Content uploaded by Michelle Rheinschmidt Same
Author content
All content in this area was uploaded by Michelle Rheinschmidt Same on Feb 16, 2018
Content may be subject to copyright.
Flash Report
Cytokine responses and math performance: The role of stereotype threat
and anxiety reappraisals
Neha A. John-Henderson ⁎, Michelle L. Rheinschmidt, Rodolfo Mendoza-Denton
Department of Psychology, University of California, Berkeley, USA
HIGHLIGHTS
•Female college students take a math exam described as gender-fair or gender-biased.
•In one condition, participants directed to reappraise physiological arousal.
•Performance on math exam and post-exam levels of the cytokine IL-6 were measured.
•Reappraisal of physiological arousal buffers inflammatory responses to exam across conditions.
•Reappraisal of arousal especially effective buffer of inflammatory responses in stereotype threat condition.
abstractarticle info
Article history:
Received 11 January 2014
Revised 1 October 2014
Available online 30 October 2014
Keywords:
Stereotype threat
Reappraisal
Stress
Gender
Inflammation
This research independently manipulated two potential attenuators of stereotype threat –reappraisal of anxiety
and test framing –to explore their independent and combined effects. Female participants took a difficult math
exam that was described as gender-biased or gender-fair and were told that anxious arousal could positively
impact performance or were given no information regarding arousal. Levels of the cytokine Interleukin-6 (IL-6),
an immune marker of inflammation, were measured in oral mucosal transudate (OMT) both before and after
the exam. Our findings indicate that directing reappraisal of physiological arousal attenuated increases in IL-6
across test framing conditions, and was especially effective under stereotype threat (i.e., gender-biased
test condition). Reappraisal also mapped onto better test performance in the threat condition. Together,
these findings provide insight into the unique and interactive effects of two situational interventions meant to
reduce stereotype threat, indexed here by both physiological and performance-based correlates of threat.
© 2014 Elsevier Inc. All rights reserved.
Introduction
A widely shared stereotype in our society alleges that women possess
weaker mathematical abilities than men (Cheryan, Plaut, Davies, &
Steele, 2009; Swim, 1994). Women risk being judged by this stereotype
and worry that they will confirm negative stereotypes when performing
math tasks; such gender-based stereotype threat has been linked to anx-
iety, cognitive resource depletion, and underperformance (e.g., Johns,
Schmader, & Martens, 2005; Mendoza-Denton, Kahn, & Chan, 2008).
Prior research has shown that reappraising anxious arousal as bene-
ficial to performance yields better performance and more adaptive phys-
iological responses relative to providing no reappraisal instructions
(Jamieson, Mendes, Blackstock, & Schmader, 2010; Jamieson, Nock, &
Mendes, 2012). Targets of negative stereotypes may be especially likely
to benefit from reappraisal, given the heightened state of anxiety and
physiological arousal inherent to the experience of stereotype threat
(Schmader,Johns,&Forbes,2008). Indeed, Johns, Inzlicht, and Schmader
(2008) observed benefits of reappraising anxiety on test performance
among participants under stereotype threat; however, to our knowledge
no research has addressed whether reappraisal strategies also reduce the
physiological consequences associated with stereotype threat. We fill
this gap in the literature by experimentally testing whether reappraisal
of anxiety, specifically among women in math contexts, diminishes both
performance impairments and increases in a marker of inflammation
associated with stereotype threat.
In this research we specifically focused on the effects of reappraisal on
the pro-inflammatory cytokine Interleukin-6 (IL-6). Pro-inflammatory
cytokines such as IL-6 are critical for orchestrating the body's inflammato-
ry response, which is crucial to fighting injury or infection (Parkin &
Cohen, 2001; Segerstrom & Miller, 2004). However, if the inflammatory
response becomes persistent or exaggerated it can lead to a host of
diseases and health conditions. Inflammation is increasingly recognized
as a risk factor for illnesses such as cardiovascular disease, autoimmune
Journal of Experimental Social Psychology 56 (2015) 203–206
⁎Corresponding author at: Department of Psychology, University of California,
Berkeley, 3210 Tolman Hall #1650, Berkeley, CA 94720-1650, USA.
