Introduction: When making a lasting first impression we often hear that image is everything. Historically, women endured the most scrutiny by society for their image, especially when women have visible tattoos or other body modifications that deviate from social constructs of beauty. In recent years, Visual Body Modifications (VBM) such as tattoos have been growing
When making a lasting first impression we often hear that image is everything. Historically, women endured the most scrutiny by society for their image, especially when women have visible tattoos or other body modifications that deviate from social constructs of beauty. In recent years, Visual Body Modifications (VBM) such as tattoos have been growing in popularity, as a nationwide poll reported in 2012 that over the past 3 years, tattoo artists have done at least 20% more tattoos on younger women than two years ago (Fong, 2012). Moreover, the state of California has seen similar trends in the frequency of VBM on young women increasing, from 3-4 a week, now to once a day (Fong, 2012). Common pejoratives such as “tramp stamp”, “target” and “bulls-eye”, are frequently used when referring to women with visible tattoos, and previous research has found that women with visible tattoos have been ranked by men as more promiscuous, more likely to abuse alcohol and lacking any motivation skills (Wohlrab, Fink, Kappeler, & Brewer, 2009). As tattoos are on the rise in young women, to what degree does one visible tattoo impact the positive personality characteristics upon her first impression ?
A recent observational study was conducted that demonstrated the negative effect tattoos had on the influencing the perception of men’s first impression of a woman. Researchers used a female model, wearing a provocative red dress, and measured what impact having a tattoo would have on the frequency of men that approached her(Vinita Mehta Ph.D., Ed.M., 2013). The female model was first observed at a bar by herself, without any VBM, and the researchers counted the amount of times heterosexual men approached the young woman and propositioned her for a date (Vinita Mehta Ph.D., Ed.M., 2013) The researchers then applied a fake VBM on the lower back of the same model, wearing the same dress and placed her at the bar, to measure what effect the tattoo would impact the quantity of times that heterosexual men would proposition her with the fake tattoo (Vinita Mehta Ph.D., Ed.M., 2013). To the researcher’s surprise, a disproportionate amount of men approached the young women with the applied VBM, compared to her previously without the presence of a VBM. The researchers later discovered by interviewing the male subjects, was that those who approached the young woman with the tattoo assumed she more likely to sleep with them on a first date (Vinita Mehta Ph.D., Ed.M., 2013). The participants also disclosed that because of the tattoo their first impression of her was that she would be a better fit for a one-night stand lacked positive personality characteristics they would be seeking in a long-term partner (Vinita Mehta Ph.D., Ed.M., 2013).
Consistent with the previous study, there is additional evidence indicating that women with visible tattoos are viewed less attractive and more negatively by men. Researchers at a community college in the Midwest asked male students to rate a photo of a woman based on her attractiveness (Resenhoeft et al., 2008). The researchers then altered one photo of the same woman, and added one small butterfly tattoo to see what effect it would have on first impression of her positive personality characteristics (Resenhoeft et al., 2008). The sample was divided in half and only exposed to one condition, one photo with a VBM and one without. The participants were asked to rate the woman in the photograph based of their first impression of 13 personality characteristics using a 5-point scale for each (Resenhoeft et al., 2008). The findings came back significant as the photo without the tattoo was rated highly more attractive, less likely to abuse alcohol and had more positive personality attributes (Resenhoeft et al., 2008). They found that participants’ exposed to the condition with the tattoo, viewed her more negatively, especially in regard to physical appearance and personality traits in comparison to the condition without the tattoo. The participants rated the model without the tattoo as significantly more athletic, attractive, motivated, honest, mysterious, religious and intelligent further substantiating the impact a tattoo has on influencing the first impression of a woman (Resenhoeft et al., 2008).
While previous studies have found strong implications that women with tattoos are viewed more negatively upon first impression, the samples that were primarily used were heterosexual, white men, from politically conservative areas. In this present study, we will be analyzing to what effect a tattoo has on the first impression of a woman, but adding categorical variables such as age, gender, sexual orientation and religion, as potential confounds neglected in previous research. This research aims to see how we as a society innately judge another person solely off their appearance. Our hypothesis leads this study to measure the main effect a VBM has on influencing the overall perception of a personality characteristics in a woman, but also looking for interaction effects caused by the added categorical factors that could contribute to mean differences between the two groups.
