Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
QUANTITATIVE APTITUDE PERMUTATION AND COMBINATION Permutation is used where arrangements of things is relevant for e.g. arrangement in a queue, words, noβs, sitting arrangements etc. Permutation of n things taken π at a time is πππ π π! ππ = (πβπ)! π ππ has r factors. For eg : 9π3 has 3 factors i.e. 9π3 = 9 × 8 × 7 10 π2 has 2 factors i.e. 10π2 = 10 × 9 Combination is used where selection is relevant for eg : Selection of a team, committee etc. Combination of n things taken r at a time is ππΆπ π π π! πΆπ = π !(πβπ)! = 10 πΆ3 = 10 π3 3! = ππ π! 10×9×8 3×2×1 Note : Fundamental Principles of counting a. Multiplication Rule β [AND] β When things are within same arrangement and all things are required then choices are multiplied. b. Addition Rule β [OR] β When there are different arrangements and either of the arrangement can work then choices are added. Factorial 0! =1 1!=1 2!=2 3!=6 4 ! = 24 5 ! = 120 6 ! = 720 7 ! = 5040 8 ! = 40320 Note : 1) Factorial cannot be calculated for a βve no. 2) Factorial cannot be calculated for a fractional no. Example β 7 Note : 1) Sum of all 4 digit no. which can be formed using a,b,c,d is (a+b+c+d) X 6666 2) Sum of all 3 digit no. which can be formed using a,b,c is (a+b+c) X 222 Circular Permutation 1) n things can be arranged in a line in n ! ways. 2) n things can be arranged in a circle in (n β 1) ! ways. 1 3) n persons can be arranged in a circle so that no person has same 2 neighbors in 2 (π β 1)! ways. 1 4) OR no, of necklaces that can be formed with n beads of different colors in2 (π β 1)! ways. Permutation with Restriction : In case of restriction in an arrangement first of all position with restriction should be considered. Note : 1) No. of ways in which n things can be arranged so that 2 particular things are never together = total ways β ways in which 2 particular things are always together. 2) Above formula cannot be applied for more than 2 things. Properties of Combination : 1) ππΆπ = ππΆπ = 1 2) ππΆπ = ππΆπβπ 3) ππΆπ = πβ1πΆπβ1 + πβ1πΆπ π πΆπ (π‘ππ‘ππ π€ππ¦π ), πβ1πΆπβ1 (πππ ππππ‘πππ’πππ π‘βπππ ππ π πππππ‘ππ), πβ1πΆπ (πππ ππππ‘πππ’πππ π‘βπππ πππ‘ π πππππ‘π Note : Permutation = Selection with arrangement Combination = Selection without arrangement Properties of Permutation π ππ = πβ1ππ + π. πβ1ππβ1 SEQUENCE AND SERIES Sequence β 1, 3, 5, 7, 9, . . . . . . . 2, 4, 8, 16, 32, . . . . . . Series β 1 + 3 + 5 + 7 + 9 .. . . . . . . 2 + 4 + 8 + 16 + 32 . . . . . . . . ARITHMETIC PROGRESSION (A.P.) a, (a+d), (a+2d), (a+3d), . . . . . . . . nth term of A.P. ππ = π + (π β 1)π π€βπππ π = πΌ π‘πππ, Sum of n terms π π ππ = 2 [2π + (π β 1)π] = 2 [π + π] π = ππππππ ππππππππππ Where π= ππ = π + (π β 1)π Arithmetic Mean (A.M.) A.M. between 2 no.s a & b is π+π 2 and if n A.M. are inserted between 2 no.s a & b then a, Aβ, Aβ, Aβ, Aβ. . . . . . . . π΄π , b are in A.P. Note : 1) In case of A.P., Tβ - Tβ = Tβ - Tβ = d 2) Tβ + Tβ = 2Tβ or Tβ + Tββ = 2Tβ and so on. 3) If 3 terms are in A.P., then terms are (a-d), a, (a+d) 4) If 4 terms are in A.P., then terms are (a-3d), (a-d), (a+d), (a+3d) 5) If 4 terms are in A.P., then terms are (a-2d), (a-d), a, (a+d), (a+2d) Note : 1) Sum of first n natural no.s 1+2+3+4.........n= π(π+1) 2 2) Sum of squares of first n natural no.s 1²+2²+3²+4² . . . . . . . . . . . n² = π(π+1)(2π+1) 6 3) Sum of cubes of first n natural no.s 1³+2³+3³+4³+ . . . . . . . . . . . n³= [ π(π+1) 2 ]² Note : When d= -ve, we get two answers for n. In such case, first value of n better option. GEOMETRIC PROGRESSION (G.P.) π, ππΎ, ππΎ 2 , ππΎ 3 , β¦ β¦ β¦ β¦ β¦ nth term of G.P. ππ = ππ πβ1 Sum of n term of G.P. ππ = ππ = π(π π β1) , πΌπ π > 1 πβ1 π(1βπ π ) , πΌπ π < 1 1βπ Sum of Infinity π πβ = 1βπ , πΌπ π < 1 Geometric Mean (G.M.) G.M. of 2 no.s a & b is βππ If n G.M. are inserted between a & b then a, Gβ, Gβ, Gβ, . . . . . . . . . . πΊπ , b are in G.P. π π Note : 1) In G.P. π2 = π3 = π 1 2 2) Tβ + Tβ = (Tβ )² π 3) 3 terms in a G.P. are π , π, ππ π π 4) 4 terms in G.P. are π 3 , π , ππ, ππ 3 π π 5) 5 terms in a G.P. are π 2 , π , ππ, ππ 2 Note : 1) For 2 equal no.s A.M. = G.M. = H.M. 2) For 2 unequal no.s A.M. > G.M. > H.M. 3) For 2 no.s A.M. β₯ G.M. β₯ H.M RATIO, PROPORTION, INDICES & LOGARITHMS RATIO β 3 : 4 β 4 : 3 Ratio has no unit π π 1) Inverse Ratio of π ππ π 2) Compound Ratio of π π π π π ππ π & π ππ π 2 3) Duplicate ratio of π ππ (π) π π π ×π π 3 π 4) Triplicate ratio of π ππ (π) 1 π π 2 5) Sub duplicate ratio of π ππ (π) 1 6) Sub triplicate ratio of π π 3 ππ ( ) π π 7) Continued ratio is ratio of more than 2 things, 2 : 3 : 4 8) Ratio of 2 integers is said to be commensurable and ratio of 2 non-integers is incommensurable. 