Revisiting Variable-Foreperiod Effects: Evaluating the Repetition Priming Account

A warning signal that precedes an imperative stimulus by a certain length of time (the foreperiod) can accelerate responses (foreperiod effect). Plotting reaction time (RT) as a function of foreperiod results in a “U”-shape curve when the foreperiod is fixed in a trial block but manipulated across blocks. When the foreperiod is varied within a block, the foreperiod-RT function is usually negative, with the foreperiod effect modulated by both the current foreperiod and the foreperiod in the prior trial (sequential foreperiod effect). This sequential effect was found to be robust at the shorter foreperiod while diminished at the longer foreperiod. Capizzi et al. (2015) used a non-aging foreperiod distribution and found an increasing foreperiod-RT function (consistent with that in a fixed-foreperiod paradigm) and a sequential effect equal for different foreperiods. They thus proposed a repetition priming account for the sequential foreperiod effect. I conducted three experiments, aiming to test this repetition priming account and to rebuild the connection between the fixed- and variable-foreperiod paradigms. Experiment 1 attempted to replicate Capizzi et al. in a choice-reaction task scenario and found an increasing foreperiod-RT function but a larger sequential effect at the shorter foreperiod. Experiment 2 examined the priming account in a short-foreperiod context and found a decreasing foreperiod-RT function with a larger sequential effect at the shorter foreperiod. Experiment 3 detected a larger sequential effect in general by increasing the difference in duration between the foreperiods that were used in Experiment 2. The current study provided converging evidence that with a non-aging foreperiod distribution the foreperiod-RT function in a variable-foreperiod paradigm shares the same direction as that in a fixed-foreperiod paradigm. However, instead of following Capizzi et al.’s account, the size of the sequential foreperiod effect in general was found to be modulated by the difference in duration between the foreperiods while the relative sizes were determined by the proportions of different foreperiods.