Some cautionary notes on sociological research: with special attention to survey studies

Samenvatting

Several critical remarks are made about sociological research, especially survey studies. People's willingness to report or admit certain feelings about themselves is affected by one biasing factor: their assessments as to the social desirability of such feelings or information. This means that the validity of certain sociological measures is called into question, & so is the validity of the findings involving the measures. One of the major reasons that sociological researchers have been unsuccessful in explaining the amount of variance is their failing to recognize the fact that the collection of social scientific data constitutes a social process. The consequences for viewing controlled data collection as a social process are that perfect control (no variables other than those of explicit theoretical & empirical interest) is not possible. The replication of procedures (not to vary anything which makes a systematic difference in peoples responses) is more difficult than is usually acknowledged. Finally then studies which are based on the utilization of procedures which are neither controlled nor replicable will now allow for empirical results that can in some way be added up or built upon. Derek L. Phillips states 2 reasons for doing "research on research," the first one is to discover the sources of error & invalidity so that they can be controlled or eliminated, & secondly to demonstrate the inadequacy of contemporary social science research & the impossibility of bias-free research in situations where human beings collect data from active thinking people like themselves. The sociologist's own views & biases will always intrude to some extent in the selection of his problem & the procedures for collecting data pertaining to that problem. The answer to this problem lies with new theories & new conceptual schemes, & one should look at the world through one's own eyes & not through scientific instruments that frequently cloud vision. P. Tucker.