| |

Received: 23 November 2021    Revised: 3 May 2022    Accepted: 30 June 2022

DOI: 10.1111/jopy.12752

ORIGINAL ARTICLE

Daily fluctuations in social status, self-­esteem, and

clinically relevant emotions: Testing hierometer theory and
social rank theory at a within-­person level

Nikhila Mahadevan1   | Aiden P. Gregg2  | Constantine Sedikides2

1
Department of Psychology, University
of Essex, Colchester, UK Abstract
2
Center for Research on Self and Introduction: Grounded in hierometer theory and social rank theory, this re-
Identity, School of Psychology, search examined how within-­person fluctuations in social status relate to within-­
University of Southampton,
Southampton, UK
person fluctuations in self-­esteem and several clinically relevant emotions. Both
hierometer theory and social rank theory postulate that particular psychological
Correspondence mechanisms help individuals to navigate social hierarchies adaptively. However,
Nikhila Mahadevan, Department
of Psychology, University of Essex, hierometer theory emphasizes self-­esteem, whereas social rank theory empha-
Wivenhoe Park, Colchester, CO4 3SQ, sizes emotions—­specifically, depression, anxiety, and shame.
UK.
Methods: We conducted a 10-­day diary study and analyzed the data using mul-
Email: [email protected]
tilevel modeling. Participants (N = 345) completed daily measures of their social
status, self-­esteem, depression, anxiety, shame, and guilt.
Results: On days when their status was higher, participants reported higher
self-­esteem and lower depression, anxiety, and shame. On days when their self-­
esteem was higher, participants reported lower depression, anxiety, and shame.
These patterns persisted after controlling for baseline individual differences.
Furthermore, multilevel mediation analyses indicated that daily self-­esteem me-
diated the links between daily status, and, individually, daily depression, anxiety,
and shame, but not guilt.
Conclusions: Supporting hierometer theory and social rank theory, self-­esteem,
and the clinically relevant emotions (except for guilt) appear to serve a status-­
tracking function. Self-­esteem plays a more primary role, accounting for the link
between status and depression, anxiety, and shame.

KEYWORDS
hierometer theory, social rank theory, social status, status, self-­esteem, emotion

1  |  I N T RO DU CT ION higher or lower in terms of their class (traditionally defined

as educational, occupational, and financial resources;
Social hierarchies are present across countries, cultures, Veblen,  1899) or power (traditionally defined as control
and groups and take a variety of forms (Sidanius & Pratto, over outcomes; Fragale et al., 2011). However, perhaps the
1999; Weber, 1946). For example, individuals can be ranked pre-­eminent form of hierarchy in human societies is status
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any
medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2022 The Authors. Journal of Personality published by Wiley Periodicals LLC.

Journal of Personality. 2022;00:1–18.  wileyonlinelibrary.com/journal/jopy  |  1

 |
2       MAHADEVAN et al.

(Anderson et al., 2015), traditionally defined as social re- relation to social situations (e.g., social status) sheds light
spect and admiration (Magee & Galinsky, 2008). The need on person-­ situation dynamics. It provides a more ho-
for status is considered to be a fundamental human mo- listic and nuanced understanding of the person within
tive (Anderson et al., 2015), alongside others, such as the the context of their social environment (Bleidorn,  2009;
need to belong (Baumeister & Leary, 1995). Accordingly, Fleeson, 2007). Third, findings at the between-­person level
status has been linked to a range of cognitive, emotional, do not automatically replicate at the within-­person level
and behavioral outcomes. For example, higher status indi- (and vice versa), and must therefore be independently
viduals receive more attention (Foulsham et al., 2010), feel addressed (Hamaker,  2012; Wilson et al.,  2017). For ex-
less threatened by others (Gregg et al., 2018), and behave ample, at the between-­person level, more anxious people
more assertively (Mahadevan et al., 2020). may also be more likely to see a therapist. That is, anxiety
and therapy may be positively correlated when assessed at
the between-­person level. However, at the within-­person
1.1  |  Within-­person variability level, a given person may be less anxious on days on which
they see a therapist. That is, anxiety and therapy may be
Research to date has predominantly examined status dif- negatively correlated when assessed at the within-­person
ferences at a between-­person level—­the extent to which level. Thus, the link between anxiety and therapy may
one individual is more respected and admired than an- differ, or even reverse, depending on whether one inves-
other on the whole. However, status differences can also tigates it at a between-­person level or at a within-­person
be examined at a within-­person level—­the extent to which level. Links at each level of analysis, therefore, need to be
the same individual is more respected and admired on independently addressed.
some occasions than on others. Yet, scarcely any studies Furthermore, the between-­ person versus within-­
have examined how within-­person differences in status person distinction has important theoretical implications.
relate to other important within-­person differences, such For example, both hierometer theory and social rank the-
as self-­esteem and clinically relevant emotions. Here, we ory (see below) postulate that psychological phenomena
sought to fill the gap. We rooted our investigation in two adaptively track one's position in a social hierarchy. But do
theories—­hierometer theory and social rank theory—­and they track one's position relative to that of other people in
tested hypotheses derived from each. Both theories pos- the hierarchy—­thereby requiring an interpersonal com-
tulate that psychological phenomena play a role in help- parison—­or do they track one's position relative to one's
ing individuals to navigate social hierarchies adaptively. own prior position in the hierarchy—­thereby requiring an
However, the two theories focus on distinct psychological intrapersonal comparison? In principle, the theories could
phenomena. Hierometer theory emphasizes self-­regard—­ be interpreted as doing either or both. But again, one pos-
principally self-­esteem (Mahadevan et al., 2016), whereas sibility is not identical to the other, and the former has
social rank theory emphasizes various clinically relevant primarily been investigated. Hence, investigation of the
emotions—­principally depression, anxiety, and shame1 latter is warranted. All else equal, the evidence for hierom-
(Gilbert, 2000). We pioneered an investigation into how, eter theory, or for social rank theory, would be stronger if
over the course of several days, within-­person fluctuations the hypothesized links between social status and the spe-
in social status relate to (a) within-­person fluctuations in cific psychological phenomenon (i.e., self-­esteem or clin-
self-­esteem, and (b) within-­person fluctuations in depres- ically relevant emotion) held both between-persons and
sion, anxiety, shame, and guilt. within-persons. Accordingly, we set out to test these rela-
We considered the investigation of within-­person vari- tions. To begin, we outline hierometer theory and social
ability worthwhile for several reasons. First, such variabil- rank theory, and the evidence for them, so as to furnish
ity is independently informative (Fleeson,  2001, 2004). the theoretical rationale for our hypotheses.
For example, daily fluctuations in stress still predicted
daily fluctuations in state self-­esteem after controlling for
baseline trait levels of stress and self-­esteem (Giacomin 1.2  |  Hierometer theory and social
& Jordan, 2016), and daily fluctuations in state extraver- rank theory: Outline and evidence
sion and neuroticism still predicted daily fluctuations in
life satisfaction after controlling for baseline trait levels 1.2.1  |  Hierometer theory
of extraversion, neuroticism, and life satisfaction (Heller
et al., 2007). Thus, assessing within-­person variability, in Outline
addition to between-­person stability, provides additional Hierometer theory postulates that self-­esteem (or self-­
information that increases predictive validity. Second, regard more generally) serves a status-­regulating func-
an examination of how self-­esteem and emotion vary in tion (Mahadevan et al., 2016). When a particular need is
MAHADEVAN et al.   
   3|

fundamental, one or more mechanisms are likely to evolve et al.,  2016). Experimental evidence, moreover, provides
to regulate its satisfaction. For example, hunger operates as more telling support for self-­esteem's posited indicative
an evolved mechanism to regulate satisfaction of the need function. Specifically, manipulating status to make it
for food (Baumeister & Leary,  1995; Leary et al.,  1995). higher or lower causes state self-­esteem to become higher
Likewise, because the need for status is fundamental, it is or lower, respectively (Mahadevan et al.,  2021). This ef-
plausible that one or more mechanisms evolved to help in- fect, moreover, occurs independently of the orthogonal
dividuals to regulate its attainment (Anderson et al., 2015; manipulation of inclusion, which exerts a parallel effect
Gregg & Mahadevan, 2014). Hierometer theory postulates on state self-­esteem (Mahadevan et al., 2019a).
that self-­esteem is a gear in one such mechanism. It helps One study also tested hierometer theory at the within-­
individuals to navigate status hierarchies adaptively. But person level (Mahadevan et al.,  2020). Adopting a daily
why might such a gear be needed? diary design, this study found that daily fluctuations in
The reason is that, although high status is desirable status moved in tandem with daily fluctuations in self-­
and offers many advantages, its pursuit entails potential esteem and assertiveness. However, this study did not aim
costs as well as benefits (Anderson et al., 2008; Frank & to test social rank theory, and so it did not examine how
Cook, 2013; Van Tilburg & Mahadevan, 2020). In particu- daily fluctuations in status relate to daily fluctuations in
lar, contests that come with the prospect of gaining status depression, anxiety, and shame. Furthermore, the study
also come with the prospect of losing it (Ridgeway, 2014; did not examine the interrelations among daily fluctu-
Ridgeway & Berger, 1986). For example, an individual who ations in status, self-­esteem, and these emotions (e.g.,
quits their job to enter a high-­stakes singing competition whether daily fluctuations in self-­esteem account for the
like American Idol risks losing, not only time and money, links between daily fluctuations in status and daily fluctu-
but also respect and admiration, if their performance is ations in depression, anxiety, and shame). Thus, how sta-
below par. Thus, the indiscriminate pursuit of status is tus, self-­esteem, and clinically relevant emotions mutually
unlikely to be beneficial; one or more mechanisms to reg- interrelate at a within-­person level remains unknown and
ulate its pursuit would be adaptive. in need of explication.
Self-­esteem, as the key gear in that mechanism, is the-
orized to serve two interrelated functions (Mahadevan
et al., 2020). First, it signals to the individual what their 1.2.2  |  Social rank theory
overall status is. This function may be termed indicative
(from the Latin indicare, “to point out”). Higher status, Outline
in the form of greater respect and admiration, is posited Social rank theory also postulates that psychological phe-
to raise self-­esteem, whereas lower status, in the form nomena play a functional role in helping individuals to
of reduced respect and admiration, is posited to lower navigate social hierarchies (Gilbert,  2000). It is one of
it. Second, self-­esteem regulates an individual's status-­ several theoretical perspectives that attempt to explain
seeking behavior. This function may be termed imperative how seemingly maladaptive emotions have continued to
(from the Latin imperare, “to command”). Higher self-­ persist within the human population (Price et al., 2007).
esteem is posited to augment readiness to compete, in the Here, the main gears in the underlying mechanism are
form of interpersonal assertiveness, whereas lower self-­ posited to be clinically relevant emotions—­such as de-
esteem is posited to diminish it, in the form of interper- pression, anxiety, and shame—­whose evolutionary roots
sonal acquiescence.2 can be plausibly traced to the ritual agonistic encounters
that occur in nonhuman animals (Price et al., 1994). Such
Evidence encounters are defined as “stereotyped interaction[s] be-
Given its relative novelty, hierometer theory has been tween two (or more) individuals that [start] with a sym-
tested predominantly at the between-­person level. Here, metric exchange of threat signals and [end] with escape
empirical evidence so far supports it. For example, in or submission by one of the individuals … In subsequent
several cross-­sectional studies, higher status (defined as encounters, the loser defers to the winner without contest-
respect and admiration) correlates strongly with higher ing the issue (Sloman & Price, 1987, p. 100).”
trait self-­esteem (r ≈ 0.60), and this link persists even after During a ritual agonistic encounter, an organism faces
controlling for social inclusion (defined as liking and ac- an ongoing choice: to compete or to concede. Many, if not
ceptance; rp ≈ 0.35; Gregg et al.,  2017; Huo et al.,  2010; most, such encounters terminate without open combat
Mahadevan et al., 2016, 2019a, 2019b).3 In addition, cross-­ when one party decides to submit and signals its submis-
sectional studies find that higher status correlates mod- sion (e.g., by baring its neck), with the other party duly ac-
erately with greater assertiveness (r ≈ 0.40), and that trait knowledging this submission and signaling its dominance
self-­
esteem statistically mediates the link (Mahadevan (e.g., by strutting; Lorenz,  1981, 1996). Such a strategy
 |
4       MAHADEVAN et al.

