A drop in performance on a fluid intelligence test due to instructed-rule mindset

Abstract

A 'mindset' is a configuration of processing resources that are made available for the task at hand as well as their suitable tuning for carrying it out. Of special interest, remote-relation abstract mindsets are introduced by activities sharing only general control processes with the task. To test the effect of a remote-relation mindset on performance on a Fluid Intelligence test (Raven's Advanced Progressive Matrices, RAPM), we induced a mindset associated with little usage of executive processing by requiring participants to execute a well-defined classification rule 12 times, a manipulation known from previous work to drastically impair rule-generation performance and associated cognitive processes. In Experiment 1, this manipulation led to a drop in RAPM performance equivalent to 10.1 IQ points. No drop was observed in a General Knowledge task. In Experiment 2, a similar drop in RAPM performance was observed (equivalent to 7.9 and 9.2 IQ points) regardless if participants were pre-informed about the upcoming RAPM test. These results indicate strong (most likely, transient) adverse effects of a remote-relation mindset on test performance. They imply that although the trait of Fluid Intelligence has probably not changed, mindsets can severely distort estimates of this trait.

Performance, Achievement, and Success in Psychological Research: Towards a More Transparent Use of the Still Ambiguous Terminology

“…However, cognitive performance was operationally defined in a unified way through scores on standardized tests (e.g. ErEl & Meiran, 2017; Golino et al., 2017) or through task-specific indicators, which were either (a) correct responses or correctly completed task-specific goals (e.g. Madzharov et al., 2018; Smith et al., 2017; Steyvers & Benjamin, 2018), (b) errors made while solving the task (e.g.…”