Research into human problems: Lower social strata and the aged underrepresented The quality of social science research is probably much lower than that of the sciences. Firstly, this is due to the quality of the observations. For example : the measurement of a temperature is encompassed in many more precautions than the measurement of attitudes, loneliness or even the number of friends. Secondly, the underrepresentation in surveys is alarming: the rate of non response amounts to 50% and more. This gives rise to distortions in the frequency distributions and in the relationships between variables as well and, also very important, to an underestimate of the problems of the lower social classes and the aged.