enables likely losers to survive and avoid serious injury examined college athletes' emotions upon winning or los-
and enables likely winners to save time and energy from ing a sporting contest, finding that dysphoria and anxiety
fighting unworthy opponents (Smith & Parker, 1976). The were higher in the latter case, particularly if participants
collective result is the formation of hierarchies in which were dispositionally prone to neuroticism or self-­criticism
differentially successful organisms occupy semi-­ stable (Sturman & Mongrain, 2007).
ranks (Broom, 2002; Schjelderup-­Ebbe, 1975).
Social rank theory proposes that the negative emotions
of depression, anxiety, and shame evolved in response to 1.3  |  Derivation of hypotheses
these ritual agonistic encounters and operate as a primi-
tive mechanism of submission. According to this theory, 1.3.1  |  The functional roles of self-­esteem,
many of the symptoms and behaviors exhibited by de- depression, anxiety, and shame
pressed, anxious, and shame-­ridden individuals resemble
the submissive behaviors observed in subordinate animals Both hierometer theory and social rank theory posit that
(e.g., social withdrawal, eye-­contact avoidance, lack of psychological phenomena help individuals to navigate
motivation, timid manner, and attempts to hide and es- social hierarchies adaptively. However, as noted above,
cape; Gilbert, 2000; Price et al., 1994, 2007). These particu- hierometer theory focuses on self-­esteem, whereas social
lar emotions are theorized to operate as part of an evolved rank theory focuses on clinically relevant emotions. In ad-
involuntary defeat strategy, which may be automatically dition, the theories differ subtly in other ways. For exam-
triggered by the prospect of competitive loss and low rank. ple, hierometer theory specifies its input variable precisely
Such an involuntary defeat strategy serves to override any (i.e., status: being respected and admired), whereas social
inclination an organism might have to keep competing, rank theory leaves its nature open (i.e., status, class, or
thereby preventing them from challenging stronger oppo- power can all be a type of social rank). Nonetheless, we
nents and risking injury or death (Price & Sloman, 1987; derive the following general hypotheses. From hierometer
Sloman,  2008). In this way, these negative emotions are theory: We hypothesize that self-­esteem will track status
purported to offer a potential adaptive benefit: They pre- at a within-­person level. That is, people will experience
vent the organism from pointlessly engaging in compet- higher self-­esteem on days when their status is higher.
itive behavior under unfavorable conditions (Bergstrom From social rank theory: We hypothesize that depression,
& Meacham,  2016; Price et al.,  2004). Thus, depression, anxiety, and shame will track status at a within-­person
anxiety, and shame are said to be—­at least sometimes and level. That is, people will experience greater depression,
in part—­adaptive responses to losing situations and low anxiety, and shame on days when their status is lower.
rank that put a brake on fruitless competitiveness (Nesse Note that both hypotheses pertain solely to the indicative
& Ellsworth, 2009; Price et al., 2004). function of the psychological phenomena in question.

Evidence
Social rank theory has been tested predominantly at the 1.3.2  |  The primacy of self-­esteem
between-­ person level. Much correlational evidence is
consistent with the theory. For example, negative social Yet an issue remains unresolved: How should status,
comparisons (e.g., rating oneself as less attractive, likable, self-­
esteem, and clinically relevant emotion mutually
and competent than others) correlate with greater de- interrelate at the within-­person level? Here, we propose
pression, social anxiety, and shame (Aderka et al., 2009; that self-­esteem operates as the more primary gear in the
Gilbert,  2000). In addition, lower socioeconomic class adaptive mechanism. In particular, we propose that fluc-
correlates with higher rates of depression and anxiety tuations in status will trigger fluctuations in self-­esteem
(Lorant et al., 2003, 2007). Some experimental and expe- before they trigger fluctuations in clinically relevant emo-
rience sampling studies also provide indirect support for tions, and, furthermore, that fluctuations in self-­esteem
the theory. For instance, success versus failure feedback will trigger those daily fluctuations in clinically relevant
in a game led to pride and shame, respectively, but only emotions. If so, then daily fluctuations in self-­esteem will
when the feedback was characterized as implying social mediate the links between daily fluctuations in status and
rank (Rebar & Conroy, 2013). In addition, an experience daily fluctuations in depression, anxiety, and shame.
sampling study found a pattern reminiscent of an invol- Why should self-­esteem be expected to play a more
untary defeat strategy: depressed participants, but not primary role? We submit that there are at least two rea-
non-­depressed ones, felt more inferior and behaved more sons: (1) self-­esteem is structurally simpler than emotion;
submissively if their partners behaved more dominantly and (2) self-­esteem is liable to operate more swiftly than
(Zuroff et al., 2007). Finally, one ecologically valid study emotion. As regards structural simplicity, self-­esteem is
MAHADEVAN et al.   
   5|

typically conceptualized as a unitary and overall evalu- leading formulation (Lewis, 1971), shame involves feeling
ation of the self, which is positive or negative to some small, inferior, and wanting to escape—­thereby prompt-
degree (Rosenberg,  1965; Sedikides & Gregg,  2003). In ing interpersonal withdrawal. In contrast, guilt involves
contrast, emotions typically involve, not only an eval- feeling remorse, empathy, and wanting to atone—­thereby
uation of some target (oneself, another person, or the prompting interpersonal reparation. Thus, given the more
world), but a mix of components: affective experience, prosocial and active character of guilt, it is a less natural
physiological changes, and facial expressions (Barrett candidate for featuring as part of the involuntary defeat
et al.,  2016; Ekman & Davidson,  1994). In general, the strategy and prompting the passive behavioral acquies-
more semi-­ connected parts something has, the more cence that would suppress competitive status-­ seeking
time and coordination is required for it to operate. As re- (Gilbert,  2003). Accordingly, we did not expect guilt to
gards swiftness of operation, global evaluations, includ- track status in the same manner as shame (and depression
ing about the self, are known to be made automatically and anxiety). We expected that daily fluctuations in status
and immediately (Chatard et al.,  2017; Ferguson,  2007; would covary negatively with daily fluctuations in shame
Gebauer et al., 2012). In contrast, the influence of emo- (after the empirical overlap with guilt was taken into ac-
tions on behavior is mostly cumulative and indirect count), but not with daily fluctuations in guilt (after the
(Baumeister et al.,  2007; Hermans et al.,  2001). Hence, empirical overlap with shame was taken into account).
self-­esteem, being likely to change first, is more likely Hence, our research approach offered the possibility of
to shape emotion than vice versa. Some longitudinal ev- examining hierometer theory and social rank theory, not
idence is consistent with this proposition. For example, only convergently in terms of the psychological variables
trait self-­esteem predicts depression considerably bet- they invoke, but also discriminantly, in terms of a cognate
ter than vice versa (i.e., validating the “vulnerability” psychological variable they do not invoke.
as opposed to the “scar” model), although the pattern
for anxiety is less pronounced (Sowislo & Orth,  2013;
Steiger et al., 2015). More recent research has replicated 2  |  OV ERVIEW AND HYPOTH E S E S
the predictive asymmetry over time for trait self-­esteem
and depression, and further found that the former pre- We tested 10 hypotheses derived from both hierometer
dicts peer victimization via the latter (Saint-­Georges & theory and social rank theory at the within-­person level.
Vaillancourt,  2020). As regards the link between self-­ Based on hierometer theory, we hypothesized that (1) on
esteem and emotion at the state level, we are unaware days when participants' status was higher (vs. lower), they
of any comparative investigations of relative directional would exhibit higher (vs. lower) self-­esteem. Based on so-
potency. What data there are merely show that manip- cial rank theory, we hypothesized that (2) on days when
ulations of self-­esteem suffice to manipulate depressive participants' status was lower (vs. higher), they would ex-
mood (Coleman,  1975; Wilson & Krane,  1980) and vice perience more (vs. less) depression, anxiety, and shame.
versa (Brown & Mankowski, 1993; Smith & Petty, 1995). We also hypothesized that (3) on days when participants'
Nonetheless, in accord with our theoretical reasoning, we status was lower (vs. higher), they would experience more
hypothesized that self-­esteem would account for the link (vs. less) shame after controlling for guilt, but that (4)
between social status and clinically relevant emotions at participants would not experience more (vs. less) guilt
the within-­person level. after controlling for shame. We further hypothesized (5)
that these effects would persist even after controlling for
between-­person (i.e., trait-­level) individual differences in
1.3.3  |  A functional role for guilt? the same constructs (Geiser, 2013; Lee, 2014).
Next, we hypothesized that (6) on days when partici-
Hierometer theory and social rank theory both postulate pants' self-­esteem was lower (vs. higher), they would exhibit
that various psychological phenomena serve as gears in more (vs. less) depression, anxiety, and shame. We also hy-
an adaptive mechanism that regulates competitive status-­ pothesized that (7) on days when participants' self-­esteem
seeking (namely, self-­esteem, on the one hand, and depres- was lower (vs. higher), they would experience more (vs. less)
sion, anxiety, and shame, on the other). Yet neither theory shame after controlling for guilt, but that (8) participants
specifies any adaptive role for guilt. Guilt is a negative would not experience more (vs. less) guilt after controlling
emotion characterized by regret, remorse, and the motiva- for shame. We further hypothesized (9) that these effects
tion to make amends for one's transgressions (Tangney & would persist even after controlling for between-­person (i.e.,
Dearing, 2002). It conceptually overlaps, and empirically trait-­level) differences in the same constructs. Finally, we
correlates, with shame (Tangney et al., 1992). Nonetheless, hypothesized that (10) the relation between daily fluctua-
shame and guilt can be distinguished. According to one tions in social status, on the one hand, and daily fluctuations
 |
6       MAHADEVAN et al.

