Self‐affirmation reduces the socioeconomic attainment gap in schools in England

Abstract

Background
Studies in the United States show that school students from some ethnic backgrounds are susceptible to stereotype threat, that this undermines their academic performance, and that a series of virtually zero‐cost self‐affirmation writing exercises can reduce these adverse effects. In England, however, socioeconomic status (SES) is a much stronger predictor of academic success than is ethnic background.

Aims
This study investigates whether self‐affirmation writing exercises can help close the SES attainment gap in England by increasing the academic performance of low‐SES (but not higher‐SES) school students.

Sample
Our sample consisted of students aged 11–14 in a secondary school in southern England (N = 562); of these, 128 were eligible for free school meals, a proxy for low SES.

Methods
Students completed three short writing exercises throughout one academic year: those randomly assigned to an affirmed condition wrote about values that were important to them, and those assigned to a control condition wrote about a neutral topic.

Results
On average, the low‐SES students had lower academic performance and reported experiencing more stereotype threat than their higher‐SES peers. The self‐affirmation raised the academic performance of the low‐SES students by 0.38 standard deviations but did not significantly affect the performance of the higher‐SES students, thus reducing the SES performance gap by 62%. The self‐affirmation also reduced the level of stress reported by the low‐SES students.

Conclusions
The benefits of this virtually zero‐cost intervention compare favourably with those of other interventions targeting the SES academic attainment gap.

Full text

Running head: SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP
Self-affirmation reduces the socioeconomic
attainment gap in schools in England
Ian Robert Hadden1,2*, Matthew John Easterbrook2, Marlon
Nieuwenhuis2, Kerry Jane Fox3, and Paul Dolan1
1 London School of Economics and Political Science, UK
2 University of Sussex, Brighton, UK
3 University of Brighton, UK
* Corresponding author information: Ian Hadden, School of Psychology, University
of Sussex, Falmer, Brighton, BN1 9QH, UK (i.hadden@sussex.ac.uk).
This is the peer reviewed version of the following article: Hadden, I. R., Easterbrook, M. J.,
Nieuwenhuis, M., Fox, K. J., & Dolan, P. (2019). Self-affirmation reduces the socioeconomic
attainment gap in schools in England. British Journal of Educational Psychology, which has
been published in final form at https://doi.org/10.1111/bjep.12291. This article may be used
for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of
Self-Archived Versions.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 2
Abstract
Background. Studies in the United States show that school students from some ethnic
backgrounds are susceptible to stereotype threat, that this undermines their academic
performance, and that a series of virtually zero-cost self-affirmation writing exercises
can reduce these adverse effects. In England, however, socioeconomic status (SES) is
a much stronger predictor of academic success than is ethnic background.
Aims. This study investigates whether self-affirmation writing exercises can help
close the SES attainment gap in England by increasing the academic performance of
low-SES (but not higher-SES) school students.
Sample. Our sample consisted of students aged 11–14 in a secondary school in
southern England (N = 562); of these, 128 were eligible for free school meals, a proxy
for low SES.
Methods. Students completed three short writing exercises throughout one academic
year: those randomly assigned to an affirmed condition wrote about values that were
important to them, and those assigned to a control condition wrote about a neutral
topic.
Results. On average, the low-SES students had lower academic performance and
reported experiencing more stereotype threat than their higher-SES peers. The self-
affirmation raised the academic performance of the low-SES students by 0.38
standard deviations but did not significantly affect the performance of the higher-SES
students, thus reducing the SES performance gap by 62%. The self-affirmation also
reduced the level of stress reported by the low-SES students.
Conclusions. The benefits of this virtually zero-cost intervention compare favorably
with those of other interventions targeting the SES academic attainment gap.
Keywords
stereotype threat; self-affirmation; socioeconomic status; attainment gap; schools

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 3
Self-affirmation reduces the socioeconomic
attainment gap in schools in England
Self-affirmation is a theoretically-precise psychological intervention that has been
shown to improve the academic performance of students who are negatively
stereotyped and thus experience stereotype threat. There is extensive evidence that
stereotype threat can inhibit the ability of members of negatively stereotyped groups
to perform well at relevant tasks (see meta-analyses Lamont, Swift, & Abrams, 2015;
Picho, Rodriguez, & Finnie, 2013). Studies in the United States have suggested that
stereotype threat can significantly reduce the academic performance of Black and
Latino school students, and is estimated to account for 20-28% of the attainment gap
between these groups and European Americans (Walton & Spencer, 2009). We are
interested in how stereotype threat translates from schools in the United States to
those in England. This is a crucial question because, in contrast with the United
States, poor educational outcomes in England are associated much more strongly with
low socioeconomic status (SES) than they are with ethnicity.
To investigate stereotype threat in England, we draw on evidence demonstrating that
self-affirmation interventions can effectively reduce its adverse effects on the
academic performance of underachieving groups (e.g. Cohen, Garcia, Apfel, &
Master, 2006; Sherman et al., 2013). We thus conduct the first ever study in a school
in England that investigates whether a self-affirmation intervention can raise the
performance of low-SES school students and thus reduce the academic attainment gap
between them and their higher-SES peers. We also investigate psychological
processes that might underpin any effects of the self-affirmation.
Background
Educational inequalities. Educational inequalities rightly receive much
attention in academic research and policy formulation. In the United States, there are
large gaps in educational outcomes between White students and lower-performing
minority ethnic groups, in particular Black and Latino students, and between students
of lower and higher socioeconomic status (SES). Both these gaps are substantial at
ages 13-14: the SES gap is of a similar magnitude to the corresponding gap between

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 4
Black students and White students (National Assessment of Educational Progress,
2016)
1
.
In England, by contrast, the largest education gap by far is associated with SES. As a
result, this gap is a major focus of educational policy spearheaded by the £2.4bn per
annum Pupil Premium initiative (Foster & Long, 2017). Only 34% of school students
in England who are eligible for free school meals (FSM), an indicator of low SES,
meet a standard threshold measure of academic attainment, compared with 61% of
those whose parents are better off, a difference of 27% (Department for Education,
2015). The corresponding differences between the four major ethnic classifications
– Black, White, Asian
2
and Mixed – are much smaller. In fact, the difference between
the lowest performing major ethnic classification (Black students) and the majority
classification (White students) is only 3% (Department for Education, 2015).
Although there are important interactions between SES and ethnicity in both countries
(Brannon, Higginbotham, & Henderson, 2017; Department for Education, 2015;
Harackiewicz, Canning, Tibbetts, Priniski, & Hyde, 2016; Strand, 2014), it is clear
that in England, SES is a significantly stronger predictor of academic success than is
ethnicity, while in the United States, it is not. Why should this be so? We suggest that
stereotype threat might be an important factor.
Stereotype threat. Although definitions vary (Shapiro & Neuberg, 2007), in
broad terms stereotype threat is the sense of threat that people feel in a given context
when they believe that they risk conforming to a negative stereotype about a group of
which they are a member (Steele & Aronson, 1995). Stereotype threat reduces
performance in a wide range of contexts including academic performance. A meta-
analysis (Walton & Spencer, 2009) showed that negatively stereotyped students who
were under conditions of stereotype threat performed on average more than 0.5
1
Details and limitations of these and the other calculations in this section are available in Appendix
S1.
2
In England, ‘Asian’ refers to students mainly of Bangladeshi, Indian and Pakistani backgrounds.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 5
standard deviations worse than those who were not (Walton & Spencer, 2009, Figure
1).
However, other studies have found more limited evidence that stereotype threat
reduces performance. For example, while a meta-analysis showed that stereotype
threat had a significant negative effect of 0.22 standard deviations on girls’
performance in mathematics tests, it also found evidence of publication bias that is
likely to have inflated the effect (Flore & Wicherts, 2015). One potential source of
such publication bias may be null effects in studies that have not taken into account
specific social contexts such as the local relative size of the stereotyped group, an area
that we explore in the following section on self-affirmation theory.
In particular, stereotype threat has been shown to reduce the academic performance of
students of low SES. In one study, French undergraduate students were given a
difficult verbal test (Croizet & Claire, 1998). Low-SES students who were told that
the test was a measure of verbal intelligence performed worse than those who were
told that it was an investigative tool for studying hypotheses about lexical processes,
whereas higher-SES students performed the same regardless of what they were told.
Other studies have found similar effects in low-SES students ranging from 6-year-
olds to college students (Browman, Destin, Carswell, & Svoboda, 2017; Désert,
Préaux, & Jund, 2009; Good, Aronson, & Inzlicht, 2003; Spencer & Castano, 2007).
Self-affirmation theory. The negative effects of stereotype threat are thought
to result from maladaptive responses aimed at protecting a person’s ‘self-integrity’,
their view of themselves as a ‘good and adequate’ person (Cohen & Sherman, 2014).
These maladaptive responses include denial of the threat (Epton, Harris, Kane,
Koningsbruggen, & Sheeran, 2015), concretization of construal levels (Sherman et al.,
2013), increased vigilance (Murphy, Steele, & Gross, 2007), increased stress
(Creswell, Dutcher, Klein, Harris, & Levine, 2013) and allocation of executive
function, thus reducing its availability for other tasks (Hall, Zhao, & Shafir, 2014;
Johns, Inzlicht, & Schmader, 2008). People can avoid these responses by
strengthening their self-integrity in domains unrelated to the threat, a process known
as ‘self-affirmation’. Studies have shown self-affirmation to be an effective buffer
against psychological threat in many situations, including in educational settings such
as colleges (see e.g. Harackiewicz et al., 2013; Kinias & Sim, 2016; Martens, Johns,