2 :. 3 is commensurable and β2 β3 is incommensurable. Proportion β An equality of 2 ratios is called proportion. π π E.g. : π = π then a, b, c, d are proportionate i.e. a : b = c : d or a : b : : c : d Product of Extreme = Product of mean π×π =π×π Mean proportion of 2 no.s a, b is βπ × π Properties of Proportion : π π 1) π = π β π × π = π × π π π π π π π π π π π π π+π π+π π π πβπ π π π π+π π π π 2) π = π β π = π 3) π = π β π 4) π = π β π 5) π = π β = = = (πΌππ£πππ‘ππππ) (π΄ππ‘πππππππ) π πβπ π π+π 6) π = π β πβπ = πβπ (πΆπππππππππ) (π·ππ£ππππππ) (πΆπππππππππ & π·ππ£ππππππ) π+π+π (π΄ππππππ) 7) π = π β π β¦ β¦ β¦ β¦ β¦. β¦ β¦ = π+π+π π π₯ Note : Third proportion of a X b is x such that π = π :. X = π2 π INDICES/INDEX Laws : 1) ππ × ππ = ππ+π 2) ππ ÷ ππ = ππβπ 3) (ππ )π = ππ×π 4) (π × π)π = ππ × π π 5) π° = 1 6) π π₯ = π π¦ β π₯ = π¦ 1 7) π π₯ = π β π = π π₯ LOGARITHM 1) Ifπ π₯ = π, then log π π = π₯ log π₯ log π π₯ { 10 } π = 2.7183 2) log π π × π = log π π + log π π 3) log π π π = log π π β log π π 4) log π (ππ ) = π log π π 5) log π π = 1 6) log π 1 = 0 7) Base changing formula log π π = logπ π logπ π common base means base = 10. EQUATION Quadratic Equation ππ₯ 2 + ππ₯ + π = 0 π₯= βπ±βπ 2 β4ππ 2π = π πππ‘π ππ πππ’ππ‘πππ 8 Nature of Roots : 1) If π 2 β 4ππ < 0 π‘βππ ππππ‘π πππ πππππππππ¦ ππ π’πππππ. 2) If π 2 β 4ππ = 0 π‘βππ ππππ‘π πππ ππππ πππ πππ’ππ. 3) If π 2 β 4ππ > 0 πππ πππππππ‘ π ππ’πππ π‘βππ ππππ‘π πππ ππππ, π’ππππ’ππ πππ πππ‘πππππ. 4) If π 2 β 4ππ > 0 πππ πππππππ‘ π ππ’πππ π‘βππ ππππ‘π πππ ππππ, π’ππππ’ππ πππ πππππ‘πππππ. Note : Irrational roots occur in pairs πβπ , π β βπ. ππ₯ 2 + ππ₯ + π = 0 Sum of 2 roots = βπ π π Products of 2 roots = π Quadratic Equation β Formation π₯ 2 β (π π’π ππ ππππ‘π )π₯ + πππππ’ππ‘ ππ ππππ‘π = 0 Cubic Equation : In case of cubic equation, factors should be made if possible. If factorization is not possible then only verification of answer should be preferred. Co β ordinate Geometry 1. Distance between 2 points A (x, y) and B (π₯2 , π¦2 ) π΄π΅ = β(π₯2 β π₯1 )2 + (π¦2 β π¦1 )2 2. Equation of straight line y = mx + c , m = slope of line π¦ βπ¦ ππ¦ π = π₯2 βπ₯1 = ππ₯ 2 1 Note : 1) Slope of a parallel lines is same. 2) Product of slope of 2 β₯ lines = -1 π1 × π2 = β1 3) Equation of line is (π¦ β π¦1 ) = π(π₯ β π₯1 ) 4) Intercept form of Equation π₯ π¦ + =1 π π Cost slope = Variable Cost per unit = π·πππππππππ ππ πππ π‘ π·πππππππππ ππ π’πππ‘π Limits and Continuity Some Important Limits 1) lim π π₯ β1 π₯ π₯β0 =1 L β Hospital Rule 0 For 0 ππ’πππ‘πππ ππ β ππ’πππ‘ππππ β π·πππππ£ππ‘π ππ ππ’πππππ‘ππ Limit = π·ππππ£ππ‘ππ£π ππ π·ππππππππ‘ππ 2) lim π₯β0 3) lim π₯β0 ππ₯ β1 = log π π₯ log(1+π₯) π₯ =1 1 4) lim [1 + x] Λ£ = e 5) lim 6) lim π₯β0 π₯ββ π₯ π βππ = π ππβ1 π₯βπ (1+π₯)βΏβ1 π₯ =π Types of functions 1) Even function : If f(-x) = f(x) 2) Odd function : If f(-x) = -f(x) 3) Neither even nor odd : f(-x) β f(x) β -f(x) Notes : A polynomial function is always continuous. STATISTICAL DESCRIPTION OF DATA Latin word βStatusβ Italian word βStastitaβ German word βStastistikβ French word βStatistiqueβ Definitition of Statistics can be in plural sense or singular sense. When satistic is used as a plural noun then it may be defined as data qualitative and quantitative. When used as singular noun then it is defined as scientific method of drawing conclusions about some important characteristic It means βScience of countingβ or βScience of averagesβ. Application of Statistics 1) Economics 2) Business management 3) Commerce and Industry 4) Health care Limitations of Statistics : 1) It deals with Aggregates (Total) 2) Statistics is concerned with quantitative data. 3) Future projections are valid under specific set of conditions. 