager to help marketing manger, formulate marketing strategies. 3.6.2 External Sources a) Census Data:This is the most comprehensive data which provides the most extensive and descriptive facts about the population, economic and the social environment. Census data is obtained at many level of aggregation. Generally starting from smaller units the data compilation goes to bigger units. 25 Like for e.g. if we talk about Data census for a Rural area, the smallest unit would be a household which will compile to a village then to tehsil then to Divisional region and finally to a district. For an Urban landscape again the smallest unit would be Household which will compile to city then to a divisional region and finally to a district. Finally the data from above are compiled in the form of States and finally total data gets compiled to make the aggregate of the country. Above given aggregation is one example of aggregation wherein the data is aggregated in various different sorts of ways. Thus Census data is the most comprehensive way of getting data which is complete in sort of requirements socially, economically and in terms of population. b) Published data sources:These are the most popular source of getting market information. The major published sources are the periodical, journals, government publications. Also the report from private sources such as trade unions, foundations, publishers and companies comes under this. c) Computer Retrieval database: Although the published data source and census data is quite comprehensive and useful in work but a search through them usually takes lot of time. So now days with advancement in Computer technology cataloging, storing and retrieving published data has became very efficient. These databases are usually accessible from personal computers as well as through shared terminals. Also there is numerous software which had made retrieval of databases quite user friendly. As a result use of these electronic information sources has expanded rapidly to facilitate almost any search for information. Summary for the secondary data research WHAT Easy to access, widely available. free(mostly) edit-and-mix source of information WHEN First approach for general information WHY USE IT Cheap , low risk, requires few resources, fast WHY NOT Not specific or tailor made, not exclusive, may be out-of date or inaccurate Table: 8 Thus Secondary research can yield good results but has shortcomings. The information can be out of date, may not quite fit the definition or may be too basic for the project. 26 3.7 Data collection through the lens of our case study? Our case study presents a very good example to understand the use of data research method. We have done data collection through primary as well as secondary data research method. This is explained below. 3.7.1 Use of Primary data research Primary data research is used here because to frame marketing strategies and to know consumer preference, it is very important to go face to face, focus group etc because these methods will help in giving insight to the consumption pattern of the area. This will also help in giving information to the researcher that what is going well for the brand in the market and what is hampering its growth prospect. This will help in formulating marketing strategy for it. Also by this, researcher can peep deep into the customer preferences and their buying habits. Thus primary data collection was done with the following groups by way of face to face interview, focus group discussion, and mystery shopping and product tests. Consumers of Product “A”,”B” and “C” (both – users and non users of “MILKY” Brand)) Retailers of “SAGAR” as well as of competitors. Distributors of “SAGAR” as well as of Competitors Private vendors of product “A”, “B” and “C” Owners of various shops and stalls which uses the above product Restaurants Sales department staff of the “SAGAR” company. 3.7.2 Use of secondary Data research As explained above this is the data which is already present with the company and has been collected for the purpose other than the research objective. But in the case of our case study, this would be of great help because it will help in giving us a fair idea about the historic sale pattern of the brand “MILKY”. By this researcher can know that in the past which marketing activity has affected the sales in what way. Also it will be of great help in telling the researcher the historic sales and consumer purchase pattern of the brand “MILKY”. For this following data sources were taken into consideration: Sales records of “SAGAR” company Sales records of the retailers and distributors of “SAGAR” Demographic data from internet for the research region. 27 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 28 CHAPTER FOUR 4.0 Designing the Questionnaire Questionnaire is another very important step in the road to achieve the desired research objectives. Questionnaire are custom built to the specification of the given research purpose, and hence they are a collection of logically framed questions. In simple terms to get meaningful answers we need the right questions. So questionnaire is a formalized, consistent set of questions used to obtain information from respondents, thus it is therefore very critical. In a questionnaire the questions are structured in a logical sequence and use simple language that will be understood by a wide range of respondents of different occupation and intelligence. The test of a good questionnaire is that questions will yield true answers that can build accurate data for the clients. The questions should never steer the respondents to give a particular answer as this will skew the result. Most problems with questionnaire analysis can be traced back to the design phase of the project. Well-defined goals are the best way to assure a good questionnaire design. When the goals of a study can be expressed in a few clear and concise sentences, the design of the questionnaire becomes considerably easier. The questionnaire is developed to directly address the goals of the study. As a general rule, with only a few exceptions, long questionnaires get less response than short questionnaires. In fact, shorter the better. Response rate is the single most important indicator of how much confidence researcher can place in the results. A low response rate can be devastating to a study. Therefore, researcher must do everything possible to maximize the response rate. One of the most effective methods of maximizing response is to shorten the questionnaire. Sometimes incentives are also provided as motivation for a properly completed questionnaire. What does the respondent get for completing questionnaire? Altruism is rarely an effective motivator. Attaching some money with the questionnaire works well. If the information researcher collecting is of interest to the respondent, offering a free summary report is also an excellent motivator. Whatever researcher chooses, it must make the respondent favorably inclined towards completion of questionnaire. 4.1 So what is a questionnaire? It’s a set of specially designed questions to which answers are written on a pre-prepared form. It tells researcher who their customer might be in demographic and psychographic terms. It tells researcher certain things about customer’s behavior and lifestyle. 29 It’s a way of finding out exactly what customer knows and need to know about the research topic. It contains up-to-date data which is not available from any other source. It helps in the construction of a text and in generating advertising to fund that text . Before designing of any questionnaire it’s very important to understand that what information researcher need’s and from whom. Questionnaire is a tool to help researcher get that data. A poor quality questionnaire will yield poor quality data. 4.2 How the answers be framed in a questionnaire In general practice, questionnaire is devised to provide three types of answers: 4.2.1 Scaled: Respondents have a choice of answers which are a matter of degree from ‘Agree strongly’ to ‘disagree strongly’. This type of scaling is the most common way of answers found in questionnaire as it gives more specific and wide knowledge about the response of a question. 4.2.2 Open-Ended: This type of question can elicit a wide range of answers. E.g. why do you go for X brand of rice? Thus respondents are free to put in the answer which suits their reason. 4.2.3 Fixed Choice/Pre-coded/Closed: These types of answers would be of the type Yes/No or multiple choice type list. This type of answers is the easiest for the respondents to understand 4.3 Qualities of a good Questionnaire Questionnaires should be clearly laid out and easy to read. Language used in questionnaire should be simple and direct. a) Keep it short; information overload will hamper the respondent to reveal all the data required for research b) Use of multiple choices or yes/no answers always make’s it easier to analyze the data c) Generally the best pattern followed in questionnaire is to start off with easier questions (age, occupation etc) and finish with the ones that have to be thought about a little more – this gives interviewee a chance to warm up and focus on the topic d) Majorly good response rate is the biggest criterion to decide the success of a questionnaire , but there are some other factor’s also depending on we can commit whether our designed questionnaire is up to the mark or not. Some of these are mentioned below : contd. 30 e) A good questionnaire in all probability must include all the possible answers which respondents can give. Asking a question that does not include all the possible answers can confuse the respondents and will lead to wrong information being collected. For e.g. consider the question Your Tractor is from which manufacturer? Sonalika Taffe Clearly, there are many problems with this question. What if the respondent doesn't own a tractor? What if he owns a tractor from some other manufacturer? What if he owns both a Sonalika as well as Taffe tractor? There are two ways to correct this kind of problem. The first way is to make each response a separate dichotomous item on the questionnaire. For example: Do you own a Sonalika tractor (circle: Yes or No) Do you own a Taffe Tractor? (Circle: Yes or No) Another way to correct the problem is to add the necessary response categories and allow multiple responses. This is the preferable method because it provides more information than the previous method. Which manufacturer’s tractor do you own? (Check all that apply) __ do not own a tractor __ Sonalika __ Taffe __ Other f) Questionnaire should have mutually exclusive options. A good question leaves no ambiguity in the mind of the respondent. There should be only one correct or appropriate choice for the respondent to make. An obvious example is: What do you drink? Tea Milk Coffee A person who depending upon his mood either drink A or B or C would find this question really stupid and meaningless. Worse this can frustrate the respondent and he may lose interest in filling the information for rest of the questions.There should be clear cut difference in the exact meanings of the responses to the respondents. Variability in response all helps in applying statistical tools to the data collected. For example How was the food? A. It's the worst food I’ve ever eaten B. It's somewhere between the worst and the best C. It's the best food I’ve ever eaten Since almost all the respondents will give choice B as their answer, hence this question will be of no help for the researcher. 31 g) It is very important that question asked in the questionnaire must not make a unwarranted assumption. This is one of the most subtle mistakes which happen while designing a questionnaire. For example Are you satisfied with your loan interest rate? This question will certainly pose a problem to a researcher who has not taken a loan. h) Good questionnaire should not use emotionally loaded or vaguely defined words. This confuses the respondents and lead to wrong collection of data. Quantifying adjectives like (most, least, majority) are generally used frequently in the questionnaires but it should be kept in mind that these adjectives mean different things to different people. i) Questions in a questionnaire should ask for an answer on only one dimension. The purpose of a survey is to find out information. A question that asks for a response on more than one dimension will not provide the information researcher is seeking. For example, a question which ask “Do you like the flavor and taste of the food dish” If a respondent answers “yes” then the researcher will not know if the researcher likes the taste or flavor , or both. Thus a good question only asks for a single bit of information in a question. By following above laid guidelines we can make a fool proof questionnaire but still there are some disadvantages associated while working with a questionnaire. These are explained below. Disadvantages of Questionnaires : The biggest disadvantage while working is the low response rates. Low response rate greatly affects the accuracy of the result. Questionnaire cannot take into account visual and verbal cues as respondents are giving details in a set formats of questions. Thus a questionnaire probing for factual information will not be affected by lack of personal contact but a questionnaire probing sensitive issue may be seriously affected. Also written questionnaire will not suit some group of people. For example in domain of village large population is still uneducated , hence if we are doing survey with the farmers for many written questionnaire would be of no use. Now for our case study, you can find below the questionnaire which has been formed taking into consideration all the above points. By making 2 different questionnaires for 2 different source of information, the researcher is taking a holistic view and covering all possible sources which will help in fulfilling the assigned research objectives. As you see it starts with the general question first and then slowly momentum is build up Wherever possible use of multiple choice or (yes/No) is done to make it easy for respondents. Also the language of questionnaire is simple and straight to the point making it very simple for the respondents. All the maximum possible solutions of a question are covered. You can see extensive use of table with the scale varying 1-3 help in giving good info about the consumer preference and purchase pattern. 