PROBLEMS OF SURVEY RESEARCH IN MODERNIZING AREAS

In: Public opinion quarterly: journal of the American Association for Public Opinion Research, Band 22, Heft 3, S. 230-234

Abstract

As research efforts have ventured away from major Ur centers, new & unexpected problems arise with a rigid use of imported techniques in underdeveloped areas. The problem areas discussed are sampling, interviewer selection & training, establishing rapport, reliability, & validity. One of the major problems in sampling is the lack of 2ndary sources of data on pop's. Other problems include: (1) Equal probability in choosing city blocks for area sampling yields larger sampling errors than controlling the selection on pop, (2) stratification of samples is impossible, (3) making pop estimates or projections is complicated, (4) there are no ready checks or controls on sample performance without knowledge of pop parameters; & breakdowns on age, sex, & marital status cannot be compared with other figures. (5) Makeshift shelters & homeless individuals make orderly prelisting of households, or routes for interviewing, or definitions of `household' or `family' difficult. (6) Random selection of R's within households may clash with local custom or familial protocol. Interviewer selection & training is difficult where the few people with the necessary skills are usually fully occupied. Housewives are often barred from employment by custom. Class & status identification and the use of more than one language or dialect pose additional complications. It is sometimes more economical to train a team & to travel around a country than to recruit & train in each locale. Personal loyalty of interviewers to the field director may lead to falsification of interviews in order not to disappoint him. Feelings of national pride may lead interviewers to make up responses to mask mass ignorance or apathy. Problems of establishing rapport are not always solved by the employment of trained nationals since there often is a gulf between the well-educated Ur person & the destitute Ur or Ru worker as between the latter & a foreigner. Courtesy calls on local leaders & token interviewers with persons not designated by random sampling are sometimes necessary. Refusal rates are usually quite low because of curiosity, fear, hospitality, & flattery at being singled out for attention. Problems of reliability & validity are the most serious & probably the most difficult to cope with esp where people are unaccustomed to voicing opinions on controversial subjects. Inexperienced interviewers are not always able to `teach' R's how to be interviewed. Either both the R & interviewer end up agreeing with each other or the interviewer accepts noncommittal or incomplete answers. J. D. Twight.