Data Mining in Social Sciences

A Decision Tree Application Using Social and Political Concepts

Bibliographic Data

ID13036618
AuthorsEfthalia Massou (0000-0003-0488-482X, University of Cambridge, corresponding author), Gerasimos Prodromitis (0000-0001-6945-2628, Panteion University), Στάμος Παπαστάμου (0000-0003-1853-2470, Panteion University)
Year2022
Volume13
Issue3
Pages297-314
Publication date2022-11-01
Peer ReviewedYes
Open AccessNo
TypeARTICLE
VenueStatistics Politics and Policy (JOURNAL)
Journal identifiersISSN: 2151-7509 • E-ISSN: 2194-6299
PublisherDe Gruyter (PUBLISHER • DE)
DOI10.1515/spp-2022-0004
OpenAlexW4311001920
LanguageEN
Citations received1
References cited27

In this paper, we investigated the utility of data mining to classify individuals into predefined categories of a target variable, based on their social and political attitude. Data collected for a social psychology study conducted in Greece in 1994 were used for this purpose. We established the theoretical background of our analysis through explanatory factor analysis. We ran the decision tree algorithm CHAID in order to build a predictive model that classifies the study participants in terms of their attitude toward physical and symbolic violence. The CHAID algorithm provided a decision tree that was easily interpreted, and which revealed meaningful predictive patterns. CHAID algorithm showed satisfactory predictive ability and promising alternatives to social psychology data analysis. To the best of our knowledge, there is no other evidence in the literature that the decision tree algorithms can be used to identify latent variables

CHAID · Data mining · Decision tree · Decision tree learning · Decision tree model · Incremental decision tree · Machine learning · Tree (set theory · Computer Science · Crime Patterns and Interventions · Cultural Differences and Values · Mathematics · Psychology · Social and Intergroup Psychology · Social Psychology · Artificial Intelligence

Tools

Open DOI