32 So by keeping into mind the importance of a good response rate for the research, questionnaire has been design to collect the data. This questionnaire was designed specifically for the household i.e normal day to day consumers as well as for the retailers/distributors. These were the two totally different segments which were very important for the research process and hence data needs to be tapped from both the sources to fulfill the basic research objective. 4.4 Questionnaire of our case study While designing the questionnaire in our case study, we have kept in mind all the above discussed points. 1. Name 2. Address 3. Respondent is a Housewife Male Member 4. Average monthly household income (a) Less than 5000 (b) 5000 to 15000 (c) more than 15000 Any other specify_____ QUESTIONS FOR PRODUCT “A” 5. What is the daily consumption in the household? _________ 6. When do you buy it? (Please tick option(s)) a. Morning b. Afternoon c. Evening d. When need arises 7. In which form do you buy the product (Please tick the appropriate ones) a. Packed b. Loose c. Powder 8. In case of packed product “A”, which brand are you consuming most? a. MILKY b. X c. Y e. others 9. At what price do you buy? ________ Rs/litre 10. Payment for the purchased a. Advance b. Daily c. Weekly d. Monthly e. No particular schedule 11. Have you changed the brand of your purchase in the past 2 years? Yes / No_______ 12. If yes, which brand were you buying? a. MILKY b. X c. Y d. Z 13. Reason for change in Brand purchase pattern? [Tick (v) whichever is applicable] Inconsistent quality Inconsistent quantity Credit problem High Price Lacking Freshness Bad taste Recommended by other Late delivery Promotions of other brand users Recommended by retailer Pack size Nearness of present source Any other, Pl specify _______________________ 33 14. Which of the following attributes influence your decision while buying product “A”:Strongly Influence Influence Do not influence Quality Price Odour (Smell) Purity Taste Freshness Packaging Convenience of purchase Consistent in quantity 15. Have you ever been exposed to any promotional campaign by brand “MILKY”? Yes/No 20. How do you rate product of “MILKY” brand on the following attributes? (Please tick) Quality Poor Satisfactory Good Price Cheap Right price Costly Odour (Smell) Bad Ok Good Purity Less pure Ok Pure Taste Bad Ok Good Freshness Not fresh Ok Fresh Packaging Bad Ok Good Convenience of Inconvenient Ok Convenient purchase Consistent in quantity No Do not know Yes For Non users of Brand “MILKY” 21. Do you want to shift to brand “MILKY”? Yes/ No 22. What will make you shift from your current Brand to brand “MILKY”? For users of Brand “MILKY” 23. What improvements / suggestions do you have for “MILKY” Brand? QUESTIONNAIRE FOR RETAILER/DISTRIBUTER (PRODUCT “A”) Name of the Outlet: ___________________________ Address: ____________________________ Location: __________________________ (commercial / residential) Nature of Outlet (Please tick one) …General Stores …Bakery …Parlour …Any other, please specify______ Age of Outlet (Year) : _______ What is the sales pattern of the product in a year? 34 PRODUCT “A” Peak Normal Lean Do you offer Home delivery of product “A”? Yes/No If yes, do you charge extra money for that? How much product “A” do you sell in a day on an average? _______ What is your preference for stocking a particular brand? Very Slightly Important Important Not important Commission Brand Name Good Quality Easily Available Sales promotion schemes offered Promotional Support Product Replacement Efficient Distribution Quality of packaging How satisfied are you being a retailer of “MILKY” Brand of products, on following attributes? Satisfied Neutral Dissatisfied Commission Brand Name Good Quality Easily Available Sales promotion schemes offered Promotional Support Product Replacement Efficient Distribution Quality of packaging What are the margins offered by competitor brands? What strategies should be adopted by “MILKY” brand to increase its sales? Thus in the same way Questionnaire were designed for the product “B” and “C” keeping into consideration the product attributes and their targeted customers segment. So I hope the above questionnaire gives you a fair bit of idea as to how the questionnaire is designed? What is the logical order of questions in a questionnaire? What are the salient points to be kept in mind while framing the questions and their responses? 35 4.5 PRETESTING Pretesting is done to check whether the questionnaire meets the researcher’s expectation in terms of the information that will be obtained. Pretest in thus in another terms can be called as a pilot test and hence respondents taken for Pretesting are a reasonable representation of the sample population. Thus pretest is of great help to find the correctness of the questionnaire for practical purpose. But there are limits to which the Pretesting can help in detecting the error in questionnaire. Sometimes it may also happen that respondents are unable to detect the flaws of a questionnaire. Thus respondents are not the only sources of insights, interviewers are also equally important for the pretesting process. Generally interviews are also taken in line with the pretesting, here interviewers’ reports their experience as well as suggestion are also asked from them. Finally pretest analysis is only complete when the researcher should take into consideration the results of it and return to the designing stage. Each question should be review once again and asked to justify its place in the questionnaire. How can the answer be used in analysis? Is the pattern of answers from the pretest sensible, or difficult to interpret? Does the question add substantial new information, or unnecessarily duplicates the results from another Question? 36 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 37 CHAPTER FIVE 5.0 Sampling Sampling is the step where the researcher defines the target population. Sampling help’s researcher in logically identifying the set of people on whom his research will be based upon. There are no strict rules to follow, and the researcher must rely on logic and judgment. The population is defined in keeping with the objectives of the study. A company has to rely on getting the views of a sample of people- and by choosing that sample correctly their views will represent the views of larger groups. It is all about statistics. A good sample provides a true result with only a small margin of error and this has been proven over the years with commercial and social research. The sampling should be done so as to get the accurate data collection from the set of population which is significant for the research. Choosing and finding the right people and deciding how many to be taken for study is the clearly very important for the research study. For example, if the research objective is “Whether Levi’s should enter sports shoe market” Sampling for the fulfillment of above objective will have inadvertently a sample size drawn majorly from the age group 15-25. As this would be the target segment which will mostly go for new sport shoe brand. What Kind of sample to interview The kind of sample will vary according to the research needs and different companies will be interested in different group of the population. Definition of the target population should be perfectly correct. When we talk about the target population it actually connotes towards segmentation. In real world people are not the same and they do not all like the same things. There is no ‘one size fits all’. The market has space for very different products and services in the same category, as well as different kinds of products competing at the same price level. Thus this leads to the concept of segmentation and marketer describes the population they want to attract to their product as their target group. Once a target/segment is clearly defined researcher select a small group of people to represent the whole. The target group is known as the ‘universe’ or ‘population’ and the small group as the ‘sample’. Sometimes, the entire population will be sufficiently small, and the researcher can include the entire population in the study. This type of research is called a census study because data is gathered on every member of the population. Usually, the population is too large for the researcher to attempt to survey all of its members. A small, but carefully chosen sample can be used to represent the population. The sample reflects the characteristics of the population from which it is drawn. Sampling methods are classified broadly in two categories : A) Probability Sampling and B) Non Probability Sampling 38 5.1 Probability Sampling In probability sampling, each member of the population has a known non-zero probability of being selected. Thus each member of the population has an equal finite probability of being selected as respondents. Thus in this techniques a group of subjects are chosen from a large group (population). Each individual is thus chosen entirely by chance and each has a known probability of being selected. The advantage of probability sampling is that sampling error can be calculated. Sampling error is the degree to which a sample might differ from the population. When inferring to the population, results are reported plus or minus the sampling error. A population can be defined as set of all objects that possess some common set of characteristics with respect to the marketing research problem. Sampling frame is the list of population members used to obtain a sample. For example this can be obtained by selecting a district, a town, a village, a hamlet and finally a farmer from the set of farmers in the hamlet Probability Sampling methods include the following : 5.1.1 Random sampling:-This is the most basic and purest for of probability sampling. In this type of sampling each member of the population has equal and known chance of being selected. When there are very large populations, it is often difficult or impossible to identify every member of the population, so the pool of available subjects becomes biased. 5.1.2 Systematic sampling: - This type of probability sampling is now a day often used instead of random sampling. It is also called an “N”Th name selection technique. After the required sample size has been calculated, every “N” th record is selected from a list of population members. As long as the list does not contain any hidden order, this sampling method is as good as the random sampling method. It advantage over the random sampling technique is its simplicity. 5.1.2 Stratified sampling: - This type of probability sampling is commonly used method that is superior to random sampling because it reduces sampling error. A stratum is a subset of the population that shares at least one common characteristic. The researcher first identifies the relevant stratums and their actual representation in the population. Random sampling is then used to select subjects from each stratum until the number of subjects in that stratum is proportional to its frequency in the population. Stratified sampling is often used when one or more of the stratums in the population have a low incidence relative to the other stratums. 5.2 Non Probability Sampling In case of non probability sampling, respondents are selected from the population in some nonrandom manner. Thus in case of non probability sampling there is no particular order or logical/scientific way in which researcher goes for respondent selection. Hence in this technique random method technique of selection is not used instead human judgment, and not chance is used to select individual for the sample. The researcher builds a sample based on defined characteristics e.g. age range, gender or socio economic group. The 39 disadvantage with this type of sampling is that researcher cannot apprehend degree to which the sample differs from population. Non Probability sampling methods include the following : 5.2.1 Convenience sampling: - Convenience non probability sampling is used in exploratory research where the researcher is interested in getting an inexpensive approximation of the truth. As the name implies, the sample is selected because they are convenient. This nonprobability method is often used during preliminary research efforts to get a gross estimate of the results, without incurring the cost or time required to select a random sample. Thus this method is used wherever researcher is just interested in getting an approximation the result. 