10, 2 SIGNIFICANCE of MEASURING CORRELATION The objecti Werf an observations or variables ie "esearch is to establish relationships between two or more sets of les to arrive at some valid conclusion. Few advantages of measuring an association (or correlation) ) between two or more variables are as under: 1. Correlati ; the men analysis contributes to the understanding of economic behaviour, aids in locating conte ios WR variables on which others depend, may reveal to the economist the Stabilizing tye, hich disturbances spread and suggest to him the paths through which ig forces may become effective, —W. A. Neiswanger See of correlation is to reduce the range of uncertainty of our prediction. The prediction ased on correlation analysis will be more reliable and near to reality. — Tippett In economic theory, an association (or correlation) between two or more variables, such as Price, supply. and quantity demanded; customers retention is related to convenience, amenities and service standards; yield of a crop is related to quantity of fertilizer applied, type of soil, quality of seeds, rainfall and so on is established. 4. In healthcare, an association (or correlation) between two or more variables such as validity and reliability of clinical measures; effect on health due to certain biological or environmental factors, blood pressure and age of a person; inter-observer reliability for two doctors who are assessing a patient's disease, and so on is established.‘orreltion Anabsis_379 eee ea ere aw Linear and Nordinear Correlations sfers to an association betwe -" ar correlation re ren two variables where variation in thei Ae proportional oF fixed. The following pattern of variation inthe snlucs of too eben ody as linear correlation. 7 i rr ee ee) 5: 0 60 8 10010 hen thes pais of values of x and y are plotted on a graph paper, the line joining these points sould bea straight lin. - Me near (or cury linear) correlation refers t an asociaton beween two variables where arom in their values is neither proportional nor fixed. The following pattern of variation in the ies of two variables x and y reveals non-linear correlation, ogg 9. 10 Dee aaa 5: 80 130 170 (150 280 560460600 When these pairs of value of andy ae plotted on graph paper, the line joining these points sould not be a straight line, rather it would be curvy linea. 443 Simple, Partial and Mutiple Correlations ‘edstncion between simple, part and multiple correlations is based upon the number of variables iaalved in the correlation analysis. TK only eo variables are chosen to study correlation between them, then such a correlation is seerred 25 imple correlation. A study on the yield ofa crop with respect to only amount of fertzer the or sales revenue with respect to amount of money spent on advertisement, are a few examples of simple corelation. "partial correlation, two variables are chosen to study the correlation between them but the cfec of other influencing variables is kept constant. For example (i yield of a crop is influenced by tie amount of fertilizer applied, whereas effect of other influencing variables such as rainfall, quality of ‘ed, ype of soil and pesticides is kept constant, and (i) sales revenue from a product is influenced by tie level of advertising expenditure, whereas effect of other influencing variables such as quality ofthe oduct, price, competitors, distribution and so on is kept constant. a tile corelation, more than two variables are chosen to study the correlation among te for example ()employer-emploee relationship in any organization may be examined wih ‘eco trdning ad development fits medi, housing and education to cilren ates ‘ey srcure; grievances handling sjstem; and so on, and (i) sales revenue from a product may examined in relation with the level of advertising expenditure, quali ‘caine! Stl wh he ig expenditure, quality of the product, price,VE ————— 11.2 ADVANTAGES OF REGRESSION ANALYSIS “The following are few advantages of regression analysis: ont 1. Regression analysis helps in developing an algebraic equation between oad ee a rie cA the given data and estimating the value of a dependent variable given the valu -pendent variable. a 2. Regression analysis helps to determine standard error of estimate to measure the variability or spread of values of a dependent variable around the regression line. Closer the pair of values &y) fall around the regression line, better the line fits the data and hence smaller the variance and error of estimate. Thus, a good estimate can be made ‘of the value of variable y when all the points fall on the line, ie. standard error of estimate equals zero. 3. If the sample size is large (n 2 30), then interval estimation for predicting the value of a dependent trai ple besed on standard error of estimate is considered to be acceptable by changing the values Sfether or. The magnitude ofr? remains the same regardless of the values of the two variables.18 ASSUMPTIONS FOR A SIMPLE LINEAR REGRESSION MoDEL foam a basis for application of simple linear aon rom which sample of observations i eager nes "ears certain assumptions about the ‘Theres a linear relationship between the dependent variable yaad ing ' atonsip canbe described by a linear regression equation y =a agree vrable x. This the deviation in the value of dependent variable, y, from is expert ens (ener ePTEsES independent variable, x. a given value of ‘he set of expected (or mean values ofthe dependent variable, forgiven values o 5 riable,, are normally distributed. The mean ofthese normally distrib independent buted values fllson the line cfregression 4. The dependent variable y is a continuous random variable, whereas values of the independent sariable x are fixed and not random. 4 The sampling error associated with the expected value of the dependent variable y is assumed to bean independent random variable distributed normally with mean zero and constant standard deviation. The amount of deviation (error) in the value of dependent variable , may be different in successive observations. The standard deviation and variance of expected values of the dependent variable about the regression ine are constant forall values ofthe independent variable x for the set of observations in asample. ‘The expected value of the dependent variable cannot be obtained for a value of an independent variable filing outside the range of values in the sample. 11.6 PARAMETERS OF SIMPLE LINEAR REGRESSION MODEL ‘The objective regression analysis is to ascertain that a regression equation (line) should provide the ‘est fit of sample data to the population data so that the error of variance is as small as possible. JR. Stockton stated thatthe device used fo estimating te values of one variable fam the valu of the other consists of ‘line though the points; drawn in sucha manner ato represent the average relationship between the two variables. Such a ine i cae line of regression. ‘The two variables x and y which are correlated can be expressed in terms of each other in the form of straight line equations called regression equations a follows: + The regression equation of y on yrathe is used for estimating the value of for given values of x + Regression equation of x on y +h is used for estimating the value of x for given values ofy Remarks 1. Regression lines coincide (overlap) when variables x and y are perfectly correlated (either positive — P) po or negative) , a>»ST | Greatolion Lien avalyse 1 Pegaso) nab veolobion este 4S Rogirassiog : ® = ee seareaiee tg Rt oe Corexe tiLion Ope ue. Hod ore vardahh 2 ue aay 2: ee < Speen ene MgB DO eens Top sete (ug ier OR sa ote 12) 10D ul thee. dh nO Bead rela Corvutalion B eae Eee eps. mares cing bop