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 6
Greenberg, & Schimel, 2006; Miyake et al., 2010; Shapiro, Williams, &
Hambarchyan, 2013; Silverman, Logel, & Cohen, 2013; Tibbetts et al., 2016) and
massive open online courses (Kizilcec, Saltarelli, Reich, & Cohen, 2017).
There is also an increasing body of evidence relating to the effectiveness of self-
affirmation in reducing stereotype threat in schools. Here, studies have found
evidence that self-affirmation increased academic performance in Black and Latino,
but not White, middle school students aged 11 to 14 (Cohen et al., 2006; Sherman et
al., 2013). Follow-up studies have found that positive effects of self-affirmation on
ethnic minority students persist over time. Remarkably, one study found benefits to
college enrolment up to 9 years after the self-affirmation intervention (Goyer et al.,
2017) and another found improvements in academic achievement up to 3 years after
the intervention, including spanning the transition from middle to high school
(Borman, Grigg, Rozek, Hanselman, & Dewey, 2018). The authors argue that this
occurs when the short-term benefits trigger a set of complex and interconnected
processes that recursively build on themselves over long periods. Initial boosts in
performance can increase confidence for future tests, reducing stress (Creswell et al.,
2005) and generating a greater sense that the individual belongs and fits in the
academic world (Cook, Purdie-Vaughns, Garcia, & Cohen, 2012; Shnabel, Purdie-
Vaughns, Cook, Garcia, & Cohen, 2013). Over time, this can change teachers’ and
peers’ perceptions of ability, raising expectations and generating support for higher
levels of academic challenge (Cohen & Sherman, 2014).
However other studies in schools have shown more nuanced effects, where self-
affirmation provided benefits to minority ethnic students by some academic measures
but not others (Borman, Grigg, & Hanselman, 2016), when certain conditions held
(Dee, 2015; Hanselman, Bruch, Gamoran, & Borman, 2014), or when used in
conjunction with enhancements to the intervention (Bowen, Wegmann, & Webber,
2013; Covarrubias, Herrmann, & Fryberg, 2016; Hernandez, Rana, Rao, & Usselman,
2017). Yet other studies, in schools with a wide range of characteristics, have shown
no significant effect of self-affirmation at all (Bratter, Rowley, & Chukhray, 2016; de
Jong, Jellesma, Koomen, & de Jong, 2016; Hanselman, Rozek, Grigg, & Borman,
2017; Protzko & Aronson, 2016).

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 7
This variation indicates the possibility that there are important moderators that vary
across school contexts of the extent to which stereotype threat is experienced by
different groups, and/or the effectiveness of self-affirmation (Borman, 2017; Ferrer &
Cohen, 2018). For example, a follow-up to the study by Borman et al. (2016) found
that self-affirmation was more effective at increasing the performance of Black and
Latino students in schools where they were relatively few in number and had a
relatively large attainment gap with White students, and where they used affirming
language (such as “care about”) in their writing exercises (Borman et al., 2018).
This brings us back to our earlier question of why, in England, SES is a significantly
stronger predictor of academic success than is ethnicity, while in the United States,
SES and ethnicity predict success with similar strength. Perhaps one factor is that
differing national cultural contexts result in differing ‘threats in the air’ (Steele, 2010)
between the two countries.
In the United States, race has historically been, and remains, a highly salient cultural
divide with deep historical roots (Gándara & Contreras, 2009; Steele, 2010), and both
race and low SES have been shown to create threat and other psychological barriers to
academic achievement (Jury et al., 2017; Stephens, Fryberg, Markus, Johnson, &
Covarrubias, 2012; Stephens, Markus, & Fryberg, 2012). In contrast, while there
certainly remain important challenges in relation to educational inequalities arising
from ethnicity in England (Equality and Human Rights Commission, 2015, 2016), the
country’s long history of striking disparities between classes is likely to mean that
societal discourses around the educational performance of different social classes are
more deeply rooted and thus more psychologically salient (Evans, 2006; Jones, 2016).
It seems plausible, therefore, that low SES generates a higher level of stereotype
threat within education in England than in the United States
3
.
3
Such an overall higher level of threat for low-SES individuals might, however, be experienced
differently by individuals of different ethnicities.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 8
Present Study
The present study tests our predictions over the course of an academic year in a
comprehensive secondary school in southern England. The school is moderately
large, is ethnically highly diverse, and has an above-average proportion of students
eligible for FSM. Students completed three short writing exercises throughout the
year: students in a self-affirmed condition wrote about values that were important to
them, and students in a control condition wrote about a neutral topic. As a primary
outcome measure, we observed the effect of the self-affirmation on scores in a
mathematics test that students undertook towards the end of the school year. To
provide a secondary outcome measure, students completed a survey near the end of
the school year measuring four psychological processes (stereotype threat
experienced, stress, self-integrity, and sense of academic fit) that might help explain
how self-affirmation improves academic performance.
We used students’ eligibility for FSM as a proxy for low SES. Eligibility for FSM is
based on current parental income, and nationally around 14% of school students are
eligible
4
(Department for Education, 2015). Although FSM has a number of
limitations as a proxy for low SES (discussed in detail later), it was the best measure
available to us (Gorard, 2012). Its key benefit is that it indicates that a student’s
family has a current low income; the student’s low SES is therefore likely to be
particularly salient and as a result create the highest levels of stereotype threat.
To summarize our argument: stereotype threat has been shown to reduce academic
performance, self-affirmation has been shown to reduce stereotype threat, and low
SES is a plausible cause of stereotype threat in England. Taking eligibility for FSM as
a proxy for low SES we therefore state our hypothesis as follows. FSM students in a
self-affirmation condition will achieve higher grades in mathematics tests than FSM
students in a control condition; however non-FSM students will be unaffected by the
self-affirmation.
4
See Appendix S1 for FSM eligibility criteria.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 9
Method
Participants
A comprehensive secondary school in southern England agreed to take part in the
study during the 2015-16 school year
5
. No significant ethical issues were raised by
the study, which therefore received ethical approval from the head of the relevant
academic programme. Students in Years 7–9 (aged 11–14) participated. At the start of
the autumn 2015 term, we randomly assigned individual students to two experimental
conditions: affirmed and control.
Of the 722 students registered at the start of the academic year, 562 were included in
the analysis (see Appendix S1 for details). Eliminations were mainly due to missing
academic baseline (N = 68) and assessment data (N = 49) and non-completion of
writing exercises (N = 36). Eliminations were not significantly heterogeneous across
experimental conditions. Our sample size was thus determined by the number and
characteristics of pupils in the participating year groups. Demographic and prior
achievement variables of the sample are provided in Appendix S1.
Procedure
Figure 1 shows the timeline for the administration of the research activities.
Figure 1. Research timeline.
5
A second school also agreed to participate but for logistical reasons was unable to provide data on the
primary outcome measure (percentage scores in academic tests), and so we were unable to include it in
our analysis.
Autumn Term
writing exercise
JanSep Oct Nov Dec Feb Mar Apr May Jun Jul
Spring Term
writing exercise
Summer Term
writing exercise
Maths
test
Autumn Term, 2015 SpringTerm, 2016 Summer Term, 2016
Student
survey