5.2.1 Judgment sampling: - Judgment non probability sampling is a common nonprobability method. Here the researcher selects the sample based on his/her own judgment. This is usually an extension of convenience sampling. For example, a researcher may decide to draw the entire sample from one "representative" city, even though the population includes all cities. When using this method, the researcher must be confident that the chosen sample is truly representative of the entire population. 5.2.2 Quota sampling: - Quota sampling is the non probability equivalent of stratified sampling. Like stratified sampling, the researcher first identifies the stratums and their proportions as they are represented in the population. Then convenience or judgment sampling is used to select the required number of subjects from each stratum. This differs from stratified sampling, where the stratums are filled by random sampling. 5.2.3 Snowball sampling: - Snowball sampling is a special non probability method used when the desired sample characteristic is rare. It may be extremely difficult or cost prohibitive to locate respondents in these situations. Snowball sampling relies on referrals from initial subjects to generate additional subjects. While this technique can dramatically lower search costs, it comes at the expense of introducing bias because the technique itself reduces the likelihood that the sample will represent a good cross section from the population. 5.3 SAMPLING PROCESS FLOW CHART Target population Sampling Frame 40 Selecting a sampling procedure Probability Sampling Random Sampling Systematic Sampling Stratified sampling Non Random Sampling Convenience Sampling Judgment Sampling Quota Sampling Snowball Sampling Deciding the Sample size for sampling Sampling process Data collection for respondents Decision making process To resolve the issue of non-response problem Flow chart-2 What should be a sample size? A sample size can be of few hundred or of few thousand people. The size will depend upon the accuracy required, variability of the group and the sample design taken for research. The key issue which is taken into consideration by researcher while deciding for the sample size is to ensure that the sample provides accurate results. Also in general it is observed that larger the sample, the smaller is the sampling error but a larger sample means increased cost, increased time and can also lead to more error if interviewers are untrained or there is question bias. However, homogeneity of the target group is also very important. A company interested in the population of parents using child care facility may find they are quite similar (homogenous universe) while a company interested in Wine drinkers may find all these consumers are quite different (Heterogeneous Universe). 41 5.4 Sampling Errors Sampling Error:-Sampling error is the difference between a measure obtained from a sample representing the population and the true measure that can be obtained from the entire population. This error can never be overcome as we cannot have perfect population as our sample size. 5.5 Non Sampling errors Non Sampling error:-This includes all other error except the sampling error. Further this can be of 4 types. Design error, administrative error, response error and non response error. 5.5.1 Design Error: - These are mainly due to flaws in the research design. 5.5.1.1 Selection error:-This error occurs when a sample obtained from a non probability sampling method is not representative of the population. 5.5.1.2 Sampling Frame Error:-A sampling frame is a directory of population member from which a sample is selected. This error occurs when the sample is drawn from an inaccurate sampling frame. 5.5.1.3 Population Specification Error:-This error occurs when an inappropriate population is chosen from which to obtain data for the research study. 5.5.1.4 Surrogate Information error:-This type of error occurs when there is difference between the information gathered by the researcher and the actual information required for research. 5.5.1.5 Measurement Error:-This type of error occur when there is difference between information sought by a researcher for a study and the information gathered by a particular measurement procedure employed by the researcher. 5.5.1.6 Experimental Error: An experiment is designed to determine the existence of any causal relationship between the two variables. Any error caused by the improper design of the experimental induces an experimental error into the study 5.5.1.7 Data Analysis Error:-This type of error occurs when the data from the questionnaire are coded or edited wrongly. 5.5.2 Administrative Errors: - These are the errors which are caused by the mistakes committed by the person administrating the questionnaire. These may be caused by following major factors: 5.5.2.1 Interference error:-This error occurs when the person taking the interview fails to follow the exact procedure while collecting data. 42 5.5.2.2 Recording error:-This error occurs because of improper recording of interviewee’s response. Generally it happens when the interviewer misinterprets the response. 5.5.2.3 Questioning error:-This error arises while addressing questions to respondents. Sometimes happen that interviewer is not able to exactly frame the question and which results in improper data collection. 5.5.3 Response Error: - These are also called as data errors. This type of error occurs when the respondents intentionally or unintentionally provides inaccurate answers to survey questions. This happens when the respondents are unable to correctly comprehend the question. 5.5.4 Non Response Errors:-This type of error can happen because of two fold reasons. Sometimes the member under survey provides an incomplete or no response to the interviewer. Secondly it may happen that some members of the data collection sample were never contracted. This may be sometimes due to the unwillingness of the interviewer to contact the respondents. 5.6 Sampling plan for our case study As explained in the previous topic, for our case study we have taken into consideration three set of population Household, Distributor and retailers. So please find attached below the sampling plan for the household population. As per the segmentation process, we have started with identifying the target population which in here is the household, distributors and retailers. Then second step is deciding about the sampling frame with the help of segmentation. Here we have done it in two ways : 5.6.1 Household Survey A survey of Household for product “A” was conducted with the help of a structured pretested questionnaire. The household were identified in consultation with the sales staff of company, and a representative sample was taken on the basis of judgement. Thus we have done non-probability judgement sampling because here in this case we need to go for a structured format of sampling to get the desired data. 5.6.1.1 By Income Income group I: Families, which are having monthly income less than Rs 5000. Income group II: Families, which have monthly income in between Rs 5000 – 15000 Income group III: Families, which have monthly income in excess of Rs 15000. 5.6.1.2 By average daily sales of product” A” 43 ‘A’ Grade towns: These are the towns where average daily sale of product “A” is in excess of 8000 lts. ‘B’ Grade towns: These are the towns where average daily sale of product “A” is less then 8000 lts but more than 1000 lts. Towns A Grade B Grade TOTAL Income Group I Group II Group III TOTAL 80 80 80 240 60 60 60 180 140 140 140 420 Table: 9 Out of four towns in A grade category and six towns in B grade category, we have taken two towns from each category which represents the whole category. Towns in which the company is selling less than 1000 lts per day have not been considered because of the low sales volume and time constraint. 30 respondents have been interviewed from each city in each income category Since the company is selling approximately 80% of its product “A” in the four A grade towns, we have taken a sample of 240 from these and 180 from B grade towns. 5.6.2 Household /Retailer survey Here in this case also we have done non-probability judgement sampling because again we need to go for a structured format of sampling to get the desired data. Survey of retailers and distributors has been carried out to find the perception of retailers and distributors regarding “MILKY” Product. It has also helped in knowing their perception for stagnant sales, what steps should be taken to increase the sales and what are the strategies that the competitors of company are employing. The sampling plan is as under: Sampling Plan for Retailers/Distributors Retailer/Dist “MILKY “ Brand Town A grade town 12 B grade town 8 Table: 1 44 Competitors 6 4 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 45 CHAPTER SIX 6.0 Preparation of Data for Data Analysis Data analysis can lead the researcher to information and insights that otherwise would not be available. It can also help avoid erroneous judgments and conclusions. Also it can provide a background to help interpret and understand analysis constructed by others. Finally, knowledge of the power of data analysis techniques can constructively influence research objectives and research designs. The ‘raw’ research data needs to be edited, tabulated and analyzed to find the results and to interpret them. -The method used may be manual or computer based. -The analysis plan follows from the research objective of the study. -Association and relationships of variables are identified and discussed in the light of the specific marketing problem Thus role of market researcher becomes very important in this analysis part, as researcher would be the person who will have an impact on the higher thinking elements of the research. The same time and effort is required to understand the client’s needs, think about the best questions, structure the survey, explore the data and ‘open ends’, develop hypothesis and validate them, and then develop intuition. Thus the challenge for the researcher is to make sense of the raw data and information so that the client- be it a government department or a drinks manufacturer-can use this knowledge to make better decisions. The researcher’s experience enables a clear distinction between the direct findings of the research, interpretation of those findings and overall recommendations. The type of data analysis required will be unique to each study. However nearly all studies involving data analysis will require the editing and coding of data, will use one or more data analysis techniques, and will have to be concerned with presenting the results effectively. Thus the whole process can be presented in the following flow of steps. It starts with data editing, coding, and statistically adjusting the data for further analysis. Then basic ways of tabulating the individual questions from a questionnaire will be developed. The discussion on tabulation will also include graphical representation of tabulated data for analysis part. Next the focus will turn to the question of tabulation among sample sub groups (cross tabulation). Then various statistical technique are decided upon case to case basis that a researcher can use in analyzing data. Thus the first step involves the preliminary preparation before making the data ready for the analysis using statistical techniques. The final analysis and the subsequent result from that will highly depend upon how well the data were prepared and converted into a form suitable for analysis. So let us start with the data editing. 46 6.1 DATA EDITING This is the first stage in the process of preparing the data for final data analysis. In this stage responses are studied by the researcher. Various errors such as ambiguities, error in the response, interviewer error and non eligible responses are taken into consideration. There are numerous alternative to solve these errors. The preferred alternative is to contact the respondents again; this will help in clearing any doubts regarding the answers given by the respondents. Second alternative which can be used is to totally discard the respondent’s questionnaire if it is quite clear that the respondent either was not able to understand the survey question or was not cooperating. Third alternative which can be used is to throw out only the problem questions and retain the rest of the questions. The reason for this is that some respondents will try to give answer for only those questions which they are interested in. Thus such an approach may simplify the data analysis without materially distorting the interpretation. 