Stephanie M. Kaplan* and Rory Michelle Sullivan***Harvard College 2010, **Harvard College 2009
This study examined whether mindfulness, or active awareness, could reduce priming effects, the subconscious effects of stimuli on one’s behavior. With a sample of 39 undergraduates, we replicated an earlier study demonstrating the effect of priming the elderly stereotype through a word task on participants’ walk speed, and extended it by introducing a mindfulness manipulation through the task instructions given. We hypothesized that participants in the elderly condition who received the mindfulness instructions would not walk more slowly. Although non-significant, our results showed an interaction in the expected direction with decreased walk-time for participants in the elderly-mindful condition, lending tentative support for our hypothesis. Future studies that address the present study’s limitations can extend these results and continue to inform our understanding of the effects of mindfulness on priming. This concept has important implications for social issues such as stereotyping, as our results imply that mindfulness may be an effective means of reducing stereotype activation.
Researchers have defined mindfulness as an effortless, simple process that consists of drawing novel distinctions and noticing new things, allowing us to become aware of how things change depending on the context and perspective from which they are viewed (Langer, 1989). Mindfulness is not about just paying attention; it is about making a conscious effort to be “in the moment” and not to ignore the environment (Langer, 1989). By being mindful, a person is more aware of the events that are happening to, and the stimuli that are acting upon, him or her, and of how those might impact his or her subsequent actions. This awareness of one’s environment allows one to make more thoughtful, calculated decisions when approaching a given situation, rather than acting without a full understanding or appreciation of it (Langer & Moldoveanu, 2000). In the current study, we sought to examine how this concept of mindfulness might inform our understanding of another concept in psychology: Priming.
Researchers in past studies have suggested that exposing people to certain stimuli can affect their subsequent behavior by increasing the accessibility of a schema, trait, or concept even without them being aware of it (Josephson, 1987). For example, researchers in one study found that children who watched a violent police film displayed more aggressive behaviors in a subsequent hockey game than did those who instead watched a nonviolent film about bike racing (Josephson, 1987). This phenomenon is called priming.
In one well-known study on priming, researchers investigated how being exposed to words associated with the elderly stereotype affected how quickly participants walked down a hallway at the conclusion of an experiment (Bargh, Chen, & Burrows, 1996). The researchers found that participants who were exposed to words related to the elderly stereotype walked more slowly than did those who were exposed to neutral words. When debriefed, participants expressed no knowledge of the connection between the words they had seen and the elderly stereotype, nor did they believe the words they saw had affected them in any way. This study demonstrates the effects of priming, as participants’ behavior was affected by the stimuli even though they were not aware of it.
However, perhaps if people are made mindful, these priming effects will not occur. There is some evidence that priming effects might be the result of participants’ mindlessness during the studies (Djikic, Langer, & Stapleton, 2008). In one study, participants who were made mindful during a priming word task were better able to judge a speaker’s characteristics than were those who received the same priming but were not made mindful (Langer & Newman, 1979). In another study, researchers found that consciousness of a priming event allowed participants to interpret subsequent stimuli more flexibly than participants who were not conscious of the priming event (Lombardi, Higgins, & Bargh, 1987).
In another study, men who were likely to sexually aggress were primed with “power”-related words, and subsequently found women more attractive than did control participants, but, unaware of the manipulation, attributed their attraction to the women’s physical characteristics (Bargh, Raymond, Pryor, & Strack, 1995). When the men’s attentions were drawn to the source of their attraction—the fact that men are sexually attracted to women whom they have authority over—the implicit effect was dampened (Bargh et al., 1995). As demonstrated by the power-primed men, implicit cognitive effects such as priming may be reduced by focusing an individual’s attention the task (Greenwald & Banaji, 1995). Schwarz and Clore (1983; as cited in Greenwald & Banaji, 1995) conducted a telephone survey inquiring about quality of life. On sunny days, reported quality of life was higher than on rainy days. However, when people were asked a question about the weather first, quality of life reports evened out. This indicates that attending, or active thinking, disrupts the influence of weak perceptual cues. Priming, then, can be interrupted by focusing on the source of the prime. These findings suggest that mindfulness may be able to reduce or even eliminate priming effects. This would mean that by being mindful, we could recognize the effects certain stimuli might have on us unconsciously, and take steps to work against this and to instead make conscious decisions.
This idea has important real-world implications for domains such as stereotyping and prejudice. Perhaps by being mindful of the existence of a stereotype, one can recognize that the stereotype is there and take steps not to be influenced by it. Findings by Shih and colleagues (2002) support this notion. When the Asian stereotype was blatantly activated for Asian participants before a math test, the stereotype did not affect their performance. That is to say, when the cue in question is cognized clearly – when it is attended to from the outset – implicit cognitive effects such as priming can be reduced or even eliminated, something which is certainly desirable with regard to stereotyping.
