Tutor HuntResources Psychology Resources

Quantitative Research Report

The effect of social-priming on the score of a general knowledge test: A replication of Dijksterhuis & Van Knippenberg’s (1998) study

Date : 14/11/2020

Author Information

Bente

Uploaded by : Bente
Uploaded on : 14/11/2020
Subject : Psychology

Abstract

According to Dijksterhuis and Van Knippenberg (1998), by just a perception of someone or a group of people, a certain mechanism is triggered for a person to behave in that same way. The study confirmed the hypothesis that priming a stereotype or trait leads to explicit behaviour which matches the primed stereotype or trait. The results showed moderate evidence to support the null hypothesis was found and thus rejecting the experimental hypothesis participants in the professor priming condition will score higher on the test compared to participants in the secretary priming group as well as the control group. No effect being found leads to the debate of the replication crisis especially with this study having been replicated many times and mostly finding no effect. After looking at confounding factors and the fact that it was not an exact replication of the original study the conclusion still leads to social priming having no effect on general knowledge as shown in the results and supported by many replication studies.

Introduction

We are often influenced by others and may act differently according to who we are surrounded by at that particular moment. According to Dijksterhuis and Van Knippenberg (1998), by just a perception of someone or a group of people, a certain mechanism is triggered for a person to behave in that same way. The study confirmed the hypothesis that priming a stereotype or trait leads to explicit behaviour which matches the primed stereotype or trait. Priming the stereotype of a professor in the study increased scores of participants when measuring general knowledge. In contrast, the priming of the stereotype soccer hooligans decreased participants performance on a scale measuring general knowledge. The results also showed that priming for a lengthy period leads to more noticeable effects on behaviour and that there is no decline in the effect for at least 15 minutes. Similarly, in Bargh, Chen, Burrows (1996) study participants who were primed with the stereotype of the elderly, as a result walked slower than participants who were not primed. This was done without awareness of the participants as they were deceived and found to not have noticed the link between exposure to priming and their behaviour. However, Doyen et al. (2012) was unable to replicate Bargh et al s (1996) automatic effect of priming on walking speed even though they used a larger sample size and adjusted their study to avoid the limitations of the original study. In another social priming study by Shariff et al. (2007) however, participants gave more money to anonymous strangers after being primed with God concepts indirectly than when they were being primed with neutral concepts or no concepts. This shows how priming God concepts can increase prosocial behaviour.

The replication crisis is prevalent in experimental sciences including psychology, and consists of two main claims (Guttinger, 2018). One being that researchers rarely replicate existing data and the other that if they do attempt to replicate it, they mostly fail. In psychology often the first study will show a statistically significant result, but the replicated study will not. This brings up arguments of whether or not the first study created false positives and if the replicated study indeed showed no effect at all (Maxwell, Lau Howard, 2015). In Maxwell, Lau and Howard s (2015) study it is suggested that this is not the case but rather the result of low statistical power in single replication studies. They highlight the importance of multiple replications to have enough power to show true effects. Stroebe and Strack (2014) suggested that in order to have a meaningful replication it is not enough to try and replace the original circumstances. Instead conditions are grasped when they reflect the theoretical variables manipulated in the original study.

In September 2012 in an open email, cited on the website of the journal Nature (3 October 2012), Daniel Kahneman challenged Social Psychologists regarding the work on social priming to replicate their studies. In his opinion the credibility of studies in this area was low due to the lack of replication as well as fraud. This study aims to replicate Dijksterhuis and Knippenberg s (1998) study within a student population in order to increase credibility and educate students about the value of replication.

The effects of priming condition (independent variable) which consisted of three levels: professor priming, secretary priming and control group, on the score of a Trivial Pursuit based general knowledge quiz (dependent variable) was tested. We hypothesized that there will be a significant effect of priming condition on the Trivial Pursuit-based general knowledge quiz. More specifically participants in the professor priming condition will score higher on the test compared to participants in the secretary priming group as well as the control group.

Method

Participants

There were 98 number participants who were recruited through opportunity sampling method. The participants were randomly allocated to the three conditions (professor priming, secretary priming and control group). The professor priming condition contained 33 participants, secretary priming 34 and control group 31 participants. The participants were MSc students from the University of Chester from either the Psychology conversion course or Family and Child Psychology course. Ethical approval was given by the University of Chester Psychology Department Ethics Committee. Participants were treated in accordance with the ethical guidelines of the British Psychological Society.

