| ] 1 Regression Analysis Oooo oe INTRODUCTION USES OF REGRESSION ANALYSIS CORRELATION AND REGRESSION ANALYSIS—A COMPARISON REGRESSION LINES REGRESSION EQUATIONS REGRESSION EQUATIONS IN CASE OF CORRELATION TABLE. STANDARD ERROR OF ESTIMATE MISCELLANEOUS ILLUSTRATIONS LIMITATIONS OF REGRESSION ANALYSISINTRODUCTION he ct that two variables are closely related we After having established Se Sine) the value of one variable given may be interested in estimating (p ee eit advertising and sales the value of another. For example, if we know that erti sal Jated we find out expected amount of sales for a given advertising a enditure or the required amount of expenditure for attaining a given amount of sales, Similarly, if we know that the yicld of rice and rainfall are closely related we may find out the amount of rain required to achieve a certain production figure. Regression analysis reveals average relationship between two variables and this makes possible estimation or prediction, . , The dictionary meaning of the term ‘regression’ is the act of returning or going back. The term ‘regression’ was first used by Sir Francis Galton in 1877 while studying the relationship between the height of fathers and sons. This term was introduced by him in the paper ‘Regression towards Mediocrity in Hereditary Stature’. His study of height of about one thousand fathers and sons revealed a very interesting relationship, i e., tall fathers tend to have tall sons and short fathers short sons, but the average height of the sons of a group of tall fathers is less than that of the fathers and the average height of the sons of a group of short fathers is greater than that of the fathers. The line describing the tendency to regress or going back was called by Galton a ‘Regression Line.’ The term is still used to describe that line drawn for a group of points to represent the trend present, but it no longer necessarily carries the original implication of “stepping back” that Galton intended. These days there is a growing tendency of the modern writers to use the term estimating line instead of regression line because the expression estimating line is more clarificatory in character. Let us examine a few definitions of the term regression. 1. “Regression is the measure of the average relationship between two or] more variables in terms of the original units of the data.” 2. The term ‘regression analysis’ refers to the methods by which estimates are made of the values of a variable from a knowledge of the values of one or more other variables and to the measurement of the errors involved in this esumation process.” 3. “One of the most frequently used technique: research, to find a relation between two or mo: causally, is regression analysis.” 4. “Regression analysis attempts to establish the ‘n between variables—that {s, to study the func! variables and thereby provide a mechanism fo —Morris Hamburg 's in economics and business re variables that are related —Taro Yamane ature of the relationship’ tional relationship between the| T prediction, or forecasting.” —Ya-Lum Chou It is clear from the above definitions that regn j is tha Gression analysis is istical device with the help of which we are in a position to estimate (or a edict) the unknown values of one variable from known values of anoth triable. The variable which ts used to predict the variable of interest 1s calle th ee dent variable or explanatory variable and the variable yee ed ne indepen is called the dependent variable le _we are trying to predict . or “explai 12 variable is denoted by X and the dependent variable, evi patie eee 3 . The analysis usecalled the simple linear regression analysis—simple because there is only one predictor or independent variable, and linear because of the assumed linear relationship between the dependent and the independent variables. The term “linear” means that an equation of a straight line of the form Y=a+ bX, where aand bare constants, is used to describe the average relationship that exists between the two variables. It should be noted that the term ‘dependent’ and ‘independent’ refer to the mathematical or functional meaning of dependence—they do not imply that there is necessarily any cause and eflect relationship betweerr the variables. What it meant ts simply that estimates of values of the dependent variable Y may be obtained for given values of the independent variable X from a mathe- matical function involving