1327169 PSPXXX10.1177/01461672251327169Personality and Social Psychology BulletinIp and Feldman research-article2025
Empirical Research Paper
The Complex Misestimation of Others’ Emotions: Underestimation of Emotional Prevalence Versus Overestimation of Emotional Intensity and Their Associations with Well-Being

Personality and Social Psychology Bulletin ﻿1–­ 19 © 2025 by the Society for Personality and Social Psychology, Inc Article reuse guidelines: sagepub.com/journals-permissions DhttOpsI://1d0o.i1.o1rg7/71/00.11147671/06174262165712322571132679169 journals.sagepub.com/home/pspb

Ho Ching Ip1 and Gilad Feldman1

Abstract Jordan et al., 2011, demonstrated that people underestimated the prevalence of others’ negative emotional experiences and that these were associated with higher well-being. We conducted a preregistered replication of Studies 1b and 3 by Jordan et al., 2011 (N = 594) with adjustments and added extensions. Building on their methodology, we examined both prevalence and intensity of emotional experiences, and our findings suggest a much more complex story with surprising effects. We found an underestimation of the prevalence of negative emotions, but also unexpectedly of an underestimation of the prevalence of positive emotions, with stronger effects for negative than for positive emotions. However, we found an opposite effect for emotional intensity; people overestimated the intensity of both positive and negative emotional experiences, again with stronger effects for negative. Surprisingly, associations between prevalence estimations and well-being were in the opposite direction to the target article’s. Materials, data, and code: https://osf.io/bwmtr/
Keywords emotional pluralistic ignorance, positive emotions, negative emotions, affect, preregistered replication, social comparison, well-being, emotional estimation error
Received August 22, 2023; revision accepted February 17, 2025

Introduction
Background
A growing body of literature has documented people’s misperceptions of others’ attitudes, beliefs, and behaviors (Miller & Prentice, 1994). These errors in perceiving others’ psyche and mental states seem at least in part related to purposeful public misrepresentations, as people misrepresent their private worlds to meet what they perceive social norms dictate. For example, people tend to underestimate their peers’ negative emotional experiences, as people tend to hide their negative emotions (Larson et al., 1982) to adhere to perceived display rules to appear positive (Ekman & Friesen, 1969).
Jordan et al. (2011) demonstrated that people underestimated the prevalence of peers’ negative emotional experiences more than their peers’ positive emotions,1 and that prevalence estimates were associated with a host of wellbeing measures. Their findings showed that those with lower prevalence estimates of negative emotions reported greater loneliness, greater rumination, and less satisfaction with life.

We aimed to revisit and extend Jordan et al. (2011)’s theory and findings. Our first goal was to conduct an independent well-powered preregistered close replication examining misestimation of emotional prevalence and their associations with psychological well-being, adopting and improving on the target article’s methodology. Our second goal was to add extensions and examine: (a) whether misestimations extend beyond prevalence to intensity, and (b) whether social comparison orientation moderates associations between prevalence and intensity and well-being.
We begin by introducing the literature on emotional misestimations and the chosen article for replication—Jordan et al. (2011). We then outline our chosen studies for replication from the target article, the target’s experimental designs, and our adaptations, improvements, and extensions.
1University of Hong Kong, Pok Fu Lam, Hong Kong SAR
Corresponding Author: Gilad Feldman, Department of Psychology, University of Hong Kong, Pok Fu Lam, Hong Kong SAR. Email: gfeldman@hku.hk; giladfel@gmail.com

2

Personality and Social Psychology Bulletin 00(0)