4) Theory of Statistics is based on Random Sampling COLLECTION OF DATA DATA: is quantitative information about particular characteristic. VARIABLE : A quantitative characteristic is called variable. ATTRIBUTE : A qualitative characteristic is called Attribute. For e.g. : gender, nationality, color etc. VARIABLES are of 2 types : A) DISCRETE VARIABLE : When a variable can assume finite no, of isolated valued then it is discrete. E.g. : Petals in a flower no. of road accidents, marks in paper etc. B) CONTINUOUS VARIABLE : When a variable can assume any value from a given interval. E.g. : Weight, height, sale etc. DATA can be of 2 types : A) PRIMARY DATA β when data is collected by the person using the data. B) SECONDARY DATA β When data is used by the person other than collecting the data sources of secondary data are : 1) International sources β WHO (World Health Organisation) ILO (International Labour Organisation) IMF (International Monetary Fund), World Bank, etc. 2) Govt. sources β CSO (Central Statistical Organisation), ministers, etc. 3) Private and Quasi Govt. organisation β RBI, NCERT 4) Research Institute and Researcher Collection of Primary Data : There are 4 methods of collection of primary data. a) Interview Method 1) Personal Interview β Natural calamities 2) Indirect Interview β Accidents 3) Telephonic Interview β Maximum non response b) Mailed Questionnaire β Maximum non response β Widest coverage c) Observation Method β Time consuming, expensive and applicable only for small population. d) Questionnaire filled by enumerator NOTE: Scrutiny of data means checking data for any possible error. Two or more series of figures which are related to each other can be compared for internal consistency. PRESENTATION OF DATA CLASSIFICATION OF DATA a) Chronological or Temporal or Time series data b) Geographical or spatial data. c) Qualitative or ordinal d) Quantitative or cardinal MODE OF PRESENTATION a) Textual presentation β Least preferred because it is monotonous. b) Tabular presentation β (Page β 10.7) 1) Table has a serial no, 2) Table can be divided into 4 parts β Caption β is upper part of table describing columns and sub columns. β Boxhead β is entire upper part including column and sub column numbers, units of measurement alongwith caption. β Stub β is left part of table β Body β is main part of table. Example : Status Member of TU M F T Status of workers Caption Non members M F T Total M F T Body Source Footnote C) Diagrammatic representation of Data 1) Line Diagram or Historiagram β Line diagrams are used when data vary over time. Logarithmic or ratio chart is used where there is wide fluctuation. Multiple line chart is used for 2 or more related time series with same units. Multiple axis line chart are used for 2 or more series with different units. 2) Bar Diagram β Horizontal Bar diagram β used for qualitative data or data varing over space (Geographical) β Vertical Bar diagram β used for quantitative data or data varing over time. β Multiple or group Bar diagram β used to compare related series. β Component or subdivided Bar diagram (Percentage) 3) Pie Chart FREQUENCY DISTRIBUTION a) Discrete or ungrouped frequency distribution (Simple) β This is used for discrete variable E.g. : Marks No. of students (f) 5 20 6 30 7 50 8 10 9 15 10 20 B) Grouped frequency Distribution β used for continuous variables. E.g. :1) Marks / Weight f 0 β 10 10 10 β 20 5 Mutually exclusive classification β used for 20 β 30 15 continuous variable 30 β 40 20 50 2) Marks / Weight 0β9 f 10 10 β 19 5 Mutually inclusive classification β used 20 β 29 30 β 39 15 20 discrete variable for CLASS LIMIT β Lower Class limit (LCL), Upper Class Limit (UCL) LCL UCL LCB UCB 10 β 20 10 20 10 20 20 β 30 20 30 20 30 LCL = LCB 30 β 40 30 40 30 40 UCL = UCB CLASS BOUNDARIES are real class limits Lower class boundary (LCB), Upper class boundary (UCB) LCL UCL LCB UCB 10 β 19 10 19 9.5 19.5 20 β 29 20 29 19.5 29.5 LCL β LCB 30 β 39 30 39 29.5 39.5 UCL β UCB Note : 29.5 β 39.5 includes 29.5 but excludes 39.5 CLASS MARK 10 10 β 20 15 10} Class width/ Interval 20 β 30 25} πΆπππ π ππππ 10 30 β 40 35 Class Mark Class width / Interval 10 β 19 14.5 9.5 β 19.5 10 20 β 29 24.5 19.5 β 29.5 10 30 β 39 34.5 29.5 