X and Y. In that sense, the values of Y are dependent upon the values of X, The X variable may or may not be causing change in the Y variable. For example. while estimating sales of a product from figures on advertising expenditures, sale is generally taken as the dependent variable. However, there may or may not be causal connection between these two factors in the sense that changes in advertising expenditures cause changes in sales. In fact, in certain cases, the cause-effect relation may be just opposite of what appears to be the obvious one. USES OF REGRESSION ANALYSIS Regression analy is a branch of statistical theory that is widely used in almost all the scientific disciplines. In economics it is the basic technique for measuring or estimating the relationship among economic variables that constitute the essence of economic theory and economic life. For example, if we know that two variables, price (X) and demand (Y), are closely related we can find out the most probable value of X for a given value of Y or the most probable value of Y for a given value of X. Similarly, if we know that the amount of tax and the rise in the price of a commodity are closely related, we can find out the expected price for a certain amount of tax levy. Thus, we find that the study of regression is of considerable help to the economists and businessmen. The uses of regression are not confined to economics and business field only. Its applications are extended to almost all the natural, physical and social sciences. The regression analysis attempts to accomplish’ the following : 1. Regression analysis provides estimates of values of the dependent variable from values of the independent variable. The device used to accomplish this estimation procedure is the regression line. The regression line describes the average relationship existing between X and Y variables, i e., it displays mean values of X for given values of Y. The equation of this line, known as the regression equation, provides estimates of the dependent variable when values of the independent variable are inserted into the equation. __ 2. A second goal of regression analysis is to obtain a measure of the error involved in using the regression line as a basis for estimation. For this purpose the standard error of estimate Is calculated. This is a measure of the scatter or spread of the observed values of Y around the corresponding values estimated from the regression line. If the line fits the data closely, that is, if there Is litde scatter of the observations around the regression line, good estimates can be made of the Y variable. On the other hand, if there is'a great deal of scatter of the observations around the fitted regression line, the line will not produce accurate esimates of the dependent variable. 3. With the help of regression coefficients we can calculate the correlation coefficient. The square of correlation coefficient (7, called coefficient of determination, measures the degree of association of correlation that existser 498 STATISTICAL METHODS between the two variables. It ses the proportion of variance in the dependent variable that has been accounted for by the regression equation. In general, the greater the value of FP the better is the fit and the more useful the regression equations as a predictive device. DIFFERENCE BETWEEN CORRELATION AND REGRESSION ANALYSIS Correlation and regression analysis are constructed under different assumptions they furnish different types of information and it is not always clear as to which measure should be used in a given problem situation. The following are the points of difference between the two. 1. Whereas coefficient is a measure of degree of covariability between X and Y. the objective of regression analysis is to study the ‘nature of relationship” between the variables so that we may be able to predict the value of one on the basis of another. The closer the relationship between two variables, the greater the confidence that may be placed in the estimates. 2. Correlation is merely a tool of ascertaining the degree of relationship between two variables and, therefore, we cannot say that one variable is the cause and other the effect. For example, a high degree of correlation between price and demand for a certain commodity or a particular point of time may not suggest which is the cause and which is the effect. However, in regression analysis one variable is taken as dependent while the other as independent— thus making it possible to study the cause and effect relationship. It should be noted that the presence of association does not imply causation, but the existence of causation always implies association. Statistical evidence can only establish the presence or absence of association between variables whether causation exists or not depends purely on reasoning. 