In our experiment, we sought to investigate whether mindfulness can reduce priming effects by examining this idea in the context of the elderly stereotype-activation study (Bargh et al., 1996). We replicated the study by Bargh and colleagues and extended it by introducing the concept of mindfulness to see whether by making participants mindful, we could reduce the priming effects the researchers found. We wanted to discover whether those participants who viewed words related to the elderly stereotype but were also made mindful would still exhibit priming effects. We hypothesized that by making these participants mindful (i.e., leading them to pay active attention to the words in the study), we could negate the priming effects discovered by Bargh and colleagues (1996) (i.e., these participants would not walk more slowly than normal). Presumably, this would be because these participants were made aware of the elderly stereotype’s presence, thus disrupting the prime. This would provide continued support for the idea that mindfulness (i.e., making active distinctions) can indeed reduce priming effects, which would allow for a greater awareness of our actions as well as the ability to make more informed decisions.
The study included 40 male (n = 8) and female (n = 32) participants who were undergraduates at Harvard College over the age of 18. Given that the study we sought to replicate and extend was conducted using college undergraduates, we did the same in the current study. The Harvard College student body was also the most easily accessible pool of potential participants. Participants were recruited through emails sent out over Harvard email lists
Informed consent was obtained from all participants. Before completing the study, participants signed a written consent form, which informed them that there would be minimal risk associated with the study and that they could choose to discontinue participation at any time, and requested their permission to use their data anonymously for future research. All procedures were approved by the Committee on the Use of Human Subjects at Harvard University.
The construct of interest to our study was the display of priming effects. We measured this as the amount of time it took each participant to walk 23.625 feet from Experimenter 1 to Experimenter 2 at the conclusion of the word task. Bargh and colleagues (1996) also used walk-time as the measure of priming effects in their study. The length of time it took the participant to walk this distance was measured by Experimenter 2, who started a stopwatch as soon as the participant crossed a fixed line on the floor, and stopped it as soon as the participant crossed a second fixed line on the floor 23.625 feet away.
The study is a 2 × 2 between-subjects design. There were four conditions, to which participants were assigned through a block randomization of four blocks with 10 random numbers. As in the study by Bargh and colleagues (1996), half of the participants were primed with the elderly stereotype. Half of that group, as well as half of the control group, were also given instructions designed to make them mindful. These “mindful” instructions read as follows: “For each set of words below, make a grammatical four word sentence and write it down in the space provided. Note the similarities and differences among the words as you unscramble the sentences. You do not need to write anything beyond the four word sentence down, but actively think about how the words in each sentence might be related to each other and about how they might be different.” The other half of each group received neutral instructions not intended to induce mindfulness; these “not-mindful” instructions read as follows: “For each set of words below, make a grammatical four word sentence and write it down in the space provided.” This design allowed us to examine the main effect of the priming condition in an attempt to replicate the results of the study by Bargh and colleagues (1996), while simultaneously introducing our mindfulness manipulation. All groups were treated the same way by the experimenters in the study; the only differences between the groups were whether the participants completed a word task that included words related to the elderly stereotype, and whether the participants received mindfulness instructions.
The study was conducted using two tables in the back of Quincy Dining Hall, on opposite sides of the dining hall. It was conducted between the hours of 10:30am and 11:30am, and between 3pm and 5pm, as the dining hall is mostly empty during these times.
To begin, Experimenter 1 was seated in a specific chair at the back-right table in Quincy Dining Hall, and Experimenter 2 was seated at a specific chair at the back-left table in Quincy Dining Hall. When the participant arrived, Experimenter 1 greeted him or her, told the participant to sit down across from her, and had the participant sign the consent form. This experimenter then handed the participant the Language Ability Task and explained the instructions.  This task was identical to that used in the original study by Bargh, Chen, and Burrows (1996), and was obtained directly from Professor John A. Bargh. Depending on the condition, the task included either mindful instructions or not-mindful instructions, and either elderly-stereotype-related words or neutral words. The participant was then told to reread the instructions silently, and also that he or she was allowed unconstrained time to complete the task. The participant’s task-time was timed by Experimenter 1 on her laptop, without the participant’s knowledge.
