An opportunity sample was used to obtain 30 participants of varying ages. 15 participants were of ethnic minority origin (black) the other fifteen were of English origin (white). Gender was of no importance in this study and therefore not recorded. Materials See the Appendix section for photos used as stimulus material and the standard questionnaire that was used for all participants to fill out.
Procedure Each participant was asked to view four photographs. Two were of a white person dressed in either a white or black shirt and the other two were of a black person wearing either a black or a white shirt. (As mentioned above; see appendix A for photos.) As they were presented with each photograph individually, they were also given a questionnaire to fill out. The questionnaire asked them to rate 3 questions on a (Likert) scale of 1-7 about the impression (with regards to aggression) the person in the photograph portrays. The same questionnaire was used for all four photos and can be found in the Appendix (B). The questions asked the participants to rate how strongly they either disagreed (1) or disagreed (7) on the “aggressiveness”, “trustworthiness” and “likeliness to start a fight” of the person in each picture.
Design The design is a repeated measures design as every participant viewed each photograph and answered the same questions. The independent variables were whether the person in the photo was wearing white or black clothes and whether they are of ethnic minority or not. The dependent variable is the score that each question obtains on the Likert scale, i.e. the aggression ratings given to each picture.
Results: See Appendix C for full SPSS output. As this is a referred piece of coursework and the original set coursework required group congruence when analysing data, copies of the raw data have been misplaced by myself and other members of my group. Therefore the only copy of the results I have is in the form of the SPSS output and am unable to insert tables and graphs in this section. Instead I will refer you to the correct table in the SPSS output which is located in the Appendix (C). Sorry for any inconvenience caused.
To test Hypothesis 1: that the photograph of the white person wearing a white shirt would have a higher perceived aggression score when rated by a person from an ethnic minority group and Hypothesis 2: that the photograph of the white person wearing a black shirt will have the highest overall perceived aggression score and Hypothesis 3: that the photograph of the black person wearing the black shirt will have a higher perceived aggression score when rated by a white participant. A within groups ANOVA was performed on each question using a 2 (Participant (PS): white, black) x2 (Clothing: white, black) x2 (Ethnic: white, black) factorial design for repeated measures. The dependent variable was the perceived aggression score that each picture was given.
Question 1 No significant main effects were found for the participant condition (F1,16 = 1.000, p = 0.332), the clothing condition (F1,16 = 1.000, p = 0.332) and for the ethnic condition (F1,16 = 3.240, p = 0.091). This suggests that perceived aggression ratings did not differ significantly as a factor of the individual conditions. A significant interaction between participant and ethnic origin was found (F1,16 = 43.56, p , 0.001). This suggests the ethnic origin of both the participant and the model had a significant effect on aggression rating, with the participants rating more favourably towards the model of the same ethnic origin (See Test of Between-Subjects Effects table on page 2 of SPSS output).
Question 2 A significant main effect for Clothing (F1,16 = 11.636, p = 0.04) was observed suggesting participants ratings of perceived trustworthiness varied significantly as a function of the clothing worn by the models in the photos. A significant interaction was also found between the ethnic origin of participants, colour of clothing and ethnic origin of models (F1,16 = 4.545, p = 0.049) showing suggesting that all three conditions had an effect on the participants’ perceived trustworthiness ratings, with a stronger interaction between the ethnic origins of the participants and models ( F1,16 = 6.545, p = 0.021), See Test of Between-Subjects Effects table on page 7 of SPSS output.
Question 3 A significant main effect for Clothing (F1,16 = 16.056, p = 0.001) was observed suggesting participants ratings of perceived likeliness to start a fight varied significantly as a function of the clothing worn by the models in the photos. A significant interaction was also found between the ethnic origin of participants, colour of clothing and ethnic origin of models (F1,16 = 16.056, p = 0.001) showing suggesting that all three conditions had an effect on the participants’ perceived “likeliness to start a fight” ratings, with stronger interaction between the ethnic origins of the participants and models ( F1,16 = 6.722, p = 0.020), and the colour of clothing worn and the ethnic origin of the model ( F1,16 = 4.500, p = 0.050), See Test of Between-Subjects Effects table on page 19 of SPSS output.