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 10
Materials. We based the wording of the writing exercises on that reported by
Sherman et al. (2013). In the affirmed condition, the exercises were designed to
invoke a self-affirmation; the student was asked to pick one or more values that are
“the most important things for you, personally” and to write about “why these things
are important to you”. In the control condition, the exercises were designed to be
psychologically neutral; the corresponding request was to write about values that are
“the least important things to you, but might be important to someone else”, or about
the student’s morning routine that day. At our request, a teacher informally pre-tested
the exercises with three Year 9 students from another school and found no issues with
comprehension. All experimental materials are available in Appendix S1.
To provide our secondary outcome measure, a short student survey was developed to
record students’ self-reports in levels of the following: stereotype threat experienced
(single item: “I worry that people judge me because of my background instead of who
I really am”), stress (single item: “I often feel stressed at school”), self-integrity
(single item: “I feel basically OK about myself”), and sense of academic fit (four
items, e.g. “I fit in well at school and really feel like I belong here”, “I feel proud of
being a student at [school name]”,
a
= .76)
6
.
These areas were drawn from emerging evidence or adapted from other measures
related to the mechanisms by which self-affirmations protect people from stereotype
threat. The stereotype threat experienced and self-integrity measures were designed to
measure the degree to which prevalent negative stereotypes resulted in a subjective
experience of threat and the degree to which the self-affirmation intervention reduced
this by bolstering a student’s self-integrity (Cohen & Sherman, 2014; Sherman et al.,
2013). The stress measure was designed to examine whether the intervention reduced
one of the maladaptive responses that stereotype threat is hypothesized to increase
(Creswell et al., 2013; Sherman et al., 2013). Finally, the sense of academic fit
measure was designed to measure the extent to which identity threat undermined
6
Items related to growth mindset were also included in the survey to provide input to a separate
initiative being undertaken by the school, but these did not form part of the present study.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 11
students’ sense of belonging in school (Cook et al., 2012; Sherman et al., 2013;
Walton & Cohen, 2007).
We used mainly single-item measures in order to enable the survey to be completed
within the constraints of the school timetable. Students were presented with a series of
statements and asked to respond on a six-point rating scale from “Strongly agree” to
“Strongly disagree”. Responses were converted to a point score of 1-6 and reverse-
coded where necessary so that higher scores reflected a higher level of the item
measured.
Self-affirmation procedure. English teachers (blind to the precise purpose of
the study) were briefed near the beginning of the school year. Administration
followed the method outlined by Sherman et al. (2013). In normal class time in
English lessons, the teachers (blind to affirmation condition) introduced the exercises
as a short piece of writing to give the opportunity to write more freely than usual,
without worrying about spelling or grammar. Each exercise took around 15-20
minutes to complete.
Student survey procedure. The student survey was administered in June
2016 during ‘house time’, a period when small groups of students from all year
groups engage in a range of activities.
Outcome Measures
Our primary outcome measure was academic performance as measured by the
school’s usual assessment procedures. Due to limited availability of baseline
measures and practical constraints in collecting percentage scores for other academic
subjects, our primary outcome measure was limited to the percentage score from the
mathematics test that each student undertook towards the end of the school year. For
Years 7 and 8, tests consisted primarily of questions from past General Certificate of
Secondary Education (GCSE) Mathematics papers. For Year 9, tests consisted of an
entire past GCSE Mathematics paper.
GCSEs are national standardized exams used to assess academic performance at age
16. They have a high level of reliability (Bramley & Dhawan, 2010; Newton, 1996).

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 12
Each test for Years 7 and 8 also included a smaller number of questions taken from
past papers of Key Stage 3 Mathematics SATs, national standardized tests undertaken
by students at the end of Year 9 (usually at age 14) until the Department for
Education (DfE) discontinued their use in 2009.
For each year group, the school assembled questions from the above sources into two
tests at differing tiers of difficulty, and each student was assigned to take one of these
two tests, based on their prior performance. No empirically validated means existed to
establish equivalent levels of performance between the two tiers of tests taken within
each year group, or between year groups. We therefore standardized the percentage
scores for each of the two test tiers within each of the three year groups, resulting in
six sets of standardized scores.
Our analytical design allowed for the possibility that there might be interactions
between test tier, FSM status and affirmation condition. For example, FSM students
assigned to the lower-difficulty tier might feel further threatened in respect of their
academic performance, making the self-affirmation intervention more effective for
them compared to FSM students assigned to the higher-difficulty tier.
In order to increase the sensitivity of the outcome measure, we used students’
mathematics scores measured in national standardized tests at the end of Key Stage 2
(Year 6, age 11) as the most reliable and relevant baseline measure of prior
mathematics performance (Opposs & He, 2011).
The school also provided the following information for each student which we used as
further covariates: year group, ethnicity, and markers for eligibility for FSM, gender
and English as an Additional Language (EAL)
7
.
Our secondary outcome measure was students’ responses to the student survey.
7
A complete record of information provided by the school is in the Supplemental Information.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 13
Results
Descriptive statistics of academic performance and raw survey responses are shown in
Table 1. Correlations are provided in Appendix S1.
TABLE 1
Descriptive statistics of academic performance and raw survey responses of the
analytical sample
All
Free school
meals
Non-free school
meals
Mean
SD
Mean
SD
Mean
SD
Academic performance
Standardized mathematics score
0.000
0.996
-0.235
0.896
0.069
1.014
Raw survey responses
Stereotype threat
3.320
1.718
3.650
1.722
3.220
1.706
Stress
3.520
1.608
3.400
1.635
3.550
1.601
Self-integrity
4.470
1.315
4.370
1.387
4.500
1.294
Sense of academic fit
4.019
0.996
3.979
1.059
4.030
0.978
Primary outcome: academic performance
We subjected the standardized end-of-year mathematics scores to a 2 (FSM status:
FSM vs. non-FSM) x 2 (affirmation condition: affirmed vs. control) x 2 (test tier:
lower difficulty vs. higher difficulty) ANCOVA. We included the following
covariates: baseline Key Stage 2 mathematics score, gender, year group (Year 8 and
Year 9, with Year 7 as reference category), ethnicity (Black, Asian, Mixed, and Other,
with White as the reference category) and the marker indicating EAL. We mean-

