Suchergebnisse

Behavioral variance among telephone interviewers (interviewer effect) was investigated using data from 11 national polls conducted by CBS News & the New York Times in 1980. The polls involved a total of 300 interviewers, of whom 245 submitted completed questionnaires; selected effects examined were generally small & tended to be inconsistent between one poll & another. Explanations for the inconsistencies were found in the political context in which the polls were held & the variable associations with the nonrandom regional distribution of Rs. Demographic variables -- sex, region, size of place, & number of household adults -- were used to test for random distribution of Ss. Implications for further research on interaction between interviewer & S characteristics are discussed. 1 Table. Modified HA.

Five political generations in the US national electorate are identified, using data drawn from the 1980 New York Times/CBS News poll of 11,312 white registered voters. Each generation of voters is hypothesized to form its political identification based on experiences gained in early adulthood. The utility of the theory of operant conditioning in the analysis of political behavior is stressed. Future research should examine the issue positions of members in each generation & the changing behavior of generations during the campaign session. 4 Tables, 41 References. Modified HA

A new methodological technique for estimating turnout in preelection studies of voters is described, & a new approach to the allocation of undecided voters presented. The estimation procedure utilizes a smaller number of predictors than past research, providing a more efficient method for telephone as well as personal interviews. A comparison of estimates based on a likelihood index & probability weighting results in equivalent predictive power, & appropriate applications for each technique are discussed. 2 Tables, 1 Appendix, 22 References. Modified AA.

Considers some methodological issues surrounding attempts to compare racial & ethnic data collected in the 2000 decennial US census, the first to allow respondents the option of selecting more than one category to designate their racial identification, with data from previous censuses & other individual-level social surveys. "Bridging estimates" calculated to analyze historical trends in data series are reviewed, cautioning that they are not always useful, nor are they required. Criteria for choosing a bridging method are outlined, & the utility of different methods is demonstrated in a comparative analysis of data from three sources: the 1993-1995 National Health Interview Surveys, the May 1995 Supplement on Race & Ethnicity to the Current Population Survey, & the 1998 Washington State Population Survey. The effects of method on outcome measures are discussed, & specific implications for the interpretation of racial data are considered. 9 Tables, 5 References. K. Hyatt Stewart

"This paper examines reporting in specific consumer item categories (or commodities) and estimates expenditure underreporting due to survey respondents who erroneously report no expenditure in a category. The author's approach for estimating underreporting errors is a two-step process. In the first step, a Markov latent class analysis is performed to estimate the proportion of consumers in various subpopulations who fail to report their actual expenditure in a particular commodity. Once this proportion is estimated, the dollar value of the missing expenditure is estimated using the mean expenditure of those in that subpopulation that did report an expenditure. Finally, the estimates are evaluated and discussed in light of external data on expenditure underreporting." (author's abstract)