Dissertation Defense - Kyosuke Kazaoka

Dissertation Defense: An Explorational Study Using an Exchange Task to Understand How we Interact with Others

Kyosuke Kazaoka

Ph.D. Candidate in Health Sciences

Concentration: Behavior Analysis

Department of Rehabilitation and Health Services

College of Health and Public Services

Date: Dec. 5, 2022

Time: 11 a.m. – 1 p.m.

Place: Chilton Hall, Room 363

Major Professor: Dr. Traci M. Cihon

Dissertation Committee: Dr. Traci M. Cihon, Dr. Shahla Ala’i, Dr. Jesús Rosales-Ruiz

Abstract: Behavioral scientists have collected a myriad of data on individual behavior in basic, applied, and practice settings; these data have helped to establish principles of behavior and technologies of behavior change based on these principles in support of the well-being of others. Because applied settings are different from highly controlled settings, applied behavior analysts must be cognizant of multiple variables interacting simultaneously, such as multiple individuals behaving under a shared environment, and the unintended consequences stemming from the application of behavioral technologies. The purpose of the current study was to explore the patterns of interactions that would develop among three participants engaged in an exchange task. Eighteen participants were divided into six groups of three and interacted with each other using an online game developed for this study. The participants interacted by giving points to each other during the experiments. The results suggest that the participants developed consistent interactions between/among themselves through the task. Although many of the participants interacted and formed cooperative relationships through the task, variabilities in when participants began forming consistent interactions between/among participants were observed. The implications of such findings and future directions are discussed.

Background: Behavioral scientists have collected a myriad of data on individual behavior in basic, applied, and practice settings; these data have helped to establish principles of behavior (e.g., Skinner, 1938, 1953) and technologies of behavior change based on these principles (e.g., Cooper et al., 1987, 2007, 2020) in support of the well-being of others. One strength of the discipline of behavior science is that the experimental analysis of behavior, applied behavior analysis, and conceptual analysis inform each other through recursive interactions among the arms of the science, working collectively to develop, refine, and improve the science and technologies that it offers (Neef & Peterson, 2003; Wilder et al., 2022). Because applied settings are different from highly controlled settings, applied behavior analysts must be cognizant of multiple variables interacting simultaneously, such as multiple individuals behaving under a shared environment, and the unintended consequences stemming from the application of behavioral technologies. Remembering that the discipline of behavior analysis has strong bonds among the basic, applied, and conceptual domains of the science, and faced with challenges common to the applied setting, it may prove beneficial to return to the basic research setting to investigate and better understand interactive, multi-level problems. Basic research on social behavior and on the metacontingency is particularly relevant.

The purpose of the current study was to address the limitations of previous research on social behavior and on the metacontingency and to explore the patterns of interactions that would form among three participants engaged in a reciprocal interchange situation (i.e., exchange tasks) in the basic experimental setting.

Method: Eighteen participants were divided into six groups of three. Participants engaged in the exchange task using an online game developed for this study. During the experiments, participants moved through a series of trials in which each participant chose a shape that produced points for themselves (independent tasks) or for one of the other participants (giving tasks). The participants did not have any shared history nor were they allowed to communicate with each other; thus, the only way for the participants to interact during the experiments was through the giving tasks. The independent variable was the change in the point values for each participant’s giving tasks. The experimental conditions were arranged following a multiple baseline across participant design (Johnston & Pennypacker, 1993) in which point values for the giving tasks were increased by 5 points for one participant at a time after every 50 trials. Each experiment consisted of no more than 350 trials spread across up to seven experimental conditions.

Data analysis: There were four dependent variables: 1) the number of times each participant chose the independent task and the two giving tasks, 2) the number of occurrences of two-person cooperation between participants, and the number of occurrences of two forms of three-person cooperation (squares or triangles), 3) the equality/inequality in earned point differences among participants, and 4) the order in which participants made choices in trials. These data were graphed to visually analyze the patterns of interactions among participants that formed through the experimental task, as well as to evaluate the effects of conditions changes on the patterns of participants’ interactions.

Significance: The results of this study showed that 14 out of 18 participants allocated their choices to the giving tasks over the independent task. Although the giving tasks involved a risk of points not being reciprocated by other participants, the participants’ allocation of their choices to the giving tasks created a condition for participants to interact with each other during the experiment. Some of their interactions evolved into two- or, albeit less frequent, three-person cooperation. When two-person cooperation occurred consistently between two participants, the remaining participants tended to allocate their choices to the independent task; thus, these participants isolated themselves from the social situations. Although there are several limitations to the current study, the findings may have some implications as to how to navigate individual behaviors as they occur in the social environment, such as encountered when providing behavioral and health care services in home and community-based settings where multiple individuals live, work, and play.

Event Date: 
Monday, December 05, 2022 - 11:00am to 01:00pm
Event Location: 
Chilton Hall, Room 363