Download Food-based approaches to fighting micronutrient

Orientation for Third Trimester Field
Practical Training
Methodologies, for Sampling,
Collection and Analysis of
Quantitative and Qualitative Data
Elvis Attakora-Amaniampong
Real Estate & Land Management Department
FPLM-UDS, Wa
12th February, 2012
Objectives
 Key Concepts: Differentiate between
quantitative and qualitative data
 Identify the tools and techniques for data
collection
 Use the identified tools and techniques to
collect data in the field
 Identify the tools and techniques for data
analysis
 Carry out simple quantitative and qualitative
data analysis
 Present the results of analysed data
2
Key Concepts
 Statistics: way to get information from data.
 A variable is any type of observation that can take
d/t values for d/t people, or for d/t times or place etc,
such as marks on an exam.
 Variable because marks obtained in each exam will
vary from student to student
 Values of a variable are the possible observations of
the variable, such as values of exam thus, any
integer between 0 and 100.
 Data are the observed values of a variable. For eg:
the marks of 10 students, which are
76 79 45 33 67 74 86 91 90 77
3
Categories/Sources of Scientific Data
1) Primary data
 Survey Data: (Censuses & Sample Survey types)
 Experimental data
2) Secondary data
 Routine Records
These categories are not all mutually exclusive.
4
Types of Variables/Data
•Variables / Data
•Quantitative
•Continuous
•e.g. height
and mass
•Discrete
•e.g. number
of students
•Qualitative
•Nominal
• e.g. colour
•Ordinal
•e.g. ranks
and grades
Hierarchy of Data/Scale of Measurement
Ratio
Applies to quantitative data only & has all the properties of interval scale,
has meaningful zero starting point and
a meaningful ratio b/n 2 numbers
Interval
Values are real numbers
All calculations are valid
Data may be treated as ordinal or nominal
Ordinal
Values must represent the ranked order of the data
Calculation based on ordering process are valid
Data may be treated as nominal but not as interval
Nominal
Values are the arbitrary numbers that represent categories
Only calculations based on the frequencies of occurrence are valid
Data may not be treated as ordinal or interval
Sampling
 Taking information based on a small set of the population
 Can be planned; preference for random surveys
Sampling Types
 Probability sampling - the selection of sampling units is
according to a probability (random & non-random) scheme.
 Non-probability sampling - selection of samples not objectively
made, but influenced a great deal by the sampler. Example –
haphazard, purposive, snowball, and convenience.
 Preference is for probability sampling, but situation may
determine otherwise
7
SYSTEMATIC SAMPLING PROCEDURE
 Sampling units are selected according to a predetermined pattern.
 Advantages of Systematic Sampling
•
•
•
•
Easy to set-up
Relative speed in data collection
Total coverage of population assured
Good base for future designs, as position of characters can
easily be mapped (with known coordinates)
• Demarcation of units not necessary, as sampling units are
defined by first unit.
8
Disadvantages of Systematic Sampling
 With only one random observation, sampling error not valid
 Unknown trend(s) in population can influence results
adversely [Examples: topography, season of sampling interval]
Avoiding the disadvantages
 The first major disadvantage on sampling error can be
rectified by introducing several multiple random starts
