The Illusion of Control: What's Luck Got To Do With It?
by Andrea Breinholt and Lynnette A. Dalrymple
The present study was designed to examine the role of one's desire for control (Burger & Schnerring, 1982) on one hand and one's belief in good luck as a controllable attribute (Darke & Freedman, 1997) on the other in determining one's susceptibility to the illusion of control phenomenon. Two hundred eighty-one undergraduate students served as participants. Each participant completed the Desirability of Control Scale and the Belief in Luck Scale immediately prior to completing an on line simulated gambling task. Participants were randomly assigned to either a high involvement or low involvement condition and either a descending or random sequence of outcomes. In general, participants in the high involvement condition wagered more than did participants in the low involvement condition. Results also showed that under low involvement conditions one's belief in luck will predict amount wagered whereas under high involvement conditions desire for control will reliably predict amount wagered.
The Illusion of Control: What's Luck Got To Do With It?
In 1975, Langer observed that under the right conditions, some people's behavior suggested a perception of outcome probability that was higher than one would expect in an objectively uncontrollable task. She proposed that the presence of certain factors usually associated with improved performance on skill-related tasks directed the participants' focus away from the chance-related aspects of the task and created an illusion of control. These skill-related factors include, but are not limited to, one's degree of active involvement in the task and the sequence in which outcomes are experienced (an early string of successes was thought to indicate greater skill).
Illusion of control research continued in a similar vein for years following Langer's initial work and has often been operationalized as higher wagers on gambling tasks. Researchers have attempted to expand on Langer's ideas and have hypothesized about the influence of various contributing factors such as individual differences in personality variables. Burger and his colleagues (Burger & Cooper, 1979; Burger & Schnerring, 1982) suggested people vary in the amount of personal control that is desired in one's life. They further suggested that illusion of control effects can be best explained by the fact that environmental stimuli that suggest a greater degree of personal control will be more salient to people with a higher desire for control over their environment. Therefore, people with a high desire for control should be more susceptible to illusion of control manipulations than people with a lower desire for control. Their hypothesis was supported in a study in which they induced an illusion of control by varying participants' level of active involvement.
More recently, Darke and Freedman (1997) suggested that certain illusion of control effects can be best explained by the fact that people will often behave as if luck was a stable, predictable factor. They hypothesized that participants might form expectations of future success at least in part on their perception that luck is present and can be used to their advantage. Specifically, they argued that certain types of illusion of control manipulations might be acting as a signal regarding the presence or absence of luck. The outcome sequence that one experiences is just such a signal. Rather than providing information about one's ability to perform well at a specific task, an early string of successes might be providing information that one is on a lucky streak. Lucky streaks are often seen to be somewhat stable and therefore one should increase one's wagers to maximize the utility of the presence of luck. This is in contrast to Burger and Schnerring's (1982) claim that wagers on certain kinds of gambling tasks are more influenced by one's desire for control than one's belief in luck.
The present study was designed to examine the effects of desire for control and belief in luck on the illusion of control phenomenon. An illusion of control was induced through a manipulation of participants' level of involvement and by varying the sequence of outcomes experienced. We hypothesized that participants with a high desire for control would wager more on the outcome of a card game when in the high involvement group than in the low involvement group. We also predicted that those with a greater belief in luck, when presented with a string of early successes would wager more.
We assessed the degree to which participants believed that luck was an influential factor in their lives from scores on the Belief in Good Luck (BIGL; Darke & Freedman, 1997) scale but. The scale consists of 12 statements to which the respondent indicates his or her level of agreement. Responses are made on a 5-point scale ranging from 1) Strongly disagree to 5) Strongly agree. Responses are summed with higher scores indicating higher belief in good luck. Participants with scores at or above the 60th percentile and at or below the 40th percentile were considered to have a high and low belief in luck, respectively.
