ABSTRACT
Research is also judged to be good if it follows certain methodological procedures. These amount, I think, to the adoption of two interlinked assumptions - about cause-effect relationships and about generalizability. Many of the rules about research have to do with
controlling extraneous sources of variance - that is, making sure that the results obtained are not an artifact of 'other variables'. These 'other variables' characteristically turn out to be such things as age, gender, race, social class. These are, of course, the very same categories which, in academic psychological texts, are credited with causal functions in human life. Yet perhaps the functions they possess represent nothing more than the power our particular society gives them
The second assumption governing judgments about the soundness of research methodology is that any conclusions drawn should be generalized beyond the particular sample studied. What this generally means is that the researchers should be able to make statements about large groups of people - statements as to the factors which make people behave in particular ways. The groups of people to be included in these generalizations are typically limited by the same categories - age, gender and so on. Within these limits, broad statements of cause and effect are expected. The underlying logic is that people basically differ only in the social categories to which they belong. Again, if this is a meaningful statement at all, it can only represent a statement about prevailing social attitudes.
