This study compares the classification accuracy of linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), logistic regression (LR), and classification and regression trees (CART) under a variety of data conditions. Past research has generally found comparable performance of LDA and LR, with relatively less research on QDA and virtually none on CART. This study uses Monte Carlo simulations to assess the crossvalidated predictive accuracy of these methods, while manipulating such factors as predictor distribution, sample size, covariance matrix inequality, group separation, and group size ratio. The results indicate that QDA performs as well as or better than the other alternatives in virtually all conditions. Suggestions for practitioners are provided.
Abedi, J. (1991). Predicting graduate academic success from undergraduate academic performance: A canonical correlation study. Educational and Psychological Measurement, 51, 151-160.
2.
Arminger, G., Enache, D., & Bonne, T. (1997). Analyzing credit risk data: A comparison of logistic discrimination, classification tree analysis and freeforward networks. Computational Statistics, 12, 293-310.
3.
Baird, L. L. (1975). Comparative prediction of first year graduate and professional school grades in six fields. Educational and Psychological Measurement, 35, 941-946.
4.
Baron, A. E. (1991). Misclassification among methods used for multiple group discrimination: The effects of distributional properties. Statistics in Medicine, 10, 757-766.
5.
Bryan, J. G. (1961). Scientific Report No. 2: Calibration of qualitative and quantitative variables for use in multiple-group discriminant analysis. Hartford, CT: The Travellers Insurance Company.
6.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.
7.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112(1), 155-159.
8.
Dey, E. L., & Astin, A. W. (1993). Statistical alternatives for studying college student retention: A comparative analysis of logit, probit, and linear regression. Research in Higher Education, 34, 569-581.
9.
Fan, X., & Wang, L. (1999). Comparing linear disriminant function with logistic regression for the twogroup classification problem. Journal of Experimental Education, 67(3), 265-286.
10.
Ferrer, A. J. A., & Wang, L. (1999, April 13-17).Comparing the classification accuracy among nonparametric, parametric discriminant analysis and logistic regression methods. Paper presented at the annual meeting of the American Educational Research Association, Montreal, Canada.
11.
Fleishman,A. I. (1978).A methodfor simulatingnon-normaldistributions. Psychometrika,43(4), 521-532.
12.
Headrick, T. C., & Sawilowsky, S. S. (1999). Simulating correlated multivariate nonnormal distributions: Extending the Fleishman power method. Psychometrika, 64, 25-35.
13.
Hess, B., Olejnik, S., & Huberty, C. J. (2001). The efficacy of two improvement-over-chance effect sizes for two-group univariate comparisons under variance heterogeneity and nonnormality. Educational and Psychological Measurement, 61(6), 909-936.
14.
Huberty, C. J. (1994). Applied discriminant analysis. New York: John Wiley.
15.
Huberty, C. J., & Curry, A. R. (1978). Linear versus quadratic multivariate classification. Multivariate Behavioral Research, 13, 237-245.
16.
Johnson, R. A., & Wichern, D. W. (2002). Applied multivariate statistical analysis. Upper Saddle River, NJ: Prentice Hall.
17.
Koffler, S. L. (1976, April). An evaluation and comparison of discriminant procedures for certain types of non-normal distributions. Paper presented at the annual meeting of the American Educational Research Association, San Francisco.
18.
Krzanowski, W. J. (1977). The performance of Fisher's linear discriminant function under non-optimal conditions. Technometrics, 19(2), 191-200.
19.
Maxwell, S. E. (1961). Canonical variate analysis when the variates are dichotomous. Educational and Psychological Measurement, 21, 259-271.
20.
McClachlan, G. J. (1992). Discriminant analysis andstatistical pattern recognition. New York: John Wiley.
21.
Meshbane, A., & Morris, J. D. (1995). A method for selecting between linear and quadratic classification models in discriminant analysis. Journal of Experimental Education, 63, 263-273.
22.
Meshbane, A., & Morris, J. D. (1996, April). Predictive discriminant analysis versus logistic regression in two-group classification problems. Paper presented at the annual meeting of the American Educational Research Association, New York.
23.
Remus, W., & Wong, C. (1982). An evaluation of five models for the admission decision. College Student Journal, 16, 53-59.
24.
Schumacher, M., Rossner, R., & Vach, W. (1996). Neural networks and logistic regression: Part I. Computational Statistics: Data Analysis, 21, 661-682.
25.
Stevens, J. (1996). Appliedmultivariate statistics for the social sciences. Mahwah, NJ: Lawrence Erlbaum.
26.
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics. Boston: Allyn & Bacon.
27.
Williams, C. J., Lee, S. S., Fisher, R. A., & Dickerman, L. H. (1999). A comparison of statistical methods for prenatalscreeningfor down syndrome. AppliedStochastic Models in Business andIndustry, 15, 89-101.
28.
Wilson, R. L., & Hardgrave, B. C. (1995). Predicting graduate student success in an MBA program: Regression versus classification. Educational and Psychological Measurement, 55(2), 186-195.
29.
Young, B. (1993, November). Quadratic versus linear rules in predictive discriminant analysis. Paper presented at the annual meeting of the Mid-South Educational Research Association, New Orleans, LA.
