Disputas: Reut Peleg

Reut Peleg's public defense of her thesis at the Ph.D programme in Behavior Analysis. Thesis title: Response Variability: A dimension of operant behavior?

Trial Lecture

Wednesday March 20 2019, at 10:00

Title: Behavioral variability: its importance in applied settings

Public defense

The candidate will defend her thesis 20th March 2019 at 12:15

We ask the audience to take their seats in good time before the public defense commences.

Ordinary opponents

Leader of the public defence

Head of Department Einar Strumse


  • Abstract

    Response variability increases when reinforcement is contingent upon it, can be placed under stimulus control (e.g., Page & Neuringer, 1985), and matches reinforcement rates (e.g., Neuringer, 1992). These three observations are the basis of what is termed operant variability, suggesting that variability may be directly reinforced. However, the term operant variability seems to contradict the very definition of reinforcement as an event that follows behaviour and subsequently increases the likelihood of reoccurrence of those responses that preceded it.

    The question that arises is how can a process that promotes repetition promote variability at the same time? This thesis seeks to provide alternative explanations for the increased variability observed under variability (lag) schedules and includes four studies. Studies 1 and 2 draw on research in the field of response sequences and suggest that the resistance to change (RTC) of different response positions within a sequence is differentially affected by reinforcement and extinction, as well as by operant and respondent processes. Study 3 includes an analysis of different procedural parameters that may contribute to the observed variability level. Study 4 compares different measurement procedures and includes a molecular analysis of the data obtained from two subjects responding on a lag reinforcement schedule.

    Based on these four studies, it is concluded that a lag schedule of reinforcement organizes responses in an orderly fashion, and together with other variables such as procedural variations and the possible involvement of respondent processes, variability is likely induced.