I have just attended the 43rd annual convention of the Association of Behavior Analysis International in Denver, Colorado. I had a great conference including the presentation of the the following poster which was based on my master’s data. I faced several inquisitive, perhaps one might say tough questions, from my colleagues, including from Bjoern Brembs. Despite the questions being tough, it always helpful to hear how another researcher interprets your research.
Abstract: Behavioral risk-taking tasks are frequently used in psychological research. These tasks measure when a subjects behavior changes from approaching reward to avoiding punishment. Typically, behavior is quantified on a unidimensional scale. In this study, a task was developed in a three-dimensional virtual environment in order to capture a multidimensional response class. In this task, exploration leads to increased reward (money), but too much exploration lead to loss of accrued rewards. Using multivariate dynamic time warping, behavioral variability was quantified by the degree to which the subjects path of exploration in one trial was similar/dissimilar to other trials. It was hypothesized that behavioral variability would decrease across trials as contact with reinforcement increased. Each subject completed a total of 60 trials between two sessions and each session was composed of two blocks of 15 trials. A mixed-effects ANCOVA compared behavioral variability between blocks. Early indication (study is ongoing) from 30 undergraduate subjects, is that this may be partially true. Blocks 2 and 3 show a decrease in variability but an increase in Block 4. This study describes a novel behavioral risk task in a virtual environment and also describes temporal shifts in a multidimensional response class across time.