Norwegian version

Public defense: Siv Kristin Nergaard

Siv Kristin Nergaard will defend her thesis "Effects of reinforcement on behavioral variability" for the PhD in Behaviour Analysis.

Trial lecture title: Application of Behavioral Theory and Techniques to Contemporary Existential Challenges.

Ordinary opponents:

Leader of the public defense is Professor Magne Arve Flaten, OsloMet.

The main supervisor is Professor Per Holth, OsloMet and the co-supervisor is Professor Ingunn Sandaker, OsloMet.

Thesis abstract

A consequence is defined as reinforcing when it increases the probability that a response class will reoccur in a similar situation.

For that reason, research on reinforcement has shown that it produces efficient and usually quite repetitive behavior. The widely accepted theory of direct reinforcement of variability, therefore, seems somewhat of an oxymoron.

Study 1

Study 1 of this thesis reviewed and discussed the research that have led to such a conclusion. The paper carefully analyses the literature, examining the experimental procedures and statistical analyses commonly adopted.

Study 1 identified five characteristics of operant variability behavior where the directly reinforced variability did not relate to independent variables in the same manner as the more typically studied repetitive behavior.

The study argues that these inconsistencies should at least cause a consideration that the variability seen to increase under variability contingencies, which it undeniably does in experiment after experiment, could be a product of experimentally unplanned reinforcement of moment-to-moment patterns of behavior.

Study 2

Study 2 investigated human rule-governed behavior, as it has been found that variability in behavior decrease when humans are presented with instructions. As of yet, the issue of reinforcement contingencies for instruction following has been somewhat overlooked.

The result of Study 2 shows that the already quite repetitive behavior under instructional control become even more repetitive in the presence of a reinforcement contingency, in the form of bonus points, contingent on responding as instructed.

This increased instructional control continued even after the bonus points were discontinued. Study 1 concluded that some of the experimental procedures in earlier variability experiments may have complicated the interpretation of the data.

Study 3

Studies 3 and 4 pursue some of these experimental procedures.

Study 3 investigated how the independent variables of a timeout and an interresponse interval affected responding in a complex, non-variability, contingency.

Under such a contingency the probability of switchovers increased notably in conditions including an interresponse interval, whereas timeout did not seem to affect response patterns.

Study 4

Study 4 investigated stepwise increases and decreases in a lag procedure using a unit consisting of one individual response on several distinct operanda. The result suggests that during increased lag schedules reinforced responses are often repeated.

In fact, several unreinforced responses to the same operandum usually occurs after a reinforced response. This intermittent extinction terminates when the repetitive responding to the same operandum ceases and responding to another operandum commences.