Ho: μ in Group 0 = μ in Group 1 The tattoo has no main effect on the perception of personality between the two groups and is independent from categorical factors.
HA : μ in Group 0 ≠ μ Group 1 The tattoo has a main effect on the perception of personality between the two groups and is dependent on the categorical factors.
Participants: The participants of this study were (n=93) residents of California (51 Women, 42 Men) coming from the University of California Berkeley Media Studies Program and the geographic regions of the Greater Bay Area, Los Angeles area and Central Valley Inland communities. The mean age of the sample was (M= 26.6 years, SD= 1.26, range 18-65) of the total sample 71 (76.3%) identified as Heterosexual, 15 (16%) identified as Gay and 7 (7.7%) identified as Lesbian. Over half the participants (54%) were not religious and just under a quarter (24.7%) were a little religious, (13.97%) some religion and the remaining (7.7%) were not religious at all (see Figure 1).
Materials: Replicating the measures utilized by (Resenhoeft et al., 2008), we took two photos of a female model, one unaltered and one altered with a butterfly tattoo on the models chest (see Figure 2) and created two online identical surveys, one for Group 0 and one for Group 1, on Qualtrics. We decided to use the butterfly tattoo in particular, as previous research has proven it to be the most neutral and popular tattoo amongst young women (Resenhoeft et al. 2008). The placement of the tattoo was also taken into consideration, as psychological research has proven that tattoos near the face, and particularly on the chest of women, are viewed the most negatively by society (Zestcott, Bean, & Stone, 2015).
To control for any confound variables that might influence the effect of the tattoo on the perception of her personality, the model made no drastic facial expressions, such as smiling or frowning. We then framed her as an up and coming actress to participants to disassociate her from having any personal ties to any social institutions, and to further control for any extraneous influences that might effect the perception of the participants. We phrased in the survey that the participants’ opinion was important when rating this “actress”, and wanted participants to view her as objectively as possible and have only the tattoo influence their first impression.
Procedures: We then recruited participants by creating two anonymous hyperlinks for the two surveys. First we had undergraduate Media Studies students, at the University of California Berkeley, complete the surveys exposed to the separate conditions. We then posted a link on Facebook that randomly assigned participants evenly between the two photos to assure random sampling occurred. The link acquired participants from a variety of geographic locations around California; including the SF/Bay Area, Los Angeles and Orange County metropolitan areas and the Sacramento metropolitan area. After the link was posted online, we continued to collect more samples by asking people to take our survey on a mobile device that generated evenly split conditions between the two photos and the researcher didn’t know which condition the sample would be taking.
Next, we stood in front of a Target on a Sunday afternoon in Walnut Creek, CA and politely asked people to take our quick survey during peak shopping hours. The sample was a mix of families and young adults, some were parents and were primarily Caucasian. We then sampled the staff at plastic surgeon’s office in Los Angeles, CA. The sample was also randomly exposed to the two photos and included nurses, doctors and receptionists who were both male and female. Finally, we exposed an older demographic to the two separate surveys at family members 95th birthday party in Orange County CA. The sample was primarily heterosexual, conservative, most were parents and came from European and Jewish background. In every setting the tattoo was not mentioned, we only asked if they could please rate the up and coming actress to maintain the consistency in the responses.
Using a 5-point scale scale similar to the Resenhoeft study (2008), participants were asked to rate the woman in the photograph on 6 common personality characteristics (Resenhoeft et al., 2008). A score of 1 (Strongly Disagree) being the most negative score on perception; and a score of 5 (Strongly Agree) was the most positive. Personality characteristics used to operationally define our measurement of first perception were; honesty/trustworthy, positive role model, attractive, intelligent, motivated and charismatic. The personality scale consisting of 6 items had measure of reliability at (α = .81) for the total, and for each personality measurement at (α > .70 (See Table A). On the demographic questionnaire, we asked participants to select and age category (18-24, 25-34, 35-44, 45-54, 55-64, 65+), indicate their gender, sexual orientation and how religious they are (ranging from none to a lot). We asked participants to fill out the demographic information after they rated the two conditions of the photo they were exposed to.