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 14
centered the baseline Key Stage 2 mathematics scores for the FSM and non-FSM
groups separately
8
.
Tests for non-influential outliers and homogeneity of variances were met but tests for
normal distribution of residuals were not
9
. We therefore amended our ANCOVA to
be robust to non-normal distribution of residuals by using the SPSS ‘bootstrap’
method with 2,000 random samples. For main effects, the standard errors (SE), 95%
bias-corrected and accelerated confidence intervals (BCa CI) and p-values reported
below are based on these bootstrapped (robust) estimates. However, SPSS does not
have the facility to report bootstrapped estimates for interactions and so these results
are not robust to non-normal distribution of residuals, and we note this where
relevant.
This analysis yielded a significant main effect of FSM status on academic
performance, F(1, 545) = 20.021, p < .001,
h
p2 = .035. The academic performance of
students who were eligible for FSM (M = -0.289, SE = 0.070, BCa CI [-0.420; -
0.140]) was significantly lower than those who were not eligible (M = 0.075, SE =
0.051, BCa CI [-0.029; 0.178]), a mean difference of Mdiff = -0.364 (BCa CI [-0.500; -
0.215]).!
There was a significant main effect of affirmation condition on academic
performance, F(1, 545) = 7.202, p = .005,
h
p2 = .013. The academic performance of
students in the affirmed condition (M = 0.000, SE = 0.065, BCa CI [-0.128; 0.137])
8
In line with Sherman et al. (2013), we wanted to avoid apparently reducing the pre-existing gap
between FSM and non-FSM students. Using unmanipulated baseline scores as a covariate could have
given a misleading impression that the gap was smaller than it actually is (Sackett, Hardison, & Cullen,
2004). However, supplementary analyses using unmanipulated baseline scores yielded substantially
similar results.
9
Additionally, the following interactions violated the assumption of homogeneity of regression slopes:
FSM status * Black ethnicity; FSM status * mixed ethnicity; and affirmation condition * Year 8.
Including these interaction terms in the ANCOVA model increased the significance of the focal effect
(the affirmation condition * FSM status interaction). Therefore, given that the model without these
interaction terms yielded substantively unchanged results while being somewhat more conservative and
parsimonious, we report the results of the simpler model in the manuscript.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 15
was significantly higher than those in the control condition (M = -0.215, SE = 0.056,
BCa CI [-0.328; -0.096]), a mean difference of Mdiff = 0.215 (BCa CI [0.077; 0.363]).
There was also a significant main effect on academic performance of test tier, F(1,
545) = 67.215, p < .001,
h
p2 = .110. The academic performance of students in the
easier tier (M = 0.328, SE = 0.072, BCa CI [0.189; 0.479]) was significantly higher
than that of those in the harder tier (M = -0.542, SE = 0.070, BCa CI [-0.685; -0.400]),
a mean difference of Mdiff = 0.869 (BCa CI [0.666; 1.072])
10
.!
In line with our hypothesis, the significant main effect of affirmation condition was
qualified by a significant two-way interaction between FSM status and affirmation
condition, F(1, 545) = 4.154, p = .042 (non-robust),
h
p2 = .008. A decomposition of
this interaction showed a significant simple main effect of affirmation condition for
FSM students, F(1, 545) = 7.192, p = .003,
h
p2 = .013, but not for non-FSM students,
F(1, 545) = 0.448, p = .497,
h
p2 = .001. Academic performance was higher for FSM
students in the affirmed condition (M = -0.099, SE = 0.099, BCa CI [-0.280; 0.113])
than for FSM students in the control condition (M = -0.478, SE = 0.088, BCa CI [-
0.645; -0.305]), a mean difference of Mdiff = 0.379 (BCa CI [0.128; 0.642]), or 0.381
SD (where SD = 0.996, Table 1). Thus the gap of 0.527 (0.529 SD) between FSM and
non-FSM students in the control condition was reduced by 62% in the affirmed
condition (see Section 5 of Appendix S1 for detailed calculations). This is shown in
Figure 2
11
.
All interactions with test tier were non-significant: these consisted of the three-way
interaction of test tier with FSM status and affirmation condition, and the two-way
interactions of test tier with FSM status and with affirmation condition (all Fs <
2.686, all ps > .102, non-robust).
10
However this effect is non-significant when covariates are not included in the model.
11
Appendix S1 contains a sensitivity analysis of these results without covariates. Unsurprisingly (since
removing the covariates reduced the power of the analysis), the p-values and confidence intervals are
expanded and, while the two-way interaction between FSM status and affirmation condition is non-
significant, the simple main effect of affirmation condition on FSM students is substantially
unchanged.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 16
Figure 2. Academic performance as a function of free school meal (FSM) status and
affirmation condition.
Secondary outcome: self-reported measures.
We subjected each of the self-reported measures of the end-of-year student survey to
a 2 (FSM status: FSM vs. non-FSM) x 2 (affirmation condition: affirmed vs. control)
x 2 (test tier: lower vs. higher) ANCOVA with 2,000 bootstrapped samples and the
same covariates as for academic performance above. We also performed robust
(bootstrapped with 2,000 samples) regression analyses using the PROCESS V2.16.3
SPSS macro (Hayes, 2013) to investigate mediation of students’ academic
performance by each of the self-reported measures.
Non-FSM FSM
Standardised mathematics score
(covariate adjusted)
-0.75
-0.25
0.00
-0.50
0.25
Control Self-affirmed

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 17
Our analysis yielded no main effects of FSM status for reported stress, self-integrity
or sense of academic fit (all Fs < 2.238, all ps > .135), but it did yield a main effect
for reported stereotype threat, F(1, 508) = 7.339, p = .009,
h
p2 =.014. FSM students
reported that they experienced significantly higher levels of stereotype threat (M =
3.699, SE = 0.160, BCa CI [3.373; 4.034]) than non-FSM students (M = 3.212, SE =
0.084, BCa CI [3.044; 3.392]), a mean difference of Mdiff = 0.487 (SE = 0.180, BCa
CI [0.147; 0.842]). No processes reported in the survey significantly mediated the
FSM gap in academic performance.
There was a non-significant three-way interaction between FSM status, affirmation
condition and test tier for all four processes reported in the survey (all Fs < 0.853, all
ps > .356). There was a marginally significant two-way interaction between
affirmation condition and test tier for reported stress, F(1, 501) = 2.859, p = .091
(non-robust),
h
p2 = .006, and a significant interaction between FSM status and test tier
for reported self-integrity, F(1, 501) = 3.872, p = .050 (non-robust),
h
p2 = .008. All
other two-way interactions between affirmation condition and test tier and between
FSM status and test tier were non-significant (all Fs < 1.342, all ps > .247). There
were non-significant two-way interactions between FSM status and affirmation
condition for reported stereotype threat, self-integrity and sense of academic fit (all Fs
< 1.003, all ps > .311). However there was a marginally significant two-way
interaction between FSM status and affirmation condition for reported stress, F(1,
501) = 3.004, p = .084 (non-robust),
h
p2 = .006. A decomposition of this interaction
showed a significant simple main effect of affirmation condition for FSM students,
F(1, 501) = 4.122, p = .032,
h
p2 = .008, but not for non-FSM students, F(1, 501) =
0.012, p = .919,
h
p2 < .001. Reported levels of stress were lower in FSM students in
the affirmed condition (M = 3.191, SE = 0.200, BCa CI [2.796; 3.590]) than FSM
students in the control condition (M = 3.793, SE = 0.214, BCa CI [3.388; 4.215]), a
mean difference of Mdiff = -0.603 (SE = 0.288, BCa CI [-1.172; -0.037]). This is
shown in Figure 3.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 18
Figure 3. Reported stress as a function of free school meal (FSM) status and
affirmation condition. Stress levels reported on a rating scale of 1-6.
No processes reported in the survey significantly mediated the interaction between
FSM status and affirmation condition on academic performance.
Discussion
Summary of results
We tested whether, over the course of an academic year, a virtually zero-cost self-
affirmation intervention reduced gaps in academic performance of school students
aged 11–14 based on their SES. We also examined whether students’ self-reports of a
set of psychological processes were associated with any academic performance gaps
and whether they helped explain any reduction in those gaps caused by the self-
affirmation intervention.
Non-FSM FSM
2.5
3.5
4.5
3.0
4.0
Level of reported stress
(covariate adjusted)
Control Self-affirmed

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 19
We observed a significant gap in academic performance between FSM and non-FSM
students, which in the control condition was 0.529 standard deviations; we also found
that FSM students reported experiencing significantly more stereotype threat than
their non-FSM peers.
As predicted, the intervention significantly improved the academic performance of the
FSM school students and we found no evidence for any effect on their non-FSM
peers. Specifically, the intervention increased the performance of FSM students by
0.381 standard deviations, reducing the FSM/non-FSM gap by 62% (Figure 2). We
also found that the self-affirmation intervention reduced the levels of stress reported
by FSM students, but not those reported by non-FSM students (Figure 3).
General discussion
This study represents the first direct evidence that self-affirmation can substantially
improve the academic performance of low-SES students in a school setting, while not
affecting their higher-SES peers
12
.
The effect size of 0.38 standard deviations for low-SES students in the present study
lies within the 0.29–0.45 range of the effects for Black and Latino students in the
earlier studies (Cohen et al., 2006; Sherman et al., 2013), but is much greater than
those of later studies where there were much smaller or no significant effects (see
Hanselman et al., 2017, Figure 1).
As we have speculated, it is possible that different countries might present different
pictures of how stereotype threat operates within them (for a broad discussion of the
effect of context, see Pettigrew, 2018). Such differences might help explain why, for
example, no effects of self-affirmation were observed in students of Moroccan and
Turkish backgrounds in a school in the Netherlands (de Jong et al., 2016). Since many
of the students chose to write about Islam, a religion about which there are popular
associations with extremism in the Netherlands (Kamans, Gordijn, Oldenhuis, &
Otten, 2009), the authors speculate that this could have actually increased their sense
12
Such evidence has, however, been found in college settings (e.g. Harackiewicz et al., 2013).

