Normality, Linearity, Multicollinearity, Homoscedasticity, Outliers and Other Assumption in SPSS

1. Univariate outliers examined by standardized z scores and need to be within ± 3.29 range.
2. Normality assumption for the responses tested through the values of skewness and kurtosis, where the values need to be found within ±1 and ±3 respectively.
3. To check the linearity assumption, regression analysis was performed and the scatter plot of residuals revealed that multivariate relationship was linear.
4. To check for multicollinearity, Multicollinearity is normally checked at Variance inflation factors (VIF) for each variable. Analysis of collinearity statistics shows this assumption has been met, as VIF scores were below 10, and tolerance scores above 0.2
5. The values of residual are independent. The Durbin-Watson statistic showed that this assumption had been met, as obtained value was close to 2
6. The variance of the residual is constant. In this study plot of standardized residual vs. standardized predicted values showed no sign of funneling, suggesting the assumption of homoscedasticity has been met. In addition to that According to Tabachnick and Fidell (2007) Homoscedasticity is related to normality and when the normality assumption is met the relationship between the variables is said to be homoscedastic

[@user-qk1ux6sd6r]

Good