The game that followed was based on the "Kings and Queens" game developed by Burger and Schnerring (1982). Participants were randomly assigned to either the high involvement or low involvement condition and either the descending or random outcome sequence condition. In the high involvement condition, the participant was allowed to shuffle and deal (by clicking buttons on the computer screen) the cards. The participant then chose the target color and the amount wagered. The participant then selected a card and the computer turned it over. After paying off a winning hand or collecting from a losing hand, the process was repeated for each of the 14 trials. The high involvement condition was designed to maximize the participants' perception of personal control.
In the low involvement condition, the computer shuffled and dealt the cards. The participant chose the amount wagered and instructed the computer to choose a card. The computer turned it over a card and settled the bet.
The descending outcome sequence was designed to maximize the illusion of control by placing the majority of successful outcomes in the first half of the trials (see Langer & Roth, 1975). In the present study the descending sequence consisted of the following outcomes:
The random outcome sequence was designed to minimize the illusion of control by placing the successful outcomes more evenly across the 14 trials:
It is important to note that the number of successes in both conditions is the same and reflects chance performance. All participants were fully debriefed after the experiment was completed.
Results and Discussion
Participants in the high involvement condition wagered more (M=47.85) than those in the low involvement condition (M=39.49), F(1, 195) = 16.12, p<.05. DFC interacted with level of involvement such that wagers did not differ reliably as a function of DFC in the low involvement condition. In contrast, in the high involvement condition, high DFC participants wagered more (M=50.96) than low DFC participants (M=43.50), F(1, 195) = 3.88, p<.05. In the low involvement condition, high (M=39.15) and low (M=38.58) DFC participants' bets did not reliably differ (see Figure 1). These findings support the presence of an illusion of control phenomenon in the traditional sense. In contrast, high BIGL participants (M=43.61) did not differ significantly in amount wagered than their low BIGL (M=43.64) counterparts, F(1, 195) = .06, p>.05. BIGL interacted with level of involvement such that in the high involvement condition, the high belief group (M=47.07) and the low belief group (M=48.59) did not differ significantly. In contrast, in the low involvement condition, the high belief group wagered more (M=40.58) than the low belief group (M=37.54), F(1, 195) = 4.41, p<.05.
Our findings did not support the hypothesis that belief in luck would interact with outcome sequence to produce illusory judgments. This is most likely due to the fact that descending and random outcome sequences were not dissimilar enough to create a differential perception of luck. In the future, the ascending sequence of outcomes proposed by Langer and Roth (1975) should be used in an attempt to produce greater contrast.
Our findings did suggest that certain induction techniques such as level of involvement can create illusory judgments. This is consistent with the work of Burger and Schnerring (1982). However, our findings also demonstrated that, in the low involvement condition where desire for control does not predict betting behavior, a belief in luck does predict amount wagered. This supports Langer's (1975) notion that people wanting control who are exposed to environmental conditions that suggest control respond accordingly. However, in situations where such environmental conditions are absent, the better predictor of wagered amounts is a belief of luck. The current study showed that while producing similar behavioral results (increased wagers), different illusion of control induction techniques tap into different processes. These findings demonstrate that what was previously thought to be a straight forward issue of effecting one's perception of personal control is a more complex phenomenon that involves many unique processes.
Burger, J. M., & Cooper, H. M. (1979). The desirability of control. Motivation and Emotion, 3, 381-393.
Burger, J. M., & Schnerring, D. A. (1982). The effects of desire for control and extrinsic rewards on the illusion of control and gambling. Motivation and Emotion, 6, 329-335.
Darke, P. R., & Freedman, J. L. (1997). Lucky events and beliefs in luck: Paradoxical effects on confidence and risk-taking. Personality and Social Psychology Bulletin, 23, 378-388.
Langer, E. J. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311-328.
Langer, E. J., & Roth, J. (1975). Heads I win, tails it's chance: The illusion of control as a function of the sequence of outcomes in a purely chance task. Journal of Personality and Social Psychology, 32, 951-955.