The Misestimation of Others’ Emotions
Misestimation of others’ views, behaviors, and experiences were initially mostly studied in the context of bystander nonintervention during emergencies (Allport, 1924; Darley & Latane, 1968). It was later developed to study people’s misinterpretations of social norms based on their observations of public behaviors, which may be misaligned with people’s true private feelings, especially when those are considered negative (Miller & Prentice, 1994).
One of the earliest demonstrations of the differences between public behaviors and private attitudes was in the 1930s, showing a misalignment between what people said in public compared to what they reported in private settings (Schanck, 1932). In a review by Sargent and Newman (2021), they summarized three areas in which this phenomenon has been demonstrated—drug use, alcohol, and sexual and dating norms. For instance, participants in the study of Miller and Prentice (1994) expressed in private that they were uncomfortable with excessive drinking, yet publicly endorsed pro-drinking policies, presumably due to misperceiving their peers’ behaviors and attitudes and being pro-drinking.
Jordan et al. (2011) extended the literature by demonstrating people’s misestimation of the prevalence of others’ negative emotions. The misestimation is likely due to people’s tendency to try to suppress their negative emotions in social contexts (Larson et al., 1982), given that they are deemed less socially appropriate and not in line with norms for social behavior (Ekman & Friesen, 1969).
Choice of Study for Replication: Jordan et al. (2011)
We aimed to revisit the classic phenomenon to examine the reproducibility and replicability of the findings with an independent, well-powered, preregistered close replication of Jordan et al. (2011) with several extensions. We answered recent calls and a growing recognition of the importance of reproducibility and replicability in psychological science (e.g., Nosek et al., 2022; Zwaan et al., 2018).
We chose the article by Jordan et al. (2011) as the target for replication based on several factors: (a) its academic and practical impact with no direct replications, (b) the potential for improvement on their methods and analyses, (c) the potential in answering new theoretical questions using an extension in examining misestimation of emotional intensity.
The article has had an impact on scholarly research in the area of social psychology, with 306 Google Scholar citations at the time of writing (August 2024). The article was followed by empirical research demonstrating misestimation and their impact in social media settings (Tandoc et al., 2015), job-seeking (Burke & Kraut, 2013), risky sexual behaviors (Young & Jordan, 2013), and impulsive buying

(Liu et al., 2019). Other research extended the original findings and further investigated the practical implications of the associations between misestimation and psychological wellbeing outcomes, such as for depressive symptoms (Steers et al., 2014) and self-esteem (Alfasi, 2019; Wang et al., 2017). To the best of our knowledge, there are currently no published direct independent replications of the chosen target article.
When we analyzed the article, we realized that some analyses were in need of revision and improvement. The analyses conducted for these studies were on an item-level, with six items for positive, and six items for negative, making it very challenging to detect any effects, which are related to what seemed like unreasonably large effects (d = 2.52 in Study 1b, and d = 3.16 in Study 3). It was not clear to us why the analyses were run only on an item-level rather than also on a participant-level. Given the target article’s null hypothesis for positive emotions, we were open to the possibility of also finding support for meaningful effects for positive emotions and that with a larger sample of items or analyses on a participant-level, the difference from the null might be more easily detected. Given our surprising findings we report below, we also sought an external expert reviewer to examine the regression analyses conducted in Study 3 of the target article (shared on https://osf.io/zy5qa/). In his analyses, the external reviewer raised the possibility of a suppression effect in the regression analyses between prevalence estimates and wellbeing (Sharpe & Roberts, 1997; Thompson & Levine, 1997). We, therefore, saw the potential in improving on the conducted analyses by also conducting participant-level analyses and focusing on raw correlations.
We also sought to go beyond examining the misestimation of emotional prevalence to also examine an important missing piece of the puzzle—the possible misestimation of emotional intensity. We built on the same methodology to examine both aspects of emotions together, resulting in a more comprehensive view of a much broader phenomenon.
Original Hypotheses and Findings in Target Article: Jordan et al. (2011)
Jordan et al. (2011) empirical work consisted of four studies, and in the current replication, we focused on Studies 1b and 3.
In Study 1b, they examined whether people would accurately predict the prevalence of others’ emotional experiences. Participants rated their own experiences and estimated the prevalence of six positive emotional experiences (such as receiving high grades or attending a fun party) and six negative emotional experiences (such as thinking about workload or having a fight/argument) in the 2 weeks prior. The researchers hypothesized that (a) people underestimate negative experiences, yet that (b) these errors are not present for positive experiences. We note that their second hypothesis regarding positive emotions was a null hypothesis, which we believe is better reframed to an

Ip and Feldman

3

Table 1.  Jordan et al. (2011) Studies 1b and 3: Summary of Findings.

