Carol H. Weiss

VALIDITY OF WELFARE MOTHERS' INTERVIEW RESPONSES

Because survey research is open to the criticism that respondents do not always tell the truth to interviewers, validation studies have long been an honored pursuit. They have led to progressive clarification of the extent and direction of response error, and to its increasing control. In recent years, as more research has focused on poor people in our society, special concerns about response error have come to the fore. Some critics claim that the poor present interviewing problems different in kind and degree from other respondent groups. Poor people, they say, misunderstand the interviewer, or defer to him, or try to con him; the interviewer stereotypes the respondent and biases his answers; or else poor people have such difficulty in knowing or remembering or verbalizing that the interview becomes a very leaky vessel indeed for taking out questions and bringing back answers.2 In an effort to test these assertions, this study validated responses of a sample of Negro welfare mothers. They were New York City residents receiving public assistance in 1966 who had been interviewed by the National Opinion Research Center for a study on the use of health services. Their responses to five questions in the original interview were compared with official records-registration status in 1964, voting in the 1964 presidential election, receipt of money from welfare,

What America's Leaders Read

I N A STUDY OF THE TOP LEADERS of American institutions, respondents were asked which newspapers, columnists, magazines, and professional journals they read. Exposure to mass media was one of the factors being examined for its possible effect on the opinion-forming and decision-making activities of the most influential Americans. The larger study also examined social background, career history, group and interpersonal context, present social position and role, and policy opinions, as these relate to the behaviors of the leadership groups. This article presents the data on what leaders read. The sample for the study was drawn in two stages. First, institutional sectors that have a broad impact on society were identified: industrial corporations, non-industrial corporations (banking, insurance, retailing, etc.), labor unions, the Congress, federal departments and independent agencies, political parties, voluntary associations, and the mass media. In addition, because of interest in the role of the very wealthy, owners of large fortunes were included. Those holding leading positions in each set of institutions were sampled: chief executive officers of major corporations; presidents of the largest unions; Senators, and chairmen and ranking members of House committees; federal political appointees at the level of secretary and assistant secretary; federal civil servants in administrative posts at grades GS 17 and i8; national committeemen and state chairmen of Republican and Democratic parties; presidents and executive directors of such voluntary associations as farm, business, professional, ethnic, religious, womens, and public affairs organizations; and publishers, editors, columnists, TV public affairs executives, and commentators., Approximately fifty persons in each sector were interviewed in 1971-72. During the interview, each respondent was asked to name other in-

Research Organizations Interview the Poor

Researchers who conducted 194 personal interview surveys of low-income populations returned mail questionnaires about their experience with middle- and lower-class interviewers. Interviewing poor people is not as difficult as expected; receptivity and completion rates are generally good. The second part of the questionnaire focussed on indigenous interviewers, i.e., those matched to respondents by class and race/ethnicity. Compared to black or white middle-class interviewers, indigenous interviewers are rated least adequate on performance of most interviewing tasks, such as asking questions and recording answers. Their main strengths are contacting, locating, and establishing rapport with respondents. Additional education and interviewing experience improve performance ratings but additional training does not. Despite performance problems, researchers who chose to employ indigenous interviewers usually plan to do so in future surveys.

Do women know which IUD they are wearing

Family Planning clinic patients wearing IUDs were asked the type they were wearing, and the actual type was identified in the pelvic examination. About 19 percent of 268 women wearing four types of IUDs did not correctly identify the type that they had. Incorrect or unknown perception that might be a medical risk was held by more than half of the wearers of the Dalkon Shield. Each user of an IUD should be given a card stating IUD type, date of insertion, date of reinsertion if applicable, and instructions in case of complications. The problem of incorrect identification may be exacerbated as the number of devices increases and with the trend toward use of medicated IUDs which must be periodically removed and replaced. Comparable studies in different populations of IUD users are needed.

Importance of correct identification of intrauterine contraceptive device.

n A study to ascertain the proportion of patients who could identify their IUDs was carried out in a large hospital. Interviews were conducted with 268 women using 4 different types of IUDs. However, only 44% of women wearing the Dalkon Shield gave the correct response; this finding is disturbing because of the evidence of association of the Dalkon Shield with fatal second-trimester abortions. It is advisable to provide every IUD user with a card showing IUD type, date of insertion, and when reinsertion is necessary, which is especially important with the medicated IUDs, which must be periodically removed and replaced.n