6.2 CODING Coding process will be different for closed ended as well as for open ended question, but the general process remains the same for both. Only difference is that in closed ended question coding process is quite straightforward whereas in case of open ended question is it’s a lengthy cumbersome process. In case of open ended questions we start with putting the questions in a computer generated sheet with the possible outcomes in front of that question. Each possible outcome is assigned a key value. Once the response values are entered a computer file, a statistical software program is employed to generate information. In case of open ended questions again the question is put in the computer generated sheet and all the possible outcomes are placed in front of that. But here this is much more complex process as for open ended questions their can be many outcomes possible. So grouping of same sort of answers in a single head are done and the onus lies on the researcher for correctly putting the same questions in a single head. After completing this process then keys are assigned to each head. The difficulty of coding and analyzing open ended responses provides a reason to avoid them in the questionnaire whenever possible for the researcher. 6.3 Adjustment of Data statistically This is done in order to enhance the quality of the data. There are numerous ways by which we can improve the quality of data so that it can be used for statistical data analysis. These are explained below Contd. 47 6.3.1 Scale transformation: - In this method of data adjustment scale values are manipulated to ensure comparability with other scales. The reason is that in the same study it may happen that different scales are employed for measuring different variables. So to have a comparison among them it is important to have some common scale. One of the most scale transformation techniques is called as standardization. It allows the researcher to compare variables that have been measured using different types of scale. For example if suppose sales is measured in Rs and price in Paisa then the actual value of the variance for the sales variable would be higher compared to price, because of unit of measurement. Thus to compare these variables both are brought down to common unit of measurement. 6.3.2 Dummy Variables: - This method is used for specifying the variables. If there are N levels of qualitative variable then, then N-1 dummy variable will be used to specify them. For example there would be 2 crop season (a qualitative variable with 2 levels). The crop cultivation time can be described by a single dummy variable. It will take value 1 if was cultivated in first half of year, and 0 if it was cultivated in second half of year. 6.3.3 Weighting: - In this method each response in database is assigned a value as per its importance. In this method categories getting less representation in sample are given higher weights while overrepresented are given lower weights. Also sometimes researcher wants to give more importance to particular response and hence gives a particular higher weight age to that response. But it is important that weight age be applied after due logical thinking with a special mention in the research proposal. 6.3.4 Variable Respecification: - This is method in which large numbers of variables are collapsed into fewer variables or existing data are modified to create new variables. The purpose of doing that is to create variables that are consistent with the study’s objective. For example if the original variable is reason for buying a particular brand of milk involves 8 categories. These might be collapsed into 3 categories: Quality, Price and availability. 6.4 TABULATION This is the first stage of data analysis wherein each question or measure is analyzed individually. This is done by way of tabulation. This simply consists of counting the number of cases that falls into the various categories. The primary use of Tabulation is to determine the frequency distribution of the variable and calculating the summary statistics, mean or percentage. Next the data are subjected to cross tabulations to assess the association between two variables. 6.4.1 Frequency Distribution Frequency distribution is way of reporting the number of responses each question received and hence is a simplest way of describing the way of distribution of variable. 48 In this way of explaining data distribution, data is divided into different classes, or groups of values, and which is turn shows the number of values falling in each set of class. Further this can in turn be showed by 2 ways one is simply in a excel sheet by showing the various classes with the percent. Second way of showing the results is visually by way of histogram. Histogram is a visual way in which the data is represented in a series of rectangle, each proportionate to in width to the range of values within a class proportional to the number of objects falling in the class. Thus the actual representation of the variable can be visualized easily through the histogram. This actual distribution then is compared to any other theoretical distribution to determine whether the data used is consistent with the prior model. Below is example of both way of representing the frequency distribution By way of percentage:Selecting Brand “A” of milk Number Percentage Yes, I will definitely go for it 57 27.0 I may be purchase Brand “A” 64 30.3 I am not sure about it 32 15.2 No, definitely not 58 27.5 By way of Histogram Figure: 3 Thus you can clearly see from above histogram that width of each bar depend upon the number of items in each category. 6.5 Descriptive Statistics These are different tools attached with the frequency distribution which helps in further analyzing the data present in the frequency distribution. These helps researcher in getting further deep into the data analysis to extract more deducible information present in the data. Broadly there are three tools which help the researcher to extract the information. These are : a) Measures of Shape (Skewness) 49 b) Measure of Dispersion (Range, Standard Deviation and coefficient of Variation) c) Measure of Central tendency (Mean, Median, Mode and percentage) Means Means is in fact the average of the data obtained just by dividing the total sum of responses to a question by a sample size (total number of respondents on which questionnaire was administered). Thus mean would be a single number to describe the response of the respondents. For example Farmers using improvised seed of cotton like BT Cotton Agreement on a scale of 1-4 on the below Mean Score Users Non users Given statements 1. I like to increase per acre yield 4.6 4.6 3.9 2. I am concerned about Environment 3.9 3.8 4.0 3. I like to grow cash crops 5.3 6.1 4.4 Sample Size 62 28 34 Percentage Percentage is the proportion of the respondents who responded in a particular way towards a question. Thus when the response is based on two alternatives, or when a single alternative of the focus of alternative, the percentage is used. For example On question that Farmers prefer BT Cotton for cultivating Cotton crop on a scale of 7. Response Frequency Percentage YES, it’s true -3 200 25 -2 100 12.5 -1 100 12.5 No, it’s False 0 150 18.7 +1 100 12.5 +2 50 6.67 +3 100 12.5 Thus Descriptive analysis can provide accurate, simple and meaningful results by getting the required information from the large set of data. But their needs to be balance between the use of frequency distribution and means, percentage. Generally frequency distribution is unwieldy but it does provide more information. Mean response can give a average response about a particular question but frequency response helps in giving the specific reason for that particular answer among the respondents. Hence in situation where the population is not likely to be clustered around the mean, the frequency distribution can be useful. Also when nominal scales are involved, frequency distribution is employed. Nominal scales are those I which numbers merely label or identify categories of objects. For example, suppose respondents were asked for their preferred mode of transport Train, Bus or two wheeler. In this example there would be no way to determine the average number to represent that sample. Thus in the above example frequency distribution be used as the correct way of data analysis. 50 Also if our initial analysis involves means (Percentage), the correct way of interpretation of data analysis would be the difference between means or percentage. If the initial analysis involves frequency distribution, then cross tabulation would be the way of analysis. 6.6 CROSS –TABULATIONS The appropriate tool for statistical analysis in case of frequency distribution is termed as cross tabulation. As the name suggest in this case relation is studied for the same sort of variable across two tables. Thus in case of cross tabulation, the sample is divided into sub groups in order to learn how the dependent variable varies from group to group. CrossTabulation serves as the basis of several statistical techniques such as log-linear analysis and chi-square. Thus in cases of cross tabulation, percentages are computed on each cell basis or by rows or column. When the computation are by rows or column, cross tabulations usually are referred to as contingency tables, because the percentages are basically contingent on the rows or column total. We’ll now see an example of cross tabulation wherein for the same question frequency distribution is done on the basis of two different set of nominal variable parameter. Question asked to the farmers was Their inclination for use of mechanized farming equipments in field. The two nominal variables on which this question was cross tabulated were a) Price of farming equipments and b) Resultant yield with the use of mechanized farming equipments. Intention to use mechanized farming equipments-By price Less than Rs 10000 Rs 10000-Rs 25000 Rs 25000 and above Interested 40 %( 40) 41 %( 45) 50 %( 50) Not interested 60 %( 60) 59 %( 65) 50 %( 50) Total 100 %( 100) 100 %( 110) 100 %( 100) 44 %( 135) 56 %( 175) 100 %( 310) Intention to use mechanized farming equipments-Resultant yield per acre with their use Less than 100-500 500 quintal 100 Quintal Quintals and above Interested 55 %( 55) Not interested 45 %( 45) Total 100 %( 100) 40 %( 80) 32 %( 32) 60 %( 120) 68 %( 68) 100 %( 200) 100 %( 100) 51 42 %( 167) 58 %( 233) 100 %( 400) 6.7 SELECTION CRITERION FOR STATISTICAL TECHNIQUES 6.7.1 Research Design: - Research design is the most important criterion which decides the choice of statistical technique. Decision which effect the researcher are regarding the number of observation per variable, number of group being analyzed and dependency of observations. Statistical test to be selected will depend upon whether research design uses dependent sample or independent sample. 6.7.1.1 Independent sample: - Suppose we want to measure the effectiveness of an education program. Also, we measure the effectiveness of that as attitude of teachers towards the educational program. For this research design is X A1 A2 Where A1 is the attitude of teacher who went through the program and A2 the set of teachers which did not went through it. Thus here A1 and A2 are entirely different and hence will not affect each other. So here t-Test for difference in means would be applied. 6.7.1.2 Dependent Sample: - Now if suppose the sample are of the format B1 X B2 Here the sample is taken from same set of teachers which have given their opinion before and after going through the educational program. Here the samples are dependent as the focus is on difference of attitude of same group of teachers between and after the exposure to educational program. Here paired difference t-Test is applied for the case. 6.7.1.3 Number of groups: - Again the number of groups is another criterion which decides the statistical method to be used. Suppose the research is to measure difference between the effectiveness of two different educational programs and the program are exposed to two different set of teacher and the third set of teacher do not go through any of the programs. Thus here there are three different set of groups. The t-test for the difference of two means would not be applicable and the best analysis method would be analysis of variance procedure. 6.7.1.4 Number of variables:-The number of variable in the study also affects the choice of Statistical method. With the increase in the variable characteristics used in the study then the method of analysis will also change. For example if suppose in the initial example of educational program the measurement is done not only on the basis of attitude of teachers towards the educational program but also on the additional characteristic of purchase of the program by the teacher. This case is example of multiple variables with same research design. So here we’ll use multivariate statistical procedure for analysis. 6.7.1.5 Variable control: - Another factor affecting analysis technique is control of variable in the design. For our previous example when we are seeing the effect of educational program on the teachers, one variable which can affect our study is the previous usage of program by the same set of teachers. So we need to keep a check on 52 this factor by matching or by randomization etc. If we are not able to reduce this error, data analysis will be incorrect. 