Up for Debate: Effects of Formal Structure on Argumentation Quality in a Crowdsourcing Platform

摘要

We examined the use of formal structure (more specifically, the Toul-min model and the use of abstraction laddering) in argument assertion templates in a crowdsourcing platform, to determine its effects on argument quality, as rated by other peer contributors. Contrary to our hypotheses, the attempt to add rigor to asserted arguments resulted in a significant decrease in quality across several measures, including the pathos, kairos, and overall level of agreement with the assertion. We found that the way participants voted (a binary outcome of supporting or dissenting) aligned more strongly with whether they agreed with the assertion (regardless of quality) rather than with the quality of the assertion. We provide multiple potential explanations for why the use of the Toulmin model was not a reliable predictor of argument quality in a crowdsourcing application.