3. In correlation analysis rg is a measure of direction and degree of linear relationship between two variables X and Y, ry and myx are symmetric (rq = Tyg. te. it is immaterial which of X and Y is dependent variable and which is independent variable. In_ regression analysis the regression coefficients by, and byx are not symmetric, i e., byy# byx and hence it definitely makes a difference as to which variable is dependent and which is independent. 4. There may be nonsense correlation between two variables which is purely due to chance and has no practical relevance such as increase in income and increase in weight of a group of people. However, there is nothing like non- sense regression, 5. Correlation coefficient {s independent of change r ig Regression coefficients are independent of change of oniain but a aise ‘lc There is something common in both regression and correlation analysis. The coefficient of correlation (7) takes the same sign as the regression coetticients (by and by. Also if the value of b is significanCat a given level of signiticance a, also signifleant at that level. ariables may have either linear or non-linear relations ariables are said to have linear relalionship when change in RS eRe ea ee (say X) by one unit leads to constant absolute change in the dependent variable (¥), When two variables have linear relaUionship, the regres He lines can be used to find out the values of dependent variable, Whee we plot two variables (say X and Y) on a scatter diagram and draw two lines of best fit which pass through the plotted points, these lines are called regression lines. In linear regression these lines are straight ones. These regression lines are based on two equations called regression equations which give best estimate of one variable when the other ts exactly known or given, _REGRESSION ANALYSIS 439 REGRESSION LINES If we take the case of two variables X and Y, we shall have two regression lines as the regression of X on Y and the regression of Y on X. The regression line of Y on X gives the most probable values of Y for given values of X and the regression line of X on Y gives the most probable values of X for given values of Y. However, when there is either perfect positive or perfect negative correlation between the two variables (r= + 1) the regression lines will coincide, é e. we will have only one line. The farther the two regression lines from each other, the lesser ts the degree of correlation and the nearer the two regression lines to each other, the higher is the degree of correlation. If the variables are independent, r is zero and the lines of regression are at right angles, £ e., parallel to OX and OY. It should be noted that the regression lines cut each other at the point of average of X and Y, ie., if from the point where both the regression lines cut each other a perpendicular is drawn on the X-axis, we will get the mean value of X and if from that point a horizontal line is drawn on the Y-axis, we will get the mean value of Y. It is important to note that the regression lines are drawn on least squares assumption which stipulates that the sum of squares of the deviations of the observed ‘Y values from the fitted line shall be minimum. The total of the squares of the deviations of the various points is minimum only from the line of best fit. The deviation from the points from the line of best fit can be measured in two ways—vertical, ie., parallel to Y-axis, and horizontal, ie., parallel to X-axis. For minimising the total of the squares separately it is essential to have two regression lines. The regression line of Y on X is drawn in such a way that it minimises total of squares of the vertical deviations and the regression line of X on Y minimises the total squares of the horizontal deviations. This can best be appreciated with the help of the following example: Height of fathers (inches) 65 63 67 64 68 62 70 66 68 67 69 71 Height of sons (inches) 68 66 68 65 69 66 68 65 71 67 68 70 ‘The two regression equations corresponding to these variables are : X=- 3:38 + 1.036 Y ld y = 35-82 + 2.476 X li By assuming any values of Y we can find out corresponding values of X from Eq. (0. For example if ¥ = 65, X would be - 3:38 + 1-036(65) = 63-96 Similarly, if Y = 70, X would be - 3.38 + 1-036(70) = 69-14 We can plot these points on the graph and obtain regréssion line of X on Y. Similarly, by assigning any values to X in Eq. (ii) we can obtain corresponding values of Y. Thus, if X = 63, Y would be : .440 STATISTICAL METHODS 35:82 + -476(63) = 65-808 or 65-81 and for X= 70, Y would be 35-82 + -476(70) = 69-14 The graph of original data and these lines would be as follows : 2 3 Regression Line of Y On x Heitht of Sons (Inches) ZZeB Boo 2 3 Regression Line of X On Y 62 63 64 65 66 67 68 69 70 71 Height of Fathers (inches) (Graph of Original Data) 88 Height of Sons (Inches) es Bf & 63 64 65 66 67 68 69 70 71 Height of Fathers (Inches) . Regression of Y on X =(¥--Yc)* IS MinimumREGRESSION ANALYSIS 441 Regression of Y on X, E (Y ~ ¥.)7 is minimum. Heitht of Sons (Inches) 63 64 65 66 67 68 69 70 71 Height of Fathers (Inches) Regression of Y on X £(¥-Yo)* 1S Minimum REGRESSION EQUATIONS Regression equations, also known as estimating equations, are algebraic expressions of the regression lines. Since there are two regression lines, there are two regression equations—the regression equation of X on Y is used to describe the variations in the values of X for given changes in Y and the regression equation of Y on X is used to describe the variation in the values of Y for given changes in X. Regression Equation of Y on X The regression equation of Y on X is expressed as follows : Y=at bX It may be noted that in this equation ‘Y is a dependent variable, ie., its value depends on X. ‘X is independent variable, ie., we can take a given value of X and compute the value of Y. ‘a’ is “Y-intercept” because its value ts the point at which the regression line Crosses the Y-axis, that 1s, the vertical axis. ‘b' is the “slope”. of line. It Tepresents change in Y variable for a unit change in X variable. ‘a’ and ‘b' in the equation are called numerical constants because for any given straight line, their value does not change. If the values of the constants ‘a’ and ‘b’ are obtained, the line is completely determined. But the question is how to obtain these values, The answer is Provided by the method of Least Squares which states that the line should be drawn through the plotted points in such a manner that the sum of the Sauares of the deviations of the actual Y values from the computed Y values is the least, or in other words, in order to obtain a line which fits the points best 2(Y~ ¥.*, should be minimum. Such a line is known as the line of ‘best fit.442 STATISTICAL METHODS, A straight line fitted by least squares has the following characteristics : 1. It gives the best fit to the data in the sense that it makes the sum of the squared deviations from the line, ¥ (Y - Y9?. smaller than they would be from any other straight line. This property accounts for the name ‘Least Squares’ 2. The deviations above the line equal those below the line, on the average. This means that the total of the positive and negative deviations is zero, or E(Y- Yo =0. 3. The straight line goes through the overall mean of the data (X, ¥). 4, When the data represent a sample from a large population the least squares line fs a ‘best’ estimate of the population regression line. With a little algebra and differential calculus it can be shown that the following two equations, if solved simultancously, will yield values of the parameters a and b such that the least squares requirement is fulfilled" : YY-Na+h3sX EXY= ab X+b¥ X* These equations are usually called the normal equations. In the equations =X, IXY. =X? indicate totals which are computed from the observed pairs of values of two variables X and Y to which the least squares estimating line is to be fitted and N is the number of observed pairs of values. Regression Equation of X on Y ‘The regression equation of X on Y is expressed as follows : Xe=a+bY To determine the values of a and b, the following two normal equations are to be solved simultaneously : EX=Nat+bzLy EXY=alY+bEyY* Mlustration 1. From the following data obtain the two regression equations : x 6 2 10 4 8 Y 9 W 5 8 7 (M. Com. Vikram Univ., 1996) Solution. OBTAINING REGRESSION EQUATIONS x Y XY x2 v2 6 9 54 36 8 2 " 22 4 10 5 50 100 8 7 Regression equation of Y on X: Yo=a+bX To deten the values of a and b the following two normal equations are to be solved. E(Y- Yq? should be minimum or £(¥-a-bX) should be mininum (Since Ye=a+bX) Let S=2(¥-a-bx) Differentiating partially with respect to a and b, OS sy-a-bX(- Ye IS yy Fart W¥-a- bX =O and SP ss(y-a-bH(-xH=0 or L(Y-a-bX)=0 or EX=Na+bEX and “Exy=aExX+bEX?REGRESSION ANALYSIS. 