When the participant informed the experimenter that the task was complete, the experimenter thanked the participant, gave a fake, partial debriefing explaining that the study was meant to examine how the brain processes language, and instructed the participant to walk over to Experimenter 2, who would finish debriefing and compensate him or her. Experimenter 2 started a stopwatch as soon as the participant crossed a fixed line on the floor, without the participant’s knowledge. This experimenter stopped the stopwatch on the laptop as soon as the participant crossed a second fixed line that was 23.625 feet away from the first line. Experimenter 2 was blind to the instructions received by the participant so as not to bias the timing. This experimenter then conducted a funneled debriefing and gave the participant his or her compensation. During this debriefing, the participants were asked what they believed the purpose of the experiment was, and all answered that they thought it had to do with language ability, demonstrating that the deception employed was effective.
We first ran t-tests between male and female participants’ walk-time, task-time, and word recall data. Next, we ran a 2 × 2 ANOVA using between-subjects factors of mindfulness (mindful instructions, not-mindful instructions) and priming (elderly stereotype-related words, neutral words). We then recoded word recall into a measure of mindfulness by splitting the data into two groups at the median and running a 2 × 2 ANOVA using the higher recall as the between-subjects mindfulness factor. We then repeated this procedure using task-time as the between-subjects mindfulness factor. To assess how the mindfulness instructions affected word recall, we ran a Pearson’s Chi-Square on the mindful/not-mindful instructions conditions and word recall, which we made into a categorical variable by splitting the groups at the median. We also ran a Pearson’s Chi-Square on mindfulness and walk-time and on mindfulness and task-time.
Preliminary analyses revealed no significant differences between male and female participants on walk-time, task-time, or word recall, so we collapsed our results across gender.
We ran a 2 (elderly/neutral) × 2 (higher word recall/lower word recall) ANOVA to test for evidence of the elderly stereotype as a prime. Participants who recalled the most words (more than the median) were recoded as primed. This ANOVA yielded a significant interaction, F(1, 39) = 5.039, p = .031, in the expected direction, with a medium to large effect size of 0.129 (Figure 1). This demonstrates that participants who received elderly words and then recalled more words were more affected by the elderly stereotype and walked slower as a result, affirming the results of the previous research by Bargh et al. for the existence of this priming effect.
To determine whether our mindfulness instructions had an effect on participant behavior, we also assessed differences in word recall and task-time between the groups that received mindfulness instructions and the groups that did not. Analyses revealed that participants who received the mindful instructions remembered more words, χ2(1, N = 39) = 4.356, p = .037, than did those who received the not-mindful instructions, and took slightly shorter time to complete the task, χ2(1, N = 39) = 1.600, p = .206, although this value did not reach the level of significance.
We ran a 2 (elderly/neutral) × 2 (mindful/not mindful) analysis of variance (ANOVA) on walk-time, which is the dependent measure. This revealed an interaction that was in the expected direction and approached significance, F(1, 39) = 2.529, p = .121, and had a medium effect size (η2 = .067; Figure 2). Participants in the elderly-not mindful condition walked more slowly than participants in the neutral-not mindful condition. Participants in the elderly-mindful condition walked more quickly than those in the elderly-not mindful condition. Participants in the neutral-not mindful condition walked slightly faster than participants in the neutral-mindful condition.
Our results provide tentative support for our hypothesis that mindfulness is capable of reducing priming effects. We were able to replicate the results of the study by Bargh and colleagues (1996), finding that participants in the elderly-not mindful condition walked more slowly than did participants in the neutral-not mindful condition. We were also able to determine that our mindfulness instructions did have an effect, as the participants who received them were able to recall more words than were those who did not, suggesting that these participants were made mindful. It is somewhat surprising that the mindful participants completed the word task more quickly, but it is possible that the mindfulness instructions made these participants more focused on the task, which would explain their lower task-times.
In support of our hypothesis, our mindfulness manipulation had the expected effect on walk-time. The participants in the elderly-mindful condition walked faster than the participants in the elderly-not mindful condition, supporting our hypothesis that mindfulness can reduce the effects of priming. Participants in the neutral-not mindful condition walked slightly faster than participants in the neutral-mindful condition, but this result is also non-significant and we expect would remain so upon further testing. Although the interaction was not significant, there was twice as large a difference in walk-times between mindful and not-mindful participants when primed than when neutral, suggesting that the mindful instructions accomplished something more in the elderly condition, as they were intended to. It is promising that with a sample size of only 39, we acquired results that were in the expected direction; this suggests that had there been more participants, the interaction might have reached significance.
Though not entirely conclusive, our study does lend support to the theory that mindfulness can indeed reduce priming effects, and prompts future research on a larger scale that incorporates more participants.