β 39.5 10 CLASS WIDTH = CLASS SIZE = CLASS INTERVAL Class Cumulative Frequency frequency interval frequency density 10 β 20 8 10 8 20 β 30 12 10 20 30 β 40 20 10 40 8 10 12 10 20 10 = .8 = 1.2 =2 πΉππππ’ππππ¦ * FREQUENCY DENSITY = πΆπππ π πΌππ‘πππ£ππ πΉππππ’ππππ¦ * RELATIVE FREQUENCY = πππ‘ππ πππππ’ππππ¦ πΉππππ’ππππ¦ * PERCENTAGE FREQUENCY = πππ‘ππ πππππ’ππππ¦ × 100 Relative frequency 8 Percentage frequency 8 40 12 40 12 40 20 40 20 40 40 × 100 × 100 × 100 GRAPHICAL REPRESENTATION OF FREQUENCY DISTRIBUTION 1) Histogram Mode β Most frequently occurring item. 2) Frequency polygon β is used for simple frequency distribution. It can be used for grouped frequency distribution provided width of class interval is same. Note : Frequency curve is limiting form of frequency polygon 3) Cumulative frequency curve or ogive (Two types) Less than ogive can be used to calculate median, quartile deciles etc. Frequency curve β It is limiting form of frequency polygon or histogram for which total area is 1. a) Bell shaped β Most common b) U β shaped MEASURE OF CENTRAL TENDENCY AND DISPERSION CENTRAL TENDENCIES ARITHMETIC MEAN (A.M.) = βππ₯ βπ ππ π΄ + βπππ₯ βπ Note : For calculation of A.M., mutually inclusive series need not be converted into mutually exclusive series. Properties of Arithmetic Mean : 1) A.M. is neither free of origin (affected by addition or subtraction of a constant) Nor free of scale (multiplication and division of a constant). 2) Combined A.M. = π1 Μ Μ Μ πβ+π2 Μ Μ Μ πβ π1 +π2 GEOMETRY MEAN (G.M.) G.M. = (π₯1 × π₯2 β¦ β¦ β¦ β¦ β¦ β¦ β¦ . π₯4 )1/4 log πΊ. π. = log(π₯1 × π₯2 β¦ β¦ β¦ β¦ β¦ β¦ β¦ . π₯π )1/π 1 log πΊ. π. = π β log π₯π 1 G.M. = A.L.[π β log π₯π ] Properties of G.M. 1) If all the observations are k then G.M. = k 2) If z = xy then G.M. of z = (πΊ. π. ππ π₯) × (πΊ. π. ππ π¦) π₯ πΊ.π.ππ π₯ 3) If z = π¦ π‘βππ πΊ. π. ππ π§ = πΊ.π.ππ π¦ HARMONIC MEAN (H.M.) = βπ π π₯ β( ) Combined H.M. of two series = πβ+πβ πβ πβ + π»β π»β E.g.: H.M. of 4, 6, 10 H.M. = 1 3 1 1 4 6 10 E.g. : x f + + = 5.77 2 3 H.M = 3 4 2 6 5 10 2 5 + + 2 4 6 E.g. : Distance (f) Speed (x) 100 km 60km/h Average of Speed = 200km 80km/h 100+200 100 200 + 60 80 Note : For average of speed, H.M. is used E.g.: Home β College β 40 km/h College β Home β 30 km/h Average speed = 2 1 1 + 30 40 Relationship between A.M., G.M., H.M.- A.M. β₯ G.M. β₯ H.M. a) For unequal no. A.M. > G.M. > H.M. b) For equal no. A.M. = G.M. = H.M. * For 2 numbers : (G.M.)² = A.M. x H.M. * G.M. cannot be calculated for negative items. * G.M. is used where compound growth is there. * A.M. is most popular central tendency. WEIGHTED MEAN = βπ€π₯ βπ€ ππ π’π ππ π€βπππ πππππππππ‘ ππ‘πππ πππ ππ πππππππππ‘ ππππππ‘ππππ. E.g. : food Rent Education x 10 20 30 340 Weighted Average = 20 w 10 6 4 20 xw 100 120 120 340 = 17 MEDIAN β If all the items are arranged in increasing order then middle item is median. π+1 π‘β Median ( in case of discrete items) = ( Median (for grouped items) = Lβ +( ) ππ‘ππ 2 π βπβ 2 π )×π Where Lβ = lower limit of median class N = Total items Nβ = Cumulative frequency of preceeding class f = frequency of median class c = class interval E.g. : f c.f. 10 β 19 10 10 20 β 29 15 25 30 β 39 25 50 β Median class 40 β 49 20 70 70 2 ( β25)×10 Median = 29.5 + 25 QUARTILE β Divide total items in 4 parts β Qβ and Qβ π+1 π+1 4 4 π1 = ( ) ππ‘ππ, π3 = ( ) × 3π‘β ππ‘ππ π+1 DECILE (10 parts) β π·π = ( 10 ) ππ‘β ππ‘ππ PERCENTILE (100 parts) MODE is most frequently occurring item. For grouped data, Mode = Lβ + (2πΉ πΉπ βπΉπ π βπΉπ βπΉπ )×π Where Lβ = lower limit of modal class πΉπ = frequency of modal class πΉπ = frequency of preceeding class πΉπ = frequency of succeeding class c = class interval Note : For moderately skewed data Mode = 3Median β 2Mean Note : All central tendencies are neither free of origin nor free of scale MEASURE OF DISPERSION β β Absolute measure Relative measure 1) Range 1) Coefficient of range 2) Mean Deviation 2) Coefficient of mean deviation 3) Standard Deviation 3) Coefficient of variation 4) Quartile deviation 4) Quartile deviation (Coefficient) Note : All measures of dispersion are free of origin but not free of scale. RANGE = L β S L = Largest S= Smallest πΏβπ β Coefficient of Range = [πΏ+π] × 100 MEAN DEVIATION about mean = β Coefficient of mean deviation + Μ | β|XβX 0π Μ | β π|XβX π ππππ πππ£ππ‘πππ ππππ βπ × 100 β(π₯βπΜ )2 STANDARD DEVIATION (π)S.D. = β π π.