443 EY=Natbyx EXVe aE xe by x? ‘Substituting the values 40=5a+30b () 214 = 304+ 220b oli) Muttiplying equation () by 6, 240 = 30 a+ 180 b (ii 214 =30 a+220b (wy Deducting equation (iv) trom (iil) - 40 b = 26 or b= - 065 ‘Substituting the value of b in equation (/) 40 = $a + 30(- 065) or 5 a= 40+ 195 =5950ra=119 Putting the values of @ and bin the equation, the regression of Yon X is 19-065 X Regression line of X on Y: Xc=a+ bY and the two normal equations ar , EX=Na+b3y EX Y=salV+bz 2 30=5a+40b (i) 214=40a+340b (ui) Multiplying equation (/) by 8: 240 = 40 a+ 320 b (it) 214 = 40 a+340b (my) From Eqn. (ii) and (iv) - 20 b= 26 or b=-13 ‘Substituting the value of bin equation (i) ; 30 = 5 a+ 40 (- 1.3) 5 a=30+52= 62 a=164 Putting the values of a and b in the equation, the regression line of Xon Vis X= 164-13 Y. Deviations Taken from Arithmetic Means of X and Y The above method of finding out regression equation is tedious. The calculations can very much be simplified if instead of dealing with the actual values of X and Y we take the deviations of X and Y series from their respective means. In cuch a case the two regression equations are written as follows : = 6, < () Regression Equation of X on Y; X- X= La (Y-Y) y X is the mean of X series ; Y is the mean of Y series r 2 is known as the regression coefficient of X on Y. ‘The regression coefficient of X on ¥ is denoted by the symbol byy or by. It measures the change in X corresponding to a unit change in Y. When deviations are taken from the means of X and Y, the regression coefficient of X on Y fs obtained as follows : Owe _Exy by, or r= = =* ye By eye Instead of finding out the value of correlation coefficient, 0 . 6 y. etc, we can find the value of regression coefficient by calculating x y and & y* and dividing the former by the latter. By (i) Regression Equation of Yon X; Y- Y=r 7 (X-X) a Ox ree EAU 8 ELY and by = UMM 2p Cy Noxoy sy ¥ oy J In case we deal with actual values of X and Y variables and not the deviations,STATISTICAL METHODS 6 eae is the regression coefficient of Y on X. It ts denoted by by or bg. It x ures the change in Y corresponding, to ¢ unit change in X. When deviations are taken from actual means, the rege icient of Y on X can be obtained as follows : rout EX u Gx Ex? It should be noted that the under-root of the product of two regression cocflicients gives us the value of correlation coefficient. Symbolically, r= Vay x byx 9, s Proof. by =r and by =r! “Sy . ox Su 2,2 yy Ox The following points should be noted about the regression coefficients : 1. Both the regression coefficients will have the same sign, ie., either they will be positive or negative. It is never possible that one of the regression coefficients is negative and the other positive. 2. Since the value of the coefficients of correlation (7) cannot exceed one, one of the regression coefficients must be less than one or, in other words, both the regression coefficients cannot be greater than one. For example, if by =1-2 and by=14 the value of correlation coefficient would be V1-2 x 1-4 = 1-296 which is not possible. 3. The coefficients of correlation will have the same sign as that of regres- sion coefficients, ie., if regression coefficients have a negative sign, r will also be negative; and if regression coefficients have a positive sign, r would also be positive. For example, if by=-08 and by=-12, r would be 8 x -1-2 =— 0-98 and not + 0:98. 4, Since byy, Bx y X bye = TSX 1 = VDyy x Dyx 6 Pa we can find out any of the four values given the other y three. For example, if we know that r= 0-6, 6.