The article looked at different rules used to decide if a model fits well in practice. It tested new rules against old ones for various fit measures. The authors found that to be sure a model fits well with the data, they need to see fit measures close to .95 for some factors and around .08 for another. They also found that using two different measures together works best to check if a model fits correctly. Some fit measures might reject good models in small samples. So, it's essential to pick the right measures depending on the sample size.