π·. β Coefficient of variation =ππππ × 100 πβπ S.D. of 2 numbers a and b = | 2 | S.D. of first n natural numbers = β β π₯2 = β π βπ₯ β ( π )² ππππππππ = (π. π·. )² βΆ. π. π·. ππ πππ£ππ β π£π π2 β 1 12 Combined S.D. of 2 groups π1 π 1 2 +π2 π 2 2 +π1 π1 2 +π2 π2 2 β π1 +π2 dβ = πΜ 1 β πΜ 12 dβ = πΜ 2 β πΜ 12 πΜ 1= mean of first group πΜ 2= mean of second group πΜ 12= combined mean πΜ 12 = π1π Μ 1 + π2 π Μ 2 π1 +π2 QUARTILE DEVIATION = π3 βπ1 2 π βπ β Coefficient of quartile deviation = [π3+π1] × 100 3 1 Note : Range is not free of units (Rs., Kg) Note : Measure of dispersion is free of origin but not free of scale. 1 Note : S.D. is 2of range of two numbers. CORRELATION AND REGRESSION Correlation and Regression β β Extent of Relationship Prediction β -1 to 1 1) MARGINAL DISTRIBUTION β There are two marginal distribution Marks in Maths (f) Marks in Stats (f) 0 β 10 7 8 10 β 20 19 24 20 β 30 32 26 2) CONDITIONAL DISTRIBUTION β There are m + n conditional distribution where m is no. of rows and n is no. of columns. Marks in Maths [If marks in stats is 10 β 20] f 0 β 10 2 10 β 20 7 20 β 30 15 Note : No. of cells = m X n CORRELATION β Correlation analysis is establishing relation between two variable (positive, negative or zero) and measuring the extent of relationship between two variables (1 to 1). Correlation can be measured by four methods : a) Scatter Diagram b) Karl Pearsonβs product moment correlation coefficient. c) Spearmanβs rank correlation coefficient. d) Coefficient of concurrent deviations. Scatter Diagram : can be used for linear as well as curvilinear variables. Exact measure of correlation cannot be measured by this method. KARL PEARSONβS PRODUCT MOMENT COEFFICIENT OF CORRELATION is the best method for finding correlation between two variable having linear relationship. πΎπ₯π¦ = πΆππ£πππππππ ππ π₯ πππ π¦ π.π·.π₯ ×π.π·.π¦ Where Covariance = β π₯π¦ = ππ.π·.π₯ ×π.π·.π¦ β(π₯βπ₯Μ )(π¦βπ¦Μ ) π = β π₯π¦ π β{ π₯ = (π₯ β π₯Μ ) } π¦ = (π¦ β π¦Μ ) Or πΎ= π β ππββ π β π βπ β π₯ 2 β(β π₯)2 βπ β π¦ 2 β(β π¦)2 Note : Coefficient of correlation is free of origin as well as free of scale. Note : r is free of unit. SPEARMANβS RANK CORRELATION : Rank correlation is used to study relationship between Qualitative variables. πΎ =1β 1 β(π‘ 3 βπ‘)] 12 π3 βπ 6[β π· 2 + Where D = Difference of Ranks t = no. of tied ranks COEFFICIENT OF CONCURRENT DEVIATION πΎ = ±β± (2πβπ) π where m = no. of pairs of deviations c = No. of +ve signs = No. of concurrent deviations REGRESSION ANALYSIS β There are two regression lines. Regression is concerned with predicting dependent variable when independent variable is known. In simple, Regression model, if y depends on x then line y on x is given by y = a + bx β β dependent Independent Regression line Y on X π¦ β π¦Μ = ππ¦π₯ (π β πΜ ) where ππ¦π₯ Regression coefficient of y on x. ππ·π¦ ππ¦π₯ = πΎ × ππ·π₯ = π β ππββ π β π π β π₯ 2 β(β π₯ 2 ) Regression line X on Y π β πΜ = ππ₯π¦ (π β πΜ ) ππ·π₯ ππ₯π¦ = πΎ × ππ·π¦ = π β ππββ π β π π β π¦ 2 β(β π¦ 2 ) Properties of Regression 1) ππ₯π¦ , ππ¦π₯ πππ πΎ βππ£π π πππ π πππ. 2) πΎ = ±βππ₯π¦ × ππ¦π₯ 3) ππ₯π¦ πππ ππ¦π₯ πππ ππππ ππ ππππππ ππ’π‘ πππ‘ ππππ ππ π ππππ. 4) Arithmetic mean of x and y are solution of two regression equations. 5) Because βππ₯π¦ × ππ¦π₯ = πΎ, βΆ. ππ₯π¦ × ππ¦π₯ β€ 1 PROBABILITY AND EXPECTED VALUE Probability can be divided into two categories : a) Subjective Probability b) Objective Probability Random Experiment β Experiment means performance of an Act. Random means that Probability of all outcomes is equal. Event β is result of and experiment a) Simple or Elementary events b) Composite or Compound events An event is simple if it cannot be decomposed into further event. Compound event are made of two or more simple event, MUTUALLY EXCLUSIVE EVENTS β When happening of an event makes happening of another event impossible then two events are mutually exclusive. EXHAUSTIVE EVENTS : are set of all possible events i.e. events have to be from set of Exhaustive events. Sum of probability of Exhaustive events is always equal to 1. EQUALLY LIKELY or Equi Probable β Events having same probability. CLASSICAL DEFINITION OF PROBABILITY ππ.