6.7.2 Type of Data: - The type of data affects the analysis technique a lot. Usually data is classified into major four types. Nominal, Ordinal, interval and ratio scaled data. Nominal scale data types of data are most basic form of data. They are the easiest type of data to be analyzed, they are just numbers assigned to objects and the objects in turn are grouped in separate categories. Since the data is very- very basic, hence no complex analysis techniques can be applied on them. The most meaningful measure applied on them is Mode. In case of ordinate scale, higher level of measurement can be applied as here numbers assigned to object also reflect an order. So median and mode both becomes the correct measure of data analysis. Third data type is interval scaled. These are the data type on which most of the analysis techniques can be applied. The mean, the median and the mode all the measures of analysis can be applied on the interval scaled data. Final data type on which analysis can be done is the ratio scaled data type. These are similar to interval data type as all the measures of central tendency mean, median and mode can be applied on them. Measure of dispersion and measurement of shapes can be also applied on them. Thus Interval data type and ratio scaled data type is generally the best form of data for the researcher as all type of analysis can be done on them resulting in all the required information for the research study. 53 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 54 CHAPTER SEVEN 7.0 Data Analysis Techniques The whole set of statistical data analysis techniques are divided in 2 sets a) Univariate techniques b) Multivariate Techniques 7.1 Univariate Statistical technique These techniques are applied when in research, for each of the variable there is only single way of measurement or even if there are multiple way of measurement for a variable still they are analyzed in isolation. These techniques will further vary depending on the type of data for analysis. Further the next level of classification is the type of sample involved i.e whether the data sample is single one or multiple. After that various techniques of data analysis can be applied like Chi square, ANOVA, Z-test, T-test, paired T-test etc. Some of these would be explained later. 7.2 Multivariate Statistical technique This technique is basically applied to draw a comparison between the two or more than two sets of measurement. This measurement can be done on each object in one or more samples of objects. Again this technique is further classified based on different logics. Classification can be done on the basis of principal point of focus i.e whether focus in on object or on the variable. Another way of classifying is whether the data can be partitioned in dependent or independent variable set. So then we can classify it on the basis of number of variables in each set. Thus use of multivariate technique results in substantial information without using too much complex techniques. Various types of multivariate statistical analysis tools are ANOVA, factor analysis, cluster analysis etc. We’ll try to learn about some of the tools in brief and other important tolls are explained in detail in the coming pages. 7.2.1 Conjoint Analysis: - Conjoint analysis is a statistical technique used to determine how people value different features that make up an individual product or service. The objective of conjoint analysis is to determine what combination of a limited number of attributes is most influential on respondent choice or decision making. A controlled set of potential products or services is shown to respondents and by analyzing how they make preferences between these products, the implicit valuation of the individual elements making up the product or service can be determined. These implicit valuations 55 can be used to create market models that estimate market share, revenue and even profitability of new designs. 7.2.2 Multidimensional scaling: - The general problem of positioning objects in an interpretable, multidimensional space is termed as multidimensional scaling (MDS). Thus it helps in determining how particular objects are perceived. Which one is perceived as similar and which one are considered as different. For example tourist can ask Is India more like Pakistan because of location and same type of habitat or India is similar to China because of economic development etc. Conjoint analysis helps in determining solution of such type of issues. 7.2.3 Factor Analysis:- Generally when a researcher is trying to find out the attitude of people regarding some particular attribute, he ask many question related to that attribute but with different connotations. Hence as a first step to analyses these responses, researcher will do a factor analysis to determine which statements belong together, in sets that are uncorrelated with other sets. The scores obtained from each set will then be used in subsequent analysis to derive required results. 7.2.4 Cluster Analysis: The term cluster analysis encompasses a number of different algorithms and methods for grouping objects of similar kind into respective categories. A general question which faces us in many areas of inquiry is how to organize observed data into meaningful structures, that is, to develop taxonomies. In other words cluster analysis is an exploratory data analysis tool which aims at sorting different objects into groups in a way that the degree of association between two objects is maximal if they belong to the same group and minimal otherwise. Given the above, cluster analysis can be used to discover structures in data without providing an explanation/interpretation. In other words, cluster analysis simply discovers structures in data without explaining why they exist. We deal with clustering in almost every aspect of daily life. For example, a group of diners sharing the same table in a restaurant may be regarded as a cluster of people etc. 7.2.5 Discriminant Analysis: - Discriminant analysis is a technique for classifying a set of observations into predefined classes. The purpose is to determine the class of an observation based on a set of variables known as predictors or input variables. The model is built based on a set of observations for which the classes are known. This set of observations is sometimes referred to as the training set. Based on the training set, the technique constructs a set of linear functions of the predictors, known as Discriminant functions, such that L = b1x1 + b2x2 + … + bnxn + c, where the b's are Discriminant coefficients, the x's are the input variables or predictors and c is a constant. 7.3 CONCEPT OF HYPOTHESIS TESTING Hypothesis testing is a way of examining at the starting stage itself the usefulness of applying time and complex techniques on the sample data set. The process of the 56 hypothesis testing begins with an assumption made on the larger population. Then data from the sample statistics is taken and information obtained from the sample statistics is judged that how likely the hypothesis on the population parameter is correct. In other words first we make an assumption on the population and then to test the validity of assumption, data from the sample is gathered and sample mean is calculated. Then the difference between the sample mean and hypothesized value of population mean is calculated. The smaller the difference is, the greater is the likelihood that the hypothesized value for population mean is correct. The larger is the difference, smaller is the probability. There are several points worth noting about hypothesis testing 1. Hypothesis testing helps in determining which population sample should be taken for further analysis and which not. Thus it helps in saving time, money for the researcher. 2. But it should be kept in mind that this test only provides a measure of sample size. A large sample will yield more statistically significant result as compared to small sample size. Hence Sample size taken into consideration also affects the test authenticity. Null and Alternative Hypothesis Null Hypothesis: Null hypothesis is hypothesis which the researcher tries to disapprove, reject or nullify. The null hypothesis is generally taken as common view of something. In research it is represented as Ho. Alternate hypothesis: This is just the opposite of the Null hypothesis. This is taken as what the researcher actually thinks is the cause of phenomenon. In research alternate hypothesis is represented as H1. Thus an Alternative hypothesis is true when null hypothesis is false and vice versa. Thus an experiment conclusion always refers to the null, rejecting or accepting Ho. For example Alternative Hypothesis can be H1:- Brinjal plants exhibits a higher rate of growth when planted in compost rather than in soil. Null Hypothesis of this would be Ho: - Brinjal plants do not exhibit a higher rate of growth when planted in compost Thus you can clearly see from the above example that Null and alternate hypothesis are opposite to each other. Now the next step in our study would be to either accept or to reject the null hypothesis. For this basically three criterion are taken, these are a) Significance level b) Number of degree of freedom and c) One or two tailed test 7.4 Significance level Significance level is the percentage of sample means which can be considered to be in variation to the null hypothesis in order to reject it. For example if we want to test the hypothesis at the 5 percent level of significance then the null hypothesis would be 57 rejected if the difference in the sample statistic and Hypothesized population parameter is greater than this. Thus this in turn will determine whether there is any relationship between the variables.. There is no common rule for selecting a significance level which is called as alpha. The most common chosen levels are 1-percent level, 5-percent level and the 10-percent level. 7.5 Statistical Tests For reasons of mathematical ease, and convention, researcher rarely use percentages to express results. A statistical significance test normally generates a result between 0 and 1, with 0 showing no correlation between variables, and 1 that there is no chance involved. Of course, values of exactly 0 or 1 are impossible, and are merely the extreme ends of the scale. There are two accepted levels of confidence in the results of significance tests. The first is the 95% level, indicating that the experimental results are 95% certain to support rejecting the null hypothesis. The second is the 99% level, which is rare, but indicates a high level of support for rejecting the null in favor of the alternative hypothesis For example, A statistical test might yield a result of 0.825. This indicates that there is a 17.5% probability that any link between the variables was due to chance. Thus, the < 0.05 level, or a statistical score of > 0.95, gives a 5% level of confidence in the results. This means that it was less than 5% chance than the result was a coincidence if the experiment was one tailed. The < 0.01, or > 0.99 level, gives only a 1% chance of randomness producing the results. If your confidence level, P, is less than 0.05 or 0.001, you should reject the null hypothesis in favor of the alternative. Alternatively, if P is greater than 0.05, you should not reject the null. 7.6 Experimental errors There are two types of error which occur while going for statistical tests with designed significant level. These are termed as Type I, Type II and Type III errors. With any scientific process, there is no such ideal as total proof or total rejection, and researchers must, by necessity, work upon probabilities. That means that, whatever level of proof was reached, there is still the possibility that the results may be wrong. This could take the form of a false rejection, or acceptance, of the null hypothesis. 7.6.1 Type I Error A Type I error is often referred to as a ’false positive’, and is the process of incorrectly rejecting the null hypothesis in favor of the alternative. For example in a case of doing HIV test on a group of patients, the null hypothesis refers to the natural state of things, stating that the patient is not HIV positive. The alternative 58 hypothesis states that the patient does carry the virus. A Type I error would indicate that the patient has the virus when they do not, a false rejection of the null. 7.6.2 Type II Error A Type II error is the opposite of a Type I error and is the false acceptance of the null hypothesis. A Type II error, also known as a false negative, would imply that the patient is free of HIV when they are not, a dangerous diagnosis. In most fields of science, Type II errors are not seen to be as problematic as a Type I error. With the Type II error, a chance to reject the null hypothesis was lost, and no conclusion is inferred from a non-rejected null. The Type I error is more serious, because you have wrongly rejected the null hypothesis. Medicine, however, is one exception; telling a patient that they are free of disease, when they are not, is potentially dangerous. 