in depression, anxiety, and shame, on the other, would be include: “… respect my achievements” and “… admire
statistically mediated by daily fluctuations in self-­esteem. me” (1 = strongly disagree, 5 = strongly agree, M = 3.20,
We conducted a 10-­day daily diary study to test these SD  =  0.72, α  =  0.89). The social status questionnaire is
hypotheses. We analyzed the data using multilevel mod- structurally validated and shows good internal consist-
eling (MLM) that incorporated multilevel mediation ency (Huo et al., 2010; Mahadevan et al., 2019a, 2019b).
(Hayes,  2013). Note that this design can assess whether
the patterns of association that emerge are consistent or
inconsistent with these hypotheses, thereby supporting or 3.2.2  |  Self-­esteem
casting doubt on them. However, it cannot establish direc-
tional causal links. We assessed trait self-­esteem with the 10-­item Rosenberg
Self-­Esteem Scale (RSES; Rosenberg,  1965). The RSES
is the most frequently used measure of global trait self-­
3  |  M ET H OD esteem (Donnellan et al., 2015). Its internal consistency,
test–­retest reliability, convergent validity, and discrimi-
3.1  |  Participants and procedure nant validity have been amply demonstrated (Schmitt
& Allik,  2005). Sample items include: “I feel that I have
We advertised the study on university notice boards and a number of good qualities” and “At times I think I am
the university's participant recruitment intranet. The no good at all (reverse-­coded)” (1  =  strongly disagree,
study was open to adults (aged 18 and above) who were 5 = strongly agree; M = 3.32, SD = 0.77, α = 0.88).
fluent in English. Participants first completed a baseline
survey online. Then, they completed short daily surveys,
also online, every day for the next 10 days. All surveys 3.2.3  | Depression
were hosted on Qualtrics™ and could be completed over
the computer or mobile phone. Participants were emailed We assessed depression with the 20-­ item Centre for
the link to each daily survey at 4 p.m. each day, followed Epidemiological Studies Depression Scale (CES-­D;
by a reminder at 9 p.m. each day. Radloff, 1977). The CES-­D is a reliable and well-­validated
Participants comprised students who took part for course measure of depression in non-­ psychiatric populations
credit, along with non-­student volunteers. In total, 345 partic- (Cosco et al.,  2017). Sample items include: “I felt de-
ipants completed at least one survey resulting in 2,831 obser- pressed” and “I felt that I could not shake off the blues
vations. Of these, 328 participants (230 women, 94 men, and even with help from my family or friend” (1  =  never or
4 unidentified) completed the initial baseline survey. Their hardly ever; 2  =  occasionally, now, and then; 3  =  a good
ages ranged from 18 to 66  years (M  =  27.50, SD  =  10.67). deal of the time; 4  =  mostly or all of the time; M  =  2.05,
Sixty-­one percent were White, 13.1% Black, 10.1% South SD = 0.64, α = 0.93).
Asian, 1.8% Hispanic, 1.2% East Asian, and 12.8% some other
ethnicity. On average, participants completed 7.26 daily sur-
veys (SD = 3.57). This sample size allowed us to detect small-­ 3.2.4  | Anxiety
to-­medium effects of approximately r = 0.07 to 0.19 with a
high power of 95% at two-­tailed α = 0.05. We assessed anxiety with the 20-­ item trait version of
the State-­ Trait Anxiety Inventory (STAI; Spielberger
et al.,  1983). The STAI is a reliable and well-­validated
3.2  |  Baseline measures measure of anxiety (Spielberger,  1989). Sample items
include: “I feel tense” and “I worry over possible mis-
Participants completed the following trait measures as fortunes” (1 = not at all; 2 = a little; 3 = somewhat; and
part of the baseline survey. We averaged item scores for 4 = very much so; M = 2.27, SD = 0.65, α = 0.94).
all measures, with higher scores indicating higher levels
of the corresponding construct.
3.2.5  |  Shame and guilt

3.2.1  |  Social status We assessed general proneness to shame and guilt with the
16-­item Guilt and Shame Proneness Scale (GASP; Cohen
We assessed overall social status with an 8-­item question- et al.,  2011). The GASP presents participants with 16 hy-
naire (Mahadevan et al.,  2016). It began with the stem, pothetical scenarios along with possible reactions to each
“Most of the time I feel that people …” Sample items one. It asks participants to indicate the likelihood that they
MAHADEVAN et al.   
   7|

would react in each way described. The GASP exhibits good Six items referred to depression: “unhappy,” “miser-
internal consistency, construct validity, and predictive valid- able,” “depressed,” “downhearted,” “sad,” and “gloomy”
ity (Cohen et al., 2011). Sample items include: “A friend tells (1  =  not at all, 7  =  extremely; M  =  2.25, SD  =  1.61,
you that you boast a great deal. What is the likelihood that αmean = 0.97, αrange = 0.96–­0.98). Seven items referred to
you would stop spending time with that friend?” and “You anxiety: “worried,” “nervous,” “anxious,” “panicky,” “on
secretly commit a felony. What is the likelihood that you edge,” “tense,” and “stressed” (1 = not at all, 7 = extremely;
would feel remorse about breaking the law?” (1 = very un- M = 2.44, SD = 1.68, αmean = 0.96, αrange = 0.93–­0.97).
likely, 7 = very likely; shame: M = 4.87, SD = 1.31, α = 0.66;
guilt: M = 5.97, SD = 1.47, α = 0.74).
3.3.4  |  Daily shame and guilt

3.3  |  Daily diary measures We assessed daily shame and guilt using the 10-­item State
Shame and Guilt Scale (SSGS; Marschall et al., 1994). Five
Participants completed the following short daily measures items referred to shame: “I want to sink into the floor and
each day. They received the instruction: “Now, we would disappear,” “I feel small,” “I feel like I am a bad person,” “I
like you to reflect upon your day today. Think about what feel humiliated, disgraced,” and “I feel worthless, power-
occurred and how you felt and acted. There are no right or less” (1 = not feeling this way at all, 7 = feeling this way very
wrong answers. Please answer honestly. Your responses strongly; M = 1.91, SD = 1.51, αmean = 0.94, αrange = 0.92–­
are strictly confidential.” 0.96). Another five items referred to guilt: “I feel remorse,
regret,” “I feel tension about something I have done,” “I
cannot stop thinking about something bad I have done,”
3.3.1  |  Daily status “I feel like apologizing, confessing,” and “I feel bad about
something I have done” (1  =  not feeling this way at all,
We assessed daily status with seven items (Giacomin & 7  =  feeling this way very strongly; M  =  1.81, SD  =  1.47,
Jordan,  2016; Mahadevan et al.,  2020): “Were you as- αmean = 0.96, αrange = 0.93–­0.98).
signed to an important role in a group?”, “Did you receive
any recognition?”, “Did you feel that people respected
you?”, “Did you feel that someone admired you?”, “Did 3.4  |  Data analytic approach
people treat you as someone important?”, “Did people
seem to think highly of your abilities and talents?”, and The data in this study followed a nested structure
“Did you feel that people saw you as someone success- such that the daily surveys (Level 1) were nested
ful?” (1 = yes, 0 = no; M = 0.54, SD = 0.36, αmean = 0.87, within participants (Level 2). Accordingly, we used
αrange = 0.77–­0.91).4 multilevel modeling (MLM) to analyze the data. This
analytic approach was also theoretically motivated.
We aimed to examine (a) whether and to what extent
3.3.2  |  Daily self-­esteem daily within-­p erson fluctuations in three types of
variable—­s ocial status, self-­e steem, and emotion—­
We assessed daily self-­esteem with three items: “How existed, and (b) whether and to what extent these
do you feel about yourself?” (1  =  very negative, 8  =  very daily within-­p erson fluctuations corresponded with
positive), “How do you feel about yourself?” (1  =  very hypotheses derived from hierometer theory and so-
bad, 8  =  very good), and “I have high self-­ esteem” cial rank theory. We did not formulate hypotheses
(1  =  strongly disagree, 8  =  strongly agree; M  =  5.41, regarding the durability or time onset of effects (e.g.,
SD = 1.85, αmean = 0.95, αrange = 0.92–­0.96). These items the impact of status on depression from one day to
have been used in previous research and found to be reli- the next), nor did we hypothesize specific trends
able and valid measures of state self-­esteem (Mahadevan over time (e.g., a linear or quadratic increase in sta-
et al., 2020; Robins et al., 2001). tus as days progressed). Accordingly, MLM repre-
sented the most appropriate analytic technique to
test our hypotheses (as opposed to, say, cross-­l agged
3.3.3  |  Daily depression and anxiety analyses or growth modeling). In addition, there
was no minimum number of daily surveys that par-
We assessed daily depression and anxiety with the Profile ticipants needed to have completed in order to be
of Mood States -­Revised (POMS-­R; McNair et al.,  1992). included in the analyses, given that MLM allows for
 |
8       MAHADEVAN et al.

the analysis of all available data and is capable of 4  |   RESULTS

handling participants with missing data (weaknesses
that can compromise other analytic techniques, such 4.1  |  Within-­person variability
as cross-­lagged analyses; Kearney,  2017; Snijders &
Bosker, 2004). The use of MLM, moreover, was con- Before proceeding to the main analyses, we quantified the
sistent with past research that has addressed simi- extent to which each of the six key constructs fluctuated
lar questions (Giacomin & Jordan, 2016; Mahadevan within-­person over the 10-­day period. A significant pro-
et al., 2020; Wilson et al., 2017). portion of the variance in all constructs occurred within-­
We used random-­ intercept MLM analyses (Singer,  ​ person (status: 50%, self-­esteem: 42%, depression: 45%,
1998), which allowed us to partition variance in a depen- anxiety: 36%, shame: 37%, guilt: 43%). Thus, participants
dent variable (e.g., anxiety) at the between-­person level experienced considerable fluctuation in their status, self-­
and at the within-­person level. Between-­person variance esteem, and emotions on a day-­to-­day basis, above and
reflects the distribution of people's scores relative to the beyond between-­ person differences in the same vari-
population average, whereas within-­person variance re- ables. We report the zero-­order correlations between the
flects the distribution of people's daily scores relative to baseline trait variables and the daily state variables in
their overall score. Specifically, random-­intercept mod- Supporting Information (Tables S1–­S10).
els, illustrated here with a single predictor, follow the
structure:
4.2  |  Main analyses I: Status, self-­esteem,
Level 1 Yij = 𝛾 0j + 𝛾 10 Xij + eij where eij ∼ N 0,𝜎 2. and emotion
( )