E-mail address: nehajohn@berkeley.edu (N.A. John-Henderson).
http://dx.doi.org/10.1016/j.jesp.2014.10.002
0022-1031/© 2014 Elsevier Inc. All rights reserved.
Contents lists available at ScienceDirect
Journal of Experimental Social Psychology
journal homepage: www.elsevier.com/locate/jesp
disorders, and some cancers (Cesari, Penninx, & Newman, 2003; Nathan
&Ding,2010).
Prior research (e.g., Dickerson, Gable, Irwin, Aziz, & Kemeny, 2009;
Murali, Hanson, & Chen, 2007) suggests shifts in inflammatory cytokine
levels in response to situational stressors (e.g., taking an exam). While
IL-6 in particular can exert both inflammatory and anti-inflammatory
effects (Scheller, Chalaris, Schmidt-Arras, & Rose-John, 2011), prior
research characterizes increases in IL-6 specifically in response to a
stressor as signaling an inflammatory response (Dickerson et al., 2009;
John-Henderson, Rheinschmidt, Mendoza-Denton, & Francis, 2014;
Slavich, Way, Eisenberger, & Taylor, 2010). In this research, we assessed
changes in IL-6 in oral mucosal transudate (OMT). While levels of
inflammatory cytokines in OMT are not a reflection of systemic inflam-
mation, which is assessed through blood (Fernandez-Botran, Miller,
Burns, & Newton, 2011), inflammatory markers in OMT have been
shown to be affected by acute situational stressors (Chiang, Eisenberger,
Seeman, & Taylor, 2012; John-Henderson et al., 2014; Slavich et al., 2010).
Method
Participants and procedure
Ninety-seven female undergraduates at UC Berkeley participated for
partial course credit. Their ages ranged from 18 to 35 years (M=20.88,
SD = 2.77), and the ethnic composition was 53.6% Asian, 21.6% White,
16.4% Hispanic, .06% other, and .02% African American.
This study crossed previously published manipulations of stereotype
threat (Spencer, Steele, & Quinn, 1999) and reappraisal of anxiety
(Jamieson et al., 2010). We manipulated stereotype threat by following
the exact procedure of Spencer et al. (1999).Morespecifically, we
described the math exam as a test of intellectual ability for solving math
problems that had or had not produced gender differences in perfor-
mance. We also varied whether participants received reappraisal instruc-
tions or no information about reappraisal. Using the wording of Jamieson
et al. (2010), the reappraisal instructions encouraged participants to view
arousal and anxiety as helpful to test performance. Participants were ran-
domly and independently assigned to condition. Participants provided
samples of oral mucosal transudate (OMT) at three time points (baseline,
post-exam, and recovery) to assess changes in IL-6 in response to the ex-
perimental manipulations. The experimenters were blind to participants'
condition, as all manipulations occurred via paper-based instructions
embedded in the math examination packet. Participants were given
30 min to complete the test.
Measures
Test performance
We measured the number of correct responses on a 17-item math
exam modeled after the Graduate Record Examination (M= 11.6,
SD = 4.3).
Inflammation measures
We assessed IL-6 levels in OMT. Participants provided a baseline OMT
sample for IL-6 measurement upon arrival to the lab (M= .45 pg/mL,
SD = .30). An OraSure collection device (Epitope, Beaverton, OR) was
placed between the lower cheek and gum for 2 min. The exam lasted
for 30 min, after which a second sample of OMT was taken using the
same method to assess changes in IL-6 (M= 1.98 pg/mL, SD = 2.52).
Athird,final sample was taken 30 min after the second sample to
gauge recovery (M= 1.82 pg/mL, SD = 2.44).
