The Psychology of Micro-Targeted Election Campaigns by Jens Koed Madsen

Author:Jens Koed Madsen
Language: eng
Format: epub, pdf
ISBN: 9783030221454
Publisher: Springer International Publishing

7.2 Modelling an Individual Voter

We are now at a point where components can begin to merge into a fuller picture of the potential data-driven political campaigns in which the campaigner attempts to build models of individual voters with the aim of persuasion or voting. This book presents an approach to doing so, namely the Bayesian models that provide a computationally specific framework for modelling voters’ subjective beliefs, their perception of the world. They also provide a predictive tool that informs how people will react to new information from more or less credible sources, how they see dependency of sources, and how best to target specific beliefs or values to best persuade or influence that particular voter (or a subset of voters).

Micro-targeted campaigns aim to get maximally relevant and accurate models of the individual voters or relevant subsets of voters to understand how they see the world, their personality, and their behavioural motivations. If the models are accurate representations, personalised models help develop and thereby optimise persuasion and influence attempts, giving a competitive edge compared to political opponents (who, presumably, will try and do the same).

Additionally, as discussed in Chap. 8, models can provide campaigns with a roadmap to the most sensitive data parameters for persuasion and influence efforts. The simplest models provide correlations between voter characteristics and traits that relate to persuasion of influence. For example, campaigns may test whether fear-based campaigns become increasingly effective and more persuasive to older rather than younger people. If the campaign finds a positive correlation between responsiveness to fear and age, they may seek out data that can identify people with that trait (for age, this is straightforward; for psychometrics and subjective beliefs, it is harder to identify the right data to parameterise the models). If the campaign has access to the relevant data, they can use it to segment the electorate. In the age and fear example, they may segment people according to age (e.g. voters 18–35, 35–60, and 60+). This segmentation can guide the design of the persuasion and influence efforts. This may result in policy messages that invoke fear for one segment (60+), but stay away from such appeals for the other groups of voters.

Going beyond correlations, heuristics, and psychometrics, more complicated models allow for deep structures of personal beliefs and belief revision. The modelling approaches used (Bayesian belief revision models, networks of source dependencies, heuristics and biases, psychometrics or a mix of all) depend on the questions the campaign is trying to answer, the resources or the campaign (both in terms of money and time), and the technical know-how of the staff on the campaign.

At the level of individually measurable outcome-oriented traits, evidence suggests psychometrics is useful to describe and potentially predict how people will respond to campaign messages. The studies on OCEAN and moral foundation facets indicated that these traits could be used for persuasion and influence attempts. Assume for the moment that people who score high on need for closure, risk aversion, and neuroticism respond strongly to fear-based framing of political ideas.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.