Level 2 𝛾 0j = 𝛾 00 + u0j where u0j ∼ N 0,𝜏 2 In the first set of analyses, we examined the links among
( )

daily status, self-­esteem, and emotion at the within-­person

We treated the participant-­level intercept as a random effect level to test hypotheses from hierometer theory and social
(i.e., we included a variance component to represent the ef- rank theory. Specifically, we tested whether daily fluctua-
fect of participants; Singer, 1998). We entered all covariates tions in status covaried positively with daily fluctuations
as fixed effects.5 in self-­esteem and negatively with daily fluctuations in de-
Following an approach recommended by Simmons pression, anxiety, and shame.6
et al.  (2011), we tested our hypotheses in three stages.
First, we computed the zero-­order daily within-­person as-
sociations by regressing the relevant outcome variables on 4.2.1  |  Zero-­order associations
the relevant predictor variables (e.g., daily anxiety on daily
status) in the multilevel analysis. Second, for daily shame Daily status covaried positively with daily self-­esteem. It
and guilt, we additionally controlled for daily fluctuations covaried negatively with daily depression, anxiety, and
in the other emotion (i.e., the association between daily shame (Table  1, upper panel). Thus, hypotheses (1) and
status and daily shame controlled for daily guilt, and the (2) received support. In addition, daily status covaried
association between daily status and daily guilt controlled negatively with daily guilt.
for daily shame). Third, we computed all the daily within-­
person associations further controlling for participants'
baseline scores on the relevant trait measures to deter- 4.2.2  |  Adjusted associations I
mine whether within-­person fluctuations in daily status
continued to predict within-­person fluctuations in daily Shame and guilt overlap considerably, both theoreti-
self-­esteem and emotion after taking these into account cally and empirically. Hence, we sought to determine
(Geiser, 2013; Lee, 2014). Finally, using multilevel medi- the independent association of daily status with daily
ation (Hayes, 2013), we tested whether daily fluctuations shame and the independent association of daily status
in self-­esteem mediated the associations between daily with daily guilt. After controlling for daily guilt, daily
fluctuations in status and daily fluctuations in depres- status still covaried negatively with daily shame. In
sion, anxiety, and shame. In all analyses, we controlled contrast, after controlling for daily shame, daily status
for measurement day (1–­10) to account for the possibil- no longer covaried negatively with daily guilt; instead,
ity that some days might be systematically linked to dif- a small positive association emerged between the two
ferent responses (e.g., people responding more positively (Table 1, middle panel). Thus, shame and guilt did not
on the last day; Bolger & Laurenceau,  2013; Snijders & relate to status in the same manner. Hypotheses (3) and
Bosker, 2004). (4) received support.
MAHADEVAN et al.   
   9 |
T A B L E 1   Within-­person associations among daily status, daily self-­esteem, and daily emotions

Covariates Predictor–­dependent variable γ SE t p 95% CI

None Daily status–­daily self-­esteem 1.80 0.09 20.28 <0.001 [1.63, 1.97]
Daily status–­daily depression −1.06 0.08 −12.82 <0.001 [−1.22, −0.90]
Daily status–­daily anxiety −0.75 0.08 −9.40 <0.001 [−0.91, −0.60]
Daily status–­daily shame −0.87 0.07 −12.11 <0.001 [−1.01, −0.73]
Daily status–­daily guilt −0.42 0.08 −5.58 <0.001 [−0.57, −0.27]
a
Daily indices Daily status–­daily shame −0.63 0.06 −11.13 <0.001 [−0.75, −0.52]
Daily status–­daily guilt 0.15 0.06 2.50 0.012 [0.03, 0.27]
Daily indices and Daily status–­daily self-­esteem 1.70 0.09 18.61 <0.001 [1.52, 1.88]
baselinesb Daily status–­daily depression −0.95 0.08 −11.33 <0.001 [−1.12, −0.79]
Daily status–­daily anxiety −0.66 0.08 −8.20 <0.001 [−0.82, −0.50]
Daily status–­daily shame −0.61 0.06 −10.29 <0.001 [−0.72, −0.49]
Daily status–­daily guilt 0.19 0.06 3.06 0.002 [0.07, 0.31]
a
Controlling for daily covariation in the concurrently assessed daily index (i.e., the association between daily status and daily shame controlled for daily guilt,
and the association between daily status and daily guilt controlled for daily shame).
b
Additionally controlling for between-­person trait levels of the relevant indices assessed at baseline (e.g., the association between daily status and daily shame
controlled for baseline levels of status, trait shame, and trait guilt).

4.2.3  |  Adjusted associations II 4.3.1  |  Zero-­order associations

We then added a further layer of statistical adjustment. Daily self-­esteem covaried negatively with daily depres-
Specifically, we examined the above set of associations, sion, anxiety, and shame (Table  2, upper panel). Thus,
but after additionally controlling for participants' base- hypothesis (6) received support. In addition, daily self-­
line scores on the relevant trait measures. Once again, esteem covaried negatively with daily guilt.
the same pattern of findings persisted (Table 1, bottom
panel). Thus, hypothesis (5) received support.
4.3.2  |  Adjusted associations I

4.2.4  | Summary Again, given the overlap between shame and guilt, we
sought to determine the independent association of
Supporting hierometer theory, on days when partici- daily self-­esteem with daily shame and the independ-
pants' status was higher (vs. lower), their self-­esteem was ent association of daily self-­esteem with daily guilt.
higher (vs. lower) too. Supporting social rank theory, on After controlling for daily guilt, daily self-­esteem still
days when participants' status was higher (vs. lower), covaried negatively with daily shame. In contrast, after
their depression, anxiety, and shame were lower (vs. controlling for daily shame, daily self-­esteem no longer
higher) too. These patterns also held after controlling for covaried negatively with daily guilt (Table  2, middle
baseline individual differences in the same constructs. panel). Thus, shame and guilt did not relate to self-­
esteem in the same manner. Hypotheses (7) and (8) re-
ceived support.
4.3  |  Main analyses II: Self-­esteem
and emotion
4.3.3  |  Adjusted associations II
In the second set of analyses, we examined the links
between daily self-­e steem and emotion at the within-­ We then added a further layer of statistical adjustment.
person level to test hypotheses from hierometer the- In particular, we examined the above set of associations,
ory and social rank theory. Specifically, we tested if but after additionally controlling for participants' base-
daily fluctuations in self-­e steem covaried negatively line scores on the relevant trait measures. Once again, the
with daily fluctuations in depression, anxiety, and same pattern of findings persisted (Table 2, bottom panel).
shame. Thus, hypothesis (9) received support.
 |
10       MAHADEVAN et al.

T A B L E 2   Within-­person associations between daily self-­esteem and daily emotions

Covariates Predictor–­dependent variable γ SE t p 95% CI

None Daily self-­esteem–­daily depression −0.49 0.02 −32.79 <0.001 [−0.52, −0.46]
Daily self-­esteem–­daily anxiety −0.39 0.02 −25.57 <0.001 [−0.42, −0.36]
Daily self-­esteem–­daily shame −0.40 0.01 −30.14 <0.001 [−0.43, −0.37]
Daily self-­esteem–­daily guilt −0.27 0.02 −18.28 <0.001 [−0.30, −0.24]
a
Daily indices Daily self-­esteem–­daily shame −0.27 0.01 −23.01 <0.001 [−0.29, −0.25]
Daily self-­esteem–­daily guilt −0.02 0.01 −1.26 0.208 [−0.05, 0.01]
Daily indices and Daily self-­esteem–­daily depression −0.46 0.02 −29.68 <0.001 [−0.49, −0.43]
baselinesb Daily self-­esteem–­daily anxiety −0.37 0.02 −23.34 <0.001 [−0.40, −0.33]
Daily self-­esteem–­daily shame −0.26 0.01 −21.28 <0.001 [−0.28, −0.23]
Daily self-­esteem–­daily guilt −0.01 0.02 −0.74 0.460 [−0.04, 0.02]
a
Controlling for daily covariation in the concurrently assessed daily index (i.e., the association between daily self-­esteem and daily shame controlled for daily
guilt, and the association between daily self-­esteem and daily guilt controlled for daily shame).
b
Additionally controlling for between-­person trait levels of the relevant indices assessed at baseline (e.g., the association between daily self-­esteem and daily
shame controlled for baseline levels of trait self-­esteem, trait shame, and trait guilt).

4.3.4  | Summary t(2,488) = −12.93, p < 0.001, 95% CI [−1.24, −0.91]. Daily

fluctuations in self-­esteem covaried negatively with daily
As hypothesized, on days when participants' self-­esteem fluctuations in depression, over and above daily fluctua-
was higher (vs. lower), their depression, anxiety, and tions in status, γ  =  −0.47, SE  =  0.02, t(2,487)  =  −28.87,
shame were lower (vs. higher) too. These patterns also p < 0.001, 95% CI [−0.50, −0.44], and vice versa, γ = −0.24,
held after controlling for baseline individual differences SE  =  0.08, t(2,487)  =  −3.01, p  =  0.003, 95% CI [−0.39,
in the same constructs.7 −0.08]. As a final step, we tested the indirect effect, indic-
ative of mediation, of daily fluctuations in status on daily
fluctuations in depression through daily fluctuations in
4.4  |  Main analyses III: Mediations by self-­esteem. The indirect effect (denoted as ab) was signif-
self-­esteem icant, ab = −0.84, 95% CI = [−0.94, −0.74].8 Thus, daily
fluctuations in self-­esteem statistically mediated the link
In the final set of analyses, we examined how daily sta- between daily fluctuations in status and daily fluctuations
tus, daily self-­esteem, and daily emotions interrelate at the in depression.
within-­person level. Specifically, using multilevel media-
tion (Hayes, 2013), we tested whether daily fluctuations in
self-­esteem statistically mediated the link between daily 4.4.2  |  Did daily self-­esteem mediate the
fluctuations in status, on the one hand, and daily fluctua- daily status–­daily anxiety link?
tions in depression, anxiety, and shame, on the other.
We then examined the mediating role of daily self-­esteem
in a model with daily status as the predictor, daily self-­
4.4.1  |  Did daily self-­esteem mediate the esteem as the mediator, and daily anxiety as the outcome
daily status–­daily depression link? variable. Daily fluctuations in status covaried negatively
with daily fluctuations in anxiety, γ =  −0.76, SE  =  0.08,
We examined the mediating role of daily self-­esteem in a t(2,463)  =  −9.49, p < 0.001, 95% CI [−0.92, −0.61]. In
model with daily status as the predictor, daily self-­esteem as addition, daily fluctuations in self-­ esteem covaried
the mediator, and daily depression as the outcome variable. negatively with daily fluctuations in anxiety, over and
We treated paths a (from the predictor to the mediator) and above daily fluctuations in status, γ  =  −0.38, SE  =  0.02,
b (from the mediator to the outcome) as fixed effects and t(2,462)  =  −23.17, p < 0.001, 95% CI [−0.42, −0.35], and
used the MCMED macro to construct 95% Monte Carlo con- vice versa, γ = −0.10, SE = 0.08, t(2,462) = −1.22, p = 0.224,
fidence intervals (CI) for the indirect effect (Hayes, 2013). 95% CI [−0.25, 0.06]. Importantly, the indirect effect was
Daily fluctuations in status covaried negatively with significant, ab  =  −0.69, 95% CI  =  [−0.78, −0.60]. Thus,
daily fluctuations in depression, γ  =  −1.08, SE  =  0.08, daily fluctuations in self-­esteem statistically mediated the
MAHADEVAN et al.    |
   11