The tables below (see Table 2-3) demonstrate through random sampling we acquired fairly similar samples in the two conditions of group 0 and group 1 (no tattoo and tattoo). Through random sampling we also acquired a fairly representative sample of the population of California in the categorical questions of gender and sexual orientation, as the state of California and our sample has roughly 51% female and 49% male, with 92% identifying as heterosexual and the other 8% identifying as gay or lesbian (Gary J. Gates et al., 2012)
|Mean||Standard Deviation||Cronbach’s Alpha|
|Honesty and Trustworthy||3.0||0.81||0.75|
|Positive Role Model||2.9||0.84||0.75|
Table 2: Group 0: Condition Without Tattoo
Table 3: Group 1: Condition with the Tattoo
Results: We first conducted an independent sample t-test to compare the overall main effect of the woman’s personality between the two groups and the conditions. We found some significance between the unaltered photo group 0 (M=3.35, SD=0.51) and the altered photo group 1 (M=3.10, SD= 0.69) with t(91)= 2.0268, p= .04561, Cohen’s D=.40. We can reject the null hypothesis and conclude the tattoo causes a significant difference in the main effect between the two groups with a Power(β) of 97.8% this was not due to chance. Since our participants were primarily within the same age range and religiousness, we discarded these from our models as they had no main effect or interaction effect with the data.
We then conducted an independent sample t-test only looking at women, both heterosexuals and lesbians between group 0 and 1. For females in group 0 (M=3.28) and for group 1 (M= 3.37), t(49)= -.57251, p = .57, 95%CI[-0.3862850,0.2149887], d= -0.16357. Fail to reject the null hypothesis and conclude there was no statistical significance on perception between the two groups of women.
Next, we conducted a second independent sample t-test looking at heterosexual and homosexual men combined exposed to the two conditions. There was significant effect found in men between group 0 as we found (M= 3.42) and in group 1 (M=2.72), t(40)= 3.8472, p=0.0042, 95%CI[0.2850824, 1.061218]. Reject the null hypothesis and conclude a significant effect the tattoo had on the perception of men, both gay and straight, with the Power (β) of 99.9% not due to chance. The effect size for this analysis (d = 1.44838) was found to exceed Cohen’s (1988) convention for a large effect (d = .80).
After these findings, we ran a bivariate and multivariate linear regression test to examine how gender and sexual orientation predicted the personality perception between the two groups. We did find some significance in the bivariate linear with a ( Multiple R2=.6022, Adjusted R2 = 0.03268 F(4,91) = 4.108, p<. 001) p-value: 0.04561) But failed to reject the null hypothesis looking at he differences in the two groups. Looking at the difference between Multiple R and the R Squared value, we found redundant co-variation between gender and sexual orientation that was moderating the effect on the personality perception indicating an interaction effect caused by gender.
We then ran a moderated regression linear model, to measure the interaction effects of gender and sexual orientation and the main effect of the tattoo on personality perception. Since our sample of gay men (n=15) and lesbians (n= 7) were small, we combined them into gender to run the test. We found a significant effect between groups, (F(3, 89)= 6.832 >4.04, p.< .01, R2= 0.1872) Reject the null hypothesis and conclude that the main effect of the tattoo is dependent on the categorical factors of gender and sexual orientation. The overall personality perception in the tattoo group was rated drastically lower by gay and straight men with (b=-.79389, 95% CI[-1.2601829 ,-0.3275894].
Finally, to analyze the interaction effect between the groups, we then ran all of the linear model through an Anova two factor design test to analyze the main effect of the tattoo, and the interaction effects between gender and sexual orientation combined. With (F (1, 89) = 11.444 > 6.96 p.< .01, p = .00107, η2= .1046) we can reject the null hypothesis and conclude the tattoo had a significant effect on men compared to women dependent on the interaction effect in gender differences in the presence of the VBM (see table below).
|Df Sum Sq Mean Sq F value Pr(>F) eta.squared|
|tat.group 1 1.490 1.490 4.729 0.03230 * .0494 Main Effect A|
|gender 1 1.362 1 .362 4.322 0.04050 * .0394 Effect of B|
|tat:gender 1 3.605 3.605 11.444 0.00107 ** .10456 Effect of A*B|
|Residuals 89 28.038 0.315|
See appendix for graphs demonstrating the adjustments in our error for accounting for the interaction effect
The findings in this study demonstrated some interesting ways we as men and women make value judgments solely based off the first impression of a woman’s appearance. Something as small as a tattoo drastically altered the positive characteristics associated with our subject mainly in gay or straight men. After running the interaction linear model, we noticed our R squared value increased and residual error decreased proving the need for the interaction test with our categorical variables. Through our Anova test we found this model to be the best fit for analysis in the data, as we saw our F max test grow with significance and a larger effect size in Eta Squared, especially in men exposed to the tattoo.