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 20
of stereotype threat. Future studies in a range of countries could help tease out these
and other such effects of national and cultural contexts.
Our examination of the psychological processes provided some intriguing results
related to SES. Firstly, FSM students reported experiencing higher levels of
stereotype threat than their non-FSM peers; however, the self-affirmation intervention
did not reduce this self-reported sense of threat. Since prior research suggests that
self-affirmation is effective because it protects against threat (see Cohen & Sherman,
2014 for a review), it is possible that the intervention did reduce the sense of threat
but that limitations in our survey measure meant that it failed to detect this reduction.
For example, our measure asks students to self-report a largely unconscious process,
is generic in relation to the source and context of threat, and consists of only a single
item. However it is also possible that the self-affirmation did not reduce levels of
stereotype threat, but rather it improved academic performance through some other
mechanism or helped the students to better cope with the threat, for example by
providing more effective access to psychological resources or by bolstering an aspect
of their self-identity. Future research could attempt to distinguish between these
possibilities, for example by using a more sophisticated measure of stereotype threat
such as daily experience sampling (Sherman et al., 2013), or by examining the extent
to which cultural mismatches rather than stereotype threat reduce low-SES students’
performance (Stephens, Fryberg, & Markus, 2012; Stephens, Hamedani, & Destin,
2014).
Secondly, the self-affirmation reduced the levels of stress reported by FSM students,
and we found no evidence that it did so for their non-FSM peers. This observation is
consistent with evidence that self-affirmation reduces stress (Creswell et al., 2005)
and improves problem-solving for chronically-stressed individuals (Creswell et al.,
2013). It is particularly encouraging since it points to the potential of self-affirmation
to reduce the chronic stress that disproportionately affects low-SES children and
young people (Evans & Schamberg, 2009). Since in England one in ten school
students has a diagnosable mental illness and three in four mental illnesses start in
childhood (MQ, 2017), with particular prevalence among lower-SES girls (Patalay &
Fitzsimons, 2018), any reduced susceptibility to such conditions arising from reduced

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 21
stress could generate important benefits that are much broader than academic
performance.
This reduced stress also suggests a potentially novel line of research. There is
extensive evidence that school transitions are particularly threatening for students
(Barber & Olsen, 2004; Rice, Frederickson, & Seymour, 2011; Riglin, Frederickson,
Shelton, & Rice, 2013; Zeedyk et al., 2003), and students have been shown to feel
even higher levels of concerns about the primary-secondary transition towards the end
of primary school than they do at the start of secondary school (Lester, Lisk, Carr,
Patrick, & Eley, 2019). If self-affirmation does reliably reduce stress, it might
therefore be particularly effective just before students make the transition to
secondary school at age 11. As far as we are aware, no study has yet examined this.
Since we needed to exclude students who performed none of the three writing
exercises, we did so by excluding those who did not complete the first (in the autumn
term). In two sensitivity analyses we tested the effect of excluding students simply by
whether they performed at least one or at least two exercises, regardless of their
timing. This slightly reduced the extent to which the SES gap was reduced, from 62%
to 60% and 56% respectively (details in Appendix S1). This might be because the
autumn exercise offered the longest period for the hypothesized recursive effects of
self-affirmation to generate benefits, and/or because it took place at the start of a
school year, a time that is likely to be particularly threatening (Sherman et al., 2013).
A supplementary analysis, in which we extended the definition of low SES to students
who had been eligible for FSM in any of the last six years, found no evidence of any
benefit (details in Appendix S1). This suggests that low SES might cause stereotype
threat only when it is currently of high salience. To the extent that this result can be
generalised to the wider school population in England, it would mean that the
potential benefits of self-affirmation would be restricted to the school population that
is eligible for FSM at any given time (currently 14%). Given the almost zero-cost of
the intervention, and its focus on the most disadvantaged students, this could
nevertheless represent an important opportunity from the perspective of educational
policy.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 22
Each student was assigned to their mathematics class based on prior performance (a
practice known as ‘tracking’ in the US and ‘setting’ in the UK), and so, almost all
students in any given class took a test of the same tier of difficulty. However, we
observed no significant two- or three-way interactions between test difficulty tier and
FSM status or affirmation condition. If, as prior research suggests, self-affirmation is
effective because it protects against threat, then this suggests that the practice of
setting did not exacerbate feelings of stereotype threat in the lower sets. This is an
encouraging finding, since it suggests that, at least in this school, the potentially
disproportionately negative impact of setting on lower-SES students (Batruch, Autin,
Bataillard, & Butera, 2018; Francis, Hodgen, Tereshchenko, & Archer, 2018) did not
include the extra burden of further stereotype threat.
Five limitations of our study could be addressed in future studies. Firstly, our primary
outcome variable was based on a single academic subject, mathematics, using tests
administered by the school. Future studies using national standardized tests, such as
GCSEs in a range of subjects at age 16, would provide results that are more robust
and generalizable. For example, self-affirmation might benefit cognitive processes
that are required more in some subjects than in others, a possibility suggested by one
study where indicative benefits of self-affirmation were found in mathematics but not
in reading (Borman et al., 2016). Secondly, our study measured academic
performance over a single academic year; a follow-up study could look at
performance over an extended period in order to test for any long-term benefits of the
complex recursive processes hypothesized by others (Borman et al., 2018; Cohen,
Garcia, Purdie-Vaughns, Apfel, & Brzustoski, 2009; Goyer et al., 2017; Sherman et
al., 2013).
Thirdly, we did not have a sufficient sample size to robustly examine moderating
factors. These include the following: whether low-SES students of different ethnic
backgrounds experience different levels of benefit from self-affirmation, an effect that
is plausible given the important interactions between SES and ethnicity in academic
performance (Harackiewicz et al., 2016; Strand, 2014); the degree of stereotype threat
present (Hanselman et al., 2014); and the extent to which students are situated in a
learning environment that supports changes in their behaviour (Cohen & Sherman,
2014; Ferrer & Cohen, 2018). A range of future studies could help build a robust

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 23
picture of the extent to which the variability observed in the effectiveness of self-
affirmation in schools is due to such moderators, or to experimental factors such as
implementation constraints (e.g. Protzko & Aronson, 2016), the timing of the
intervention (Cook et al., 2012; Critcher, Dunning, & Armor, 2010; Ferrer & Cohen,
2018) and sampling variation (see Gelman & Carlin, 2014; Hanselman et al., 2017).
Fourthly, although eligibility for FSM is a reasonable proxy for current low SES
(Gorard, 2012), it is an imperfect one. For example, for technical reasons some
families outside the lowest income brackets are not eligible (Hobbs & Vignoles,
2010) and an estimated 14% of families who are eligible are not flagged as such
(Department for Education, 2018; Iniesta-Martinez & Evans, 2012). It also does not
take into account any cumulative effects of chronic poverty over an extended period
(Andrews, Robinson, & Hutchinson, 2017; Michelmore & Dynarski, 2017) or
important non-financial elements of SES such as parental education and parental
occupation (Harwell & LeBeau, 2010; Reardon, 2011). Further studies could use
more sophisticated indicators to examine the effect of these distinct components of
SES.
Finally, in the context of long-term trends in the skills required for the future
workforce (Chui, Manyika, & Miremadi, 2016; EY, 2015), evidence about whether
self-affirmation improves non-cognitive abilities such as persistence and creativity
will increasingly be of value to social and economic policymakers (Garcia, 2014;
Gutman & Schoon, 2013).
Additionally, while we confirm that our hypothesis was set prior to gathering our
outcome measures, we did not pre-register the study. We report analyses for all other
relevant measures obtained from the school in Appendix S1 and would welcome
future pre-registered studies investigating the role of SES in stereotype threat and
self-affirmation in England.
Conclusion. With an effect size of 0.38 standard deviations on the academic
performance of the most economically disadvantaged school students, a virtually
zero-cost self-affirmation intervention compares favorably with other interventions
aimed at reducing the SES gap (Education Endowment Foundation, n.d.). However,
evidence from other studies is mixed and we are still at an early stage of

