What is it about?
The authors' model explored how the U.S. presidential candidates’ popularity rates as observational units can be explained differently by citizens’ economic voting and money sources depending on the parties as an experimental unit. The partial effects of these variables were investigated by controlling for individual contributions, super PACs' donations, consumer sentiment, and the levels of ideology from parties.
Featured Image
Photo by Arnaud Jaegers on Unsplash
Why is it important?
Because of the bipartisan system, it was inevitable to examine the factors in terms of a multi-level approach based on a nested structure. This is a methodological study to show how hierarchical linear modeling (HLM) is used with the nested data in the U.S. presidential candidates with party membership.
Perspectives
Based on the analysis of the fixed-effect model, individual contributions as a level-1 predictor had the most significant effect on candidates’ popularity rates. Subsequently, the random slope of individual contributions had a different magnitude of effect on the popularity rates depending on the candidates’ party affiliation. Nevertheless, the cross-level interaction between level-1 and level-2 variables did not exist. For a methodological suggestion, the authors took note that stratifying variables at multi-levels is more accurate in analysis than the use of ordinary least squares (OLS) because an HLM analysis can prevent the overestimation of the statistical significances.
Methodologist in Predictive Analytics in Public Affairs Jiwon Speers
Florida State University
Read the Original
This page is a summary of: Revisiting Institutional Choice in Presidential Campaign Finance- A Hierarchical Linear Model -, Korean Journal of Local Government & Administration Studies, June 2016, The Korea Association for Local Government and Administration Studies,
DOI: 10.18398/kjlgas.2016.30.2.139.
You can read the full text:
Contributors
The following have contributed to this page