7.6.3 Type III Errors Many statisticians are now adopting a third type of error, a type III, which is where the null hypothesis was rejected for the wrong reason. In an experiment, a researcher might postulate a hypothesis and perform research. After analyzing the results statistically, the null is rejected. The problem is that there may be some relationship between the variables, but it could be for a different reason than stated in the hypothesis. 7.7 Replication Thus as we have seen that whatever may be the accuracy or correctness in doing a significance test, still there are chances of occurrence of errors. Hence this is the reason why scientific experiments must be replicable, and other scientists must be able to follow the exact methodology. Even if the highest level of proof, where P < 0.01 (probability is less than 1%), is reached, out of every 100 experiments, there will be one false result. To a certain extent, duplicate or triplicate samples reduce the chance of error, but may still mask chance if the error causing variable is present in all samples. If however, other researchers, using the same equipment, replicate the experiment and find that the results are the same, the chances of 5 or 10 experiments giving false results is unbelievably small. This is how science regulates, and minimizes, the potential for Type I and Type II errors. Of course, in non-replicable experiments and medical diagnosis, replication is not always possible, so the possibility of Type I and II errors is always a factor. 7.8 Degrees of freedom Degrees of freedom refer to Number of unconstrained data used in calculating a sample statistic. Normally in research we refer to them as n-k, where n refer to total number of 59 information available with the researcher and k refers to the number of linear constraints or restrictions required while calculating a sample statistic. For research purposes, the more degree of freedom there are, the greater is the likelihood of observing differences or relationships among variables. 7.9 One or Two Tail Test In case of one tail test, researcher is concerned only about a particular population parameter larger or smaller than a predefined value. So In this case only one critical value and region of the test statistic would be used. But in case of two tailed test, researcher determines the likelihood that a population parameter is within certain upper and lower limits. Hence in two tailed test two regions would be considered for the test. 7.10 Hypothesis testing of mean, proportion and Anova Generally the most important marketing question researcher faces while drawing a comparison between two data sets is whether the two proportions are different from one another-in statistical terms, are these two proportions significantly different from each other, given the sample information and significance level. To solve this query researchers take help of Hypothesis testing of means and proportions. 7.10.1 Hypothesis testing about single mean Testing of hypothesized value of the population mean helps the researcher in making judgments regarding the population mean. Choice of probability distribution in hypothesis testing depends upon size of sample, purpose of hypothesis sampling and whether the population standard deviation is known to researcher. If the population standard deviation is known then normal distribution or z-tables are used for testing else size of sample became the main criterion for the choice of probability standard distribution. We’ll discuss both the cases and will start with taking case where we know the standard deviation. 7.10.1.1 Case where Population standard deviation is known To understand the logic of both the cases we’ll work with the help of an example Agri harvest is a company which manufactures good quality seeds for the farmers. The company has to harvest for the upcoming season cash crop seeds that can sustain the quality rating of 10,000 points which is what their competitors are also offering. But any further improvement in the quality of seeds will result in increased R&D Cost and hence competitive disadvantage in terms of Pricing. Based on Agri harvest researcher experience, the company knows that standard deviation of quality rating is 300. Now the management wants to check whether their produced seeds are competitive enough. So for this purpose they took a sample of 225 farmers and find that the mean 60 quality rating is 9960. Now based on this sample management, want to test this hypothesis at .05 level of significance. So Let us start with the hypothesis Ho: - Population Sample means confirm to the quality standard of 10000 points H1:- Population sample mean do not confirm to the quality standard of 10000 points. Sample Size: 200 Sample mean: 9960 Population Standard Deviation: 300 Significance level: .05 Now in this case we’ll treat the size of sample as infinite and apply normal distribution. So we’ll now calculate the standard error of the mean. Standard error =Standard Deviation/ Square root of sample size σx =σ/√n =300/15=20 Now this is a two tailed test with a significance level of .05, so by using normal distribution table, Z value for .975(1-(.05)/2) pf the area under the curve comes out as 1.96. Now Calculated Z-Score would be -µ Z= /σx Z= 9960-10000/15=-2.66 Now to reject Null Hypothesis |Z| calculated> Z Value from table. Here 2.66>1.96 Hence Agri harvest company’s produced seeds do not meet the quality ratings. 61 7.10.1.2 One tailed test Now suppose take a related case with the above that Mr. Arvind who is a head of NGO working for good farming yields wants to buy cash crop seeds from Agri harvest but the quality ratings should be 10000. But he is not sure of the quality claim of Agri Harvest. So to convince him Agri harvest has commissioned a survey of 36 farmers which gives quality rating mean of 9960. Now Mr. Arvind wants to check the results of this survey and also to have chance of occurrence of Type II error to be as low as possible, he wants to do the test on .01 level scale of significance. Thus is a case of one tailed test and the data can be represented as Ho: The quality ratings of the cash crop seeds are equal to or greater than 10000 quality points H1: The quality ratings of cash crop seeds are less than 10000 quality points. Sample Size=36 Sample mean=9960 Population standard deviation=300 Significance level=.01 Standard error of mean=300/6=50 Now as per the question we know that this is one tailed test with significance level of .01, hence using the normal distribution table, the Z value of .990 comes as 2.33. Z calculate= 9960-10000/50=-0.80 Now in case of one tailed test in order to reject null hypothesis Z calculated<-Z from the table Since -.80>-2.33. Hence, we fail to reject the null hypothesis. 62 7.10.1.3 Samples with Population standard deviation not known Taking the previous example forward lets us take a case that Agri harvest do not have the knowledge of population standard deviation and sample standard deviation of a sample of 36 is 294. Also the mean quality ratings from the sample were 9960. So for this case Ho: The quality ratings of the cash crop seeds are equal to or greater than 10000 quality points H1: The quality ratings of cash crop seeds are less than 10000 quality points. Sample size=31 Sample mean=9960 Sample standard deviation=294 Significance level= .01 Now in this case we do not know the population standard deviation, so we’ll take sample standard deviation as the estimate population standard deviation. Again since we are taking estimate, hence we’ll calculate an estimate of standard error of mean. Sx = 294/6= 49 In the case when Population standard deviation is not known to the researcher, the probability distribution taken into consideration will be the t-distribution. The tdistribution for this case will have n-1 Degrees of freedom which is 50 here. Now we will see from the t-table value under .01 column and 35 degree of freedom row. The t-value obtained is 2.457 Now t-calculated is given as (Sample mean-hypothesized mean)/estimated standard error of mean =9960-10000/49=-0.81 To reject the null hypothesis, we require that 63 t calculated<-t (from table) Here -.81>-2.457 Hence Mr Arvind will fail to reject the null hypothesis that the mean quality ratings of Agri harvest cash crop seeds is greater than or equal to 10000 points. 7.10.2 Hypothesis testing of proportions Sometimes it happens that while making business decisions, management is more concerned about the proportions than the means, so in those cases we apply hypothesis testing of proportions. Again we’ll try to understand it by an example Take a case of a company which manufactures sprinkler systems. Now manager of the firm assesses that the 95 percent of the sprinklers manufactured by the company are defect free. But the head of the company checks a random sample of 225 and finds that 87 percent of the sprinklers to be defect free. So the head now want to test the hypothesis at the level of .05 of significance that 95 percent of the sprinklers manufactured by the company are defect free. Now numerically we can say that Po: .95 hypothesized value o the proportion of defect free sprinkler system Qo: .05 Hypothesized value of proportion of defective sprinkler system P: .87 sample proportion of defect free sprinkler system Q: .13 Sample proportion of defective sprinkler system Null hypothesis: p=.95 Alternative hypothesis: p not equal to .95 Sample size: n=225 Significance level: .05 So standard error of proportion is = square root of (.95*.05/225)=.0145 Since here np and nq both are greater than 5, hence normal approximation of binomial distribution will be applied. So the Z-Value for the .975 of the area under the curve would be obtained from z-tables 1.96. So the limit of acceptance region will be Po+-1.96*(Standard error of proportion) =.95+-(1.96*.0145) = (.922, .978) So as the sample proportion of defect free sprinkler system, .87, does not fall within the acceptance region, the head would reject the claim of the manager. 7.11 Effect of Sample size and Test results Sample size in itself is very important criterion for interpreting the results and in hypothesis test. The p-value is very sensitive to sample size, so if sample size increases pvalue will become smaller and smaller. This may in cases can totally reverse the result. But in cases where the sample size too large, the variation in the hypothesis test and the 64 subsequent interpretation reduces. In case of too small a sample the chances of interpreting the results in variance of actual condition increases. Thus if a sample size is large, a low p-value should be expected. If the sample size is small, a high p-value is more likely. 7.12 ANALYSIS OF VARIANCE (ANOVA) The procedure known as the Analysis of Variance or ANOVA is used to test hypotheses concerning means when we have several populations. In its simplest form ANOVA gives a statistical test of whether the means of several groups are all equal, and therefore generalizes two-sample t-test to more than two groups. Thus it is a general technique that can be used to test the hypothesis that the means among two or more groups are equal, under the assumption that the sampled populations are normally distributed. There are three conceptual ways in which the ANOVA is modeled. Fixed-effects models: - In this model assumption is that the data came from normal populations which may differ only in their means. Random effects models: - This model assumes that the data describe a hierarchy of different populations whose differences are constrained by the hierarchy. Mixed-effect models: - This model describes situations where both fixed and random effects are present. In practice, there are several types of ANOVA depending on the number of treatments and the way they are applied to the subjects in the experiment: One-way ANOVA is used to test for differences among two or more independent groups. Typically, however, the one-way ANOVA is used to test for differences among at least three groups, since the two-group case can be covered by a t-test. When there are only two means to compare, the t-test and the f-test are equivalent; the relation between ANOVA and t is given by F = t2. Factorial ANOVA is used when the experimenter wants to study the effects of two or more treatment variables. The most commonly used type of factorial ANOVA is the 2×2 (“two by two", as you would a matrix) design, where there are two independent variables and each variable has two levels or distinct values. Factorial ANOVA can also be multilevel such as 3×3, etc. or higher order such as 2×2×2, etc. but analyses with higher numbers of factors are rarely done by hand because the calculations are lengthy. However, since the introduction of data analytic software, the utilization of higher order designs and analyses has become quite common. When one wishes to test two or more independent groups subjecting the subjects to repeated measures, one may perform a factorial mixed-design ANOVA, in which one factor is a between-subjects variable and the other is within-subjects variable. This is a type of mixed-effect model. Multivariate analysis of variance (MANOVA) is used when there is more than one dependent variable. 