 Our students now we are going to do another exercise of multiple linear regression in SPSS software. In previous module we did the exercise with enter method. Now we are going to do the perform an analysis with the stepwise method. Students I have opened the file of SPSS. Here we are going to the tab of regression and after clicking it we are going to the linear regression and in linear regression I have asked the SPSS to take the same variables but just change the method of doing the analysis. Now I'm doing the analysis with stepwise method. So what will happen in stepwise method? More than one, your regression equations will be formed and models will be formed. And among these strong predictors will come on the first number stepwise and the least strong predictor will come on the end and the variable predict is not doing. Dependent variable will be automatically excluded which was not in the enter method. So we will take the statistics, we will take the R-squared change and we will take the descriptives as well. We will continue, we will take the plots and we will okay it. So here we see that first we will look at the colliderity, we will look at the colliderity as well that our variables are IVs. So there is no significant relationship in all of them, a strong relationship which is of 0.80 level. So there is no multicolliderity on this indicator. Now here you see how many models have you made? Five models have you made which stepwise is that regression equations are made more by adding one by one the variables, independent variables. Look at this, your first predictor told that there is self-expression along with constant. Second, self-expression with making new social ties. Third, self-expression new social ties, seeking and sharing information. And after that self-expression make new social ties, seeking and sharing information, maintain existing social ties. And the last model in that, you have personal information, disclosure be added. No dependent variable is our offline bonding social capital. So automatically we found out that our strongest predictor is self-expression and our least predictor is personal information disclosure. Now along with this, R-square is also giving that your first variable is 10.3%, 2% is explaining your dependent variable. After that, you get to know from R-square that 3% 1% and 4.5% is explaining your dependent variable. Actually, the main predictor is your self-expression which is explaining 10%. In the models, all these equations are theoretically significant. And if we go to coefficients, then you see that the coefficient of self-expression is 0.498, it is standardized and it is 0.321. But as soon as we added our second variable self-expression, then with making new social ties, our standardized coefficient decreased. When we added our third variable, then it decreased even more. And when we added our fourth variable, we found out that it decreased even more. And in the fifth one, we can see that it decreased even more. So what happens with this is that you get to know exactly from the step-wise method that what is your actual predictor and when other variables are added, then the actual predictor has a lot of impact. That is, actually, in model 1, what we were saying is 0.321, that is actually 0.174. And the rest of these variables were explaining it. Talking about VIF, tolerance values are okay, VIF values are okay too. And with this, we can conclude that our model regression equation is significant. And this 16.3% variance explains the independent variable in which self-expression is the strongest predictor in explaining the bonding social capital of the youth.