All OMT samples were immediately frozen and stored at −80°°C.
IL-6 concentrations were determined by an enzyme-linkedimmunosor-
bent assay using commercially available kits (R&D systems, Minneapolis,
MN). The intra-assay coefficient of variation (CV) was 6.5% and the
inter-assay CV was 8.7%. Shapiro–Wilk tests (Shapiro & Wilk, 1965)re-
vealed that IL-6 values at each time point were not normally distributed
(baseline: W= .91, pb.001, post-exam: W= .67, pb.001, recovery:
W=.65,pb.001). Thus, following prior research (John-Henderson,
Jacobs, Mendoza-Denton, & Francis, 2013)weaddedaconstantof
one to all raw values (see Osborne, 2002) before applying a log-
transformation to IL-6 values. In addition to measuring IL-6 levels, we
measured levels of total protein in each OMT sample using the BCA
protein assay with bovine serum albumin as the standard (Thermo-
scientific, Rockford, IL) to control for individual differences in salivary
flow rate (Dickerson, Kemeney, Aziz, Kim, & Fahey, 2004). Salivary
flow rate was specific to each sample timepoint, given documented
fluctuations in these rates under acute stress (Bakke et al., 2004). All
total protein samples were run in triplicate following kit instructions.
Body mass index (BMI)
Given its relationship with levels of IL-6 in previous research
(Khaodhiar, Ling, Blackburn, & Bistrian, 2004), we calculated partici-
pants' BMI (M= 21.96, SD = 3.10) from their self-reported height
and weight for use as a covariate in the IL-6 analyses.
Results
IL-6 reactivity
We conducted a general linear model analysis predicting IL-6 levels
as a function of the sample timepoint (3 within-participant levels: base-
line, post-exam, recovery) and our between-participant factors of
test framing (2 levels: gender-biased, gender-fair) and reappraisal
instructions (2 levels: instructions, no instructions). We contrast-coded
condition assignments prior to analyses (for reappraisal instructions:
−.5 = no mention of arousal, .5 = arousal helps performance; for test
framing: −.5 = gender-fair, .5 = gender-biased). We included total
protein levels at each timepoint and BMI as covariates; all were mean-
centered. Based on Mauchly's test of sphericity, χ
2
(2) = 42.79,
pb.001, we applied a Greenhouse–Geisser correction (ε= .72) to esti-
mate our overall effect more conservatively. We report adjusted degrees
of freedom below.
Our analysis revealed the predicted 3-way interaction of time,
test framing, and reappraisal instructions, F(1.43, 121.50) = 10.25,
pb.001 η
p
2
=.11.
1
BMI accounted for a significant amount of variance
between participants, F(1, 85) = 4.39, p=.04,η
p
2
= .05, but total pro-
tein levels at baseline, post-exam, and recovery did not, p'sN.34, all
η
p
2
b01. We broke down the 3-way interaction by looking at the effects
of reappraisal instructions within test framing condition. Specifically,
we examined whether changes in IL-6 could be characterized by linear
and/or quadratic trends over time as a function of reappraisal condition.
In the gender-biased test condition, we observed a significant
simple interaction between time and reappraisal instruction condition
F(1, 85) = 59.75, pb.001, η
p
2
= .41, for the linear effect of time, and
F(1, 85) = 65.32, pb.001, η
p
2
= .44, for the quadratic effect of time.
We conducted simple comparisons at each timepoint to identify signif-
icant differences in IL-6 levels by reappraisal condition, applying a
Bonferroni correction to reduce Type 1 error. As illustrated in Panel A
of Fig. 1, these comparisons indicate no IL-6 differences at baseline,
F(1, 85) = 1.17, p=.28,η
p
2
= .01. However, women in the reappraisal
instruction condition (vs. no instructions condition) had significantly
lower levels of IL-6 during the post-exam, F(1, 85) = 61.30, pb.001,
η
p
2
= .42, and recovery periods, F(1, 85) = 54.15, pb.001, η
p
2
= .39.