through stratification of the population
 The second problem of trend is more difficult but simply
relates to the choice of the sampling interval.
9
SIMPLE/UNRESTRICTED RANDOM SAMPLING
 Unlike the systematic sampling, sampling units
need not be equally spaced.
Types
 STRATIFIED RANDOM SAMPLING
 CLUSTER SAMPLING
10
STRATIFIED RANDOM SAMPLING
 Requires dividing the population into non-overlapping
homogeneous units, which we are called STRATA.
 SRS is then applied to each stratum, hence stratified random
sampling (STRS).
 Examples of strata types or criteria are ages of plantation,
species types, aspect, topography/ altitude, farm types, habitat
 Dividing the population into such homogeneous units usually
leads to better estimates of the desired population parameters
11
CLUSTER SAMPLING
 This is similar to SRS, except that the sampling unit
is cluster
 Unlike STRS, all units within selected clusters are
observed.
 Advantages
 natural aggregation automatically defines sampling
unit.
 provides information on both clusters and unit
constituting clusters.
 variation within clusters either eliminated or reduced
as all units within selected clusters are observed.
12
Disadvantages
 requires knowledge of total number of cluster, (to
establish the frame)
 variation in clusters and clusters sizes could affect
estimation of cluster-based parameters.
 complex computational formulae for unequal or
naturally occurring cluster.
 impossible to observe all units if the cluster sizes are
too large.
13
ANALYSING QUALITATIVE DATA
 Qualitative data are essentially labels of a
categorical variable
 Statistical Analyses involve totals, percentages
and conversion to pie-charts and bar charts
(bar-graphs).
 Sophisticated analyses include categorical
modelling
14
QUALITATIVE TOOLS AND TECHNIQUES
Tools and techniques of Participatory Rural
Appraisal (PRA)
Broadly categorised into 3
1. Interviews and discussions
2. Diagrams
3. Observation
15
Interviews and Discussions
Interview type: semi-structured.
 A guided interviewing or conversation with
some predetermined topics or questions.
 Instrument thus described as checklist but not
formal questionnaire and new questions not on
the checklist can be asked during the interview.
 Questions are open-ended. They should not be
‘yes’ or ‘no’ questions.
16
Types of semi-structured interview
 1. Individual interview
 2. Key informant interview; e.g. model
farmers/innovators, educators/school teachers,
government officials, religious leaders, women’s
leaders, etc.
 3. Group interviews: group size of 20-25.
 4. Focus group interviews: 6-12 knowledgeable
or interested people.
17
Hints for carrying out semi-structured interview
 Be sensitive and respectful
 Use same language or interpreter
 Use dialogue. Be interested in what is
important to the respondent.
 Observe non-verbal indicators, e.g. body
language, facial expression, tone of voice and
eye contact.
 Start questions with who, what, whom, when,
where, why, how, etc.
18
Hints for carrying out semi-structured interview-cont.
 Ask one clear, short and unambiguous question at a
time.
 Ask no leading questions, e. g. do you plant cassava?
 Do not conclude statements for respondents.
 Listen, don’t lecture.
 Probe for more details. Use any of the following.
A. Nod your head or say ‘yes’.
B. Repeat questions in slightly different ways.
C. Use questions such as: “could you tell me more about that?”,
“could you give me an example?”, “could you explain that to
me?” and Use ‘why’ sparingly.

