Investigating single-process and dual-process theories of transitive reasoning: Applying Signal Detection Theory and Signed Difference Analysis.

Abstract

According to influential dual-process theories, reasoning is driven by distinct Type 1 and 2 processing. Type 1 processing is characterised as fast, intuitive and heuristics based, while Type 2 processing is thought to be more effortful, deliberate and requires working memory. However, the dual-process view has faced an increasing amount of criticism over recent years. Single-process theories offer an alternative account, suggesting that reasoning across a range of contexts is reliant on a common assessment of inference strength. The current experiment tested the competing theories using a transitive reasoning task. Key factors relevant to dual-process accounts were manipulated, including premise integration time and working memory demands via premise ordering. Results showed that validity ratings were higher for valid than for invalid arguments, and for believable than for unbelievable conclusions. An interaction between premise ordering (unscrambled vs. scrambled premises) and validity was also observed. These results were consistent with dual-process theories, however, quantitative models were then compared to investigate whether the results were inconsistent with single-process theories. Signal detection theory was applied and dual- and single-process accounts were instantiated as two-dimensional and one-dimensional models, respectively. Model testing via signed difference analysis showed that the observed data do not rule out the simpler single-process, one-dimensional model. This suggests that such single-process models offer a viable account in explaining the underlying cognitive processing intransitive reasoning.