Conceptual fluency in inductive reasoning

Abstract

Psychological effects connected with fluent processing are called fluency effects. In a sample of 403 participants we test whether conceptual fluency effects can be found in the context of inductive reasoning, a context that has not been investigated before. As a conceptual manipulation we vary the use of symbols (persons and crosses) in reasoning tasks. These symbols were chosen to provide hints for the solution of the implemented tasks and thus manipulate fluency. We found evidence that these hints influence ease of processing. The proportion of solved tasks increased by 11% on average in the condition with conceptual hints, F(1,399) = 13.47, partial η2 = .033, p < .001. However, we did not find an effect of the conceptual manipulation on the temporal perception of the task. In a second study (n = 62) we strengthened our findings by investigating solution strategies for the tasks in more detail, 79% of the participants described the tasks in a way they were intended. Our results illustrate the advantages of the separation of ease of processing, fluency experience, and judgments.

Fig 1. Model of assumed generalizability of fluency effects [4].

Fig 2. Important parameters that can be explicitly or implicitly varied in a fluency study.

Fig 3. Inductive task example.

“On which fields do four points appear in the next step?” Correct response: “DHLP”.

Fig 4. Filling task.

“In which field does the next cross appear?” Correct response: “J”.

Fig 5. Movement task.

“In which fields are two persons in the following picture?” Correct response: “HO”.

Fig 6. Filling task with persons (less conceptually congruent condition).

Fig 7. Movement task with crosses (less conceptually congruent condition).

Fig 8. Parallel version of movement task with crosses (less conceptually congruent condition).

Fig 9. Proportions of solved tasks for the conceptually congruent and the less conceptually congruent condition.

Task 1 = Filling Task, original version; task 2 = Filling Task, parallel version; task 3 = Movement Task, original version; task 4 = Movement Task parallel version. Total = mean proportion. Error bars reflect the 95% confidence intervals for the mean difference of the dichotomous tasks and their score.

Fig 10. Mean task durations in seconds for each condition.

Task 1 = Filling Task, original version; task 2 = Filling Task, parallel version; task 3 = Movement Task, original version; task 4 = Movement Task parallel version. Total = mean task duration. Error bars reflect the 95% confidence intervals for each mean difference.

Fig 11. Mean temporal perception ratings for each condition.

Y-axis simultaneously represents visual analogue scale and seven-point Likert scale (range: 0–6). Error bars reflect the 95% confidence intervals for each mean difference.