Study 1b

Study 3

Experiences

Estimation errora

Average estimation

Estimation Average estimation

error (%)

t-statistics error

error (%)

t-statistics

Negative experiences   Had fight/argument    Thought about distant friends/
family   Thought about enormous
workload   Was rejected by boy/girl    Received low grade   Thought about bad personal
health habits Positive experiences
 Received high grade  Attended fun party  Participated in athletics Went out with friends  Talked to distant friends/family  Had great meal

−13.0*** −28.2***
−12.2***
−8.9*** −15.9*** −24.1***
−3.0 +20.9*** +13.7*** +12.6***
−8.3*** −2.3

−17.2 +5.6

5.47** 1.18

−13.8*** −26.3***
−11.3***
−18.4*** −23.3*** −35.0***
−0.3 +13.2***
+7.6*** +11.5***
−9.9*** +0.5

−21.4 +3.8

5.99**    
 
     
1.06            

Source. The table was adopted from Jordan et al. (2011, pp. 126, 130). aA positive number indicates an overestimation and a negative number indicates an underestimation. **p < .01. ***p < .001.

interaction hypothesis: (b-reframed) underestimation errors are stronger for negative experiences than for positive experiences. Study 2 showed similar results using a re-analysis of a dataset by Srivastava et al. (2009) with an underestimation of negative emotions and an overestimation of positive emotions.
In Study 3, the researchers examined the associations between prevalence estimations and psychological wellbeing. Participants completed scales regarding loneliness, rumination, depressive symptoms, life satisfaction, subjective happiness, and self-reported their number of friends, in addition to rating the same emotional events from Study 1b. They hypothesized that (c) higher prevalence estimates of others’ negative emotional experiences are associated with poorer psychological outcomes: greater loneliness, rumination, depressive symptoms, and lower life satisfaction and subjective happiness. They also hypothesized that (d) having fewer peers is associated with stronger misestimation.
We summarized the findings of Study 1b and Study 3 in Table 1 and the hypotheses of the target article in Table 2 (under “Replication: Prevalence estimations”).
Extensions
Intensity of Emotional Experiences.  The authors of the target article acknowledged a limitation in their design, specifically that they did not differentiate between prevalence and intensity. For instance, participants were asked to rate the prevalence of their peers “receiving a bad grade and felt really bad about it,” making it difficult to distinguish between their

estimation of the prevalence of receiving a bad grade and their estimation of the intensity of the emotions associated with that event. As such, we were unsure whether participants underestimated the prevalence of their peers’ negative emotions because they believed their peers rarely experienced such events (prevalence), or because they believed they these events would not elicit strong emotional responses (intensity).
Therefore, we revised the questions to inquire about both the prevalence of the events and the intensity of the emotional experiences. By expanding upon the original article, our aim was to examine the interplay between prevalence and intensity estimates and gain a clearer understanding of the source of the misestimation.
Social Comparison Orientation (Exploratory). We aimed to investigate the role of social comparison orientation, the tendency to compare oneself with others, as a potential predictor of misestimation and well-being measures. Gibbons and Buunk (1999) found that individuals with a higher social comparison orientation exhibited greater uncertainty and instability regarding their self-worth. Consequently, they tended to evaluate themselves based on how others were faring in public. This has the potential to be even more pronounced in an era in which social media provides ample opportunities for social comparison.
Accumulating evidence suggests a positive link between social comparison orientation and threats to psychological wellbeing online, as many social media users tend to selectively present mostly positive aspects of their lives (Vogel et al., 2014).