Nevertheless, the study suffered from certain limitations. For one, we were unable to replicate the conditions of the study by Bargh and colleagues (1996) exactly. In the 1996 study, walking distance was 9.75 meters (31.9882 feet) and located in a hallway between the experimenter and an elevator. In the present study, walking distance was 7.2009 meters (23.625 feet) and located in a mostly empty dining hall between two experimenters. Though the closest conditions available were used, these discrepancies could have interfered with our results. First, a longer stretch of hallway may allow for a greater range of measured walking times and therefore greater variability between participants. Second, the present study required participants to walk toward an experimenter, which may have created an environment differing significantly from one in which the participant believed he or she is merely leaving the room. Future studies that mirror the conditions of the study by Bargh and colleagues (1996) more exactly may provide more conclusive results than ours did. However, researchers Djikic, Langer, and Stapleton (2008) found, using a similar design to the present study in which after a partial debriefing participants walked between two experimenters, that when participants were primed with photos of elderly people, they walked significantly faster when encouraged to be mindful.
It is also possible that the present study’s mindfulness instructions were not powerful enough, since some participants mentioned upon debriefing that they only remembered to “notice similarities and differences” after the first few scrambled sentences. Nevertheless, the fact that such a subtle manipulation, coupled with small variability in walking time because of the short distance, yielded a medium effect size in the expected direction is promising for future studies that can address the above limitations.
Future research should be conducted with more participants to bolster the theory that mindfulness can reduce priming effects. It should also be extended to assess mindfulness’s effect on other versions of stereotype activation, including stereotype threat. It would be desirable to determine how subtle mindfulness instructions can be while still being effective, as the more general the instructions can be made, the more contexts to which they can be applied, and the more pervasive the mindfulness mindset can become. It is when mindfulness becomes a pervasive mindset that we can be protected from subtle stereotype activation, which will particularly benefit those subject to stereotype threat. By practicing mindfulness, we may find ourselves able to make more conscious decisions in our daily lives.
This research was made possible by a stipend from the Harvard University Psychology Department for students enrolled in Psychology 1901: Methods of Behavioral Research. We also thank Berryline, which supported our research by granting forty $2 gift certificates within our $50 budget limitation to be used as compensation for participants. We would like to thank Professor John A. Bargh for providing us with easy access to his scrambled sentence task. We also thank Professor Ellen Langer and the students of her mindfulness lab for their helpful comments on experimental design and task instructions, Dr. Mariann Weierich and Michelle Wedig for all of their help, and especially Michelle Wedig for her assistance and suggestions in analyzing and interpreting the data.
Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior: Direct effects of trait construct and stereotype activation on action. Journal of Personality and Social Psychology, 71(2), 230-244.
Bargh, J., Raymond, P., Pryor, J. & Strack, F. (1995). Attractiveness of the underling: An automatic power—sex association and its consequences for sexual harassment in addition, aggression. Journal of Personality and Social Psychology, 68, 768-781.
DeMarree, K. G., Wheeler, S. C., Petty, R. E. (2005). Priming a new identity: Self-monitoring moderates the effects of nonself primes on self-judgments and behavior. Journal of Personality and Social Psychology, 89(5), 657-671.
Djikic, M., Langer, E. J., & Stapleton, S. F. (2008). Reducing stereotyping through mindfulness: Effects on automatic stereotype-activated behaviors. Unpublished manuscript, University of Toronto and Harvard University, Cambridge.
Greenwald, A & Banaji, M (1995) Implicit social cognition. Psychological Review, 102(1),1024-1027.
Josephson, W.D. (1987). Television violence and children’s aggression: Testing the priming, social script, and disinhibition prediction. Journal of Personality and Social Psychology, 53, 882-890.
Langer, E.J. (1989). Mindfulness. Cambridge, MA: Da Cappo Press.
Langer, E. J. & Moldoveanu, M. (2000). The construct of mindfulness. Journal of Social Issues, (56), 1-9.
Langer, E. J. & Newman, H. (1979). The role of mindlessness in a typical social psychology experiment. Personality and Social Psychology Bulletin, (5), 295-298.
Lombardi, W.J., Higgins, E.T., & Bargh, J.A. (1987). The role of consciousness in priming effects on categorization: Assimilation versus contrast as a function of awareness of the priming task. Personality and Social Psychology Bulletin, 13(3), 411-429.
Shih, M., Ambady, N., Richeson, J. A., Fujita, K., & Gray, H. M. (2002). Stereotype performance boosts: the impact of self-relevance and the manner of stereotype activation. Journal of Personality and Social Psychology, 83 (3), 638-647.
-  Copies of the tasks are available from the authors upon request.↩