Discussion: As can be seen from the results there is clear evidence to show that both skin colour and the colour of one’s clothing do indeed have an impact on the impressions formed by others with regards to perceived aggression and trustworthiness. The results show that there is a highly significant difference in the perceived aggression scores of the picture showing the white person wearing the white shirt when rated by participants from the two different ethnic groups, with black participants rating the model as most aggressive.
Hypothesis 1 is therefore accepted; suggesting that ethnic background is a factor when forming impressions concerning aggression, trustworthiness and likeliness of appearing in a fight. This supports Brown’s (1995) finding that racial stereotypes and fear of different races are still a factor when we form our impressions of people. It is interesting to note that the mean perceived aggression score rated by White participants was the highest for the black model wearing a black shirt (a mean score of 6, with 7 being most aggressive, on the Likert Scale); and the mean perceived aggression score rated by Black participants was identical (mean score of 6), but instead, for the white model wearing a white shirt, thus allowing acceptance of Hypothesis 3.
Hypothesis 2 was rejected; the photograph of the white person wearing the black shirt did not have the highest perceived aggression score. It can also be observed that, in total, the models wearing the black shirt were rated the most aggressive by all participants (mean, 4.0; S.D. 1.907) with the difference between them, and the models wearing the white shirt being 0.417. This supports Vrij’s (1997) claim that the colour of a person’s clothing has a stereotypical effect on perceived aggression and also our notion that age-old stereotypes regarding black as being “the colour of evil” are still in use today.
In addition, it was discovered from our results that, overall, the models (of both ethnic origins) wearing the white shirt were regarded as more trustworthy than the models wearing the black shirt. As can be expected in light of the above findings, White participants rated Picture 1 as the person they would trust the most (Mean, 5.0; S.D. 1.0). One of the main limitations of this study is that ratings themselves, without textual accompaniment supporting reasons for participants’ choices, do not give us full insight into the reasoning of the participants when making their choices.
The participants tended to be friends chosen at random some of whom were psychology students conducting similar studies and were not typical of the population of the world, although we do acknowledge that obtaining a truly typical representative of the population would be next to impossible. Even the friends who were not students were made aware when completing the informed consent form that they were partaking in a study about impression formation and since stereotypes are universal, may have engineered their ratings to resemble what they thought we wanted them to say.
For example, they may not have found any of the pictures particularly aggressive, however, because there were only two models of different ethnic origin being compared, opted for the one they assumed to be the “correct” answer. A way to combat this would be to rate models from a variety of different ethnic origins and maybe to even have two models of each race within the stimuli. Also, to have a greater variety of coloured shirts and have the participants rate them. This study suggests that regardless of the ever-changing opinions of society and conscious global effort to remain politically correct, stereotypes regarding race and colour in general (Williams et al., 1970), are still very much used with regards to aggression and trustworthiness.
The study conducted gave sufficient results for the research questions but there are other areas that could be investigated if the study were to be replicated. In addition to increasing the types of ethnicity, including more colours of shirt would further determine if a significant difference is maintained and whether the colour of one’s shirt has an effect on impression formation. Furthermore, the quality of the stimuli needs to be addressed. The photos of the black model in particular were not clear and blurry and this may have had a negative effect on participants’ ratings. Also the pictures were taken in a bedroom setting and the background in the photos may have affected the participants’ impressions. Having a clear background will thwart that effect. It may also be of interest to separate participants in future studies by gender and age and see if stereotypes are either gender or age specific.
In conclusion, this study has indicated that colour of clothing certainly has an effect, as does skin colour, on impression formation. There is also evidence that people of differing ethnic origins are more likely to trust people of the same ethnic origin and perceive people from different ethnic origins as more aggressive. The origins of impression formation are easily linked to traditional stereotypical ideas of aggression and fits in well with previous research based on similar topics.
Brehm, S., Kassin, S.M., & Fein, S., (2002). Social Psychology, 5th Edition. New York: Houghton Mifflin Company.
Gleitman, H., Fridlund, A. J., & Reisberg, D. (1999). Psychology, 5th Edition. New York: W. W. Norton & Company.
Vrij, A. (1997). Wearing black clothes: The impact of offenders’ and suspects’ clothing on impression formation. Applied Cognitive Psychology, 11, 47-53.