RESUMEN
Testing that an experiment works as intended is critical for identifying design problems and catching technical errors that could invalidate the results. Testing is also time-consuming because of the need to manually run the experiment. This makes testing the experiment costly for researchers, and therefore testing is less comprehensive than in other kinds of software development where tools to automate and speed up the testing process are widely used. In this paper, we describe an approach that substantially reduces the time required to test behavioral experiments: automated simulation of participant behavior. We describe how software that is used to build experiments can use information contained in the experiment's code to automatically generate plausible participant behavior. We demonstrate this through an implementation using jsPsych. We then describe four potential scenarios where automated simulation of participant behavior can improve the way researchers build experiments. Each scenario includes a demo and accompanying code. The full set of examples can be found at https://jspsych.github.io/simulation-examples/ .
Asunto(s)
Programas Informáticos , Humanos , Simulación por ComputadorRESUMEN
One of the recent major advances in cognitive psychology research has been the option of web-based in addition to lab-based experimental research. This option fosters experimental research by increasing the pace and size of collecting data sets. Importantly, web-based research profits heavily from integrating tasks that are frequently applied in cognitive psychology into open access software. For instance, an open access random-dot kinematogram (RDK) plugin has recently been integrated into the jsPsych software for web-based research. This plugin allows researchers to implement experimental tasks with varying coherence levels (with that varying task difficulty) of moving dots or varying signal to noise ratios of colored dots. Here, we introduce the random-object kinematogram (ROK) plugin for the jsPsych software which, among other new features, enables researchers to include oriented objects (e.g., triangles or arrows) instead of dots as stimuli. This permits experiments with feature congruency (e.g., upwards-moving triangles pointing upwards) or incongruency (e.g., upwards-moving triangles pointing downwards), allowing to induce gradual degrees of stimulus interference, in addition to gradual degrees of task difficulty. We elaborate on possible set-ups with this plugin in two experiments examining participants' RTs and error rates on different combinations of coherence and congruency levels. Results showed increased RTs and error rates on trials with lower coherence percentages, and on trials with lower congruency levels. We discuss other new features of the ROK plugin and conclude that the possibility of gradually varying the coherence level and congruency level independently from each other offers novel possibilities when conducting web-based experiments.
Asunto(s)
Investigación , Programas Informáticos , Humanos , Estimulación Luminosa , Investigadores , InternetRESUMEN
Cognitive control refers to the ability to make correct decisions concurrent to distracting information, and to adapt to conflicting stimulus configurations, eventually promoting goal-directed behavior. Previous research has linked individual differences in cognitive control to psychopathological conditions such as anxiety. However, a link with uncertainty tolerance (UT) has not been tested so far, although both constructs describe cognitive and behavioral performance in ambiguous situations, thus they share some similarities. We probed cognitive control in web-based experimentation (jsPsych) with a simple flanker task (N = 111) and a version without confounds in episodic memory (N = 116). Both experiments revealed two well-established behavioral indices: congruency effects (CEs) and congruency-sequence effects (CSEs). Only small-to-zero correlations emerged between CEs, UT, and need for cognitive closure (NCC), a personality trait inversely related to UT. A subtle correlation (r = .18) was noted in Experiment 2 between NCC and CSE. Throughout, Bayesian analyses provided anecdotal-to-moderate evidence for the null-hypotheses.