Suchergebnisse

A report of a survey to determine the extent to which the survey method may be valid in other cultures. A survey questionnaire was administered to 984 M R's in a geographically stratified sample. Specific reference was made to the reliability & validity of the data & to the influence of interviewer affiliation on responses. The reliability check asked R's to say whether or not they had ever participated (a) in any kind of voluntary work & (b) in a Shramdan (a voluntary contribution to public labor). 95% of R's who reported participation in Shramdan also reported participation in voluntary work while 95 of those who never participated in Shramdan deny participation in any kind of voluntary work. Internal consistency,was shown by the fact that 50% of R's who participated in Shramdan recognized the term whereas only 20% of nonparticipants recognized it, early in the interview. 5 checks of interviewer bias who were both gov related &/or Cornell U affiliates were made: (1) little effect of the affiliation of interviewers on their rating of R's capacities & comprehensions, such as intelligence, outspokenness, precision of responses to questions, friendliness, & sincerity, were observed. (2) Since it is difficult for a stranger to remain inconspicuous in Indian villages, PRAI have a greater tendency than Cornell interviewers to report the more desirable kind of interview situation without bystanders & uninvited participants. (3) The distribution of interviewer ratings of R's age, educ, occup & caste by PRAI interviewers differed signif'ly from the equivalent distribution by Cornell interviewers. PRAI interviewers tended to rate R's as better educated, younger, more likely to be small landowners in the intermediate caste group. The sampling technique leaves little reason to assume that distribution of these traits differed in R's reached by both crews. (4) 3 factors are given as possibly important in the finding that PRAI interviewers were most likely to leave blank the question 'What are your greatest worries & difficulties?' or to report R's as saying 'I have no cares or worries:' (a) PRAI interviewers may have been less skilled in their ability to elicit replies; however, the incidence of `no answers' did not differ for the 2 groups of interviewers on all other questions. (b) Some tendency was noted in the responses to conceal worries about the gov - repayment of loans - from the gov affiliated PRAI interviewers but not from the Cornell interviewers. (c) PRAI interviewers expressed misgivings about the question during the training seminar & expectations of failure may have influenced the kind of replies obtained. (5) Villagers appeared to be reluctant to tell gov representatives that the gov rather than the villages should build roads. A test showed that the r between educ & occup held for both sets of interviews so that analytic conclusions reached by the survey's r analysis & statements of co-variation are affected only slightly by affiliation bias. J. D. Twight.