65 7.12.1 Use of ANOVA in our case study To further illustrate the use of ANOVA in case of Practical cases let us take our basic case study and try to see how we have applied ANOVA to it. Here in our case study through the initial data analysis we’ll try to see the results with the help of ANOVA as the basic calculation part will be covered in your statistics course. In our basic data analysis part we came to the conclusion that 2 factors are very important for the Sales of Product “A”. a) Price and b) Quality Now through ANOVA we’ll try to see the importance to price and quality across income groups (a) Importance attached to Price across Income Groups ANOVA was conducted to test the difference between the means of the importance attached to price across the three age groups. The hypotheses are as follows. Ho: null hypothesis: there is not significant difference in the importance attached to price across the three income groups H1: alternative hypothesis: there is significant difference in the importance attached to price across the three income groups The results of the ANOVA for testing difference in the importance attached to the price across the three income groups of the sample are as shown. Price Sum of Squares df 124.433 2 Mean Square 62.217 F 271.587 Sig. .000 Between Groups Within Groups 95.529 417 .229 Total 219.962 419 For 2 degrees of freedom in numerator and 419 degrees of freedom in denominator, the significance value of .000 is less than our significance level of .05. Hence we reject the null hypothesis and conclude that there is a difference in importance attached to price across various income groups (b) Importance attached to Quality across Income Groups ANOVA was conducted to test the difference between the means of the importance attached to quality across the three income groups. The hypotheses are as follows. Ho: null hypothesis: there is not significant difference in the importance attached to quality across the three income groups H1: alternative hypothesis: there is significant difference in the importance attached to quality across the three income groups Quality Sum of Squares df Mean Square F Sig. Between .129 2 .064 .252 .778 Groups 66 Within 106.500 417 .255 Groups Total 106.629 419 For 2 degrees of freedom in numerator and 419 degrees of freedom in denominator, the significance value of .778 is more than our significance level of .05. Hence we fail to reject the null hypothesis and conclude that there is no difference in importance attached to quality across the three income groups. Thus we can see from the above example that ANOVA is a really good tool to get complex results when we are trying to compare a mean across two or more groups under a normally distributed population. Except the above mentioned techniques, there are various ways in which we can further do the data analysis but are used only in case where we need to do complex statistical calculations. 67 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 68 CHAPTER EIGHT 8.0 Presentation, Application and role of marketing research Presentation of the study is the final step in the whole process of Data research. Effective communication between research user and research professional is extremely important to the research process. The formal presentation plays a key role in the communication effort. In general a presentation should keep following points in consideration. 1. Presentation should be structured 2. Presentation should be visual and specific 3. Presentation should be made kept in mind that it is for specific audience 4. It should be interesting so that audience interest be maintained 5. It should address validity and reliability results Reporting thus consist of a written document and a presentation. The size and scope of these will both vary according to the scale of the research project. A simple brand awareness project may have a four page document and a forty five minute presentation and discussion. A national habits survey with many long term implications for a company’s strategy may require a series of presentations including the video footage from focus groups. This may be tailored to different functions within the company, from new product development through to the board room. The key for researcher when looking at what to report to the client is not merely to distinguish the wood from the trees, but to resist the temptation to describe every leaf on the tree. The more the market researcher understands the particular market or society’s problem, the more likely they are to produce a meaningful report and the more successful and engaging the presentation. The researcher’s experience enables them to make a clear distinction between the direct findings of the research, the interpretation of those findings and the overall recommendations. 8.1 Application of marketing Research Marketing decisions are traditionally divided into 4 P’s product, price, place and promotion. So the marketing researches usually done are based to make the changes in marketing strategy of the company depending on the changes in Price, promotion, place and promotion. Broadly the application of market research to the company is discussed below. 8.1.1 New product Research: - New product launch is the area where the companies are not clear what to launch and in what way to launch. New product launch is an area which contains lot of uncertainty and ambiguity. Thus a large part of marketing research is done to reduce the uncertainty associate with the new product launch. Marketing research helps in new product launch in multi folded way. It helps in identifying the 69 needs of the market, which is done by the way of segmentation and quantitative studies of the market. For the purpose of new product research criticisms of existing product present in the market are taken from the potential prospect and based on that screening of ideas are done. Subsequently the companies develop the products which are responsive to the criticism. Another way is that marketing research monitor the environment systematically to learn of the technological or competitive developments that may suggest new concepts. Usually market research helps in properly framing a new concept which in turn can be communicated to marketing specialist of the company to work in the direction of development of new product. Finally marketing research also helps in getting the data in response to exposure of the people towards the concept, Thus helps in taking the final decision of whether to produce that product or not. 8.1.2 Test marketing: - Test marketing allows the researcher to test the impact of total marketing program with all its interdependencies in actual market context. The purpose of test marketing is to gain an insight regarding the advantages and disadvantages of a program before fully implementing in the market. It also helps in predicting the actual market outcome of a marketing program when applied in the market. The test marketing is done in two ways. Sell in test markets are cities in which the product is sold just as it would be in a national launch. Controlled distribution scanner markets are cities for which distribution is prearranged and the purchase of a panel of customers are monitored. 8.1.3 Pricing Research: - Marketing research can be also used to evaluate alternative price approaches for products. This can be done both for new products before launch or for proposed changes in product already on the markets. In case of new product launches a product concept cannot be tested fully, without indicating its price, so when the product is ready to be introduced, a decision is made about its specific price. Decision on price changes will then need to be made over the product’s life cycle. Pricing strategy further can be of two types. Skimming strategy in which the objective is to generate as much as profit as possible. The other is a share penetration strategy, whose objective is to capture an increasingly larger market share by offering a lower price. 8.1.4 Distribution research: - Traditionally, the distribution decisions in marketing strategy involves the number and locations of salespersons, retail outlets, warehouse and the size of discount to be offered. The discount to be offered to the members in the channel of distribution is usually determined by what is being offered existing or similar products, and also whether the firm wants to follow a push or a pull strategy. Marketing research thus plays an important part in the number and location in decision about the numbers and location. 8.1.5 Promotional Research: - The decision for the promotion part of the strategy is divided into two parts advertising and sales promotion. Sales promotion affects the company in the short term and advertising decisions have a long term effect. Companies spend more time and resources on advertising research than on sales promotion research because of greater risk and uncertainty in advertising research. 70 a) Advertising research: - Most promotion research companies concentrate on advertising because advertising decisions are more costly and risky than sales promotion decisions. Advertising decisions generally involves generating information for making decisions in the awareness, preference, recognition and purchasing stages. Marketing research helps in to determine how effective the advertisement will be for the intended audience. b) Sales Promotion Research: - There are three major types of sales promotion: consumer promotion, retailer promotions and trade promotions. In general, the consumer or end user is the ultimate target of all sales promotional activity. In consumer promotion manufacturer offer the promotion directly to consumers, whereas retail promotion involve promotions by retailers to consumers. Trade promotion involves manufacturers offering promotions to retailers or other trade entities. Trade entities can also promote to each other. For example a distributor can offer a steep temporary price cut to retailers in order to sell excessive inventory. 8.1.6 Assessing Competitive advantage:- Now a days with cut throat competition every company tries to gain a upper hand then the rivals, marketing research helps a company to gain distinctive competency at the lowest delivered cost or to achieve differentiation through superior value. Marketing research is done to shape the company marketing strategy based on Porter five force models which is present of current competitors in the market, Threat of new entrant, threat of new substitute in the same product category, the bargaining power of customers and the bargaining power of suppliers. We can present in a table the ways of assessing competitive advantage by the help of marketing research. Evaluation of share in market This is done through Retail Management audit Evaluation of recall share of the product This is done through analysis of relative cost of various products Evaluation of advertising share This is done through comparison of winning versus losing competition Evaluation of R&D Share This is done through identifying high leverage phenomena Table: 10 8.1.7 Measuring Brand Equity: - Brand equity can be explained as total set of positives and negatives linked to a brand that add or subtract the total value of a product or service to a company. The brand equity or the appraisal of the brand is based on five major factors and marketing research help in judging a brand on these parameters. These five dimensions are termed as follows. a) Brand Loyalty: What is the level of loyalty of customers for a particular brand? b) Brand Awareness: What is company’s brand awareness level as compared to that of competitor? c) Brand Perceived Quality: What is the perceived value of the brand in the market? How do the customer, market perceive it in the market. 71 d) Brand Associations: What is the image does a brand stimulate. How are the brand and its competitors positioned in the market? e) Other brand Values: Are sustainable competitive advantages attached to the brand name that are not reflected in the other four equity dimensions. 