Suchergebnisse

The failure of the preelection polls in Nicaragua to predict the 1990 election results & the wide divergence of poll results are examined. Though several simple explanations for poll results exist, including fraud & systematically flawed polling methods, data collection problems, primarily measurement of voter intention, seem more responsible for their inaccuracies. Measurement was difficult because of the unprecedented nature of the election, the novelty of the polling process, the extreme political polarization in the country, intimidation of voters, & the tenuousness of the secret ballot concept. In such a climate, voter intention likely varied according to the perceived political affiliation of the polling firm. 1 Table, 8 References. Modified AA

An examination of the efficacy of record check studies involving the comparison of survey responses with external record evidence. Analysis of match rates between residents (N = 570) surveyed in Oct & Nov 1981, & victims of crime taken from the Peoria, Ill, Police Dept records from Jan to Sept 1981 (N = 1,300) reveal marked differences in the level of survey-record correspondence. It is suggested that since the level of match rate appears highly variable -- depending on the definition of match -- previous record check studies be reexamined; such research methods should be employed in constructing alternative indicators of phenomena to be measured, rather than as the arbiter of survey response quality. 2 Tables, 14 References. Modified HA

The absence of visual aids in telephone interviews has generated a search for methods of asking questions normally accompanied by such aids in personal interviews. Telephone interview data from 4,300 households are used to compare 2 approaches to asking 7-point-scale attitude questions in a national health survey: single-step numerical selection tasks, or 2-stage tasks proceeding from general statements of attitude to more detailed specifications. The 2-question forms produced comparable univariate distributions of results, but items administered with the single-step form showed slightly higher intercorrelations. Overall, the results are more comparable than would have been expected. 4 Tables, 1 Figure, 12 References. Modified HA.

Dissatisfaction with economic circumstances is commonly thought to increase with temporal or cultural distance from Mexico, if, over time, immigrants or their offspring fail to experience substantial mobility within the US. However, interview data from Wc M heads of household in Ariz, Calif, Colo, & Tex (N = 48, 77, 60, & 119, respectively) do not support this argument. Immigrants & the less acculturated & assimilated do not perceive themselves as less deprived than other Mexican Americans. Actual income was found to be the primary factor influencing levels of reported economic well-being. 4 Tables, 13 References. Modified HA.

We use this article to introduce the Militarized Interstate Confrontation (MIC) dataset, 1816-2014—a new dataset for international conflict with a host of innovative features. The MIC data corrects thousands of errors in existing interstate conflict data and provides fatality ranges for all conflicts, with meaningful fatality estimates and no missing fatality values. Thus, the MIC data fixes missing data problems that have precluded researchers from analyzing escalation and related issues because of the lack of integrated conflict and war data. We also identify and distribute separate datasets for state-versus-citizen actions that are protest-dependent. These are attacks on shipping, fishing boats, and rebels, which were previously included in the data because the sovereign of those private citizens protested. We discuss our systematic search for new conflict cases and the 108 new conflicts we found, and we provide analyses and summaries that demonstrate the usefulness of our MIC data. Finally, we use our new data to create the first ever dataset of truly dyadic, directed dyad-year data, with highest actions and fatalities that vary appropriately within conflicts by both year and dyad. We believe these datasets will be useful for a host of studies but especially those interested in how conflicts evolve over time.