link between daily fluctuations in status and daily fluctua- interrelate at the within-­person level. Specifically, they in-
tions in anxiety. dicate that daily social status relates to daily depression,
anxiety, and shame via daily self-­esteem. Thus, hypothesis
(10) received support. The pattern is consistent with self-­
4.4.3  |  Did daily self-­esteem mediate the esteem acting as the primary gear, and clinically relevant
daily status–­daily shame link? emotions as the subsequent ones, in the regulatory mecha-
nism jointly implied by hierometer theory and social rank
We then examined the mediating role of daily self-­esteem theory. The findings show that self-­esteem and emotion
in a model with daily status as the predictor, daily self-­ not only relate to social status as these theories would pre-
esteem as the mediator, and daily shame as the outcome dict, but also that self-­esteem and these clinically relevant
variable. Moreover, to examine the independent influ- emotions relate to one another in theoretically meaning-
ences of shame and guilt, we controlled for daily guilt in ful ways, with self-­esteem accounting for the link between
all analyses. Daily fluctuations in status covaried negatively social status and these clinically relevant emotions at the
with daily fluctuations in shame, γ  =  −0.64, SE  =  0.06, within-­person level.
t(2,488) = −11.16, p < 0.001, 95% CI [−0.75, −0.52]. Daily
fluctuations in self-­esteem covaried negatively with daily
fluctuations in shame, over and above daily fluctuations in 5  |  GENERAL DISC USSION
status, γ = −0.25, SE = 0.01, t(2,487) = −19.94, p < 0.001,
95% CI [−0.27, −0.23], and vice versa, γ = −0.22, SE = 0.06, The need for status, regarded as a fundamental human
t(2,487)  =  −3.92, p  =  0.003, 95% CI [−0.33, −0.11]. motive, has been linked to a range of cognitive, emo-
Importantly, the indirect effect was significant, ab = −0.41, tional, and behavioral outcomes (Anderson et al.,  2015;
95% CI  =  [−0.47, −0.35]. Thus, daily fluctuations in self-­ Magee & Galinsky, 2008). However, research to date has
esteem statistically mediated the link between daily fluc- focused predominantly on between-­person differences
tuations in status and daily fluctuations in shame. in status (i.e., the extent to which one person is, on the
whole, more respected and admired than another) as op-
posed to within-­person differences in status (i.e., the ex-
4.4.4  |  Did daily self-­esteem mediate the tent to which the same person is, on some occasions than
daily status–­daily guilt link? on others, more respected and admired). Here, we ex-
amined, for the first time, how social status, self-­esteem,
Finally, for exploratory purposes, we examined the mediat- and several emotions—­depression, anxiety, shame, and
ing role of daily self-­esteem in a model with daily status as guilt—­interrelate at the within-­person level. Moreover,
the predictor, daily self-­esteem as the mediator, and daily we rooted our investigation in two theories pertaining
guilt as the outcome variable. Moreover, to examine the to social hierarchies—­hierometer theory and social rank
independent influences of shame and guilt, we controlled theory—­and tested hypotheses derived from each.
for daily shame in all analyses. Daily fluctuations in status
did not covary negatively with daily fluctuations in guilt;
instead, a positive association emerged, γ = 0.15, SE = 0.06, 5.1  |  Summary of results
t(2,488) = 2.53, p = 0.011, 95% CI [0.04, 0.27]. Daily fluc-
tuations in self-­esteem also did not covary negatively with The results of a 10-­day diary study showed that status, self-­
daily fluctuations in guilt, over and above daily fluctuations esteem, depression, anxiety, shame, and guilt all exhibited
in status, γ = −0.03, SE = 0.02, t(2,487) = −1.79, p = 0.074, considerable within-­ person fluctuation across different
95% CI [−0.06, −0.003], nor vice versa, γ = 0.19, SE = 0.06, days. Importantly, these daily within-­person fluctuations
t(2,487)  =  2.96, p  =  0.003, 95% CI [0.06, 0.32]. Finally, covaried in a manner consistent with both hierometer
the indirect effect was not significant, ab  =  −0.04, 95% theory and social rank theory. Specifically, in line with
CI = [−0.07, 0.003]. Thus, daily fluctuations in self-­esteem hierometer theory, daily fluctuations in status covaried
did not statistically mediate the link between daily fluctua- positively with daily fluctuations in self-­esteem. Also, in
tions in status and daily fluctuations in guilt. line with social rank theory, daily fluctuations in status
covaried negatively with daily fluctuations in depression,
anxiety, and shame. In addition, providing some discrimi-
4.4.5  | Summary nant, as well as convergent validation, daily fluctuations
in status covaried negatively with daily fluctuations in
These mediational findings provide insight into how shame (after accounting for daily fluctuations in guilt),
daily status, self-­esteem, and clinically relevant emotions but not with daily fluctuations in guilt (after accounting
 |
12       MAHADEVAN et al.

for daily fluctuations in shame). Moreover, all foregoing established over the course of evolution, and triggered by
patterns remained even after controlling for baseline indi- competitive losses that lower one's rank in some hierar-
vidual differences in the relevant constructs. Finally, tests chy. Thus, as one's rank declines, levels of demotivating
of multilevel mediation indicated that daily fluctuations depression, anxiety, and shame should grow. To date, both
in self-­esteem explained the links between daily fluctua- theories have been tested mostly at a between-­person level
tions in status, on the one hand, and daily fluctuations (Mahadevan et al., 2016; Gilbert, 2000; but see Mahadevan
in depression, anxiety, and shame (but not guilt), on the et al.,  2020). One might argue, however, that an equally
other. This is consistent with self-­esteem playing the more stringent test of their validity can be achieved by looking
primary role as a tracker of status. at the same person across time rather than at different
people at the same time. Accordingly, our research con-
ducted such a within-­person test of both hierometer the-
5.2  |  Implications ory and social rank theory.
As it happened, the results that emerged fully sup-
Our findings highlight the fruitfulness of examining ported both hierometer theory and social rank theory at
status, self-­esteem, and clinically relevant emotions at the within-­person level (given that status is one key form
a within-­person level. First, each of our six constructs of social rank). Specifically, on days when their status
exhibited considerable observable fluctuation across dif- was higher or lower, the same participants in our study
ferent days (about one-­third to one-­half of the available experienced higher or lower levels of self-­esteem—­just
variance). This indicates that there is meaningful within-­ as in prior research on hierometer theory, different par-
person variability present in each construct, available to ticipants, whose status had been higher or lower, had ex-
be examined.9 Second, as we illustrated empirically, such hibited higher or lower levels of self-­esteem (Mahadevan
within-­person variability can prove predictive above and et al., 2016, 2019a, 2019b, 2021). Moreover, on days when
beyond between-­ person differences (i.e., Hypothesis their status was higher or lower, the same participants
[5] was confirmed). Third, our findings show that, at experienced lower or higher levels of depression, anxi-
a within-­person level, daily fluctuations in status, self-­ ety, and shame—­just as in prior research on social rank
esteem, and clinically relevant emotions are far from in- theory, different participants, who had ranked themselves
coherent; rather, just like their between-­person analogs, higher or lower relative to others on a range of dimen-
they exhibit intelligible patterns of covariation. Fourth, sions (e.g., attractiveness and likeability), had reported
these intelligible patterns are theoretically relevant. In lower and higher levels of depression, anxiety, and shame
particular, our findings add meaningfully to the grow- (Gilbert, 2000).
ing body of evidence that self-­esteem and some clinically Moreover, our findings have further implications, be-
relevant emotions play a functional role as intrapsychic cause they also addressed, for the first time at the within-­
trackers of social status. This is especially important person level, how status, self-­ esteem, and clinically
because—­as highlighted in the Introduction—­findings relevant emotions interrelate, thereby helping to theoret-
at a within-­person level need not replicate findings at ically integrate hierometer theory and social rank theory.
a between-­person level. Accordingly, parallel findings Specifically, we posited that self-­esteem—­being structur-
at each level are logically non-­redundant and mutually ally simpler than clinically relevant emotions and liable
confirming. to operate more swiftly—­would respond to levels of sta-
As it turned out, our findings provided good indepen- tus earlier, and would trigger clinically relevant emotions
dent support for both hierometer theory and social rank later. Stated otherwise, we posited that, although self-­
theory. To reiterate, hierometer theory posits that self-­ esteem and clinically relevant emotions would both ulti-
esteem plays a functional role in helping individuals to mately bridge that gap between social status or rank, and
navigate social hierarchies. It assumes that current levels interpersonal assertiveness or submission, self-­ esteem
of status—­specifically defined as respect and admiration—­ would serve as the first part of that bridge, and clinically
are a good predictor of the utility of additional status pur- relevant emotions as the second. Consistent with self-­
suit. Accordingly, it predicts that higher levels of status esteem playing this more primary role, we found that daily
will ultimately augment assertiveness, and lower levels fluctuations in self-­esteem statistically mediated the links
diminish it.10 As a precondition for achieving this imper- between daily fluctuations in status and daily fluctuations
ative function, however, self-­esteem must function to in- in depression, anxiety, and shame. Although experimental
dicatively track status. Likewise, social rank theory posits designs are required to settle the issue, our current find-
that emotions like depression, anxiety, and shame play a ings nonetheless add usefully to the cumulative findings
similar functional role. Specifically, it assumes that these bearing on the causal priority of self-­esteem and clinically
emotions operate as part of an involuntary defeat strategy, relevant emotions (Sowislo & Orth, 2013).
MAHADEVAN et al.    |
   13

Finally, our findings suggest that the emotions of such as other reports or observational methods. In addition,
shame and guilt, despite their similarities, are unlikely we did not assess or control for the number of social interac-
to serve the same psychological function. Shame related tions that participants had on a given day. It is possible that
to status and self-­esteem in the same way as depression the number of social interactions people have bears some
and anxiety (i.e., covaried negatively with it, day by day), relationship to the constructs we measured, and that having
whereas guilt showed a different pattern, at least when social interactions might be a vehicle that allows people to
controlling for its overlap with shame. One possibility is obtain the status that subsequently grants them higher self-­
that shame is a functional emotion when it comes to the esteem and more positive emotional states. Accordingly,
regulation of status or rank because it promotes interper- we recommend that future studies take this into account.
sonal withdrawal, whereas guilt is a functional emotion Another limitation was that the daily diary design, not being
when it comes to the regulation of inclusion or belong- experimental, precluded conclusions about causality. That
ingness because it promotes interpersonal reconciliation is, although the correlational and mediational patterns we
(Lewis,  1971; Tangney & Dearing,  2002). That is, where observed were consistent with the causal hypotheses derived
shame may be triggered by a loss of status, in a manner from hierometer and social rank theory, alternative explana-
addressed by social rank theory, guilt may be triggered by tions involving reverse causality or causal confounds cannot
a loss of inclusion, in a manner addressed by sociometer be ruled out. Note, however, that although correlational and
theory (Leary et al., 1995). This theory (at least in its orig- mediational tests cannot conclusively test causal hypothe-
inal form: see Mahadevan et al., 2016, 2019a, 2019b, 2020, ses (Maxwell & Cole, 2007), they can assist in corroborating
for further discussion) postulates that self-­esteem indic- them, especially when potentially falsifiable patterns are
atively tracks inclusion and then imperatively prompts theoretically specified in advance (Baumeister et al., 2003;
affiliative behavior if inclusion falls below a critical Fiedler et al., 2011). Finally, the cross-­cultural generalizabil-
value. Prima facie then, guilt would appear well-­suited ity of our research was limited. Although our study included
to act as an emotional gear in a sociometer mechanism participants of different ages, genders, and ethnicities, the
(Baumeister et al., 1994), perhaps intervening causally be- majority of its participants were White, female, and resident
tween decreases in social inclusion and increases in inter- in Western countries. Future research could address these
personal affiliation. associations in non-­White and non-­Western samples to im-
prove generalizability.

