Data for the manuscript: Business as Usual? A Social Capital Approach to Understanding Interactions with Journalists on Twitter

Published: 01-03-2019| Version 1 | DOI: 10.17632/dk6df432np.1
Contributor:
Matthew Barnidge

Description

Datasets used in for the article entitled, ""Business as Usual? A Social Capital Approach to Understanding Interactions with Journalists on Twitter." Both files (IWJMI11 and IWJMI21) are .rdat files containing five imputed datasets (see manuscript for imputation details). IWJMI11 is the imputation for the cross-sectional data, and IWJMI21 is the imputation for the longitudinal data. This study relies on a two-wave, online panel survey of adult, internet users who are residents of the United States. The first wave was collected between September 19-29, 2018, six weeks before the 2018 U.S. Midterm Elections, and the second wave was collected during the month after the Elections, from November 7-December 5, 2018. The survey was administered by a private survey firm, Survey Sampling International (SSI), which uses a three-stage sampling process. First, subjects are randomly selected from an online panel constructed by SSI using geographic and demographic quotas based on age, gender, race, and census region, in such a way that they are comparable to the U.S. Census statistics for the population of interest. Next, subjects were randomly presented with screening questions asking whether respondents are over the age of 18, whether they are U.S. residents, and whether they had Internet access, in order to determine their eligibility for the study. Finally, subjects were randomly invited to take the study based on their likelihood to complete it based on their past completion of surveys. This final step is taken to maximize the likelihood of obtaining complete responses. The first survey wave has a sample size of N = 1,493 and a cooperation rate (an appropriate metric when traditional response rates cannot be reported because parameters of sample invitations are unknown) of approximately 70% (AAPOR, 2016; CR3). The second survey wave has a sample size of N = 576 and a 39% retention rate. The first-wave sample is broadly reflective of the population of interest (see Appendix A), with an average age of 48.39 (SD = 16.18), 51% women, 77.2% white, and 75% reporting affiliation with a religion. The average respondent has an associate’s or bachelor’s degree (M = 4.38, SD = 1.71, where 1 = Some high school and 7 = Post-graduate degree) and lives in a household that makes between $45,000 and $75,000 per year (M = 4.84, SD = 2.14, where 1 = Less than $15,000 and 8 = More than $150,000). A multiple imputation technique (predictive mean matching) was used to impute missing values. Separate imputations were performed for the first and second waves. Predictive mean matching works as follows: First, cases with complete data were used to predict values of variables with missing data, producing a set of coefficients. Next, a random draw was taken from the predictive posterior distribution to produce a new set of coefficients, which were then used to compute predicted values for all cases with at least one missing value. Finally, an observed value close to the predicted

Files