Materials

Materials that were prepared before the experiment ran were the Trivial Pursuit questions. The questions were taken from Trivial Pursuit but changed into multiple choice questions specifically for this study. There was a total of 42 questions with an equal number of questions from each question category and 4 multiple choice options.

Example of multiple-choice questions

5. What might a NASA astronaut condense to the initials LEO?

A. Late entry officer

B. Light energy orange

C. Liquid energy oil

D. Low earth orbit

11. What is Britain s largest, wild land carnivore?

A. Red deer

B. Bear

C. Red fox

D. Badger

See PS7301 Moodle Page for full list

Procedure

The study took place in a teaching room at The University of Chester. Participants were given the information sheet followed by the consent form which they were asked to read and sign to provide informed consent. There was an element of deception when informed about the aim of the study as participants were told that the aim of the study was to acquire two separate sets of data, one for our lab report and one for a module in second semester but in fact both were for the lab report with part 1 being levels of priming. This was done as it was necessary for the priming to be unconscious. In part 1 of the study the participants were randomly assigned to conditions by handing out the booklets of different conditions around the room (professor priming, secretary priming and control group). The participants in the professor priming condition were asked to imagine a typical professor and write about it for 5 minutes. In the secretary priming condition, the participants were asked to do the same but for a typical secretary and in the control group the participants were asked to doodle for 5 minutes. In part 2 of the study participants were told that the data was going to be used to construct a general knowledge scale with five subscales, each containing 42 questions. They were told the subscales range from very easy (1) to very difficult (5) and that the differences in difficulty between the five subscales were being tested. The participants were then told that they had been assigned to the most difficult subscale. The participants were asked to answer all questions and there were no time constraints. When finished they were informed to wait quietly till everyone else finished and not turn to the next page. When everyone had completed their questions, the researcher read out the correct answers and everyone tallied and recorder their score. The scores were submitted anonymously via a reporting sheet, not revealing any names or personal details in order to maintain anonymity and confidentiality. The participants were then debriefed, and the real aim of the study was revealed as well as the reasons behind the deception. The participants were given resources to access for support if they needed it.

Analysis and Design

The dependent variable was the score (0-42) on a Trivial Pursuit-based general knowledge quiz. The independent variable was the priming condition which consisted of three levels: professor priming, secretary priming and control group (doodling). The study was a between subjects design. Data was analyzed using SPSS. The normality of data was assessed using a Shapiro-Wilk test and the homogeneity of variance was assessed using a Levene test. A one-way ANOVA was carried out to show whether the mean scores differed significantly or not.

Results

The dependent variable (score from 0-42 on a Trivial Pursuit-based general knowledge quiz) was counted and recorded by the participants and inserted into a table by the researcher, ready for analysis.

To assess the normality of the data, a Shapiro-Wilk test was carried out. The result of the Shapiro-Wilk (S-W) test showed that the sample distribution of (control group) was not statistically significantly different from a normal distribution (SW = .96 df = 31 , p = .34). Similarly the professor priming condition (SW = .97 df = 34 , p = .36) and secretary priming condition (SW = .98 df = 34 , p = .72) was also not statistically significant from a normal distribution. Next, a Levene test for homogeneity of variance was carried out and showed that there was not an issue with homogeneity of variance, F(2, 95) = 2.13 , p = .12

Table 1

Descri ptive statistics number of participants, means and standard deviations for results of Trivial Pursuit test

Priming condition

Number of participants (N)

Mean (M)

Standard Deviation (SD)

Professor Priming

33

19.03

4.47

Secretary Priming

34

18.35

3.60

Control group

31

19.45

4.69

Based on the means shown in table 1, the general knowledge scores for professor priming were lower than the means for the control group but higher than the means for the secretary priming condition. The results are all quite similar, not as would be expected if the original pattern (Dijksterhuis van Knippenberg) were to be replicated.

A one-way ANOVA with three levels (Professor Priming, Secretary Priming, Control) was carried out and showed that the means did not differ significantly, F( 2, 95) = .55, p .58 , 2 = .01 There is a small effect size, shown by the eta squared result (Cohen, 1988). To explore the null effect further, a Bayes Factor was generated using JASP s Bayesian ANOVA, which showed BF01 = 6.87. What this analysis shows is that the observed data are 6.87 times more likely under the null hypothesis than under the experimental hypothesis. This showed moderate evidence support for the null hypothesis over the experimental hypothesis.