= 4 and by, = 0'8, we can find oy. Ox bys 06x40 224 Sy w" 08” 5. Regression coefficients are independent of change of origin but not of scale. Proof. Let U= X- Aand V= Y-B where A and Bare arbitrary means Substituting the given values : 0:8 = Y=B+v _Y=B+V w-Y=(V-/Y oi = 03 Hence change of origin has no effect on the regression coefficient. E xy Sy EY Aso by = COVEY) = - Su Gx Noxoy ox yx? on Ox In case we deal with actual values of X and Y variables and not the deviations x ZX) (EY)REGRESSION ANALYSIS 7 Illustration 2. From the data of Illustration 1, calculate the regression equations taking deviation of items from the mean of X and Y series. Solution : CALCULATION OF REGRESSION EQUATIONS (X=) 7 “( Y- Y xe ee 2 a v 6 oO 0 9 +1 oO 2 “4 16 W +3 -12 10 +4 16 5 -3 —12 4 -2 4 8 ° 0 8 __ +2 4 u -1 -2 EX=30 Ex=0 sy2-49 2¥=40 Ly=0 Exy=-26 Regression Equation of X on Y : X- La 1:3(Y-8)=-13 +104 -13Y+164 of X=164-13Y Hence, Regression Equation of Yon X: Y- V= rex) x oy _Exy_- 26 __ Ose 407 0-65 Y-8=-065 (X-6) =- 065 X+39 Y=-065X+119 of Y=119-065X Thus, we find that the answer is the same as obtained earlier. However, the calculations are very much simplified without the use of normal equations. Deviations Taken from Assumed Means When actual means of X and Y variables are in fractions the calculations can be simplified by taking the deviations from the assumed means. When deviations are taken from assumed means the entire procedure of finding regression equations remains the same—the only difference is that instead of taking deviations from actual means, we take the deviations from assumed means. The two regression equations are : x- =r %(y-Y) oy The value of r S¥ will now be obtained as follows : y Ox NE de dy~(E de X¥ dy Gy NE dj -(2 a)” d,=(X- A) and dy=(Y~ A)STATISTICAL METHODS 446 Similarly, the regression equation of Yon X is 6 Y-¥ na (X= X) oy NS dydy-Ldy¥ dy rYe 4 e Sx NE AZ - (Sy It should be noted that in both the cases the numerator 1s the same, the only diflerence ts in the denominator. When the regression coefficients are calculated from correlation table their value obtained as follows : NE fit ty ~ ¥ fle ¥ fly be = class interval of Y variable NE fay dy-EdyE fay i oy NEL indy ly BL yy y Ox. NE fag - (& fay) be As is clear from above the formulae for calculating regression coefficients in a correlation table are the same—the only difference is that in a correlation table we are given frequencies also and hence we have multiplied every value bys. Mlustration 3. From the data of Illustration 1 calculate regression equations by taking devia- Similarly, tions of X series from § and of Y series from 7. x (x= 5) — ¥ (¥7) ox a? _ iY Ca x dy 6 4 1 9 +2 4 +2 2 -3 9 W +4 16 12 10 +5 25 5 -2 4 -10 4 1 1 8 ra] 1 a 8 +3 9 7 0 o 0 EX=30 Sae=+5 sa?-45 LY=40 La=5 Sop =25 = dkdy=- 21 Regression equation of X on ¥: X- X= r 2% (Y-Y) oy] 13 (5) (25) ~ (6)? 13(¥-8) 13Y+104 or X=164-13Y Regression Equation of Yon X: y--y= 2% (x X) Ox 1 2t NE dy dy—E dy Edy NE de (5 dy? = 3221) ~(5) (5) _ -105- 25 (5)(45)-(5)2 200 765 Y¥-8=-065 (x-6) Y-8=-065X+39 of Y=119-065xSION ANALYSIS 447 REGRI It is clear from this example that answer would come out to be the same whether we take deviations from actual means or assumed means. Graphing Regression Lines It is quite casy to graph the regression lines once they have been computed. All one has to do ts to (a) choose any two values (preferably well apart) for the unknown variable on the righthand side of the equation, (b) compute the other variable, (co) plot the two pairs of values, and (a) draw a straight line through the plotted points. Nlustration 4. Show graphically the regression equations of illustration 3. Solution. (a) Regression Line of Yon X[ Y= 11.9 - 063 X]. (@ Let X=2, ¥=119-065(2) =11-9-13= 106. () Let X= 10, Y= 11-9 -065x 10=54. These points and the regression line through them are shown on the graph on the next page. (b) Regression line of Xon Y(X = 16-4 - 1:3 Y) w Let Y-10 Es X= 16-4 - 13(10) = 164-13 =34 ( Let Y=6 6-4-1-3(6) =164-78=86 Mlustration 5. From the data of illustration 1, obtain regression equations taking deviations from 5 in case of X and 7 in case of Y These points and the regression line through them are shown in the graph below : Regression Line of X on Regression Line of YonX x ‘Thus the value of regression coefficient comes out to be the same Ilustration 6. The following data relate to the scores obtained by 9 salesmen of a com pany in an intelligence test and thew weekly sales in thousand rupeos : Salesmen Ineligence: A B C D0 E F G oH | Test Scores: 50 60 50 60 80 50 80 40 70 Weekly Sales: 30 60 40 50 60 30 70 50 60 telligence test scores of the salsemen (a) Obtain the regression equation of sales on intellige ne salsemen {b) If the inteligence test score of a salesman in 65, what would be his expoctty wash) sales?STATISTICAL METHODS 448 Solution. Let intelligence test score be denote CALCULATION OF REGRESSION EQUATIONS _ d by X and weekly sales by Y. XX (X= 60) y (Y= 50) 2 » x x? YE “so 10 wo 80 =20 400 +200 60 0 0 60 +10 100 0 50 -10 100 40 -10 100 +100 60 0 0 50 oO oO 0 80 +20 400 60 +10 100 +200 50 -10 100 30 -20 400 +200 80 +20 400 70 +20 400 +400 40 20 400 50 0 0 oO 70 +10 100 60 +10 100 +100 “EX=540 Ex=0 sx221600 2¥=450 Ey=0 sy% 1600 =xy=1200 Regression equation of ¥ on X: ¥-¥= FX (X—%) rf Sy _ Exy _ 1200 _ 9. ox ye ~ 16007°"° 75X-45 or Y=5+0-75X Expected weekly sales when intelligence test score of a salesman is 65 +0-75 X. Putting X = 65 75 x (65) + 5 = 48-75 + 5 = 53-75 Illustration 7. The following table shows the ages (X) and blood pressure of 8 persons. 