4

Personality and Social Psychology Bulletin 00(0)

Table 2.  Replication and Extensions: Summary of Hypotheses.

Replication: prevalence estimations

No

Hypothesis

2a People underestimate the prevalence of others’ negative emotional experiences.

2b People do not underestimate the prevalence and extent of others’ positive emotional experiences.

[Our reframing of the target’s null hypothesis: Prevalence underestimation errors are stronger for negative

experiences than for positive experiences.]

4

4a) There is a positive association between the estimation of the prevalence of negative emotional experiences and

well-being.a

4b) There is a negative association between the estimation of the prevalence of positive emotional experiences and well-being.a

 

Negative indicators—higher well-being: lower depressive symptoms, lower loneliness, lower rumination

4-1 4-1a) There is a negative association between the estimation of the prevalence of negative emotional experiences and

depressive symptoms.

4-1b) There is a positive association between the estimation of the prevalence of positive emotional experiences and

depressive symptoms.

4-2 4-2a) There is a negative association between the estimation of the prevalence of negative emotional experiences and

loneliness.

4-2b) There is a positive association between the estimation of the prevalence of positive emotional experiences and

loneliness.

4-3 4-3a) There is a negative association between the estimation of the prevalence of negative emotional experiences and

rumination.

4-3b) There is a positive association between the estimation of the prevalence of positive emotional experiences and

rumination.

 

Positive indicators—higher well-being: higher life-satisfaction, higher subjective happiness.

4-4 4-4a) There is a positive association between the estimation of the prevalence of negative emotional experiences and

life satisfaction.

4-4b) There is a negative association between the estimation of the prevalence of negative emotional experiences and

life satisfaction.

4-5 4-5a) There is a positive association between the estimation of the prevalence of negative emotional experiences and subjective happiness.b

4-5b) There is a negative association between the estimation of the prevalence of negative emotional experiences and subjective happiness.b

Extensions: intensity estimations

No

Hypothesis

2c People underestimate the intensity of others’ negative emotional experiences.c

2d People overestimate the intensity of others’ positive emotional experiences.c

5

Social comparison orientation interacts with misestimation in predicting well-being:

The higher the social comparison orientation, the stronger the associations in Hypothesis 4

(Hypotheses 4-1a to 4-5b).

6

Social comparison orientation is negatively associated with well-being.

 

Negative indicators—higher well-being: lower depressive symptoms, lower loneliness, lower rumination

6a Social comparison orientation is positively associated with depressive symptoms.

6b Social comparison orientation is positively associated with loneliness.

6c Social comparison orientation is positively associated with rumination.

 

Positive indicators—higher well-being: higher life-satisfaction, higher subjective happiness.

6d Social comparison orientation is negatively associated with life satisfaction.

6e Social comparison orientation is negatively associated with subjective happiness.

aThe target article sometimes shifts from referring to misestimation in its predictions to tests regarding prevalence estimations. We aligned the hypotheses with the tests conducted and the findings reported, rather than the versions of theory and hypothesis referring to misestimation. In addition, we specified separate hypotheses for estimations of negative emotions and estimations of positive emotions, to make things clearer. bThe hypothesis was not supported in the original target article, yet is included due the possibility of lacking power given the item-level analysis and small number of items. cWe had no specific predictions for the associations between intensity estimations of positive and negative emotional experience, and these should therefore be treated as exploratory.