In the gender-fair test condition, we observed a significant simple
interaction between time and the reappraisal instruction manipula-
tion, F(1, 85) = 6.85, p= .01, η
p
2
= .08, for the linear effect of time,
and F(1, 85) = 6.32, p= .01, η
p
2
= .07, for the quadratic effect of time.
1
We confirmed that our models hold when using age, subjecti ve general health ,
current illness (1 = yes; 0 = no), and chronic health conditions (1 = yes; 0 = no) as
covariates. Further, we repeated the anal yses while excluding anyone with a physical
illness and/or chronic health condition, and our results hold.
204 N.A. John-Henderson et al. / Journal of Experimental Social Psychology 56 (2015) 203–206
We isolated these divergent trajectories by conducting simple compar-
isons at each timepoint, again using a Bonferroni correction (see Fig. 1,
Panel B). We found no statistically significant IL-6 differences at base-
line according to reappraisal instruction condition, F(1, 85) = 2.91,
p=.09,η
p
2
= .03. However, at the post-exam and recovery timepoints,
receiving reappraisal instructions yielded lower levels of IL-6 relative
to receiving no instructions, F(1, 85) = 10.83, p= .001, η
p
2
= .11, for
post-exam, and F(1, 85) = 12.66, p=.001,η
p
2
= .13, for recovery.
Looking within women who did not receive reappraisal instructions,
we observed a significant simple interaction between time and test
framing condition, F(1, 85) = 57.58, pb.001, η
p
2
= .40, for the linear
effect of time, and F(1, 85) = 64.16, pb.001, η
p
2
= .43, for the quadratic
effect of time. Women in the gender-biased condition had higher levels
of IL-6 at post-exam, F(1, 85) = 71.23, pb.001, η
p
2
= .46, and recovery,
F(1, 85) = 60.93, pb.001, η
p
2
= .42, than those in the gender-fair
condition.
Among women who received reappraisal instructions, we observed
a significant simple interaction between time and testing framing
condition, F(1, 85) = 6.90, p= .01, η
p
2
= .08, for the linear effect of
time, and F(1, 85) = 6.74, p= .01, η
p
2
= .07, for the quadratic effect of
time. However, we found smaller effects of test framing at post-exam,
F(1, 85) = 15.65, pb.001, η
p
2
= .16, and recovery, F(1, 85) = 16.47,
pb.001, η
p
2
= .16, than we did in the no reappraisal condition.
In summary, across both test framing conditions, women who
received reappraisal instructions (vs. no instructions) showed lower
levels of IL-6 following a difficult math exam. We observed larger effects
of reappraisal in the gender-biased condition. Further, we observed
larger effects of stereotype threat when no reappraisal instructions
were given.In the gender-biased condition, women who did not receive
reappraisal instructions had higher peak levels of IL-6 than all other
women (see Fig. 1), suggesting that these women were especially
vulnerable to rises in IL-6 following the exam.
Test performance
We observed a significant zero-order correlation between post-exam
inflammation levels and exam performance, r(97) = −.49, pb.001. We
observed a non-significant partial correlation between post-exam
inflammation levels and exam performance, after controlling for our
experimental conditions, BMI, IL-6 levels at baseline, and total protein
levels at baseline and post-exam, r(86) = −.08, p= .46.
We conducted an ANOVA with test framing condition and reap-
praisal instruction condition as independent variables to see whether
our experimental manipulations interacted to predict test performance.
We observed main effects of both the test framing manipulation,
F(1, 93) = 47.38, pb.001, η
p
2
= .34, and the reappraisal instruction
manipulation, F(1, 93) = 9.34, p= .003, η
p
2
= .09. These main effects
were qualified by a marginally significant interaction between the two
manipulations, F(1, 93) = 3.34, p= .07, η
p
2
= .04. As illustrated in
Fig. 2, participants performed significantly better under stereotype threat
when given reappraisal instructions versus no instructions, F(1, 93) =
12.05, p= .001, η
p
2
= .12. Women's performance in the gender-fair test
framing condition did not differ as a function of reappraisal instructions,
F(1, 93) = .75, p= .39, η
p
2
= .008. Thus, reappraisal was most beneficial
to performance under stereotype threat.