ππ πΉππ£ππ’πππππ πΈπ£πππ‘π P (A) = πππ‘ππ ππ.ππ πππ’ππππ¦ ππππππ¦ ππ£πππ‘π Note : 1) It is applicable when total no. of events is finite. 2) It can be used only when events are equally likely. 3) Above definition is also termed as βa prioriβ and is useful only in coin tossing, dice throwing etc. Note : E.g. : In one coin H T 1 1 2 2 In two coins 0H 1H 2H 1 1 1 4 2 4 In three coins 0H 1H 2H 1 3 3 1 8 8 8 8 Sum of no. on two dice 2 3 4 5 3H 6 7 8 9 10 11 1 2 3 4 5 6 5 4 3 2 12 1 36 36 36 36 36 36 36 36 36 36 36 STATISTICAL DEFINITION OF PROBABILITY If a random experiment is repeated a very good no. of times say n under identical conditions and an event occurs πΉπ΄ times then ratio of πΉπ΄ and n when n tends to infinity is defined as statistical definition of probability. P= πΉπ΄ π π€βπππ π β β 1) π(π΄ βͺ π΅) = π(π΄ ππ π΅) = π (π΄) + π(π΅) β π(π΄ β© π΅) 2) π(π΄ β© π΅) = π (π΄ πππ π΅) For Independent events π(π΄ β© π΅) = π(π΄) × π(π΅βπ΄) = π(π΅) × π(π΄βπ΅ ) π(π΅βπ΄) = ππππππππππ‘π¦ ππ π΅ π π’πβ π‘βππ‘ ππ£πππ‘ π΄ βππ π‘ππππ πππππ 3) π(π΄ βͺ π΅ βͺ πΆ) = π(π΄) + π(π΅) + π(πΆ) β π(π΄ β© π΅) β π(π΄ β© πΆ) β π(π΅ β© πΆ) + π(π΄ β© π΅ β© πΆ) 4) π(π΄ β© π΅1 ) = π(π΄ πππ πππ‘ π΅) = π(π΄) β π(π΄ β© π΅) π(π΄) = π(π΄ β© π΅) + π(π΄ β© π΅1 ) 5) π(π΄1 β© π΅1 ) = 1 β π(π΄ βͺ π΅) * Only A + only B + A or B + A¹ and B¹ = 1 6) P (A¹) = 1 β P(A) Mutually exclusive events β π(π΄ β© π΅) = 0 Exhaustive β π(π΄) + π(π΅) + π(πΆ) = 1 Equally/likely β π(π΄) = π(π΅) = π(πΆ) π(π΅βπ΄) = π(π΄βπ΅ ) = π(π΄β©π΅) π(π΄) } π(π΄β©π΅) Conditional Probability π(π΅) EXPECTED VALUE is sum of product of different values and its probability. π(π₯) = ππ₯ × π(π₯) = ππππ π(π₯ 2 ) = ππ₯ 2 × π(π₯) Variance = (S.D.)² = π(π₯ 2 ) β [π(π₯)]2 = π(π β πΜ )2 Properties of Expected value : 1) π(π₯ + π¦) = π(π₯) + π(π¦) 2) π(π. π₯) = ππ(π₯) π€βπππ π ππ π ππππ π‘πππ‘ 3) π(π₯. π¦) = π(π₯). π(π¦)π€βππππ£ππ π₯ πππ π¦ πππ πππππππππππ‘ Mean = π = π₯Μ = π(π₯) = β π₯ π(π₯) β(π₯βπ₯Μ )2 S.D. = β π β π₯2 or β π βπ₯ 2 β( π ) β π(π₯) = 1 V = π(π₯ β π₯Μ )2 = β(π₯ β π₯Μ )2 × π(π₯) V = π(π₯ 2 ) β [π(π₯)]2 = β π₯ 2 π(π₯) β [β π₯ π(π₯)]2 THEORETICAL DISTRIBUTION Theoretical Distribution exists only in theory. It is theoretical probability distribution of experiments. BINOMIAL DISTRIBUTION : Features are 1) There are 2 possible outcomes. 2) Trials are independent. 3) n is small 4) It is Biparametric Distribution (n, p) 5) Probability of πΎ successes out of n trials is π(πΎ) = ππΆπΎ . ππΎ . π πβπΎ 6) Mean = np, variance = npq 7) It can be unimodal or bimodal Mode = Largest integer in (n+1) P if (n+1) p is a non integer. = (n+1) P and (n+1)p β 1 if (n+1)p is integer. π 8) Variance is maximum when p = q = .5 Maximum variance = 4 Additive Property If x and y are two independent variables such that π~π΅(π1 , π) π~π΅(π2 , π) X + Y ~π΅(π1 + π2 , π) * {p = success , q = failure} POISION DISTRIBUTION : Properties 1) Probability of success in very small time interval (t, t+ dt) is kt where k is constant. 2) Probability of success is independent. 3) Probability of success in this time interval is very small. 4) This distribution is called distribution of rare events. 5) n is large and p is very small such that n X p is finite. 6) Mean = Variance = n X p = m 7) It is uniparametric distribution (m). 8) There are two possible outcomes. 9) Probability of r success out of n π(π) = π βπ .ππ π! 10) Mode is largest integer in m if m is non integer Mode is m and m β 1 if m is an integer. Additive Property : If x and y are two independent poisson variables and z=x+y x ~π(π1 ) y ~π(π2 ) z = x + y~π(π1 + π2 ) NORMAL OR GAUSSIAN DISTRIBUTION P(x) = f(x) = Probability density function f(x) =π 1 β2π ×π β(π₯βπ)² 2π² m = mean π = S.D. Properties of Normal Distribution 1) Mean = median = mode 2) Mean deviation = .8 S.D. 3) First Quartile = Qβ = Mean β .675 S.D. Third Quartile = Qβ = Mean + .675 S.D. 4) Point of inflexion are a) Mean β S.D. b) Mean + S.D. 5) In standard normal variate, Mean = 0 , S.D. = 1 6) Area under normal curve πβπΜ z =| π.π·. | CHI-SQUARE DISTRIBUTION (X²-DISTRIBUTION) Properties : 1) It is continuous, positively skewed probability distribution. 