Ip and Feldman

5

For example, Vogel et al. (2015) found that people with higher social comparison orientation were more susceptible to deleterious consequences such as lower self-esteem and poorer affect balance when using social media. Similarly, Wang et al. (2017) found that social comparison orientation was associated with increased usage of social networks, which in turn predicted lower life satisfaction and higher levels of jealousy and envy.
We therefore considered the possibility that this trait encompasses aspects related to social misestimation, as a greater inclination to compare oneself against others may also make individuals more vulnerable to potential discrepancies between their own experiences and those of others. Therefore, our aim was to examine the association between social comparison orientation, misestimation of others’ emotional experiences, and psychological well-being. Furthermore, we sought to explore the interaction between social comparison orientation and misestimation of emotional experiences in relation to well-being measures, with the exploratory prediction that the higher the social comparison orientation, the stronger the negative link between misestimation of emotional experiences and well-being measures.
Replication and Extension: Overview. We summarized the hypotheses for our adjusted replication and extensions in Table 2. In hypotheses 2c and 2d, we built on the target article’s hypotheses 2a and 2b to separately assess the prevalence and emotional intensity of emotional events. We also introduced social comparison orientation as a new measure to examine its association with the misperception of peers’ emotional experiences and psychological well-being in Hypotheses 5 and 6.
Data Availability, Preregistration, and Open-Science Disclosures.  We provided all materials, data, and code on the Open Science Framework (OSF): https://osf.io/bwmtr/. We first preregistered the experiment on the OSF (https://osf.io/ qda7j/) and data collection was launched later that week. All measures, manipulations, exclusions are reported, and data collection was completed before analyses. The preregistration and manuscript were written based on a template by Feldman (2023).

which does not seem to matter for the recruitment of participants, and with the small number of items requiring a very large effect size on the item-level to be able to detect any effects. This is related to other issues with some of the effects reported in the target (e.g., d = 2.52 in Study 1b and d = 3.16 in Study 3) being much larger than typical effects in the field (see Jané et al., 2024).
Therefore, instead of using the target article’s effects, we aimed for a sample size that would allow for participantlevel detection of paired-sample t-test d(z) = 0.2 (N = 272), and correlations of r = .15 (N = 472). These effect sizes are far smaller than any of the effects reported in the target article and are considered weak to medium effects in social psychology (Jané et al., 2024). This approach also aligns with the conservative estimate of 460 according to the rule of thumb recommendations by Simonsohn (2015) for a sample size 2.5 times larger than the combined samples in the original article (80 in the original Study 1b and 104 in the original Study 3), though we note that these recommendations were intended for participant-level between subject-designs, and not for item-level analyses with fixed number of items. To account for possible exclusions in case of a failed replication (see data analysis strategy section) and multiple analyses on the same sample, we exceeded the planned sample by aiming for a sample of 600.
Participants
We recruited a total of 594 U.S. American college students on Prolific (Palan & Schitter, 2018; 290 males, 285 females, 19 other/did not disclose; using the Prolific internal qualifier for students). We used Prolific’s filters, by restricting the location to the United States using “standard sample,” we set it to “Nationality: United States,” “Country of birth: United States,” “Student status: Yes,” “Minimum Approval Rate: 95, Maximum Approval Rate: 100,” “Minimum Submissions: 50, Maximum Submissions: 100,000,” “Total times a participant can complete your study: Once.” We summarized a comparison of the target article sample and the replication sample in Table 3. We outlined the details of the pay and survey procedures in the Supplemental Material.

Method
Power Analysis and Sensitivity Test
We first calculated the effect sizes of the findings reported in the target article with the help of a guide by Jané et al. (2024). We then conducted an a priori power analysis (power = 0.95, alpha = .05), with an upward adjustment. The calculated effect sizes are summarized in the Supplemental Materials.
There were several issues with using the target article’s effects as a basis for power analyses. The target conducted their analyses on an item-level with an analysis on 12 items,

Design and Procedure
We summarized the experimental design and all measures in Table 4 (more details are available in the Supplemental Materials).
Participants rated their own emotional experiences and others’ emotional experiences in random order. Misestimations were measured as participants’ estimates of others’ emotional experiences minus the average of all participants’ own emotional experiences. In the target article’s Study 1b, the prevalence and intensity were confounded, which we adjusted to measure prevalence and intensity separately (see “Measures” subsection).

