Michael Zhang and Dr. Harold Miller, Psychology
To begin, I believe that the process of grant proposal writing was very important to both formulating my research idea and design as well as requesting the necessary assistance for the realization of this project. The idea which was developed has its origins with prominent psychologist, George Ainslie—personal rules are conceptually known as verbal commitments which we generate ourselves in order to control our behavior, e.g., we may tell ourselves that we will never cheat on an examination and this rule is effective if we follow through on it; however, the delineation of their effectiveness was only ever explained theoretically—the purpose of this project was to show how people use personal rules: if and when they keep and break them.
Bubble It Up is a computer game which was designed for behavioral experiments with purposes consistent with mine. For this design, the game was programmed to provide only rewarding contingencies and no punishment. The participants were rewarded 3 nickels on average every twenty seconds by clicking on bubbles on either the left side or the right side of the screen. In this design, the program only rewarded participants 50% of the time a reward was due. This could have been a developmental flaw, however, which remains to be examined. In future administrations of the game, we will consider how participants are affected by this.
The three conditions of the study were designed to test the efficacy of personal rules. These conditions included a control group and two experimental groups. In the control group, participants were asked to earn as many nickels as they could. The first experimental group was asked to articulate a personal rule in order to help them to earn as many nickels as they could. Participants in the second experimental group were given the same instructions as those in the first experimental group but instructed that they would earn rewards twice as often on the left side of the screen compared to the right side (a deceptive instruction designed to hinder their performance). We expected the first experimental group to perform optimally since they would have a personal rule to help them achieve their goal. We also expected the second experimental group to break their personal rule once they realize that it does not help them to earn as many nickels.
With the formulation and articulation of personal rules for the game, there were several unforeseen problems. For instance, some participants articulated personal rules which were very vague— “using an eclectic, anything-goes approach,” and some participants articulated personal rules which by design could not help them to maximize their outcome— “always click on the right unless there are only bubbles on the left.” There was little consistency in the articulation of personal rules. Assignment of rules would not have helped either since it would no longer be personal. Furthermore, the optimal rules could only be articulated if knowledge of the design’s game was present. One of the participants had an optimum rule and earned more nickels as a result— “clicking bubbles when they just come up and alternating sides.” However, since this rule was not based on the actual design of the game, its use and effectiveness can be questioned.
There were several limitations in this study—one of which prevented the timely collection of all the data. Attrition was a problem because participants were not motivated to play the game if they were not rewarded as much money as they deemed worthwhile continuing. Future administrations of the game should include a completion bonus. Due to attrition, we still require a few participants in order to complete the three comparable groups.
Another limitation was the method of data analysis. The frequency of bubble bursts on the left side was compared to the right side but this does not consider the timing of these bursts. For instance, if two participants burst the same amount of bubbles on each side, their rewards (one large and one low, for example) might not be comparable since the timing of their clicks is not examined. Furthermore, individual differences could not be carefully considered using the agreed upon quantitative method of analysis. For instance, one participant clicked on the right side for the duration of the study and obtained almost no rewards the entire time. However, the participant kept the personal rule to only click on the right side. In other instances, participants departed from their personal rules but to fewer rewards. These individual differences were not forecasted in the study’s original design.
I would suggest either a different experimental design or a different mode of analysis for this sort of behavioral research. Personal rule research requires a context of moral or ethical decisions in order to be conducted. Another appropriate dimension would be life or lifestyle altering decisions. A simple game of reward is not sufficient to reach the limits to which personal rules apply. Furthermore, the proposed quantitative method of analysis for this study is very limiting since the following of different personal rules should not be compared. A qualitative analysis might work better for when personal rules are kept or broken and for what reasons.
This project would not have been possible or efficacious without the support and assistance of my colleagues. Dr. Harold Miller provided an extensive bank of knowledge as well as a very resourceful laboratory for the study. Jason van der Horst designed the computer game which was used and manipulated it for the purposes of this study. JeLyn Whicker contributed countless hours organizing with and supervising the participants. I am very grateful for their help.
This project was presented in poster form at the inaugural 2010 Psychology Undergraduate Research Conference at the University of California, Berkeley.