# Correlation analysis-meaning & types of correlation, Karl Person¶s coefficient of correlation and spearman¶s rank correlation; regression analysis

- meaning and two lines of regression; relationship between correlation and regression coefficient. Times series analysis- measurement of trend and seasonal variations; time series and forecasting.

Unit 2 A S Raheja asraheja@gmail.com 9717871155

Regression  Examples  Price and supply  Income & Expenditure  Correlation analysis is used to measure strength of the association (linear relationship) between two variables  Correlation is only concerned with strength of the relationship.  No causal effect is implied with correlation.  Reduces range of uncertainty.  Most widely used and widely abused statistical measure. .Correlation vs.

Types of Correlations  Issues with Correlation  Chance Coincidence  Influence of one on other or mutual dependence  Both being influenced by third variable  Methods  Scatter diagram  Karl Pearson¶s Coefficient of Correlation  Rank Method .

r<0 Y Y X No Correlation. r = 0 X X .Scatter Diagram + Correlation. r>0 Y Y Non Linear correlation X .Correlation.

Type of Correlation  A measure of the linear association between variables  ± Positive Correlation indicates positive linear relationship  ± Negative Correlation indicates a negative linear relationship  ± Values close to zero indicates no linear relationship  It not affected by the units of measurement for x and y variables  ± Pearson product moment correlation coefficient or Sample correlation coefficient. it is also referred to as Bi variate . r (used in case data is continuous)  It is a numerical index that reflects the linear relationship between two variables  The values of the descriptive statistic range between a value -1 (perfect ±ve correlation) to +1(perfect positive correlation) .

xy = sample covariance x = sample standard deviation of x y = sample standard deviation of y .Correlation Formula 1 W r ! WW xy xy x y _ _ 1 W XY ! n § ( X  X )(Y  Y ) 2 1 W x ! n §( X  X ) 1 ! W y n §(Y Y) 2 Where r = sample correlation coefficient.

r Covariance = 11.Calculating from Covariance Knowing the covariance and the standard deviations of each variable we can compute the sample correlation coefficient. y = 7.93 .93 So Pearson r = 11/(1.49 x 7.93) = 0.49. x = 1.

§ Y n 1 Y 2  (§ Y ) 2 § n 1 X 2  (§ X ) 2 § n .Correlation Formula 2 § XY  r! 1 § X .

Solution .

Assumed mean § xy  r! 1 § x.§ y n 1 y 2  (§ y ) 2 § n 1 x 2  (§ x) 2 § n x ! X a y !Y b .

Spearman¶s Rank Correlation  On some occasion it is not possible to measure the variables quantitatively or exact magnitude of the variable can not be determined  Numerical value of beauty  Measure taste of wine rs ! 1  6§ D 2 N ( N 2  1) .

Rules  When Rank are given ± above formula is sufficient  When Ranks are not given ± ranks are assigned by taking either the highest value as 1 or the lowest value as 1  Equal Ranks. in such case each individual is given an average rank . rank may be equal .in some cases.

Exercise 1 .

Exercise 2 .

Exercise 3 .