6

Personality and Social Psychology Bulletin 00(0)

Table 3.  Comparison of Target Article Versus Replication: Differences and Similarities.

Study

Jordan et al. (2011) Study 1b

Jordan et al. (2011) Study 3

Replication

Sample size Geographic origin Gender
Medium (location) Compensation

80 U.S. American students 35 males, 45 females, 0 other/did
not disclose Paper-and-pen N/A

Year

2011 or earlier

104 U.S. American students 51 males, 54 females, 0 other/did
not disclose Computer (online) Nominal payment
2011 or earlier

594 U.S. American students on prolific 290 males, 285 females, 19 other/did not
disclose Computer (online) Nominal payment (2 British pounds for an
estimated 12 min completion time) 2023

We conducted the study using Qualtrics. All participants first indicated their consent. They then rated the prevalence of six positive and six negative emotional experiences for themselves and for other U.S. student participants taking this survey, in counterbalanced order.
Prior to each rating task, we added comprehension checks to ensure that participants were paying attention to the type of rating (prevalence or intensity) and who they are rating (self vs. other). Participants had to answer these checks correctly in order to proceed to the rating task. We note that this is a deviation from the target article’s procedure to ensure that participants were attentive and knew who and what they were rating.
For exploratory purposes, we included three questions that asked participants to provide short sentences regarding their perceptions of the test items they had previously rated, to help us better understand participants’ mindsets when answering some of the items from the target article, to address the possibility of puzzling or surprising findings using these items.
Participants then completed measures assessing their loneliness, rumination tendency, depressive symptoms, life satisfaction, subjective happiness, and social comparison orientation. We randomized the order of the six well-being and trait measures. Finally, participants answered a funneling section, provided demographic information, and were debriefed.
Measures
Prevalence and Intensity of Own Emotional Experiences (Replication and Extension). In the target article, participants only rated whether they had experienced an emotional event or not. We adjusted and extended the measures to also include intensity:
For each of the following emotional experiences, please indicate whether you have experienced those sometime in the past 2 weeks and, if you have, the intensity of the emotion.
0 means: You have NOT experienced this emotion in the past two weeks.
1–100 means: You have experienced this emotion at least once in the past two weeks.

1 = lowest emotional intensity, and 100 = highest emotional intensity.
In this way, we measured both prevalence and intensity in the same question: a value of 0 indicated that the participant has not experienced this event, and a value of 1 to 100 indicated that the participant has experienced it with a rating of the intensity of the experience.
Prevalence of Others’ Emotional Experiences (Replication).  We closely followed the target’s measure of participants’ estimates of others’ emotional experiences with: “Please estimate the percentage of other U.S. American student participants on Prolific taking the survey like you who had had, sometime in the past 2 weeks, each of the following emotional experiences. (0%–100%)”
Intensity of Others’ Emotional Experiences (Extension). We extended the target’s measure of prevalence (above) with a measure of intensity: “Please try and estimate the emotional intensity for other U.S. American student participants on Prolific taking the survey like you who have experienced this emotion (1 = Lowest emotional intensity, and 100 = Highest emotional intensity).”
Misestimations of Prevalence and Intensity.  We first calculated the actual prevalence of emotional experiences by counting for each experience the number of participants who indicated having that experience, and then converted that into a percentage. We then calculated for each participant, and for each emotional experience, the misestimation of the prevalence of that emotional experience as the estimation of prevalence minus the actual prevalence of that emotional experience.
We calculated the actual intensity of emotional experience by calculating the average of all self-reported intensity for all those who reported having that experience. We then calculated for each participant and for each emotional experience, the misestimation of the intensity as the participant’s estimation of the intensity of that emotional experience minus the actual intensity of that emotional experience.
This means that for both prevalence and intensity, a misestimation score higher than zero indicates an overestimation, and a score lower than zero indicates an underestimation.

