However, it would be necessary to investigate this constraint that a set of common parameter estimates is sufficient to capture the population characteristics. Regression mixture models, on the other hand, places the regression structure in a much more flexible way.