2) Mean = n 3) S.D. = β2π 4) When n is large it follows normal distribution. STUDENTS T β DISTRIBUTION 1) It is a continuous symmetrical distribution. 2) Mean = 0 π 3) S.D. = βπβ2 ππ π > 2 4) For large n (n>30) t distribution is identical to z distribution. F Distribution 1) F Distribution is positively skewed. 2) It is continuous. Note : z Area under the normal curve 1 .6826 1.64 .90 (90%) 1.96 .95 (95%) 2.33 .98 (98%) 2.58 .99 (99%) 3 .9973 (99.73%) SAMPLING BASIC PRINCIPLES OF SAMPLE SURVEY : 1) Law of Statistical regularity : States a large sample drawn at random from population possesses characteristics of population at an average. 2) Principle of inertia : States that results drawn from samples are likely to be more reliable, accurate and precise as sample size increases, other things remaining same. 3) Principle of Optimization : means maximum efficiency at given cost or minimum cost for optimum level of efficiency. 4) Principle of Validity : States that sampling designs is valid only if it is possible to obtain valid results. Probabilistic sampling ensures this validity. COMPARISON BETWEEN SAMPLE AND COMPLETE ENUMERATION (CENSUS) 1) Speed 2) Cost 3) Reliability 4) Accuracy 5) Necessity ERRORS IN SAMPLING : 1) Sampling errors a) Errors due to defective sampling design. b) Errors arising out due to substitution. c) Errors due to faulty demarcation. d) Errors due to wrong choice of statistic. e) Variability in the population. 2) Non-Sampling Errors : Memory lapse, preference for certain digits, ignorance, psychological factors like vanity, non-response etc. SAMPLING DISTRIBUTION AND STANDARD ERROR OF A STATISTIC Standard Error (SE) is standard deviation of sample statistic. π 1) SE of means = SE (π₯Μ ) = π(with replacements) β where π= population S.D. 2) SE (π₯Μ )(πππ ) = 3) SE (π₯Μ ) = 4) SE (π₯Μ ) = π βπβ1 β πβπ πβ1 where N = population size s = Sample S.D. βπβ1 π π βπ n = Sample size πβπ β πβ1 π×(1βπ) 5) Standard error of proportion = β π×(1βπ) 6) SE (P) (WOR) = β π π πβπ β πβ1 INTERVAL ESTIMATION OF MEAN AND PROPORTION 1) Interval estimation of mean =πΜ ± π × ππΈ(π₯Μ ) 2) Interval estimation of mean if a) Sample size is less than 30. b) Population S.D. is not known. then interval estimation of mean πΜ ± π‘ × ππΈ(πΜ ) 3) Interval estimation of proportion = π ± π × ππΈ(π) Sample size π×π§ 1) Sample size for population mean = | 2) Sample size for proportion = π |² π×π×π§ 2 π2 TYPES OF SAMPLING I) PROBABILITY SAMPLING II) NON-PROBABILITY SAMPLING III) MIXED SAMPLING PROBABILITY SAMPLING : When each member of population has equal chance of selection. a) Simple Random Sampling (SRS) β should be used when 1) Population is small. 2) Sample size is not large. 3) Population is homogeneous. b) Stratified Sampling (strata = layer) : is used when 1) Population is large. 2) Population is heterogeneous c) Multi stage Sampling : If population is very large then it is cost effective and flexible system of sampling.E.g. : Estimation of foodgrain production in India. PURPOSIVE OR JUDGEMENT OR NON PROBABILISTIC SAMPLING : Probability of selection of each unit is not equal. MIXED SAMPLING : (Systematic Sampling) is partly probabilistic and partly non probabilistic. E.g. : Sampling done by Auditors. Systematic Sampling has a drawback β If there is unknown or undetected periodicity in sampling frame and sampling interval is multiple of that period then we get most biased samples. CRITERIA FOR IDEAL ESTIMATION 1) Unbiasedness and minimum variance (MVUE) Minimum variance unbiased estimation 2) Consistency and Efficiency 3) Sufficiency INDEX NUMBERS Index Numbers is an average of ratios expressed as percentage. Two or more time periods are involved one of which is base time period. The value at the time of base period serves as standard of comparison. INDEX NUMBER ARE OF FOLLOWING TYPES : 1) PRICE INDEX 2) QUANTITY INDEX 3) VALUE INDEX 4) COST OF LIVING INDEX OR CONSUMER PRICE INDEX METHODS OF CONSTRUCTION OF INDEX NUMBER (PRICE INDEX) 1) SIMPLE AGGREGATIVE METHOD (0 is base year, 1 is current year) π° 1 = β π1 β π° × 100 π 2) SIMPLE AVERAGE OF PRICE RELATIVE(π1 ) ° π° 1 = π β( 1 ) π° π × 100 3) WEIGHTED