63 45 36 72 65 25 Va 62 53 51 25 * 79 43 a 33 Obtain the regression equation of Y on X and find the expected blood pressure of a person who is 49 years old. (B. Com. Bombay Univ... 1996) Solution. CALCULATION OF REGRESSION EQUATION OF YON x xX (X-50) Y (Y- 50) - de a dg a? dy 52 42 4 62 +12 144 ———SSO 63 +13 169 53 +3 9 +39 45 5 25 51 rs 1 -5 36 “14 196 25 -25 625 +350 72 +22 484 79 +29 841 +638 65 +15 225 43, -7 49 =105 47 -3 9 +10 100 -30 17289 +425 Ed2-2068 Edad 1505 25 -25 625REGRESSION ANALYSIS 449 y EY _ 406 LEX _ 405 _ Yo ty tg 79078, X= "= OS . 50.625 Oy NEG dy 3 dh ¥ dy | (8) (1336) - (5) (6) _ 10688 - 30 _ 4 r6g Ox NY ae (Lay? (8) (1737) - (5)? 13896 - 26 Y¥ - 50.75 = 0.768 (X - 50625) Y- 50.75 = 0768 X 88 oor Y= 1187+0-768 x Yao = 11.87 + 0-768 (49) = 49.502 Thus, the expected blood pressure of a person who is 49 yonrs old shall be 49 5. tlustration 8. In a correlation study the following values are obtained x Y Mean 65 67 Standard Deviation 25 as Coetticient of Correlation 08 Find the two regression equations that are associated with the above values. (B. com., Kashmir Univ., 1996; MBA, HPU, 2000) Solution. The two regression equations are : = rey Oy X= 65, r= 08, 0x=25, y= 3.5, ¥= 67 Substituting the values : X-65 = 8 28 (y— 67) X65 = 6714 (Y- 67) X~65 =-5714 Y- 38.28 or X= 26-72 + 05714 Y Regression Equation of ¥ on X: Y- Y= rH (X—R) _ Regression Equation of X on Y: X-X = 23 (y_ Y- 67 = 852 (X- 65) 12 (X- 65) 12 X-728 of Y=-5841-12X. Mlustration 9. In a partially destroyed laboratory record of an an: the following results only are legible : Variance of X= 9 Regression equations 8 X- 10 ¥+66=0 40 X-18 Y=214 Find on the basis of the above information () The mean values of X and Y, (1 Coetticient of correlation between X and Y, and (ui) Standard deviation of Y. (B. Sc., Madurat-Kamaraj Univ.. | M. Com, M.D. Univ..: M. Com., Bhopal, 1999) ‘alysis of correlation data, Solution. () The Mean values of X and Y: 8X - 10 Y=~ 66 Ai 40X-18 ¥=214 Oy Muluplying equation (i) by 40 X- 50 Y= - 330 40X-18 Y=214 Y=17 or 7 Substituting the value of Yin eq, (i) :8X - 10x 17 =~ 66 8X=~-66+170 a 8X = 104 X=130rX=13 ch of the two regression equations is the equation @ ¥ on ¥. we make © eq, (/) as the regression equation of X on Y.STATISTICAL ME Tog, 450 gX--66+10Y mn x2 884 0; or bya e125 ~18Y=214 we can calculate byx a0 X med =18 Y=214-40X or, 214 40 yo by = 40 Ye~"1e * 18 18 Since both the regression coolticionts are exceeding 1, our assumption is wrong. Hence, the first equation is equation of Yon X. From eq. (i) From eq. (i) > Bye OE Ox=N9=3; by= ro 3 18 045-65 or O45 0y=18 or oy=-7 y Hence, standard deviation of Yis 4. Mlustration 10. For 50 students of a class the regression equation of marks in Statistics (X) on the marks in Accountancy (Y) is 3 Y-5 X+180=0. The mean marks in Account ancy is 44 and variance of marks in Statistics is 9/16th of the variance of marks in Ac- countancy. Find the mean marks in Statistics and the coefficient of correlation between marks in the two subjects. IM. Com., Madurai, 1993 ; B. Com. (H), Dethi Univ., 1994) Solution. We are given 3Y-5X+180=0 or 3Y+180-5X X represents marks in Statistics and Y marks in Accountancy. When Y= 44, X will be given by 5 X= (3) (44) + 180 =0; 5 X= 192 + 180 or X= S12 - 62.4 Hence, the mean marks in Statistics are 62-4. For calculating coefficient of correlation we know that Oe by=r Regression coefficient of X on Y from the given equation is 5X=3 ¥+180 or X=06 Y+36 by= 06:12 given oy" Vie fe 06s ie or O6= a Hence, 3r=24 * rs +08. Mlustration 11. You are given the following data : x Y Arithmetic mean 36 85 Standard Deviation "1 8 Correlation coefficient between X and ¥ = 0. = 066 () Find the two Regression Equations, and (i) Estimate the value of X when Y = 75. (IB. A. {Hons.). Econ. Dethi Univ. R. Cam Canwahati Unive