5.3  |  Strengths, limitations, and

future directions 6  |  CONC LUSIONS

The current research had several strengths. Theoretically, Some individuals generally enjoy higher status than other
it began by synthesizing disparate literature on social sta- individuals, but each individual also enjoys higher status on
tus, self-­esteem, and emotion, to provide a theoretically-­ some occasions than on others. The current research exam-
grounded investigation of how they interrelate. It ined how these within-­person fluctuations in status relate
proceeded to concurrently test hypotheses systematically to within-­person fluctuations in self-­esteem and clinically
derived from two theories—­hierometer theory and social relevant emotions. The patterns obtained were consistent
rank theory—­at a novel within-­person level, including with the predictions of both hierometer theory and social
after controlling for baseline individual differences in a rank theory: daily levels of self-­esteem, depression, anxiety,
multilevel model. The daily diary design also provided and shame rose and fell in tandem with daily levels of status,
an ecologically valid way to study the links among social suggestive of a functional tracking mechanism. Moreover,
status, self-­esteem, and emotion as they unfolded in the daily levels of self-­esteem accounted for the link between
course of people's everyday lives. Furthermore, the study status and these clinically relevant emotions, consistent
was highly powered. with self-­esteem playing a more primary role. Accordingly,
Yet the current research was also not without its lim- ups and downs in how much respect and admiration others
itations. The daily diary design relied on self-­reports that accord us go hand in hand with, and may be responsible for,
can be vulnerable to assorted response biases, including ups and downs in our self-­esteem and emotions.
social desirability and shared method variance (Podsakoff
et al., 2003). Although we attempted to minimize the like- AUTHOR CONTRIBUTIONS
lihood of such response biases (e.g., by emphasizing con- Nikhila Mahadevan and Aiden P. Gregg developed
fidentiality of responses and adopting different response the original theoretical framework with input from
formats for different measures), future research would do Constantine Sedikides. Nikhila Mahadevan designed and
well to implement additional methods beyond self-­reports, ran the study, analyzed the data, and wrote the first draft
 |
14       MAHADEVAN et al.

to consolidate an individual's position within the hierarchy. In con-

of the manuscript. Nikhila Mahadevan, Aiden P. Gregg, trast, sociometer theory posits that self-­esteem serves an inclusion-­
and Constantine Sedikides all contributed to revisions of regulating function, and that this function is compensatory. Lower
the manuscript. inclusion promotes lower self-­esteem, which fosters greater affilia-
tiveness, which serves to restore an individual's level of inclusion to
ACKNOWLEDGMENTS an optimum level. (See Mahadevan et al., 2016, 2019a, 2020, for an
We would like to thank Wijnand van Tilburg and the extensive discussion and series of parallel tests of hierometer theory
anonymous reviewers, whose thoughtful comments and and sociometer theory.)

suggestions significantly improved the manuscript. 3 Cross-­sectional studies also show that higher inclusion correlates
strongly with higher trait self-­esteem (r  ≈ 0.60), and that this link
persists after controlling for status (rp  ≈ 0.30; Huo et al.,  2010;
FUNDING INFORMATION
Mahadevan et al., 2016, 2019a, 2019b).
This research did not receive any specific grant from fund-
4 We calculated Cronbach's alphas separately for each daily survey.
ing agencies in the public, commercial, or not-­for-­profit
We report the average alpha (across the 10 daily surveys) along with
sectors. the range.
5 Level 1 of this model deals with within-­person differences. Yij refers
CONFLICT OF INTEREST to observed scores on day i of participant j. This observed score is
All authors consented to the submission of this manuscript. estimated with a regression function where γ0j is the intercept of
The authors declare no conflict of interest. The reported re- that participant and γ10 represents the regression slope for the pre-
search and analyses were not preregistered in an independ- dictor variable, Xij. The random-­effect (error) term eij captures the
ent, institutional registry. The data and analyses code will deviation from the predicted score for a specific day and person,
be deposited in a trusted digital repository upon publication. normally distributed with an average of 0 and variance of σ2. Level
2 of this model deals with between-­person differences. Each per-
son's regression intercept γ0j is a function of the overall regression
ETHICS STATEMENT
intercept γ00 plus a person's deviation from this overall intercept
All study procedures involving human participants expressed as u0j. This deviation, also called the random-­intercept, is
were in accordance with the ethical standards of the normally distributed with a mean of 0 and variance of τ2.
Institutional Research Board and with the 1964 Helsinki 6 Prior research has established—­on both theoretical and empirical
declaration and its later amendments or comparable ethi- grounds—­that social status and self-­esteem are distinct constructs
cal standards. Informed consent was obtained from all (Anderson et al.,  2015; Huo et al.,  2010). For example, the status
participants. measures focus on others, whereas the self-­esteem measures focus
on the self. These items also factor analyze on separate factors
ORCID (Mahadevan et al.,  2021). Nonetheless, we examined whether the
status and self-­esteem measures used in the current study were em-
Nikhila Mahadevan  https://orcid.
pirically distinct. We conducted two exploratory factor analyses: on
org/0000-0003-2626-1324 the baseline trait measures of status and self-­esteem, and on the daily
state measures of status and self-­esteem (Day 1). We used principal
ENDNOTES axis factoring with direct oblimin rotation—­a method that avoids
1 We use the terms depression and anxiety to refer to individual dif- artificially imposing a fixed number of factors and factorial indepen-
ferences that are continuous rather than to clinical conditions that dence (Costello & Osborne, 2005). For the baseline trait measures,
are categorical. The results of taxometric analyses generally support three factors emerged with eigenvalues greater than one. All the sta-
such a continuous conceptualization (Hankin et al., 2005; Sowislo tus items loaded highest on Factor 1 (average = 0.71, range = 0.58 to
& Orth, 2013). The social rank literature, which, for example, refers 0.80), whereas all the self-­esteem items loaded highest on Factors 2
to depression as depressed mood, depressive personality, and depres- and 3 (average = 0.64, range = 0.43 to 0.84). (Note: the precise factor
sive illness, evidently shares this conceptualization (Price et al., 1994, structure of the Rosenberg Self-­Esteem scale can vary, with more
2007). than one factor sometimes emerging [e.g., for forward-­scored vs.
2 Hierometer theory differs from another theory of self-­esteem's func- reverse-­scored items; Marsh et al., 2010]). Cross-­loadings were low
tion, namely, sociometer theory (Leary et al.,  1995). Whereas hier- (status items on Factors 2 and 3: average = 0.01, range = −0.12 to
ometer theory focuses on status (i.e., respect and admiration) and 0.11; self-­esteem items on Factor 1: average = 0.05, range = −0.06
status-­optimizing behavior (i.e., assertiveness), sociometer theory to 0.14). For the daily state measures, two factors emerged with ei-
focuses on inclusion (i.e., liking and acceptance) and inclusion-­ genvalues greater than one. All the daily status items loaded highest
optimizing behavior (i.e., affiliativeness; Leary et al.,  1998). Status on Factor 1 (average = 0.55, range = 0.37 to 0.70), whereas all the
and inclusion are conceptually and empirically distinct (Anderson daily self-­esteem items loaded highest on Factor 2 (average = 0.90,
et al., 2015). Status is agentic—­it involves ‘getting ahead’—­whereas range  =  0.75 to 0.98). Cross-­loadings were low (status items on
inclusion is communal—­it involves ‘getting along’ (“The Big Two”; Factor 2: average = −0.04, range = −0.25 to 0.06; self-­esteem items
Abele & Wojciszke, 2014; Mahadevan & Jordan, 2022). Hierometer on Factor 1: average = −0.01, range = −0.05 to 0.06). Thus, in line
theory posits that self-­esteem serves a status-­regulating function, with previous findings, the status and self-­esteem measures loaded
and that this function is consolidatory. Higher status promotes on separate factors, consistent with their being empirically distinct
higher self-­esteem, which fosters greater assertiveness, which serves constructs.
MAHADEVAN et al.    |
   15