Due to the one-way ANOVA not being significant, no further comparisons were carried out.

Discussion

From the results, moderate evidence to support the null hypothesis was found and thus rejecting the experimental hypothesis participants in the professor priming condition will score higher on the test compared to participants in the secretary priming group as well as the control group. However, even though there was no effect detected, this does not necessarily mean that there was no effect. Null results can be ambiguous. It can either be the cause of a type 2 error or there is indeed no effect. In the case of this study, there have been a lot of replications done on this social priming experiment, many of which showed no effect. Shanks et al. (2013) replicated Dijksterhuis and Van Knippenberg s (1998) study and results showed support of the null hypothesis with none of the experiments showing an effect. Similarly, O Donnell et al. (2018) replication study showed no overall difference in trivia performance between participants primed with professor and those primed with hooligan . There are many more studies that have replicated this experiment and have found no effect. This repeated failure to replicate indicates a stronger doubt about the original finding than in a single instance for example.

However on the other hand there are studies that have successfully replicated the social priming experiment. For example, in Bargh, Chen, Burrows (1996) study participants who were primed with the stereotype of the elderly, as a result walked slower than participants who were not primed. However, this study was not an exact replication of Dijksterhuis and Van Knippenberg s (1998) study as different priming conditions were used and therefore it is not a fair comparison. The reason for finding no effect could be due to confounding factors. It was not a complete replication of the original study with slight differences in methodology. For example, the participants were all sat together in a classroom completing the tasks at the same time rather than individually in a lab. This may have caused the participants to behave differently. In addition in our study all participants were psychology students and therefore may have been aware of the deception. Several students confirmed afterwards that they knew they were being deceived because they had taken part in many experiments before. This would have altered the way they felt and the priming may not have had its effect. With the psychology students possibly being more aware of the reasons behind deception as well as knowing they are being observed, they may have acted in a certain way which could have changed the results. In order to get rid of these confounding factors an exact replication or as close to the original study as possible is necessary. Another change that could be made is to make the participants do a general knowledge test before without priming, followed by a test after priming. This would make sure the results aren t due to initial differences in general knowledge between participants from the start.

After looking at confounding factors and the fact that it was not an exact replication of the original study the conclusion still leads to social priming having no effect on general knowledge as shown in the results and supported by many replication studies.

Future work that needs to be done is another replication study within the same setting as the original study, with a larger sample size and not with psychology students to rule out these factors and confirm that the same conclusion still stands.

References

Bargh JA, 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: 230 244.

Dijksterhuis, A. van Knippenberg, A. J. Pers. Soc. Psych. 74, 865-877 (1998).

Doyen, S., Klein, O., Pichon, C., Cleeremans, A. (2012). Behavioural priming: It s all in the mind, but whose mind? PloS One, 7(1)

Ed Yong, Nobel laureate challenges psychologists to clean up their act, (03 October 2012). Nature, Retrieved from https://www.nature.com/news/nobel-laureate-challenges-psychologists-to-clean-up-their-act-1.11535

Guttinger, S. (2018). Replications everywhere: Why the replication crisis might be less severe than it seems at first. BioEssays : News and Reviews in Molecular, Cellular and Developmental Biology, 40(7)

Maxwell, S. E., Lau, M. Y., Howard, G. S. (2015). Is psychology suffering from a replication crisis? What does failure to replicate really mean? American Psychologist, 70(6), 487-498.

Shanks, D. R., Newell, B. R., Lee, E. H., Balakrishnan, D., Ekelund, L., Cenac, Z., ... Moore, C. (2013). Priming intelligent behavior: An elusive phenomenon. PloS one, 8(4),

Shariff, A. F., Norenzayan, A. (2007). God is watching you: Priming god concepts increases prosocial behavior in an anonymous economic game. Psychological Science, 18(9), 803-809.

Stroebe, W., Strack, F. (2014). The Alleged Crisis and the Illusion of Exact Replication. Perspectives on Psychological Science, 9(1), 59 71.

O Donnell, M., Nelson, L. D., Ackermann, E., Aczel, B., Akhtar, A., Aldrovandi, S., Zrubka, M. (2018). Registered Replication Report: Dijksterhuis and van Knippenberg (1998). Perspectives on Psychological Science, 13(2), 268 294


This resource was uploaded by: Bente

Other articles by this author