Should we always choose a nonparametric test when comparing two apparently nonnormal distributions?

References 

1. Welch BL. The significance of the difference between two means when the population variances are unequal. Biometrika. 1937;29:350–362
   • CrossRef
2. Chand U. Distributions related to comparison of two means and two regression coefficients. Ann Math Stat. 1950;21:507–522
3. Wetherill GB. The Wilcoxon test and non-null hypotheses. J Roy Statist Soc (Series B). 1960;22:402–418
4. Van der Vaart HR. On the robustness of Wilcoxon's two-sample test. In:  Jonge HD editors. Quantitative methods in pharmacology. New York: Interscience; 1961;
5. Pratt JW. Robustness of some procedures for the two-sample location problem. JASA. 1964;59:665–680
6. Lehmann EL. Nonparametrics (statistical methods based on ranks). San Francisco: Holden-Day; 1975;
7. Fenstad GU. A comparison between the U and V tests in the Behrens-Fisher problem. Biometrika. 1983;70:300–302
   • CrossRef
8. Zimmerman DW, Zumbo BD. Parametric alternatives to the student t test under violation of normality and homogeneity of variance. Percept Mot Skills. 1992;74:835–844
   • CrossRef
9. Bridge PD, Sawilowsky SS. Increasing physicians' awareness of the impact of statistics on research outcomes (comparative power of the t-test and Wilcoxon rank-sum test in small samples applied research). J Clin Epidemiol. 1999;552:229–235
10. Aurlien E, Holte H, Pharo A, et al.  Combination chemotherapy with mitoguazon, ifosfamide, MTX, etoposide (MIME) and G-CSF can efficiently mobilize PBPC in patients with Hodgkin's and non-Hodgkin's lymphoma. Bone Marrow Transplant. 1998;21:873–878
    • MEDLINE
11. Birtwistle GM, Dahl OJ, Myhrhaug B, Nygaard K. Simula begin. Bromley: Chartwell-Bratt Ltd; 1983;
12. Fenstad GU, Skovlund E. The Behrens-Fisher problem: a comparison of test properties when observations are not necessarily normal. Statistical Research Report 10. University of Oslo: Department of Mathematics, 1992.