Studies where results are limited to only group differences and/or correlations between variables cannot go beyond probabilisitc prediction. The nature of a relationship or a difference between groups cannot be understood unless all exceptions are understood. So, for example, all reports of sex differences, race differences, or differences between cultures should not be attributed to sex, race, or culture, respectively, unless the exceptions are addressed. (p. 13)
Toomela, A. (2007). Culture of science: Strange history of the methodological thinking in psychology. Integrative Psychological and Behavioral Science, 41, 6-20.