8.2 Marketing Research in 21st Century In the 21st century marketing research has gained an ever important role in the framework of company’s winning strategy. In this century, rather than a presentation and report which look back at technique used to find the information, the emphasis is now far more on Market focused approach which looks forward to the application and use of the information. Market research has value only if its results are successfully used to make better decisions- to make a difference to organization, societies and importantly people. This final stage is therefore critical to ensure that the results are correctly interpreted and communicated to the client. In market research – as with so many aspects of business and life today- communication is everything. A researcher studies, analyses and interpret a situation but they also have to transfer a vision to others in order to trigger action. Researcher need to understand the past, interpret the present and forecast the future. And, most importantly, have the knowledge and courage to voice their views so that organizations can make sound decisions based on objective, reliable data and insight. The more complex the problem- and the research project – the more the researcher has to exercise his judgment, experience, intutition and creativity to organize and summaries the data to draw conclusions that are most relevant to the decision and actions. The final decision on how to apply the findings and implement the recommendation rest with the client- be they marketing man or head of a government office –but market researcher play a vital role in contribution to those decisions. As decision making in business becomes more complex and risky, market research is shifting from description, through diagnosis, to consultancy. A holistic approach, which focuses not just on data but the wider context, will ensure that organizations provide a place for market research in 21st century board room. 72 NOTES ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ ________________________________________________________________________ 73 CHAPTER NINE 9.0 Marketing Research and Rural Marketing In recent time due to growth in Indian economy, the purchasing power of rural communities have increase manifold. This has been noticed by all the major companies operating in India and globally. Now the rural areas are consuming the ever highest urban and industry manufactured product. This vying for space in rural market has resulted in the enhanced role of marketing research and a new strategy termed “rural marketing” has became of prime importance for all the companies. To launch a product in rural areas the preferences and consumption pattern knowledge is being increasing asked from marketing researcher as this has been a neglected area till now? 9.1 Salient features making Rural Market as an attractive pit stop. 1) Huge size of the market wherein FMCG itself comprises of 65000 crore 2) Of all the mobile connection issued by BSNL half are in small tons or villages. 3) Huge Investment making capacity of the rural households e.g. 6.6 million areas of rural households have done some sort of investment. So above mentioned are some of the attractive feature of rural market which is encouraging the companied to jump the wagon. But flip side of this is the very scarce information available about the rural market. Marketing manager has in the past not given due importance to the rural market analysis and hence the role of market researcher becomes of paramount importance. 9.2 Role of Marketing Researcher in Rural Market With the focus area of companies being the rural Market, marketing researcher are being employed in greater numbers by the companies to help them. The role of marketing researcher in rural spectrum is quite complex and varied. This is uncharted territory and the future growth of the rural market as well as the marketing strategy of various companies will be largely dependent on the recommendations and findings of marketing researchers. Their role would be to give insights to the companies regarding the purchasing pattern, preferences and buying process of the consumers like for example following points came through marketing research study only which were previously not known to the companies. 74 a) Indian rural market is not a homogenous mess as thought about previously, in fact it is a conglomeration of a huge population with various tiers based on Income, land holdings etc. So each segment needs to be targeted with different sort of strategy. b) Rural disposable income of the rural households is comparable with their urban counterparts. Previously companies were viewing rural areas as collection of households with very low purchasing capacity. c) Again it the collective decision of the whole family not a individual decision when the buying is done in rural households. Thus the concept of rural marketing is still at an infant stage and the sector poses a variety of challenges to the marketer which can be only tackled by the help of marketing research. High Distribution cost, non availability of retail outlets and no sure shot formula of brand success has made the role of marketing research of paramount importance for all the companies. So the unique consumption pattern, taste and need of the rural market can be only catered by the companies working on the plan developed on the basis of marketing research. 9.2.1 Ways in which marketing research can help rural manager 1) To help the product manager of the company by advising him about the new products to be launched in the rural market. This will largely depend upon the needs and preference of rural mass. 2) To help the company on the proper placement of the product in rural shops like Kirana etc so that it should have good availability and lifting from the shop floor. 3) Getting a through understanding about the present market preposition vis a vis the competitor. What kind of product generally rural mass buys, reason for buying particular type of product like for example in rural areas sachet pack of shampoo has been a total success. What is the level of liking of consumers in regard to promotions? 4) Since there is lot of uncertainty and ambiguity of supply chain maintenance, hence marketing research also helps in advising companies in making a forecast of the future demands. This is very important because proper forecast plan will help in reduced distribution cost, proper presence of the product at shop floor and hence enhances sales. 5) Marketing research helps marketing manager in formulation of marketing strategy catering to rural areas. Analysis of customers needs, competitor and market environment leads to set of market related decisions covering 4P’s and product, packaging features. 6) Market research also helps the companies in making them aware regarding their marketing share and their sales pattern. Whether there share is increasing, decreasing or stagnant. Which strategies are helping them in increasing their market share? How has the competitor performed in the same market environment? 75 7) Finally marketing research insights helps marketing manager in formulating sales strategy, promotional strategy, distribution strategy and finally the broad marketing program with the marketing strategy. Thus above mentioned are some of the major areas in which marketing research is playing a key role in helping the marketing managers by making them understand better the rural markets. Thus rural market offers lot of opportunity and the future would be very encouraging for the marketing researcher which understands the rural market well and can exploits them to their advantage. With the huge development taking place in India, rural market will be the areas showing the maximum growth in the times to come and there is lot of scope for all the companies ready to bend their back in challenging market conditions. If the products are cost effective and marginal profit going to manufacturer, it will be a win-win situation for both consumers as well as for the producers. As rural markets are extremely price sensitive and vital for survival for the companies because urban markets are getting saturated in India, products have a great future in rural areas 9.3 Some Rural Success stories a) E-Chaupal Model of ITC: - This is predominantly a supply chain model but the real success lies in the research done by ITC before going ahead with it. ITC did for many years extensive research looking for the best possible win-win preposition for both rural households as well as for its operation. The result have been phenomenal good planning backed up by extensive market research has helped ITC in becoming the number one organized private buyer of rural output. This is turn has helped ITC to enter into confectionary category and now its flagship brand Sunfeast in giving it rich dividends. b) Sonalika tractors: - Tractors are very important for farmers in their farming operation but there is a cut throat competition in Tractor segment as lot of Indian as well multinational brands are vying for the space. Sonlika tractors have made their presence felt in the rural market riding on the good promotional strategy backed by marketing research. They have also entered into virgin territories with good package promotion so as to gain a foothold into it. This again has happened only after they have gone for extensive market research trying to understand the needs of farmers, reasons because of which they go for particular brand of tractor, how to bring non user into the fold of users so as to increase the total base. c) Success of Hindustan Lever Ltd in Rural areas: - Hindustan Lever Ltd., India's most admired FMCG Company and whose products have among the deepest penetration in rural India, saw sales of about Rs 9,954 crore in year 2008, in the rural market. Reason for this can be again attributed to good marketing strategy backed by the marketing research. HLL by the help of marketing research estimated the ample opportunity present in the rural market. So it moved in the rural market swiftly following closely the pulse of 76 rural consumers. Thus Hindustan Lever, became the first MNC's to realize the potential of India's rural market, and has launched a variant of its largest selling soap brand, Lifebuoy at Rs.2 for 50 gm with great success d) Success of Cavin Kare: - The success of Cavin Kare in rural market has become a very notable case study. It is a company that began in a small way. It started the Chic shampoo sachet for 50 paisa when shampoo was available at Re.1, and it revolutionized the market. The company was successful because it tapped the major marketing research finding of the rural consumption pattern that Rural polity are price sensitive and prefer buying in small quantity. e) Success of Coca-Cola in rural sectors: - There was always an affordability issue associated with the cold drink in the rural India. Coca cola has addressed the affordability issue by introducing the returnable 200-ml glass bottle priced at Rs.5. The initiative has paid off: Eighty per cent of new drinkers for coke now come from the rural markets. Also coca cola has adapted better to the reality of rural households. Lack of electricity AND Refrigerator in rural areas was the major findings of study which were hampering the sales of cold drink. Coca-Cola has solved this issue by providing low-cost ice boxes — a tin box for new outlets and thermocol box for seasonal outlets. This has resulted in boosting the sales of cola giant vis a vis competitor. f) Success of LIC and LG in rural Area: - Another important finding of the various studies conducted in rural areas emphasized the importance of adaptability required for success in rural markets. Two notable success stories are LIC and LG. LG has adapted to the rural segment by developed a customized TV for the rural market and was a runway hit selling 100,000 sets in the very first year. Similarly LIC has entered the rural market with tailor made policies and small premium amount with huge success. Now a day’s 55 percent of total policies of LIC are coming through rural market. With all the above facts & figures the question is, can we afford to ignore rural India and move ahead? Well the answer is, no one will be able to survive without rural India in future! It is estimated that the rural India will consume 60% of the goods produced in the country. For most of marketing managers, rural India is an unknown entity even today, and it calls for a lot of investment. Initially, the ratio between investment and returns will not be the same as we see in urban India. But eventually, we will get the returns. In today's world, all the entrepreneurs and managers, looks for quick returns and depend on their quarterly results. They only look at what gives them immediate success. I think freebies have no meaning in rural India. Indian rural market is undoubtedly complex but there are some simple truths that everyone need’s to accept. The rural consumers are very valueconscious. They may be poor, but they are not backward and they can make a difference to our products and brands. Gone were the days when a rural consumer had to go to a nearby town or city to buy a branded product. The growing power of the rural consumer is an opportunity for all the managers to flock to the rural markets. At the same time, they also threw up major challenges. Gandhiji believed that India's future lay in her villages and it goes without 77 saying those among us who can bring innovations in marketing mix with rural markets in mind will take away the “fortune at the bottom of the pyramid “! REFERENCES 1. Aaker, Kumar, Dey (Seventh edition). “Marketing Research” John Wiley and sons (Asia) Pvt. Ltd. 2. Oliveto, Veronique (Second Edition). Marketing Research Explained Esomar Research Pvt Ltd 3. Sahu,Saswat Kumar.”Rural Marketing: The future battlegrounds” 4. Gary, T Henry, (Ed 1990). “Practical Sampling” Applied Social Research Methods Series Vol. 21, Sage Publications, London. 5. Green E. Paul, Tull S. Donald and Albaum Gerald (Ed 1999). “Research for Marketing Decisions” Prentice-Hall of India Pvt. Ltd., N. Delhi. 6. Jain, D. K. and Sharma A. K. “Consumer Opinion for Purchase of Milk and Milk Products”, Indian Dairyman, August 1999, p. 13-18 7. Kotler Philip (2003). “Marketing Management”. Pearson Education, Singapore. 8. Rafeek,Shamim. “Success through Intelligence” 78