METHODS a) Laspeyreβs Price Index (L) π° 1 = β(π1 ×π° ) β(π° ×π° ) × 100 b) Passcheβs Price Index (P) π° 1 = β(π1 ×π1 ) β(π° ×π1 ) × 100 c) Marshall βEdge worth π° 1 = β π1 (π° +π1 ) β π° (π° +π1 ) × 100 d) Fisherβs ideal Index no. = βπΏ × π e) Bowleyβs Index no. = πΏ+π 2 4) WEIGHTED AVERAGE OF PRICE RELATIVE METHOD π° 1 = β π1 ×π° β π° ×π° × 100 [Similar to Laspeyreβs Index no.] 5) Chain Index = πΏπππ πππππ‘ππ£π ππ ππ’πππππ‘ π¦πππ×πβπππ πππππ₯ ππ πππ π‘ π¦πππ 100 ππ Link Relative =π πβ1 × 100 QUANTITY INDEX β REPLACE P BY Q AND Q BY P IN PRICE INDEX FORMULA. VALUE INDEX π°1 = β π1 ×π° β π° ×π° × 100 COST OF LIVING INDEX (CLT) OR CONSUMER PRICE INDEX (CPI) = weighted average of indices] or β π1 ×π° β π° ×π° β ππΌ βπ [is × 100 [same as Laspeyreβs] DEFLATING πΆπ’πππππ‘ π£πππ’π ×100 Deflated Value = πππππ πππππ₯ ππ ππ’πππππ‘ π¦πππ or πΆπ’πππππ‘ π£πππ’π ×π΅ππ π π¦πππ πππππ4 πΆπ’πππππ‘ π¦πππ ππππππ BASE SHIFTING ππππππππ πππππ πΌππππ₯ ×100 Shifting Price Index = πππππ πΌππππ₯ ππ π¦πππ ππ π€βππβ ππ‘ ππ π βπππ‘ππ Note : Index no. of base year is always 100. 325 Real wages of Base year = 110 × 100 = 295.45 500 Real wages in Current year = 200 × 100 = 250 Decrease in Real wages = 295.45 β 250 = 45.45 TEST OF ADEQUACY 1) UNIT TEST β This test requires index no. to be free of unit. All index no. expect simple aggregative method satisfy this test. 2) TIME REVERSAL TEST is cleared if π° 1 × π1 ° = 1 a) Fisher Index no. b) Marshall Edgeworth sastisfy this text. 3) FACTOR REVERSAL TEST is satisfied if π° 1 × π° 1 = π° 1 a) Fisher Index no. b) Simple Aggregative method satisfy this text. 4) CIRCULAR TEST β is satisfied if π° 1 × π12 × π23 = π° 3 ππ π° 1 × π12 × π20 = 1 It is extension of time reversal test a) simple Geometric mean of Price relative b) Weighted Aggregative with fixed weight c) simple aggregative meet this test. DIFFERENTIAL CALCULUS Derivative of y w.Ι£.t. x is change in y w.r.t. x. when change in x is very small ππ¦ ππ₯ βπ¦ = lim βπ₯β0 βπ₯ π(π₯+β)β π(π₯) = lim β ββ0 Standard Results π 1) ππ₯ (π₯ π ) = π π₯ πβ1 π 2) ππ₯ (π π₯ ) = π π₯ π 3) ππ₯ (πΆπππ π‘πππ‘) = 0 π 4) ππ₯ (π π₯ ) = π π₯ log π π 5) ππ₯ (log π₯) = 1 π₯ Product Rule π ππ₯ (πΌ × πΌπΌ) = πΌ × π ππ₯ π × (πΌπΌ) + πΌπΌ × ππ₯ (πΌ) Quotient Rule π πΌ ( )= πΌπΌ× ππ₯ πΌπΌ π π (πΌ)βπΌ× (πΌπΌ) ππ₯ ππ₯ (πΌπΌ)2 ππ¦ Gradient = Derivative = ππ₯ * Ordinate = value of x Abscissa = value of y INTEGRATION π 1 π₯βπ π 1 π+π₯ 1) β« π₯ 2 βπ2 = 2π log (π₯+π) + πΆ 2) β« π2 βπ₯2 = 2π log (πβπ₯) + πΆ π 3) β« βπ₯ 2 +π2 π 4) β« βπ₯ 2 βπ2 = log(π₯ + βπ₯ 2 + π2 ) + πΆ = log(π₯ + βπ₯ 2 β π2 ) + πΆ 5) β« π π₯ (π(π₯) + π 1 (π₯)) = π π₯ π(π₯) π₯ 6) β« βπ₯ 2 + π2 ππ₯ = 2 βπ₯ 2 + π2 + π2 2 log[π₯ + βπ₯ 2 + π2 ] + πΆ SETS, FUNCTIONS AND RELATIONS Sets : A set is defined as collection of well defined distinct objects. For e.g. : A = {a,e,i,o,u} = {x : x is a vowel in the alphabets} B = {2, 4, 6, 8, 10} ={ x : x = 2m and m is an integer o<m<6} Note : Each set has 2π subsets where n are no. of items in a set. Each set has2π β 1 proper subsets. For e.g. : A={1, 2, 3}, Subsets are {1},{3},{2},{1,2},{1,3},{2,3},{1,2,3} { } = 8 = 2³ If 1,2,3 is excluded then remaining subsets are proper subsets. Null set = = β π(π΄ βͺ π΅) = π (π΄) + π (π΅) β π (π΄ β© π΅) 1) Singleton set β contains only one element a 2) Equal set β A = {1, 3, 4} B = {1, 3, 4} 3) Equivalent set β A = {1, 3, 4} B = {2, 3, 5} n (A) = n (B) :. A & B are Equivalent set. 4) Power set is collection of all possible subsets of a set. Cartesian product of Sets β If A and B are 2 non empty sets then A x B is set of all ordered pairs (a, b) such that a belongs to A and b belongs to B. For E.g. : A ={1, 3, 6} , B ={3, 5} A x B = {(1, 3), (1, 5), (3, 3), (3, 5), (6, 3), (6, 5) n (A x B) = n (A) x n (B) Domain of a function y = f (x) is all possible values of x. Range of a function y = f (x) is all possible values of y. Relations 1) Reflexive relation is a relation if a is related to a. [a=a] 2) Symmetric relation : If a = b and b = a then a and b has symmetric relation. 3) Transitive relation : If a = b and b = c then If a = c then relation is transitive. 4) Equivalence relation : If a relation is reflexive, symmetric and transitive then it is equivalence relation.