7 We additionally analyzed these data using different analytical ap- Bleidorn, W. (2009). Linking personality states, current social roles
proaches and report the results of these analyses in Supporting and major life goals. European Journal of Personality, 23(6),
Information. Specifically, in one set of analyses, we analyzed the 509–­530. https://doi.org/10.1002/per.731
data using multilevel modeling with all predictor variables person Bolger, N., & Laurenceau, J. P. (2013). Intensive longitudinal meth-
mean-­centered. In the other set of analyses, we analyzed the data ods: An introduction to diary and experience sampling research.
using dynamic structural equation modeling with autoregressions. Guilford Press.
In both cases, the same pattern of findings emerged. Broom, M. (2002). A unified model of dominance hierarchy forma-
8 We treated the paths a and b as fixed effects; therefore, there was no tion and maintenance. Journal of Theoretical Biology, 219(1),
Level 2 covariance between these parameters. The simple ab prod- 63–­72. https://doi.org/10.1006/yjtbi.3109
uct was sufficient to quantify the indirect effect (Hayes, 2013). Brown, J. D., & Mankowski, T. A. (1993). Self-­esteem, mood, and
self-­evaluation: Changes in mood and the way you see you.
9 Or, to misquote John Donne, they illustrate how “No man is a con-
Journal of Personality and Social Psychology, 64(3), 421–­430.
stant, fixed in itself; each is a variable, deviating from the mean.”
https://doi.org/10.1037/0022-­3514.64.3.4211
10 The consolidatory dynamic postulated here is reminiscent of that Chatard, A., Bocage-­Barthélémy, Y., Selimbegovic, L., & Guimond,
articulated in Luke 13:12 (King James' Version): “For whosoever S. (2017). The woman who wasn't there: Converging evidence
hath, to him shall be given, and he shall have more abundance: but that subliminal social comparison affects self-­ evaluation.
whosoever hath not, from him shall be taken away even that he Journal of Experimental Social Psychology, 73, 1–­13. https://doi.
hath.” org/10.1016/j.jesp.2017.05.005
Cohen, T. R., Wolf, S. T., Panter, A. T., & Insko, C. A. (2011).
Introducing the GASP scale: A new measure of guilt and shame
REFERENCES proneness. Journal of Personality and Social Psychology, 100(5),
Abele, A. E., & Wojciszke, B. (2014). Communal and agentic con- 947–­966. https://doi.org/10.1037/a0022641
tent in social cognition: A dual perspective model. Advances Coleman, R. E. (1975). Manipulation of self-­ esteem as a deter-
in Experimental Social Psychology, 50, 195–­255. https://doi. minant of mood of elated and depressed women. Journal
org/10.1016/B978-­0-­12-­80028​4-­1.00004​-­7 of Abnormal Psychology, 84(6), 693–­700. https://doi.
Aderka, I. M., Weisman, O., Shahar, G., & Gilboa-­Schechtman, E. org/10.1037/0021-­843X.84.6.6933
(2009). The roles of the social rank and attachment systems Cosco, T. D., Prina, A. M., Stubbs, B., & Wu, Y.-­T. (2017). Reliability
in social anxiety. Personality and Individual Differences, 47(4), and validity of the Center for Epidemiologic Studies Depression
284–­288. https://doi.org/10.1016/j.paid.2009.03.014 Scale (CES-­ D) in a population-­ based cohort of US adults.
Anderson, C., Ames, D. R., & Gosling, S. D. (2008). Punishing hubris: Journal of Nursing Measurement, 25(3), 476–­485. https://doi.or
The perils of overestimating one's status in a group. Personality g/10.1891/1061-­3749.25.3.476
and Social Psychology Bulletin, 34(1), 90–­101. https://doi. Costello, A. B., & Osborne, J. W. (2005). Best practices in explor-
org/10.1177/01461​67207​307489 atory factor analysis: Four recommendations for getting the
Anderson, C., Hildreth, J. A. D., & Howland, L. (2015). Is the desire most from your analysis. Practical Assessment, Research &
for status a fundamental human motive? A review of the empir- Evaluation, 10(7), 1–­9. https://doi.org/10.7275/jyj1-­4868
ical literature. Psychological Bulletin, 141(3), 574–­601. https:// Donnellan, M. B., Trzesniewski, K. H., & Robins, R. W. (2015).
doi.org/10.1037/a0038781 Measures of self-­esteem. In G. J. Boyle, D. H. Saklofske, & G.
Barrett, L. F., Lewis, M., & Haviland-­Jones, J. M. (Eds.). (2016). Matthews (Eds.), Measures of personality and social psychologi-
Handbook of emotions (4th ed.). Guilford Press. cal constructs (pp. 131–­157). Elsevier Academic Press.
Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. Ekman, P., & Davidson, R. J. (Eds.). (1994). Series in affective sci-
(2003). Does high self-­ esteem cause better performance, ence. The nature of emotion: Fundamental questions. Oxford
interpersonal success, happiness, or healthier lifestyles? University Press.
Psychological Science in the Public Interest, 4(1), 1–­44. https:// Ferguson, M. J. (2007). The automaticity of evaluation. In J. A. Bargh
doi.org/10.1111/1529-­1006.01431 (Ed.), Frontiers of social psychology. Social psychology and the
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: Desire unconscious: The automaticity of higher mental processes (pp.
for interpersonal attachments as a fundamental human motiva- 219–­264). Psychology Press.
tion. Psychological Bulletin, 117(3), 497–­529. https://doi.org/10. Fiedler, K., Schott, M., & Meiser, T. (2011). What mediation analysis
1037/0033-­2909.117.3.497 can (not) do. Journal of Experimental Social Psychology, 47(6),
Baumeister, R. F., Stillwell, A. M., & Heatherton, T. F. (1994). Guilt: 1231–­1236. https://doi.org/10.1016/j.jesp.2011.05.007
An interpersonal approach. Psychological Bulletin, 115(2), 243–­ Fleeson, W. (2001). Toward a structure-­and process-­integrated view
267. https://doi.org/10.1037/0033-­2909.115.2.243 of personality: Traits as density distributions of states. Journal
Baumeister, R. F., Vohs, K. D., DeWall, C. N., & Zhang, L. (2007). of Personality and Social Psychology, 80(6), 1011–­1027. https://
How emotion shapes behavior: Feedback, anticipation, doi.org/10.1037/0022-­3514.80.6.1011
and reflection, rather than direct causation. Personality Fleeson, W. (2004). Moving personality beyond the person-­situation
and Social Psychology Review, 11(2), 167–­203. https://doi. debate: The challenge and the opportunity of within-­person
org/10.1177/10888​68307​301033 variability. Current Directions in Psychological Science, 13(2),
Bergstrom, C. T., & Meacham, F. (2016). Depression and anxiety: 83–­87. https://doi.org/10.1111/j.0963-­7214.2004.00280.x
Maladaptive by-­products of adaptive mechanisms. Evolution, Fleeson, W. (2007). Situation-­based contingencies underlying trait
Medicine, and Public Health, 2016(1), 214–­218. https://doi. content manifestation in behaviour. Journal of Personality, 75(4),
org/10.1093/emph/eow019 825–­862. https://doi.org/10.1111/j.1467-­6494.2007.00458.x
 |
16       MAHADEVAN et al.

Foulsham, T., Cheng, J. T., Tracy, J. L., Henrich, J., & Kingstone, A. Hermans, D., De Houwer, J., & Eelen, P. (2001). A time course analy-
(2010). Gaze allocation in a dynamic situation: Effects of social sis of the affective priming effect. Cognition and Emotion, 15(1),
status and speaking. Cognition, 117(3), 319–­331. https://doi. 143–­165. https://doi.org/10.1080/02699​93012​5768
org/10.1016/j.cogni​tion.2010.09.003 Huo, Y. J., Binning, K. R., & Molina, L. E. (2010). Testing an integra-
Fragale, A. R., Overbeck, J. R., & Neale, M. A. (2011). Resources ver- tive model of respect: Implications for social engagement and
sus respect: Social judgments based on targets' power and sta- well-­being. Personality and Social Psychology Bulletin, 36(2),
tus positions. Journal of Experimental Social Psychology, 47(4), 200–­212. https://doi.org/10.1177/01461​67209​356787
767–­775. https://doi.org/10.1016/j.jesp.2011.03.006 Kearney, M. W. (2017). Cross-­lagged panel analysis. In M. Allen (Ed.),
Frank, R. H., & Cook, P. J. (2013). Winner-­take-­all-­markets. Studies in The sage encyclopedia of communication research methods (pp.
Microeconomics, 1(2), 131–­154. https://doi.org/10.1177/23210​ 313–­314). Sage. https://doi.org/10.4135/97814​83381411
22213​501254 Leary, M. R., Haupt, A. L., Strausser, K. S., & Chokel, J. T. (1998).
Gebauer, J. E., Göritz, A. S., Hofmann, W., & Sedikides, C. (2012). Calibrating the sociometer: The relationship between interper-
Self-­
love or other-­ love? Explicit other-­ preference but im- sonal appraisals and the state self-­esteem. Journal of Personality
plicit self-­preference. PLOS ONE, 7(7), e41789. https://doi. and Social Psychology, 74(5), 1290–­1299. https://doi.org/10.103
org/10.1177/09567​97611​427045 7/0022-­3514.74.5.1290
Geiser, C. (2013). Methodology in the social sciences. Data analysis Leary, M. R., Tambor, E. S., Terdal, S. K., & Downs, D. L. (1995). Self-­
with Mplus. Guilford Press. esteem as an interpersonal monitor: The sociometer hypothe-
Giacomin, M., & Jordan, C. H. (2016). The wax and wane of nar- sis. Journal of Personality and Social Psychology, 68(3), 518–­530.
cissism: Grandiose narcissism as a process or state. Journal of https://doi.org/10.1037/0022-­3514.68.3.518
Personality, 84(2), 154–­164. https://doi.org/10.1111/jopy.12148 Lee, E. J. (2014). The relationship between unstable self-­esteem and
Gilbert, P. (2000). The relationship of shame, social anxiety aggression: Differences in reactive and proactive aggression.
and depression: The role of the evaluation of social rank. The Journal of Early Adolescence, 34(8), 1075–­1093. https://doi.
Clinical Psychology and Psychotherapy, 7(3), 174–­189. https:// org/10.1177/02724​31613​518973
doi.org/10.1002/1099-­0 879(20000​ 7 )7:3<174::AID-­C PP23​ Lewis, H. B. (1971). Shame and guilt in neurosis. International
6>3.0.CO;2-­U Universities Press.
Gilbert, P. (2003). Evolution, social roles and the differences Lorant, V., Croux, C., Weich, S., Deliège, D., Mackenbach, J., &
in shame and guilt. Social Research: A Quarterly Journal, Ansseau, M. (2007). Depression and socio-­economic factors: 7-­
70(4), 1205–­1230. https://doi.org/10.1002/1099-­0879(20000​ year longitudinal population study. British Journal of Psychiatry,
7)7:3<174::AID-­CPP23​6>3.0.CO;2-­U 190(4), 293–­298. https://doi.org/10.1192/bjp.bp.105.020040
Gregg, A. P., & Mahadevan, N. (2014). Intellectual arrogance and in- Lorant, V., Deliège, D., Eaton, W., Robert, A., Philippot, P., &
tellectual humility: An evolutionary-­epistemological account. Ansseau, M. (2003). Socioeconomic inequalities in depression:
Journal of Psychology and Theology, 42(1), 7–­18. https://doi. A meta-­analysis. American Journal of Epidemiology, 157(2), 98–­
org/10.1177/00916​47114​04200102 112. https://doi.org/10.1093/aje/kwf182
Gregg, A. P., Mahadevan, N., & Sedikides, C. (2017). Intellectual Lorenz, K. (1981). The foundations of ethology. Springer Science &
arrogance and intellectual humility: Correlational evidence Business Media.
for an evolutionary-­ embodied-­ epistemological account. Lorenz, K. (1996). On aggression. Routledge Press (Originally pub-
The Journal of Positive Psychology, 12(1), 59–­73. https://doi. lished 1963).
org/10.1080/17439​760.2016.1167942 Magee, J. C., & Galinsky, A. D. (2008). Social hierarchy: The
Gregg, A. P., Mahadevan, N., & Sedikides, C. (2018). Taking the high self-­reinforcing nature of power and status. The Academy
ground: The impact of social status on the derogation of ideo- of Management Annals, 2(1), 351–­398. https://doi.
logical opponents. Social Cognition, 36(1), 43–­77. https://doi. org/10.1080/19416​52080​2211628
org/10.1521/soco.2018.36.1.43 Mahadevan, N., Gregg, A. P., & Sedikides, C. (2019a). Is self-­regard
Hamaker, E. L. (2012). Why researchers should think “within-­ a sociometer or a hierometer? Self-­esteem tracks status and
person”: A paradigmatic rationale. In M. R. Mehl & T. S. Conner inclusion, narcissism tracks status. Journal of Personality and
(Eds.), Handbook of research methods for studying daily life (pp. Social Psychology, 116(3), 444–­466. https://doi.org/10.1037/
43–­61). Guilford Press. pspp0​000189
Hankin, B. L., Kassel, J. D., & Abela, J. R. Z. (2005). Adult attachment Mahadevan, N., Gregg, A. P., & Sedikides, C. (2019b). Where I am
styles and specificity of emotional distress: Prospective inves- and where I want to be: Perceptions of and aspirations for sta-
tigations of cognitive risk and interpersonal stress generation tus and inclusion differentially predict psychological health.
as mediating mechanisms. Personality and Social Psychology Personality and Individual Differences, 139, 170–­174. https://
Bulletin, 31(1), 136–­151. https://doi.org/10.1177/01461​67204​ doi.org/10.1016/j.paid.2018.10.041
271324 Mahadevan, N., Gregg, A. P., & Sedikides, C. (2020). The ups and
Hayes, A. F. (2013). Introduction to mediation, moderation, and downs of social life: Within-­person variations in daily status
process control analysis: A regression-­based approach. Guilford and inclusion differentially predict self-­regard and interper-
Press. sonal behavior. Journal of Personality, 88(6), 1111–­1128. https://
Heller, D., Komar, J., & Lee, W. B. (2007). The dynamics of personality doi.org/10.1111/jopy.12559
states, goals, and well-­being. Personality and Social Psychology Mahadevan, N., Gregg, A. P., & Sedikides, C. (2021). Self-­esteem as a
Bulletin, 33(6), 898–­910. https://doi.org/10.1177/01461​67207​ hierometer: Sociometric status is a more potent and proximate
301010 predictor of self-­esteem than socioeconomic status. Journal of
MAHADEVAN et al.    |
   17