Ip and Feldman

7

Table 4.  Replication and Extension: Experimental Design and Measures.

IV1: positive vs. negative emotions (within; all participants rated both)
IV2: self vs. others rating (within; all participants rated both)
Self ratings Participants were asked to estimate
based on their own emotional experiences.
Others ratings Participants were asked to make
estimations of U.S. American student participants on Prolific taking the survey.
Exploratory open-ended
Well-being measures and traits

IV1: estimation of positive emotional events IV1: estimation of negative emotional events

“Felt happy because they. . .”

“Felt sad because they. . .”

1.  Received high grades

1.  Had a fight or argument

2.  Attended fun party

2.  Thought about distant friends or family

3.  Participated in athletics

3.  Thought about enormous workload

4.  They went out with friends

4.  Were rejected by someone

5.  They talked to distant friends or family

5.  Received a low grade

6.  Had great meal

6.  Thought about bad personal health habits

DV1: prevalence and intensity of own emotional experiences [replication + extension]

“For each of the following emotional experiences, please indicate whether you have experienced those

sometime in the past 2 weeks and if you have—the intensity of the emotion.

0 means: You have NOT experienced this emotion in the past two weeks.

1–100 means: You have experienced this emotion at least once in the past two weeks”

(0 = Not experienced; 1 = Lowest emotional intensity; 100 = Highest emotional intensity.)

DV1a: prevalence of others’ emotional experiences [replication]

“Please estimate the percentage of other U.S. American student participants on Prolific taking the survey like

you who had had, sometime in the past 2 weeks, each of the following emotional experiences.

(0% to 100%)

DV1b: intensity of others’ emotional experiences [extension]

“Please try and estimate the emotional intensity for other U.S. American student participants on Prolific

taking the survey like you who have experienced this emotion

(1 = Lowest emotional intensity; 100 = Highest emotional intensity)”

Perception of vague test items:

Participants briefly wrote about their understanding of vague test items.

“What is a “bad” grade? How do you classify a grade as being a bad grade? Can you give a quick example?

(1–2 sentences)”

“What is an ‘enormous’ workload? How do you classify workload as being an enormous workload? Can you

give a quick example? (1–2 sentences)”

“What is a ‘bad’ personal health habit? How do you classify personal health habit as being bad? Can you give

a quick example? (1–2 sentences)”

(presented in random order)

Loneliness [replication]

UCLA Loneliness Scale (Hays & DiMatteo, 1987, p. 4), 8-item short-form version.

Sample items: “There is no one I can turn to.” and “I can find companionship when I want it.”

(1 = Never to 4 = Often)

Rumination [replication]

Brooding subscale (5-item) of the Ruminative Responses to Depression Questionnaire (Nolen-Hoeksema,

1991; Treynor et al., 2003). Sample items: “Think What am I doing to deserve this?” and “Think about a

recent situation, wishing it had gone better.” etc.

(1 = Almost never to 4 = Almost always)

Depressive symptoms [replication]

Center for Epidemiologic Studies Depression Scale 10-item short-form version (Cole et al., 2004; Radloff,

1977). Sample items: “I was bothered by things that usually don’t bother me.” and “I had trouble keeping

my mind on what I was doing.” etc.

(1 = Rarely or none of the time (less than 1 day) to 4 = All of the time (5–7 days))

Life satisfaction [replication]

Participants were asked to complete the 5-item SWLS (Diener et al., 1985). Sample items: “In most ways my

life is close to my ideal.” and “If I could live my life over, I would change almost nothing.” etc.

(1 = Strongly disagree to 7 = Strongly agree)

Subjective happiness [replication]

SHS 4-item (Lyubomirsky & Lepper, 1999). Sample items:

“In general, I consider myself. . .” and “Compared to most of my peers, I consider myself. . .” etc.