While these findings found significance and rejected our initial hypothesis, it was only within men and they weren’t much different from the previous studies mentioned earlier (Guéguen, 2013). These findings could implicate the importance men place on exterior beauty and shed some light onto the unconscious way in which they view the opposite sex, gay or straight. Since the sample came of men and women came California, known to be more progressive, the results in this study could imply that we as a society have more work to do in the equal treatment of women.
Another interesting finding was not only the significance in men, but an unexpected trend in women as the mean in Group 1 had a slight increase. While we can’t make conclusive evidence about this trend, it would be interesting to explore further as well as many other factors, leading to improvements to our study.
Limitations: While this study produced significant results, there were some major limitations and weaknesses in our categorical data. Most of our sample came from the same age demographics between 18-35. This categorical had to be omitted as it wasn’t relevant to the research provided. Another major limitation was in our sexual orientation sample, as only (n=7) lesbians responded and (n=15) gay men. The majority of our sample were not religious at all, which also had to be admitted as it was over fitting our models. In future research accounting for these differences could provide stronger inferential statistics and more complexity, where our data could not.
Other categorical data we failed to collect was demographic information regarding ethnic backgrounds. Our subject was a white, young female. Our study failed to recognize the variability in which other cultures value beauty in a woman, or how the evaluate first impressions in a larger social context. Background information about participants or more qualitative questions and research would have strengthened our findings post. Further research exploring these variables could also yield more variation in data and stronger inferential statistics that couldn’t be implemented in this study.
Given the limitations, this study could provide some insight into the way we as a society still make value judgments based off a woman’s appearance. This is important research to continue as women in modern society are still struggling to receive equal opportunity. Our study only framed the subject as an actress; but suggested future research could explore an observation study and how a tattoo effects or appearance effects a more than qualified woman in a business or educational from obtaining employment. Future research could look not only into this, but also measure the differences between men and women in the same setting, to see if men are effected by having a VBM as much as women in obtaining employment.
This study has made us see that first impressions do matter and the extent we make unconscious value judgments could be problematic. If one tattoo could drastically alter someone’s positive personality characteristic, it would be relevant to see the other superficial ways we make value judgments that could be limiting a whole group of people from having a chance to make that coveted lasting impression.
Cash, T. F. (1990). The psychology of physical appearance: Aesthetics, attributes, and images. In T. F. Cash & T. Pruzinsky (Eds.), Body images: Development, deviance, and change. New York: Guilford Press
Fong, M. (2012, May 24). Hipster Jeans, Crop Tops Boost Lower-Back Tattoos. The Wall Street Journal.
Gates, G. (2012). Geography of the LGBT Population. International Handbook on the Demography of Sexuality International Handbooks of Population, 229-242.
Guéguen, N. (2013). Effects of a Tattoo on Men’s Behavior and Attitudes Towards Women: An Experimental Field Study. Archives of Sexual Behavior, 42(8), 1517–1524. http://doi.org/10.1007/s10508-013-0104-2
Mehta Ph.D, V. (2013). How do people view women with tattoos? Psychology Today. Retrieved December 5, 2015, from https://www.psychologytoday.com/blog/head-games/201305/how-do-people-view-women-tattoos
Perez, J. (2011). Timeless Ideals: Images of Women in Magazine Advertising. International Journal Of The Image, 1(3), 187-199.
Resenhoeft, A., Villa, J., & Wiseman, D. (2008). Tattoos Can Harm Perceptions: A Study and Suggestions. Journal Of American College Health, 56(5), 593
Swami, V., & Furnham, A. (2007). Unattractive, promiscuous and heavy drinkers: perceptions of women with tattoos. Body Image, 4(4), 343–52. http://doi.org/10.1016/j.bodyim.2007.06.005
Wohlrab, S., Fink, B., Kappeler, P. M., & Brewer, G. (2009). Perception of human body modification. Personality and Individual Differences, 46(2), 202–206. http://doi.org/10.1016/j.paid.2008.09.031
Zestcott, C. A., Bean, M. G., & Stone, J. (2015). Evidence of negative implicit attitudes toward individuals with a tattoo near the face. Group Processes & Intergroup Relations. http://doi.org/10.1177/1368430215603459