In OLS Regression, we shall make the following assumptions:

1-The residuals are independent of each other.

2- The residuals have a common variance.

3- It is sometimes additionally assumed that the errors have the normal distribution. For example, if the residuals are normal, the correlation test can replace the independence test.

After calculating the latent variables, other software measures the relationship between them using OLS or bootstrap OLS method. The results will be distorted when the assumptions of the OLS method are not confirmed.

Furthermore, the independent latent variables may have a significant influence on each other. This issue is known in the statistical literature as multicollinearity.

The  PLS and PC methods are included in the SEM-PLS software to solve these problems.

For example, suppose we want to examine the effect of the amount of training and performance on the point of the game for a football team. In this example, Performance and Training are independent variables, and Point is dependent. But it may distort the strong relationship between the two variables Performance and Training of the model results.  In addition, the assumptions of the OLS method may not be confirmed.

You can measure multicollinearity in the SEM-PLS software (Path Analysis, D8.2.8). Also, you can use the PLS and PC methods to solve these problems (See PLS  Regression,  PC  Regression, and Path Analysis).