SmartPLS

The posts follow the two stages of a PLS-SEM analysis in the order you run them. The measurement model comes first: outer loadings, composite reliability, average variance extracted, and discriminant validity by both the Fornell-Larcker criterion and HTMT. The structural model follows: collinearity with VIF, path coefficients and bootstrapping, R squared, f squared, Q squared, and then mediation, moderation, and multigroup analysis when the study calls for them.

Each post points to the exact place in SmartPLS 4, explains the results table it produces, gives its threshold with a source, and closes with a paragraph on writing that result into the results chapter. The troubleshooting posts cover the three problems that come up most: an outer loading too low to keep, an HTMT value over the limit, and a path that does not reach significance on the hypothesis you cared about.

If you are still choosing between PLS-SEM and covariance-based SEM in AMOS, one post compares them plainly, including the cases where PLS is the weaker option. Background on SEM and on the statistics underneath it sits in the Statistics category.

Where to start. Read the introduction to PLS-SEM and the comparison with CB-SEM first, so you are confident SmartPLS suits your model rather than that it was the tool with the most tutorials. Then follow the real run order: the measurement model post covering loadings, CR and AVE, then discriminant validity and HTMT, then the structural model with bootstrapping and path coefficients. The mediation and moderation posts matter only if your model contains them. Save the troubleshooting posts for the moment you hit the problem, since reading them in advance mostly teaches you to expect failures you may never have.