19
Diagram
 Pictorial and symbolic representation of information.
Very useful when dealing with illiterate and semiliterate people
 Provide focus for attention during discussions.
 Stimulate discussions.
 Represent complex issues or processes simply.
 Used in crosschecking thus provoking effective group
work.
 Stimulates memory about past and present situations.
 Reinforce spoken or written word.
 Assist in decision making and monitoring.
20
Commonly used diagrams






Local histories/historical profiles/time line
Transects
Trend analysis
Seasonal calendars
Social/resource maps
Ranking
• preference ranking
• pairwise ranking
• matrix ranking
21
Time lines
 For recall of important historical events in the community.
 Help to gather information on causes of problems.
How?
 Organise elders/leaders into groups.
 Explain process to them.
 Draw vertical line and record events with dates.
Example of a time line
• 2001----This drought resulted in purchasing maize for 5,000
cedis per bowl for the first time
• 1999----The drought led to the invasion of army worms that
destroyed all the crops thus resulting in severe hunger.
• 1997/98--People travelled to Burkina Faso to purchase
grains.
22
Transect
 Diagram provides information on land use and
compares features (topography, vegetation, trees,
soils, drainage) resources, problems, opportunities,
etc.
How to do a transect walk
 Find knowledgeable and willing community members.
 Discuss what is to be included in the transect with
them.
 Choose the starting point and route with the people.
 Walk the transect
 Observe and ask questions for clarification.
23
How to do a transect walk-cont.
 Listen and take down notes.
 Discuss problems and opportunities,
 Identify the main natural and agricultural zones and
sketch the distinguishing features
 For each zone describe the soils, crops, livestock,
topography, cropping pattern, drainage, socioeconomic indicators, problems, etc.
 Draw the transect on paper at the end of the walk and
cross-check with the community members.
24
Trend analysis
 Used to trace changes that have occurred in resources, diseases,
nutritional status, etc.
 The period of evaluation spans childhood to the future.
How to do trend analysis
 Draw a table with 5 columns and 2 rows.
 In the columns indicate the resources(s), childhood days, today,
expected future and desired future. (Symbols can be used for
these based on preference)
 List the resources or use symbols to represent them in the
resource column.
 Let them indicate the changes pictorially for the different
periods.
25
Seasonal calendar
 It shows the main activities, problems, rainfall pattern, labour
availability, periods of diseases, pest infestation, opportunities,
birth (livestock and human), weaning, type of food consumed,
prices of goods, etc. in the community. There can be one or a
series of diagrams on a single sheet of paper.
How to prepare a seasonal calendar
 Draw the scale of the months of the year at the top of a sheet of paper or on
the ground. Allow the community members to use their local scale of
months. (To make it more practicable let them use symbols to represent the
months)
 Allow the participants to indicate the information on the scale
 Local materials like stones, seeds, etc. can be used to indicate the intensities
of activities and occurrences on the calendar. Put more stones where activity
is very intensive and less stones where less intensive.
 Copy calendar on a sheet of paper.
26
Social and resource maps
 A map indicating resources and where they are
in the community. E. g. hills, valleys, wells,
rivers, roads, places of worship, schools,
markets, clinics, herbalists, priests, etc.
 Let the members of the community draw an
outline of the community and insert the desired
features. Use a benchmark if available
27
Ranking
 Pairwise ranking: compare resources or problems with each
other to determine which is more important or serious so as to
rank and prioritise them.
How to do pairwise ranking
 List resources, problems, etc.
 Make a table and divide into rows and columns according to
number of items.
 Write items horizontally in the top row and vertically in the
first column.
 Compare 2 items at a time and ask respondents to indicate the
preferred one.
 Count the number of time each item appears and record then
rank.
28
Matrix ranking
 Comparing resources with established criteria in a
tabular form.
 Make a table with resources in rows and criteria in
columns.
 Compare each resource to criteria to determine their
importance. Use stones.
 Count the number of stones against each resource and
record.
29
Resource flow diagrams
 Show how resources are used in the community.
Draw arrows from one resource to another to
indicate which resource is used for what
purpose.
Observation
 Directly observe objects, events, processes,
relationships, etc. in the field and record them
mentally and in notes or diagrams.
30
ANALYSING QUANTITATIVE DATA
 Basic analyses involve determining the
CENTRE and SPREAD of data.
 Inferential, probability and non-probability
based on measuring central tendencies
31
Numerical Description of Data
 Measures of Center and Location
• Mean, median, mode, geometric mean,
midrange
 Other measures of Location
• Weighted mean, percentiles, quartiles
 Measures of Variation
• Range, interquartile range, variance and
standard deviation, coefficient of variation
•32
5/24/2017
Summary Measures
•Describing Data Numerically
•Center and Location
•Mea
n
•Median
•Mod
e
•Weighted Mean
•Other Measures
of Location
•Variation
•Range
•Percentiles
•Interquartile Range
•Quartiles
•Variance
•Standard Deviation
•Coefficient of
Variation
•33
5/24/2017
Measures of Center and Location
•Overview
•Center and Location
•Mean
•Median
•Mode
•Weighted Mean
n
x
x
i 1
i
XW
n

i 1
N
i
i
i
N
x
wx


w
wx


w
i
W
i
i
i
•34
5/24/2017
Mean (Arithmetic Average)
•n = Sample Size
n
Sample mean
x
x
i1
x1  x 2    x n

n
i
n
•N = Population Size
N
Population mean
x
x1  x 2    xN


N
N
i1
i
•35
5/24/2017
Mean (Arithmetic Average)
•(continued)
 The most common measure of central tendency
 Mean = sum of values divided by the number of values
 Affected by extreme values (outliers)
• 0 1 2 3 4 5 6 7 8 9 10
•Mean = 3
1  2  3  4  5 15

