A NOVEL MEASURE OF HABITS AND GOAL-DIRECTED CONTROL

Abstract

Repetitive habitual behaviour is thought to occur even when this directly conflicts with goals. This is termed an action slip. The current research aimed to present a simple procedure that captures action slips and is easy to implement and interpret. Chapter 1 reviews the current literature. Chapters 2-4 then report nine experiments that aimed to test our novel procedure. A congruency effect, where participants’ performance was good on congruent trials but comparatively poor on incongruent trials, was consistently found in our experiments (with one exception), providing evidence of action slips. Chapter 2 details three experiments that used an Outcome (O)-Stimulus (S) delay manipulation. Although evidence for the effect of O-S delay was not strong, the numerical pattern was consistent across experiments, where shorter delays showed a bigger congruency effect on accuracy. Chapter 3 explores a load manipulation in three experiments, which aimed to reduce participants’ working memory capacity during the experiments, and the effects of time pressure in a fourth experiment. The results provide some evidence that the congruency effect can be increased by reducing working memory capacity. Chapter 4 explores a devaluation version of the experiment, followed by a further experiment that manipulates the amount of training. When the devaluation procedure was used, the congruency effect was not observed. When we returned to the standard testing procedure and manipulated the amount of training, we observed a congruency effect, even after a short amount of training. This is inconsistent with the S-R account and dual process theory. Chapter 5 explores whether individual differences are associated with performance in our experiments: self-reported habitual behaviour (COHS), goal-directed control (HSCQ), impulsivity (BIS-11), perceived stress (PSS), depression (PHQ-9), and anxiety (GAD-7). Our results found no association between action slip scores and self-reported habitual behaviour, nor any of our other individual differences measures. Participants with high non-planning impulsivity scores performed less accurately across the experimental trials, which could be interpreted as a lack of goal-directed control. Chapter 6 discusses the question of whether the congruency effect seen in our Experiments is evidence to support a dual-process account of habitual behaviour.