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 24
understanding the mechanisms by which self-affirmation works, and the contexts in
which it is likely to be most effective. This is an exciting time for research into an
intervention that has the potential to make a substantial difference to the lives of many
young people from the most disadvantaged backgrounds.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 25
References
Andrews, J., Robinson, D., & Hutchinson, J. (2017). Closing the Gap? Trends in
Educational Attainment and Disadvantage. Retrieved from
https://epi.org.uk/publications-and-research/closing-gap-trends-educational-
attainment-disadvantage/
Barber, B. K., & Olsen, J. A. (2004). Assessing the Transitions to Middle and High
School. Journal of Adolescent Research, 19(1), 3–30.
https://doi.org/10.1177/0743558403258113
Batruch, A., Autin, F., Bataillard, F., & Butera, F. (2018). School Selection and the
Social Class Divide: How Tracking Contributes to the Reproduction of
Inequalities. Personality and Social Psychology Bulletin, 45, 477–490.
https://doi.org/10.1177/0146167218791804
Borman, G. D. (2017). Advancing values affirmation as a scalable strategy for
mitigating identity threats and narrowing national achievement gaps.
Proceedings of the National Academy of Sciences, 201708813.
https://doi.org/10.1073/pnas.1708813114
Borman, G. D., Grigg, J., & Hanselman, P. (2016). An Effort to Close Achievement
Gaps at Scale Through Self-Affirmation. Educational Evaluation and Policy
Analysis, 38(1), 21–42. https://doi.org/10.3102/0162373715581709
Borman, G. D., Grigg, J., Rozek, C. S., Hanselman, P., & Dewey, N. A. (2018). Self-
Affirmation Effects Are Produced by School Context, Student Engagement With
the Intervention, and Time: Lessons From a District-Wide Implementation.
Psychological Science, 29(11), 1773–1784.
https://doi.org/10.1177/0956797618784016
Bowen, N. K., Wegmann, K. M., & Webber, K. C. (2013). Enhancing a brief writing
intervention to combat stereotype threat among middle-school students. Journal
of Educational Psychology, 105(2), 427–435. https://doi.org/10.1037/a0031177
Bramley, T., & Dhawan, V. (2010). Estimates of Reliability of Qualifications.
Retrieved from

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 26
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/57
8868/2011-03-16-estimates-of-reliability-of-qualifications.pdf
Brannon, T. N., Higginbotham, G. D., & Henderson, K. (2017). Class advantages and
disadvantages are not so Black and White: intersectionality impacts rank and
selves. Current Opinion in Psychology, 18, 117–122.
https://doi.org/10.1016/j.copsyc.2017.08.029
Bratter, J. L., Rowley, K. J., & Chukhray, I. (2016). Does a Self-Affirmation
Intervention Reduce Stereotype Threat in Black and Hispanic High Schools?
Race and Social Problems, 8(4), 340–356. https://doi.org/10.1007/s12552-016-
9187-4
Browman, A. S., Destin, M., Carswell, K. L., & Svoboda, R. C. (2017). Perceptions
of socioeconomic mobility influence academic persistence among low
socioeconomic status students. Journal of Experimental Social Psychology, 72,
45–52. https://doi.org/10.1016/j.jesp.2017.03.006
Chui, M., Manyika, J., & Miremadi, M. (2016). Where machines could replace
humans —and where they can’t (yet). McKinsey Quarterly, (July). Retrieved
from http://www.mckinsey.com/business-functions/digital-mckinsey/our-
insights/where-machines-could-replace-humans-and-where-they-cant-yet
Cohen, G. L., Garcia, J., Apfel, N., & Master, A. (2006). Reducing the Racial
Achievement Gap: A Social-Psychological Intervention. Science, 313(5791),
1307–1310. https://doi.org/10.1126/science.1128317
Cohen, G. L., Garcia, J., Purdie-Vaughns, V., Apfel, N., & Brzustoski, P. (2009).
Recursive Processes in Self-Affirmation: Intervening to Close the Minority
Achievement Gap. Science, 324(17 April), 400–403.
https://doi.org/10.1126/science.1170769
Cohen, G. L., & Sherman, D. K. (2014). The Psychology of Change: Self-Affirmation
and Social Psychological Intervention. Annual Review of Psychology, 65(1),
333–371. https://doi.org/10.1146/annurev-psych-010213-115137
Cook, J. E., Purdie-Vaughns, V., Garcia, J., & Cohen, G. L. (2012). Chronic threat

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 27
and contingent belonging: Protective benefits of values affirmation on identity
development. Journal of Personality and Social Psychology, 102(3), 479–496.
https://doi.org/10.1037/a0026312
Covarrubias, R., Herrmann, S. D., & Fryberg, S. A. (2016). Affirming the
Interdependent Self: Implications for Latino Student Performance. Basic and
Applied Social Psychology, 38(1), 47–57.
https://doi.org/10.1080/01973533.2015.1129609
Creswell, J. D., Dutcher, J. M., Klein, W. M. P., Harris, P. R., & Levine, J. M. (2013).
Self-Affirmation Improves Problem-Solving under Stress. PLoS ONE, 8(5), 1–7.
https://doi.org/10.1371/journal.pone.0062593
Creswell, J. D., Welch, W. T., Taylor, S. E., Sherman, D. K., Gruenewald, T. L., &
Mann, T. (2005). Affirmation of Personal Values Buffers Neuroendocrine and
Psychological Stress Responses. Psychological Science, 16(11), 846–851.
Critcher, C. R., Dunning, D., & Armor, D. A. (2010). When Self-Affirmations
Reduce Defensiveness: Timing Is Key. Personality and Social Psychology
Bulletin, 36(7), 947–959. https://doi.org/10.1177/0146167210369557
Croizet, J.-C., & Claire, T. (1998). Extending the Concept of Stereotype Threat to
Social Class: The Intellectual Underperformance of Students from Low
Socioeconomic Backgrounds. Personality and Social Psychology Bulletin, 24(6),
588–594. https://doi.org/10.1177/0146167298246003
de Jong, E. M., Jellesma, F. C., Koomen, H. M. Y., & de Jong, P. F. (2016). A values-
affirmation intervention does not benefit negatively stereotyped immigrant
students in the Netherlands. Frontiers in Psychology, 7, 1–30.
https://doi.org/10.3389/fpsyg.2016.00691
Dee, T. S. (2015). Social Identity and Achievement Gaps: Evidence From an
Affirmation Intervention. Journal of Research on Educational Effectiveness,
8(2), 149–168. https://doi.org/10.1080/19345747.2014.906009
Department for Education. (2015). SFR 06/2015. GCSE and equivalent attainment by
pupil characteristics, 2013 to 2014 (Revised). Retrieved from

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 28
https://www.gov.uk/government/statistics/gcse-and-equivalent-attainment-by-
pupil-characteristics-2014
Department for Education. (2018). SFR 01/2018: Revised GCSE and equivalent
results in England, 2016 to 2017. Department for Education. Retrieved from
https://www.gov.uk/government/statistics/revised-gcse-and-equivalent-results-
in-england-2016-to-2017
Désert, M., Préaux, M., & Jund, R. (2009). So young and already victims of
stereotype threat: Socio-economic status and performance of 6 to 9 years old
children on Raven’s progressive matrices. European Journal of Psychology of
Education, 24(2), 207–218.
Education Endowment Foundation. (n.d.). Teaching and learning toolkit: An
accessible summary of the international evidence on teaching 5-16 year-olds.
Retrieved from https://educationendowmentfoundation.org.uk/evidence-
summaries/teaching-learning-toolkit
Epton, T., Harris, P. R., Kane, R., Koningsbruggen, G. M. Van, & Sheeran, P. (2015).
The Impact of Self-Affirmation on Health-Behavior Change: A Meta-Analysis.
Health Psychology, 34(3), 187–196. https://doi.org/10.1037/hea0000116
Equality and Human Rights Commission. (2015). Is Britain Fairer? The state of
equality and human rights 2015.
Equality and Human Rights Commission. (2016). Healing a divided Britain: the need
for a comprehensive race equality strategy.
Evans, G. (2006). Educational Failure and Working Class White Children in Britain.
Palgrave Macmillan.
Evans, G. W., & Schamberg, M. A. (2009). Childhood poverty, chronic stress, and
adult working memory. Proceedings of the National Academy of Sciences,
106(16), 6545–6549. https://doi.org/10.1073/pnas.0811910106
EY. (2015). EY transforms its recruitment selection process for graduates,
undergraduates and school leavers. Retrieved from
http://www.ey.com/uk/en/newsroom/news-releases/15-08-03---ey-transforms-