3
5
5
•
0 1 2 3 4 5 6 7 8 9 10
•Mean = 4
1  2  3  4  10 20

4
5
5
•36
5/24/2017
Median
 Not affected by extreme values
• 0 1 2
3
4
5 6 7 8 9 10
•Median = 3
• 0 1
2
3
4
5
6
7
8
9
10
•Median = 3
 In an ordered array, the median is the “middle”
number
» If n or N is odd, the median is the middle number
» If n or N is even, the median is the average of the two middle
numbers
•37
5/24/2017
Mode






A measure of central tendency
Value that occurs most often
Not affected by extreme values
Used for either numerical or categorical data
There may be no mode
There may be several modes
•0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
•0 1 2 3 4 5 6
•Mode = 6
•No Mode
•38
5/24/2017
 Review Example
 Five houses on a
$2,000 K
•House Prices:
hill by the beach
¢2,000,000
500,000
300,000
100,000
100,000
$500 K
$300 K
$100 K
$100 K
•39
5/24/2017
•House Prices:
Summary Statistics
•
 Mean: ($3,000,000/5)
= $600,000
$2,000,000
500,000
300,000
100,000
100,000
•Sum 3,000,000
 Median: middle value of ranked data
= $300,000
 Mode: most frequent value
= $100 ,000

.
•40
5/24/2017
Measures of Variation/Spread
•Variation
•Range
•Interquartile
•Range
•Variance
•Standard Deviation
•Population
•Variance
•Population
•Standard
•Deviation
•Sample
•Variance
•Sample
•Standard
•Deviation
•Coefficient of
Variation
MSAD
•MSAD
•SE
Mean of the sum of the absolute
deviation
Standard error
•41
5/24/2017
Variation
 Measures of variation give information
on the spread or variability of the data
values.
•Same center,
•different variation
•42
5/24/2017
Range
 Simplest measure of variation
 Difference between the largest and the
smallest observations:
•Range = xmaximum – xminimum
•Example:
•0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
•Range = 14 - 1 = 13
•43
5/24/2017
Disadvantages of the Range
 Ignores the way in which data are distributed
•7 8 9 10 11
12 •Range = 12 - 7 = 5
•7
12
8
9
10
11
•Range = 12 - 7 = 5
 Sensitive to outliers
•
1,1,1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,3,3,3,3,4,5
•Range = 5 - 1 = 4
•
1,1,1,1,1,1,1,1,1,1,1,2,2,2,2,2,2,2,2,3,3,3,3,4,120
•Range = 120 - 1 = 119
•44
5/24/2017
Interquartile Range
 Can eliminate some outlier problems by using the
interquartile range
 Eliminate some high-and low-valued observations
and calculate the range from the remaining values.
 Interquartile range = 3rd quartile – 1st quartile
•45
5/24/2017
Interquartile Range
•Example:
•X
•minimum
•25%
•12
•Q1
25%
30
•Median
•Q3
•(Q2)
25%
45
•X
•maximum
25%
57
70
•Interquartile range
• = 57 – 30 = 27
•46
5/24/2017
Variance
 Average of squared deviations of values
from the mean
• Sample variance:
n
s2 
• Population variance:
σ2 
 (x
i1
i
 x)
2
n -1
N
 (x
i1
i
 μ)
2
N
•47
5/24/2017
Standard Deviation
 Most commonly used measure of variation
 Shows variation about the mean
 Has the same units as the original data
n
• Sample standard deviation:
s
2
(x

x
)
 i
i1
n -1
N
• Population standard deviation: σ 
2
(x

μ)
 i
i1
•48
N
5/24/2017
Calculation Example:
Sample Standard Deviation
•Sample
Data (Xi) :
10
12
14
•n=8
s 
15 17 18 18 24
Mean = x = 16
(10  x ) 2  (12  x ) 2  (14  x ) 2    (24  x ) 2
n 1

(10  16) 2  (12  16) 2  (14  16) 2    (24  16) 2
8 1

126
7

4.2426
•49
5/24/2017
Measures of association
 Chi-square
 Correlation analysis (Pearson’s rho)
 Linear regression
Presentation of Results
 Tables
• Summary of counts, frequencies with %
 Diagrams/graphs
• Pie Chart (in only percentages), Histogram
• Line graph , bar graph( count or percentages)

50
Thank you
Questions and Discussions