A two-wave longitudinal panel dataset on United States public acceptance of Precision Livestock Farming technology in swine production

This data article describes a two-wave longitudinal panel dataset that supports research on the public acceptance of Precision Livestock Farming (PLF) technology in swine production among adult residents of three U.S. pork-producing states. The dataset comprises two structured comma-separated-value files — panel_long_analysis.csv (n = 2,565 person-wave observations; 37 variables) and panel_wide_analysis.csv (n = 345 balanced-panel respondents; 64 variables) — alongside four fully documented R analysis scripts. Data were collected via a two-wave online and mail survey administered by the Michigan State University Office for Survey Research (MSU-OSR) to adult residents of Iowa, Michigan, and North Carolina: Wave 1 in Fall 2022 (n = 1,287) and Wave 2 in Fall 2023 through Spring 2024 (n = 1,278). These three states together account for the majority of U.S. pork production but their residents constitute approximately 7% of the U.S. population; the deposit is therefore best characterized as a regional U.S. dataset on publics most proximate to active swine production, not as a nationally representative sample. Sampling used a 50/50 urban–rural address stratification within each state. The survey instrument measured PLF acceptance via an eight-item belief index (plf_index) and general attitudes toward agriculture and technology via a ten-item index (att_index), together with binary livestock familiarity indicators, social proximity to farmers, and standard sociodemographic covariates. A balanced sub-panel of 345 respondents who completed both waves enables within-person analysis of attitude dynamics. To recover the theoretical n = 690 person-wave panel, the deposited analysis pipeline applies multiple imputation by chained equations (MICE; m = 50, predictive mean matching for continuous and Likert items, logistic regression for binary covariates) and pools regression estimates via Rubin's rules; a listwise-deletion sample of n = 414 person-wave observations is preserved as a sensitivity benchmark. The deposited data support pooled ordinary least squares, individual fixed-effects, state × wave fixed-effects, random-effects, and first-difference panel model estimation, as well as principal components analysis, reliability assessment, distribution diagnostics, cross-sectional sub-group comparisons, and geographic analyses. All files are structured in UTF-8 CSV format with a detailed variable codebook documented herein. To the best of our knowledge, this is the first publicly available longitudinal panel dataset on U.S. general public acceptance of PLF technology in swine production. All files are deposited on Zenodo under a CC BY 4.0 license.