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 29
its-recruitment-selection-process-for-graduates-undergraduates-and-school-
leavers
Ferrer, R. A., & Cohen, G. L. (2018). Reconceptualizing self-affirmation with the
“Trigger and Channel” framework: Lessons from the health domain. Personality
and Social Psychology Bulletin. https://doi.org/10.1177/014616728173007
Flore, P. C., & Wicherts, J. M. (2015). Does stereotype threat influence performance
of girls in stereotyped domains? A meta-analysis. Journal of School Psychology,
53(1), 25–44. https://doi.org/10.1016/j.jsp.2014.10.002
Foster, D., & Long, R. (2017). The Pupil Premium. Retrieved from
www.ofsted.gov.uk/resources/120197
Francis, B., Hodgen, J., Tereshchenko, A., & Archer, L. (2018). Dos and Don’ts of
attainment grouping. UCL Institute of Education.
https://doi.org/10.1080/0305764X.2018.1441372
Gándara, P., & Contreras, F. (2009). The Latino Education Crisis: The consequences
of failed social policies. Harvard University Press.
Garcia, E. (2014). The Need To Address Noncognitive Skills In The Education Policy
Agenda. Retrieved from https://www.epi.org/publication/the-need-to-address-
noncognitive-skills-in-the-education-policy-agenda/
Gelman, A., & Carlin, J. (2014). Beyond Power Calculations: Assessing Type S
(Sign) and Type M (Magnitude) Errors. Perspectives on Psychological Science,
9(6), 641–651. https://doi.org/10.1177/1745691614551642
Good, C., Aronson, J., & Inzlicht, M. (2003). Improving adolescents’ standardized
test performance: An intervention to reduce the effects of stereotype threat.
Journal of Applied Developmental Psychology, 24(6), 645–662.
https://doi.org/10.1016/j.appdev.2003.09.002
Gorard, S. (2012). Who is eligible for free school meals? Characterising free school
meals as a measure of disadvantage in England. British Educational Research
Journal, 38(6), 1003–1017. https://doi.org/10.1080/01411926.2011.608118

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 30
Goyer, J. P., Garcia, J., Purdie-Vaughns, V., Binning, K. R., Cook, J. E., Reeves, S.
L., … Cohen, G. L. (2017). Self-affirmation facilitates minority middle
schoolers’ progress along college trajectories. Proceedings of the National
Academy of Sciences, 114(29), 7594–7599.
https://doi.org/10.1073/pnas.1617923114
Gutman, L. M., & Schoon, I. (2013). The impact of non-cognitive skills on outcomes
for young people: Literature review. Retrieved from
https://educationendowmentfoundation.org.uk/public/files/Publications/EEF_Lit
_Review_Non-CognitiveSkills.pdf
Hall, C. C., Zhao, J., & Shafir, E. (2014). Self-Affirmation Among the Poor:
Cognitive and Behavioral Implications. Psychological Science, 25(2), 619–625.
https://doi.org/10.1177/0956797613510949
Hanselman, P., Bruch, S. K., Gamoran, A., & Borman, G. D. (2014). Threat in
Context- School Moderation of the Impact of Social Identity Threat on
Racial/Ethnic Achievement Gaps. Sociology of Education, 87(2), 106–124.
https://doi.org/10.1177/0038040714525970
Hanselman, P., Rozek, C. S., Grigg, J., & Borman, G. D. (2017). New Evidence on
Self-Affirmation Effects and Theorized Success of Heterogeneity from Large-
scale Replications. Journal of Educational Psychology, 109(3), 405–424.
Harackiewicz, J. M., Canning, E. A., Tibbetts, Y., Giffen, C. J., Blair, S. S., Rouse, D.
I., & Hyde, J. S. (2013). Closing the social class achievement gap for first-
generation students in undergraduate biology. Journal of Educational
Psychology, 106(2), 375–389. https://doi.org/10.1037/a0034679
Harackiewicz, J. M., Canning, E. A., Tibbetts, Y., Priniski, S. J., & Hyde, J. S.
(2016). Closing achievement gaps with a utility-value intervention:
Disentangling race and social class. Journal of Personality and Social
Psychology, 111(5), 745–765. https://doi.org/10.1037/pspp0000075
Harwell, M., & LeBeau, B. (2010). Student Eligibility for a Free Lunch as an SES
Measure in Education Research. Educational Researcher, 39(2), 120–131.
https://doi.org/10.3102/0013189X10362578

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 31
Hayes, A. F. (2013). Introduction to Mediation, Moderation, and Conditional Process
Analysis: A Regression-Based Approach (1st ed.). Guilford Press.
Hernandez, D., Rana, S., Rao, A., & Usselman, M. (2017). Dismantling Stereotypes
About Latinos in STEM. Hispanic Journal of Behavioral Sciences, 39, 436–451.
https://doi.org/10.1177/0739986317731100
Hobbs, G., & Vignoles, A. (2010). Is children’s free school meal ‘eligibility’ a good
proxy for family income? British Educational Research Journal, 36(4), 673–
690. https://doi.org/10.1080/01411920903083111
Iniesta-Martinez, S., & Evans, H. (2012). Pupils not claiming free school meals.
Retrieved from https://dera.ioe.ac.uk//16039/
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(4), 691–705.
https://doi.org/10.1037/a0013834
Jones, O. (2016). Chavs: The Demonization of the Working Class (First). Verso
Books.
Jury, M., Smeding, A., Stephens, N. M., Nelson, J. E., Aelenei, C., & Darnon, C.
(2017). The Experience of Low-SES Students in Higher Education:
Psychological Barriers to Success and Interventions to Reduce Social-Class
Inequality. Journal of Social Issues, 73(1), 23–41.
https://doi.org/10.1111/josi.12202
Kamans, E., Gordijn, E. H., Oldenhuis, H., & Otten, S. (2009). What I think you see
is what you get: Influence of prejudice on assimilation to negative meta-
stereotypes among Dutch Moroccan teenagers. European Journal of Social
Psychology, 39, 842–851. https://doi.org/10.1002/ejsp
Kinias, Z., & Sim, J. (2016). Facilitating women’s success in business: Interrupting
the process of stereotype threat through affirmation of personal values. Journal
of Applied Psychology, 101(11), 1585–1597. https://doi.org/10.1037/apl0000139
Kizilcec, R. F., Saltarelli, A. J., Reich, J., & Cohen, G. L. (2017). Closing global

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 32
achievement gaps in MOOCs. Science, 355(6322), 251–252.
https://doi.org/10.1126/science.aag2063
Lamont, R. A., Swift, H. J., & Abrams, D. (2015). A review and meta-analysis of age-
based stereotype threat: Negative stereotypes, not facts, do the damage.
Psychology and Aging, 30(1), 180–193. https://doi.org/10.1037/a0038586
Lester, K. J., Lisk, S. C., Carr, E., Patrick, F., & Eley, T. C. (2019). Associations
Between Attentional Bias and Interpretation Bias and Change in School
Concerns and Anxiety Symptoms During the Transition from Primary to
Secondary School. Journal of Abnormal Child Psychology.
https://doi.org/10.1007/s10802-019-00528-3
Martens, A., Johns, M., Greenberg, J., & Schimel, J. (2006). Combating stereotype
threat: The effect of self-affirmation on women’s intellectual performance.
Journal of Experimental Social Psychology, 42(2), 236–243.
https://doi.org/10.1016/j.jesp.2005.04.010
Michelmore, K., & Dynarski, S. (2017). The Gap Within the Gap: Using Longitudinal
Data to Understand Income Differences in Educational Outcomes. AERA Open,
3(1), 1–18. https://doi.org/10.1177/2332858417692958
Miyake, A., Kost-Smith, L. E., Finkelstein, N. D., Pollock, S. J., Cohen, G. L., & Ito,
T. A. (2010). Reducing the Gender Achievement Gap in College Science: A
Classroom Study of Values Affirmation. Science, 330(6008), 1234–1237.
https://doi.org/10.1126/science.1195996
MQ. (2017). 12 statistics to get you thinking about mental health in young people.
Retrieved from https://www.mqmentalhealth.org/news-blog/post/12-statistics
Murphy, M. C., Steele, C. M., & Gross, J. J. (2007). Signaling Threat: How
Situational Cues Affect Women in Math, Science, and Engineering Settings
Mary. Psychological Science, 18(10), 879–885. https://doi.org/10.1111/j.1467-
9280.2007.01995.x
National Assessment of Educational Progress. (2016). Extracts from reports generated
by the Nations Report Card website. Retrieved from