We also tested whether our stereotype threat manipulation was more
potent in the absence of reappraisal instructions. Looking at women who
did not receive reappraisal instructions, we observed worse performance
in the gender-biased condition compared to the gender-fair condition,
F(1, 93) = 38.33, pb.001, η
p
2
= .29. This difference by test framing
condition was indeed less pronounced among women who received
reappraisal instructions, F(1, 93) = 12.65, p= .001, η
p
2
=.12.
General discussion
Our experimental findings provide evidence that reappraisal of
anxiety yields better performance and lower IL-6 activation under
stereotype threat relative to receiving no reappraisal instructions. We
extend Jamieson et al. (2010) by exploring the benefits of reappraisal
among negatively stereotyped individuals, who have known vulnerabil-
ities to under-performance and increases in IL-6 under stereotype threat
(John-Henderson et al., 2014).
We might have expected that reappraisal helps performance in any
testing situation, given that reappraisal of anxiety has been found to
help even members of non-stigmatized groups (e.g., men; Jamieson
et al., 2010). Here, we did not find that reappraisal boosted performance
relative to no reappraisal instructions when the test was framed as
gender-fair. The discrepancy in our findings might be explained by the
fact that Jamieson et al.'s (2010) research did not make stereotypes
or gender differences salient. By contrast, our findings suggest that a
Fig. 1. Average IL-6 levels by time of sample (x-axis), reappraisal instructions (separate
lines), and test framing (Panels A vs. B). IL-6 values have been transformed by adding a
constant of 1 and then performing a log-transformation. Slopes between timepoints are
not statistically significant (pN.05)unless notedabove the slope.All pvalues were adjusted
with a Bonferroni correction. Error bars represent standard error. *** pb.001.
Fig. 2. Performance ona mathematics examinationas a function of the reappraisal instruc-
tion condition and the test framing condition. Error bars represent standard error.
205N.A. John-Henderson et al. / Journal of Experimental Social Psychology 56 (2015) 203–206
gender-fair message may be enough, even without reappraisal, to release
women from the cognitive intrusions associated with stereotype threat.
Limitations and future directions
A limitation of our research is that we did not inquire about the
smoking behavior of our participants, a possible predictor of oral
inflammation
2
.Givenourfindings, it will be important for future research
to examine whether the patterns observed here emerge via measure-
ment of inflammation in blood plasma, so as to examine more directly
the implications of this research for longer-term health outcomes and
vulnerabilities.
Our findings might also align with work showing increases in sali-
vary alpha amylase (sAA), reflecting increased sympathetic nervous
system activation when reappraisal instructions are given in testing
contexts (Jamieson et al., 2010). In addition, based on a large body of
literature showing that cortisol levels increase in response to naturalis-
tic acute stressors (e.g., Weekes et al., 2006), we would expect to find
that changes in cortisol would parallel the increase in IL-6 in this
research. An important extension of this work will be to examine the
simultaneous or differential trajectories of these physiological systems.
This type of multi-modal investigation would allow for amore thorough
understanding of the mechanisms that explain the variance in IL-6
levels documented here.
Acknowledgments
The first two authors contributed equally to the development of this
project and the ideas within it. We are grateful for the datacollection as-
sistance of Leeran Baraness, Hardev Chhokar, Mary Cunningham, and
Meital Mashash. We would also like to thank Darlene Francis for her
feedback during initial stages of the project and for her support in
conducting the assays for this research. This work was supported by
the National Science Foundation Graduate Research Fellowship
awarded to Michelle L. Rheinschmidt and by the National Science Foun-
dation Award HRD-1306709.
