Robustness of individual differences in temporal interference effects.

Abstract

Magnitudes or quantities of the different dimensions that define a stimulus (e.g., space, speed or numerosity) influence the perceived duration of that stimulus, a phenomenon known as (temporal) interference effects. This complicates studying the neurobiological foundation of the perception of time, as any signatures of temporal processing are tainted by interfering dimensions. In earlier work, in which judgements on either time or numerosity were made while EEG was recorded, we used Maximum Likelihood Estimation (MLE) to estimate, for each participant separately, the influence of temporal and numerical information on making duration or numerosity judgements. We found large individual differences in the estimated magnitudes, but ML-estimates allowed us to partial out interference effects. However, for such analyses, it is essential that estimates are meaningful and stable. Therefore, in the current study, we examined the reliability of the MLE procedure by comparing the interference magnitudes estimated in two sessions, spread a week apart. In addition to the standard paradigm, we also presented task variants in which the interfering dimension was manipulated, to assess which aspects of the numerosity dimension exert the largest influence on temporal processing. The results indicate that individual interference magnitudes are stable, both between sessions and over tasks. Further, the ML-estimates of the time-numerosity judgement tasks were predictive of performance in a standard temporal judgement task. Thus, how much temporal information participants use in time estimations tasks seems to be a stable trait that can be captured with the MLE procedure. ML-estimates are, however, not predictive of performance in other interference-tasks, here operationalized by a numerical Stroop task. Taken together, the MLE procedure is a reliable tool to quantify individual differences in magnitude interference effects and can therefore reliably inform the analysis of neuroimaging data when contrasts are needed between the accumulation of a temporal and an interfering dimension.