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 33
http://www.nationsreportcard.gov/reading_math_2015/
Newton, P. E. (1996). The Reliability of Marking of General Certificate of Secondary
Education Scripts: Mathematics and English. British Educational Research
Journal, 22(4), 405–420.
Opposs, D., & He, Q. (2011). The reliability programme: Final report. Retrieved
from https://dera.ioe.ac.uk//2568/
Patalay, P., & Fitzsimons, E. (2018). Mental ill-health and wellbeing at age 14 –
Initial findings from the Millennium Cohort Study Age 14 Survey. Retrieved from
https://cls.ucl.ac.uk/cls_research/initial-findings-from-the-millennium-cohort-
study-age-14-survey/
Pettigrew, T. F. (2018). The Emergence of Contextual Social Psychology. Personality
and Social Psychology Bulletin, 44, 963–971.
https://doi.org/10.1177/0146167218756033
Picho, K., Rodriguez, A., & Finnie, L. (2013). Exploring the moderating role of
context on the mathematics performance of females under stereotype threat: A
meta-analysis. Journal of Social Psychology, 153(3), 299–333.
https://doi.org/10.1080/00224545.2012.737380
Protzko, J., & Aronson, J. (2016). Context Moderates Affirmation Effects on the
Ethnic Achievement Gap. Social Psychological and Personality Science, 7(6),
500–507. https://doi.org/10.1177/1948550616646426
Reardon, S. F. (2011). The Widening Academic Achievement Gap Between the Rich
and the Poor: New Evidence and Possible Explanations. In G. J. Duncan & R. J.
Murnane (Eds.), Whither Opportunity? Rising Inequality, Schools and
Children’s Life Chances (pp. 91–116). Russell Sage Foundation.
Rice, F., Frederickson, N., & Seymour, J. (2011). Assessing pupil concerns about
transition to secondary school. British Journal of Educational Psychology, 81(2),
244–263. https://doi.org/10.1348/000709910X519333
Riglin, L., Frederickson, N., Shelton, K. H., & Rice, F. (2013). A longitudinal study
of psychological functioning and academic attainment at the transition to

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 34
secondary school. Journal of Adolescence, 36(3), 507–517.
https://doi.org/10.1016/j.adolescence.2013.03.002
Sackett, P. R., Hardison, C. M., & Cullen, M. J. (2004). On interpreting stereotype
threat as accounting for African American-White differences on cognitive tests.
The American Psychologist, 59(1), 7–13. https://doi.org/10.1037/h0087868
Shapiro, J. R., & Neuberg, S. L. (2007). From Stereotype Threat to Stereotype
Threats: Implications of a Multi-Threat Framework for Causes, Moderators,
Mediators, Consequences, and Interventions. Personality and Social Psychology
Review, 11(2), 107–130. https://doi.org/10.1177/1088868306294790
Shapiro, J. R., Williams, A. M., & Hambarchyan, M. (2013). Are all interventions
created equal? A multi-threat approach to tailoring stereotype threat
interventions. Journal of Personality and Social Psychology, 104(2), 277–288.
https://doi.org/10.1037/a0030461
Sherman, D. K., Hartson, K. A., Binning, K. R., Purdie-Vaughns, V., Garcia, J.,
Taborsky-Barba, S., … Cohen, G. L. (2013). Deflecting the trajectory and
changing the narrative: How self-affirmation affects academic performance and
motivation under identity threat. Journal of Personality and Social Psychology,
104(4), 591–618. https://doi.org/10.1037/a0031495
Shnabel, N., Purdie-Vaughns, V., Cook, J. E., Garcia, J., & Cohen, G. L. (2013).
Demystifying Values-Affirmation Interventions: Writing About Social
Belonging Is a Key to Buffering Against Identity Threat. Personality and Social
Psychology Bulletin, 39(5), 663–676.
https://doi.org/10.1177/0146167213480816
Silverman, A., Logel, C., & Cohen, G. L. (2013). Self-affirmation as a deliberate
coping strategy: The moderating role of choice. Journal of Experimental Social
Psychology, 49(1), 93–98. https://doi.org/10.1016/j.jesp.2012.08.005
Spencer, B., & Castano, E. (2007). Social class is dead. Long live social class!
Stereotype threat among low socioeconomic status individuals. Social Justice
Research, 20(4), 418–432. https://doi.org/10.1007/s11211-007-0047-7

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 35
Steele, C. M. (2010). Whistling Vivaldi. W. W. Norton and Company.
Steele, C. M., & Aronson, J. (1995). Stereotype threat and the intellectual test
performance of African Americans. Journal of Personality and Social
Psychology, 69(5), 797–811. https://doi.org/10.1037/0022-3514.69.5.797
Stephens, N. M., Fryberg, S. A., & Markus, H. R. (2012). It’s your choice: How the
middle-class model of independence disadvantages working-class Americans. In
S. T. Fiske & H. R. Markus (Eds.), Facing social class: How societal rank
influences interaction (pp. 87–106). New York, NY, US: Russell Sage
Foundation.
Stephens, N. M., Fryberg, S. A., Markus, H. R., Johnson, C. S., & Covarrubias, R.
(2012). Unseen Disadvantage: How American Universities’ Focus on
Independence Undermines the Academic Performance of First-Generation
College Students. Journal of Personality and Social Psychology, 102(6), 1178–
1197. https://doi.org/10.1037/a0027143
Stephens, N. M., Hamedani, M. G., & Destin, M. (2014). Closing the Social-Class
Achievement Gap: A Difference-Education Intervention Improves First-
Generation Students’ Academic Performance and All Students’ College
Transition. Psychological Science, 25(4), 943–953.
https://doi.org/10.1177/0956797613518349
Stephens, N. M., Markus, H. R., & Fryberg, S. A. (2012). Social Class Disparities in
Health and Education: Reducing Inequality by Applying a Sociocultural Self
Model of Behavior. Psychological Review, 119(4), 723–744.
https://doi.org/10.1037/a0029028
Strand, S. (2014). Ethnicity, gender, social class and achievement gaps at age 16:
intersectionality and ‘getting it’ for the white working class. Research Papers in
Education, 29(2), 131–171. https://doi.org/10.1080/02671522.2013.767370
Tibbetts, Y., Harackiewicz, J. M., Canning, E. A., Boston, J. S., Priniski, S. J., &
Hyde, J. S. (2016). Affirming Independence: Exploring Mechanisms Underlying
a Values Affirmation Intervention for First-Generation Students. Journal of
Personality and Social Psychology, 110(5), 635–659.

SELF-AFFIRMATION REDUCES THE SOCIOECONOMIC GAP 36
https://doi.org/10.1037/pspa0000049
Walton, G. M., & Cohen, G. L. (2007). A question of belonging: Race, social fit, and
achievement. Journal of Personality and Social Psychology, 92(1), 82–96.
https://doi.org/10.1037/0022-3514.92.1.82
Walton, G. M., & Spencer, S. J. (2009). Latent Ability. Grades and Test Scores
Systematically Underestimate the Intellectual Ability of Negatively Stereotyped
Students. Psychological Science, 20(9), 1132–1139.
https://doi.org/10.1111/j.1467-9280.2009.02417.x
Zeedyk, M. S., Gallacher, J., Henderson, M., Hope, G., Husband, B., & Lindsay, K.
(2003). Negotiating the transition from primary to secondary school: Perceptions
of Pupils, Parents and Teachers. School Psychology International, 24(1), 67–79.
https://doi.org/10.1177/0143034303024001010