Sorting Methods for Online Deliberation: Towards a Principled Approach

TL;DR

This paper evaluates different sorting methods for online deliberation platforms, proposing a conceptual framework, criticizing common practices like approval-based sorting, and advocating for more principled, integrated approaches.

Contribution

It introduces a framework for classifying sorting methods, critiques common ad hoc practices, and advocates for more justified and integrated sorting strategies in online deliberation.

Findings

Approval-based sorting should consider multiple parameters.

Common approval sorting methods are often ad hoc and unjustified.

More appropriate sorting methods exist beyond approval counts.

Abstract

Recent years have seen an increase in the use of online deliberation platforms (DPs). One of the main objectives of DPs is to enhance democratic participation, by allowing citizens to post, comment, and vote on policy proposals. But in what order should these proposals be listed? This paper makes a start with the principled evaluation of sorting methods on DPs. First, we introduce a conceptual framework that allows us to classify and compare sorting methods in terms of their purpose and the parameters they take into account. Second, we observe that the choice for a sorting method is often ad hoc and rarely justified. Third and last, we criticise sorting by number of approvals ('likes'), a method that is very common in practice. On the one hand, we show that if approvals are used for sorting, this should be done in an integrated way, also taking into account other parameters. On the other hand, we argue that even if proposals are on a par in terms of those other parameters, there are other, more appropriate ways to sort proposals in light of the approvals they have received.