5 Conclusion

A method based on the $\chi^2$ merit function is presented to compare "observed" with synthetic single stellar populations. Monte-Carlo simulations have been performed to display the diagnostic power from a CMD containing the points for which no corresponding synthetic point was found within a reasonable error ellipse. The simulations indicate that the CMDs of residual points might provide hints about model parameters to be improved. The simulations further indicate that one ought to be cautious with the analysis of stellar luminosity functions and that strong hints can be obtained about the shape of the initial mass function.

Acknowledgements

I. Saviane and L. Portinari are thanked for reading of the manuscript. G. Bertelli and C. Chiosi are thanked for their stimulating discussions. In particular, G. Bertelli is acknowledged for making available the synthetic stellar population generator (HRD-ZVAR), which produced the stellar populations used for this paper. The research of Ng is supported by TMR grant ERBFMRX-CT96-0086 from the European Community.