The similarities and differences of free and forced choice tasks

Abstract

Free choice tasks are tasks in which more than one response is considered correct, while in forced choice tasks only one response is considered correct. They are often used in conjunction to investigate differences between self-generated (free choice) and externally triggered (forced choice) actions. The general purpose of the present work was to investigate what free choice tasks are, both in themselves and in contrast with forced choice tasks. This was investigated over the course of three studies. Study 1 was a follow-up study to Naefgen, Caissie, and Janczyk (2017), which in turn was an investigation of the mechanisms behind the backward crosstalk effect (BCE). The BCE is an interference effect that appears in dual-tasking situations and refers to the phenomenon that response times in the first task are influenced by whether the two tasks are compatible or incompatible on (theoretically) any dimension. Naefgen, Caissie et al. investigated the role of stimulus-response links in the BCE, finding reduced BCEs when one of the tasks was a free choice task. The alternative explanation for these results that Study 1 investigated was that this reduction was due to conflict adaptation in response to the presence of free choice tasks. As the BCE in Study 1 was reduced neither in trials following free choice Task 1 trials nor with higher proportions of free choice Task 1 trials in a block, the alternative explanation was rejected. In Study 2, the common observation that free choice tasks have slower responses than forced choice tasks was investigated. Within a sequential sampling framework, in which evidence is noisily accumulated towards decision thresholds, the crossing of which causes a response to be emitted, the mean response time difference was sought to be attributed to either differences in the speed of evidence accumulation or differences in the time of non-accumulation time. This was done by manipulating the decision thresholds with proportions of catch trials in Experiment 1 and time pressure in Experiments 2 and 3. If the difference is due to different evidence accumulation speeds, the response time difference should change. As it did not, the difference was attributed to a difference in non-accumulation time, possibly suggesting that free choice tasks involve an additional process. In Study 3, the question whether free choice tasks are random generation tasks was investigated. This was done by manipulating the working memory load and observing whether the randomness of the choices changes in a manner consistent with random generation tasks. As both a manipulation supporting the working memory and one adding working memory load had effects consistent with random generation tasks, it was concluded that free choice tasks are at least similar to random generation tasks. The implications of the results of all three studies for free choice tasks and their uses are discussed in the General Discussion.