Researchers compared different methods for determining the correct number of groups in a type of statistical model. They tested these methods using computer simulations with different conditions like sample size and complexity of the model. The results showed that some methods were better at identifying the correct number of groups than others. In general, as the number of people or items in the model increased, the accuracy of the methods also improved. The Bayesian information criterion (BIC) and other similar methods performed well, while others like Akaike's information criterion (AIC) did not work as effectively. Overall, the study found that certain methods were more reliable for identifying the correct number of groups in the model.