The price of fear: Developing a behavioural assessment of fear-related avoidance incorporating dynamic response measures

Introduction: Fears (e.g., of spiders, or heights) are maladaptive to the extent that they can cause us to avoid opportunities that could enrich our lives. Thus, avoidance is central to fear constructs. However, the roles that avoidance motivations play in fears are complex and remain partially understood. Much research of avoidance has considered it to be the result of subjective fear and this viewpoint tends to construe avoidance exclusively as problematic behaviour. Even though avoidance can significantly impair an individual’s life in extreme cases, it is a crucial component of adaptive behaviour. Early researchers understood that the balance between avoidance motivations and approach motivations was critical to identifying maladaptive behaviour and modifying it. Recently, there has been a revival of interest in developing experimental paradigms based on theories that explicitly postulate a conflict between coexisting motivations to approach and avoid stimuli. The present work connects with the early work on approach-avoidance conflict (AAC) and is informed by recent empirical developments. Aims: The primary aims of the current research programme were to: (a) characterise the state of the art of fear measures; (b) develop novel measures of maladaptive fears using approach-avoidance paradigms; (c) explore approach-avoidance conflicts whilst responding; (d) provide a preliminary dynamical model of approach-avoidance conflict. In accordance with the foregoing, a review of current fear measures was conducted. Next, a series of novel empirical measures of approach-avoidance conflict were developed and tested. Then, the implications of these findings were incorporated within a dynamical systems model of the dynamic interplay between approach and avoidance motivations within a context. Finally, some implications of the present work for the conceptualization of maladaptive avoidance were briefly discussed. Methodology: Study 1 presented a scoping review examining the strength of association between performance on exposure-based behavioural approach tasks (BATs), and two types of measures of fears/anxiety: (a) self-report scales, and (b) implicit response time tasks. The procedural consistency of BATs across studies was critically evaluated. Experiment 1 and 2 tested two novel AAC-based measures of spider fear. Participants chose between higher or lower points rewards, when the higher points were paired with spiders. Based on how many more points were needed for participants to choose the spider option, a price was calculated as an avoidance-based index of spider fear. This price was then compared to self-report measures of spider fear and against performance in a BAT procedure in Experiment 2. Experiments 3 and 4 investigated whether AAC could be observed in movement while participants made decisions. Participants chose to win or lose varying numbers of points. In 20% of trials wining the points incurred a 20% risk of mild electrical shock. Mouse cursor movements were recorded as participants registered their decision. Results: Study 1 revealed that the majority of studies explored spider fear, suggesting an underrepresentation of other fears in the experimental literature. In comparison to implicit measures of fears, explicit (questionnaire) measures related more strongly to performance in BATs. Overall, there is a lack of standardization in the BAT procedures implemented in the sampled studies. Experiments 1 and 2 report significant and non-significant correlations between the price index and self-reported fear, compromised by strong floor effects. The price index was a better predictor of performance in the BAT than self-reported fear. Experiments 3 and 4 demonstrated that AACs are indeed visible in the response trajectories whilst deciding between the options. Moreover, conflict was highest when participants chose between equally valued response options. Finally, within-trial analyses revealed that different experimental arrangements facilitate the manifestation of motor and cognitive defaults when deciding. Conclusions: Study 1 suggests that the relationship between indirect (explicit/implicit) measures of fears and BATs cannot be generalised beyond arachnophobia. Developing BATs that target a wider range of phobias will contribute to the ecological representativeness of fear measures and research on maladaptive avoidance. The experimental approach-avoidance studies (Experiments 1 to 4) provide evidence in support of this methodology to investigate the processes underlying maladaptive avoidance, with potential ecological validity and clinical utility. Furthermore, these experiments provide evidence for the importance of considering the dynamic interplay between approach and avoidance motivations in the conceptualization of avoidance; and the role of approach contingencies in modifying maladaptive patterns of avoidance. Experiments 3 and 4 put forth mouse-tracking as an ideal methodology to investigate the time course of competing approach-avoidance motivations during decisions. The data typified avoidance responses as more complex than approach responses, and demonstrated that response trajectories change depending on the level of conflict induced. Counterbalancing the response options of decision-making tasks across trials, instead of across trial-blocks as frequently done, helps to dissociate potential cognitive and motor defaults