Experimental Psychology: General, 150(12), 2613–­2635. https:// Ridgeway, C. L., & Berger, J. (1986). Expectations, legitimation, and
doi.org/10.1037/xge00​01056 dominance behavior in task groups. American Sociological
Mahadevan, N., Gregg, A. P., Sedikides, C., & de Waal-­Andrews, W. Review, 51(5), 603–­617. https://doi.org/10.2307/2095487
G. (2016). Winners, losers, insiders, and outsiders: Comparing Robins, R. W., Hendin, H. M., & Trzesniewski, K. H. (2001).
hierometer and sociometer theories of self-­regard. Frontiers in Measuring global self-­esteem: Construct validation of a single-­
Psychology, 7, 334. https://doi.org/10.3389/fpsyg.2016.00334 item measure and the Rosenberg self-­esteem scale. Personality
Mahadevan, N., & Jordan, C. (2022). Desperately seeking status: and Social Psychology Bulletin, 27(2), 151–­161. https://doi.
How desires for, and perceived attainment of, status and inclu- org/10.1177/01461​67201​272002
sion relate to grandiose and vulnerable narcissism. Personality Rosenberg, M. (1965). Society and the adolescent self-­image. Princeton
and Social Psychology Bulletin, 48(5), 704–­717. https://doi. University Press.
org/10.1177/01461​67221​1021189 Saint-­Georges, Z., & Vaillancourt, T. (2020). The temporal sequence
Marschall, D., Sanftner, J., & Tangney, J. P. (1994). The state shame of depressive symptoms, peer victimization, and self-­esteem
and guilt scale. George Mason University. across adolescence: Evidence for an integrated self-­perception
Marsh, H. W., Scalas, L. F., & Nagengast, B. (2010). Longitudinal driven model. Development and Psychopathology, 32(3), 975–­
tests of competing factor structures for the Rosenberg self-­ 984. https://doi.org/10.1017/S0954​57941​90008655
esteem scale: Traits, ephemeral artifacts, and stable response Schjelderup-­ Ebbe, T. (1975). Beitrage zur sozialpsychologie des
styles. Psychological Assessment, 22(2), 366–­381. https://doi. Haushuhns (contributions the social psychology of domestic
org/10.1037/a0019225 chickens) [M. Schleidt & W. M. Schleidt, trans]. In M. W. Schein
Maxwell, S. E., & Cole, D. A. (2007). Bias in cross-­sectional analyses (Ed.), Social hierarchy and dominance. Benchmark papers in
of longitudinal mediation. Psychological Methods, 12(1), 23–­44. animal behavior (Vol. 3, pp. 35–­49). Dowden, Hutchinson and
https://doi.org/10.1037/1082-­989X.12.1.23 Ross.
McNair, D. M., Lorr, M., & Droppleman, L. F. (1992). Profile of Schmitt, D. P., & Allik, J. (2005). Simultaneous administration of
mood states (POMS)–­revised manual. Education and Industrial the Rosenberg self-­esteem scale in 53 nations: Exploring the
Testing Service. universal and culture-­specific features of global self-­esteem.
Nesse, R. M., & Ellsworth, P. C. (2009). Evolution, emotions, and Journal of Personality and Social Psychology, 89(4), 623–­642.
emotional disorders. American Psychologist, 64(2), 129–­139. https://doi.org/10.1037/0022-­3514.89.4.623
https://doi.org/10.1037/a0013503 Sedikides, C., & Gregg, A. P. (2003). Portraits of the self. In M. A.
Podsakoff, P. M., MacKenzie, S. B., Lee, J.-­Y., & Podsakoff, N. P. Hogg & J. Cooper (Eds.), Sage handbook of social psychology
(2003). Common method biases in behavioral research: A (pp. 110–­138). Sage.
critical review of the literature and recommended remedies. Sidanius, J., & Pratto, F. (1999). Social dominance: An intergroup the-
Journal of Applied Psychology, 88(5), 879–­903. https://doi.org/ ory of social hierarchy and oppression. Cambridge University
10.1037/0021-­9010.88.5.879 Press.
Price, J. S., Gardner, R., Jr., Wilson, D. R., Sloman, L., Rohde, P., & Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2011). False-­positive
Erickson, M. (2007). Territory, rank and mental health: The his- psychology: Undisclosed flexibility in data collection and anal-
tory of an idea. Evolutionary Psychology, 5(3), 531–­554. https:// ysis allows presenting anything as significant. Psychological
doi.org/10.1177/14747​04907​00500305 Science, 22(11), 1359–­1366. https://doi.org/10.1177/09567​
Price, J. S., Gardner, R., & Erickson, M. (2004). Can depression, 97611​417632
anxiety and somatisation be understood as appeasement dis- Singer, J. D. (1998). Using SAS PROC MIXED to fit multilevel models,
plays? Journal of Affective Disorders, 79(1–­3), 1–­11. https://doi. hierarchical models, and individual growth models. Journal of
org/10.1016/S0165​-­0327(02)00452​-­4 Educational and Behavioral Statistics, 23(4), 323–­355. https://
Price, J. S., & Sloman, L. (1987). Depression as yielding behaviour: doi.org/10.3102/10769​98602​3004323
An animal model based on Schjelderup-­Ebbe's pecking order. Sloman, L. (2008). A new comprehensive evolutionary model of de-
Ethology and Sociobiology, 1(Supplement 1), 85–­98. https://doi. pression and anxiety. Journal of Affective Disorders, 106(3), 219–­
org/10.1016/0162-­3095(87)90021​-­5 228. https://doi.org/10.1016/j.jad.2007.07.008
Price, J. S., Sloman, L., Gilbert, P., Gardner, R., & Rohde, P. (1994). Sloman, L., & Price, J. (1987). Losing behaviour (yielding subrou-
The social competition hypothesis of depression. British tine) and human depression: Proximate and selective mecha-
Journal of Psychiatry, 164(3), 309–­315. https://doi.org/10.1192/ nisms. Ethology and Sociobiology, 8(Supplement 1), 99–­109.
bjp.164.3.309 https://doi.org/10.1016/0162-­3095(87)90022​-­7
Radloff, L. S. (1977). The CES-­D scale: A self-­report depression scale Smith, J. M., & Parker, G. A. (1976). The logic of asymmetric contests.
for research in the general population. Applied Psychological Animal Behaviour, 24(1), 159–­175. https://doi.org/10.1016/
Measurement, 1(3), 385–­401. https://doi.org/10.1177/01466​ S0003​-­3472(76)80110​-­8
21677​00100306 Smith, S. M., & Petty, R. E. (1995). Personality moderators of mood
Rebar, A. L., & Conroy, D. E. (2013). Experimentally manipulated congruency effects on cognition: The role of self-­esteem and
achievement goal state fluctuations regulate self-­ conscious negative mood regulation. Journal of Personality and Social
emotional responses to feedback. Sport, Exercise, and Psychology, 68(6), 1092–­1107. https://doi.org/10.1037//002
Performance Psychology, 2(4), 233–­249. https://doi.org/10.1037/ 2-­3514.68.6.10922
a0034645 Snijders, T. A. B., & Bosker, R. J. (2004). Multilevel analysis: An intro-
Ridgeway, C. L. (2014). Why status matters for inequality. American duction to basic and advanced multilevel modeling. Sage.
Sociological Review, 79(1), 1–­16. https://doi.org/10.1177/00031​ Sowislo, J. F., & Orth, U. (2013). Does low self-­esteem predict de-
22413​515997 pression and anxiety? A meta-­analysis of longitudinal studies.
 |
18       MAHADEVAN et al.

Psychological Bulletin, 139(1), 213–­240. https://doi.org/10.1037/ Wilson, A. R., & Krane, R. V. (1980). Change in self-­esteem and its
a0028931 effects on symptoms of depression. Cognitive Research and
Spielberger, C. D. (1989). State-­trait anxiety inventory: Bibliography Therapy, 4(4), 419–­421. https://doi.org/10.1007/BF011​782177
(2nd ed.). Consulting Psychologists Press. Wilson, R. E., Thompson, R. J., & Vazire, S. (2017). Are fluctuations
Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., & Jacobs, in personality states more than fluctuations in affect? Journal of
G. A. (1983). State-­trait anxiety inventory for adults. Consulting Research in Personality, 69, 110–­123. https://doi.org/10.1016/j.
Psychologists Press. jrp.2016.06.006
Steiger, A. E., Fend, H. A., & Allemand, M. (2015). Testing the vul- Zuroff, D. C., Fournier, M. A., & Moskowitz, D. S. (2007). Depression,
nerability and scar models of self-esteem and depressive symp- perceived inferiority, and interpersonal behavior: Evidence
toms from adolescence to middle adulthood and across gener- for the involuntary defeat strategy. Journal of Social and
ations. Developmental Psychology, 51(2), 236–­247. https://doi. Clinical Psychology, 26(7), 751–­778. https://doi.org/10.1521/
org/10.1037/a0038478 jscp.2007.26.7.751
Sturman, E. D., & Mongrain, M. (2007). The role of personality
in defeat: A revised social rank model. European Journal of
Personality, 22(1), 55–­79. https://doi.org/10.1002/per.653 SUPPORTING INFORMATION
Tangney, J. P., & Dearing, R. (2002). Shame and guilt in interpersonal Additional supporting information can be found online
relationships. Guilford Press.
in the Supporting Information section at the end of this
Tangney, J. P., Wagner, P., & Gramzow, R. (1992). Proneness to
article.
shame, proneness to guilt, and psychopathology. Journal
of Abnormal Psychology, 101(3), 469–­478. https://doi.
org/10.1037/0021-­843X.101.3.469
Van Tilburg, W. A. P., & Mahadevan, N. (2020). When imitating
successful others fails: Accidentally successful exemplars in- How to cite this article: Mahadevan, N., Gregg,
spire risky decisions and can hamper performance. Quarterly A. P., & Sedikides, C. (2022). Daily fluctuations in
Journal of Experimental Psychology, 73(6), 941–­956. https://doi. social status, self-­esteem, and clinically relevant
org/10.1177/17470​21819​895705 emotions: Testing hierometer theory and social
Veblen, T. (1899). The theory of the leisure class: An economic study in
rank theory at a within-­person level. Journal of
the evolution of institutions. The Macmillan Company.
Personality, 00, 1–18. https://doi.org/10.1111/
Weber, M. (1946). Class, status, party. In H. H. Gerth & C. W. Mills
(Eds.), From Max Weber: Essays in sociology (pp. 180–­ 195). jopy.12752
Oxford University Press.