References
American College Health Association (2013). American College Health Association-National
College Health Assessment (ACHA-NCHA). Available from. http://www.acha-ncha.org/
Bakke, M., Tuxen, A., Thomsen, C.E., Bardow, A., Alkjaer, T., & Jensen, B.R. (2004). Salivary
cortisol level, salivary flow rate and masticatory muscle activity in response to acute
mental stress: A comparison between a ged and young women. Gerontology,50,
383–392. http://dx.doi.org/10.1159/000080176.
Cesari, M., Penninx, B.W., & Newman, A.B. (2003). Inflammatory markers and cardiovascular
disease (The Health, Aging and Body Composition [Health ABC] Study). American Journa l
of Cardiology,92, 522–528. http://dx.doi.org/10.1016/S0002-9149(03)00718-5.
Cheryan, S., Plaut, V.C., Davies, P.G., & Steele, C.M. (2009). Ambient belonging: How stereo-
typical cues impact gender participation in computer science. Journal of Personality
and Social Psychology,97(6), 1045–1060. http://dx.doi.org/10.1037/a0016239.
Chiang, J.J., Eisenberger, N.I., Seeman, T.E., & Taylor, S.E. (2012). Negative and competitive
social interactions are related to heightened proinflammatory cytokine activity.
Proceedings of the National Academy of Sciences,109, 1878–1882. http://dx.doi.org/
10.1073/pnas.1120972109.
Dickerson, S.S., Gable, S.L., Irwin, M.R., Aziz, N., & Kemeny, M.E. (2009). Social-evaluative
threat and proinflammatory cytokine regulation: An experimental laboratory investi-
gation. Psychological Science,20, 1237–1244. http://dx.doi.org/10.1111/j.1467-9280.
2009.02437.x.
Dickerson, S.S., Kemeney, M.E., Aziz, N., Kim, K.H., & Fahey, J.L. (2004). Immunological ef-
fects of ind uced shame and g uilt. Psychosomatic Medicine,66,124–131.
Fernandez-Botran, R., Miller, J.J., Burns, V.E., & Newton, T.L.(2011). Correlations among in-
flammatory markers in plasma, saliva, and oral mucosal transudate in post-
menopausal women w ith past intimate partner violence. Brain, Behavior, and
Immunity,25,314–321. http://dx.doi.org/10.1016/j.bbi.2010.09.023.
Jamieson, J.P., Mendes, W.B., Blackstock, E., & Schmader, T. (2010). Turning the knots in
your stomach into bows: Reappraising arousal improves performance on the GRE.
Journal of Experimental Social Psychology,46,208–212. http://dx.doi.org/10.1016/
j.jesp.2009.08.015.
Jamieson,J.P., Nock, M.K., & Mendes, W.B. (2012). Mindover matter: Reappraising arousal
improves cardiovascular and cognitive response to stress. Journal of Experimental
Psychology: General,141,417–422. http://dx.doi.org/10.1037/a0025719.
John-Henderson, N., Jacobs, E.G., Mendoza-Denton, R., & Francis, D.D. (2013). Wealth,
health and the moderating role of impli cit social class bias. Annals of Behavioral
Medicine,45,173–179. http://dx.doi.org/10.1007/s12160-012-9443-9.
John-Henderson, N., Rheinschmidt, M.L., Mendoza-Denton, R., & Francis, D.D. (2014). Per-
formance and inflammation outcomes are predicted by different facets of SES. Social
Psychological and Personality Science,5,301–309. http ://dx.doi.org/10.1177/
1948550613494226.
Johns, M., Inzlicht, M., & Schmader, T. (2008). Stereotype threat and executive resource
depletion: Examining the influence of emotion regulation. Journal of Experimental
Psychology: General,137,691–705. http://dx.doi.org/10.1037/a0013834.
Johns, M.,Schmader, T., & Martens, A. (2005). Knowing ishalf the battle: Teaching stereo-
type threat as a me ans of improving women's math performance. Psychological
Science,16,175–179. http://dx.doi.org/10.1111/j.0956-7976.2005.00799.x.
Khaodhiar, L., Li ng, P.R., Blackb urn, G.L., & Bistr ian, B.R. (2004) . Serum levels of
interleukin-6 an d C-reactive protein correlate with bod y mass index across the
broad range of obesity. Journal of Parental and Enteral Nutrition,28,410–415. http://
dx.doi.org/10.1177/0148607 104028000410.
Mendoza-Denton, R., Kahn, K., & Chan, W. (2008). Can fixed views of ability boost perfor-
mance in the context of favorable stereotypes? Jour nal of Experimental Social
Psychology,44,1187–1193. http://dx.doi.org/10.1016/j.jesp.2008.03.005.
Murali, R., Hanson, M.D., & Chen, E. (2007). Psychological stress and its relationship to
cytokine s and inflammatory disease. In N.P. Plotnikoff, R.E. Faith, & A.J. Murgo (Eds.),
Cytokines: Stress and Immunity(p p. 29–49) (2nd ed.). Boca Raton, FL: Taylor & Francis
Group.
Nathan, C., & Ding, A. (2010). Nonresolving inflammation. Cell,140,871–882. http://dx.
doi.org/10.1016/j.cell.2010.02.029.
Osborne, J. (2002). Notes on the use of data transformations. Practical Assessment,
Research & Evaluation,8(6) (Retrieved June 4, 2014 from http:/ /PAREonline.net/
getvn.asp?v=8&n=6).
Parkin, J., & Cohen, B. (2001). An overview of the immune system. Lancet,357, 1777–1789.
Scheller, J., Chalaris, A., Schmidt-Arras, D., & Rose-John, S. (2011). The pro- and anti-
inflammatory effects of the cytokine Interleukin-6. Biochimica et Biophysica Acta,
1813,878–888. http://dx.doi.org/10.1016/j.bbamcr.2011.01.034.
Schmader, T., Johns, M., & Forbes, C. (2008). An integrated process model of stereotype
threat effects on performance. Psychological Review,115(2), 336. http://dx.doi.org/
10.1037/0033-295X.115.2.336.
Segerstrom, S.C., & Miller, G.E. (2004). Psychological stress and the immune system: A
meta-analytic study of 30 years of inquiry. Psycho logical Bulleti n,130,601–630.
http://dx.doi.org/10.1037/0033-2909.130.4.601.
Shapiro, S.S., & Wilk, M.B. (1965). An analysis of variance test for normality. Biometrika,
52,591–611.
Slavich, G.M., Way, B.M., Eisenberger, N.I., & Taylor, S.E. (2010).Neural sensitivity tosocial
rejection is associated with inflammatory responses to social stress. Proceedings of the
National Academy of Sciences of the United States of America,107, 14817–14822. http://
dx.doi.org/10.1073/pnas.1009164107.
Spencer,S.J., Steele, C.M.,& Quinn,D.M. (1999). Stereotype threat and women's mathper-
formance. Journal of Experimental Social Psychology,35,4–28.
Swim, J.K. (1994). Pe rceived versus meta-analytic effe ct sizes: An asses sment of the
accuracyof gender stereotypes. Journal of Personality and Social Psychology,66,21–36.
Weekes, N., Lewis, R., Patel, F., Garrison-Jakel, J., Berger, D.E., & Lupien, S.J. (2006). Exam-
ination stress as an ecol ogical inducer of cor tisol and psychological responses to
stress in undergraduate students. Stress,9,199
–206. http://dx.doi.org/10.1080/
1025389060102.
2
Only 2.3% of female undergraduates on campus report using cigarettes 10 or more
days per month (American College Health Association, 2013).
206 N.A. John-Henderson et al. / Journal of Experimental Social Psychology 56 (2015) 203–206