(1 = Not at all/Less happy; 7 = A very happy person/More happy/A great deal)

Social comparison orientation [extension]

The INCOM (Gibbons & Buunk, 1999)

Sample items: “I often compare myself with others with respect to what I have accomplished in life” and “If I

want to learn more about something I try to find out what others think about it”

(1 = Strongly disagree to 5 = Strongly agree)

Note. All materials are provided in the Qualtrics survey export provided in the OSF folder. INCOM = Iowa-Netherlands comparison orientation measure; SWLS = satisfaction with life scale; SHS = subjective happiness scale.

Depressive Symptoms. We measured depressive symptoms using the 10-item short-form version of the Center for Epidemiologic Studies Depression Scale (Radloff, 1977). Respondents indicated the frequency of events in the week prior on

items such as “My sleep was restless,” and “I had trouble keeping my mind on what I was doing” (0 = “Rarely or none of the time (less than 1 day)”; 3 = “All of the time (5–7 days)”; Cole et al., 2004; α = .88).

8

Personality and Social Psychology Bulletin 00(0)

Brooding/Rumination.  We measured Brooding/Rumination using the Brooding subscale (5-item) of the Ruminative Responses to Depression Questionnaire (Nolen-Hoeksema, 1991; Treynor et al., 2003). Participants rated items such as “Why can’t I handle things better,” and “What am I doing to deserve this” (1 = almost never; 4 = almost always), scoring from 5 to 20 (α = .87).
Loneliness.  We measured participants’loneliness with the 8-item short-form version of the UCLA Loneliness Scale (Hays & DiMatteo, 1987; Russell et al., 1980). Participants rated their agreement with statements such as “I feel left out” and “I lack companionship” (1 = Never; 4 = Always; α = .88).
Life Satisfaction. We measured participants’ life satisfaction with the 5-item Satisfaction with Life Scale (SWLS; Diener et al., 1985). Participants rated five statements (1 = Strongly disagree; 7 = Strongly agree), such as “So far I have gotten the important things I want in life,” and “The conditions of my life are excellent.” (α = .92).
Subjective Happiness.  We measured participants’ overall happiness using the 4-item Subjective Happiness Scale (SHS). Participants rated four statements on a 7-point Likert scale, with items such as “In general, I consider myself. . .” ranging from “not a very happy person” to “a very happy person” (Lyubomirsky & Lepper, 1999) (α = .88).
Number of Friends.  Participants indicated a rough number of friends with whom they feel comfortable talking to concerning their personal emotional experiences.
Social Comparison Orientation (Extension). We measured social comparison orientation using the 11-item Iowa-Netherlands Comparison Orientation Measure (INCOM) (Gibbons & Buunk, 1999), with items such as “I always like to know what others in a similar situation would do” and “I often like to talk with others about mutual opinions and experiences” (5-point scale; 1 = Strongly disagree; 5 = Strongly agree; α = .73).
Deviations and Replication Closeness Evaluation.  We made several adjustments to the target article’s study design. We summarized our deviations with a comparison of study design of the target article with our replication using the LeBel et al. (2018) replication closeness evaluation criteria in Table 5. We categorized the replication as a close to far replication (see “replication closeness evaluation” in the Supplemental Material).
Data Analysis Strategy
Replication Measures Prevalence Estimation Error: Replication Item-Level Analy-
ses.  We followed the target’s analysis and used this as our criteria for a successful replication.

Prevalence Estimation Error: Item-Level One Sample t-Test for Each Item.  In the target article, the authors first aggregated means of the self-rating of emotional experiences for each of the six negative and six positive items. They then calculated an estimation error comparing participants’ estimation of others’ emotional experiences against the aggregated mean of all participants’ self-ratings. They then conducted a series of 12 one-sample t-tests to examine if participants overestimated by comparing estimation error to 0.
Prevalence Estimation Error: Item-Level One-Sample t-Test for Positive and Negative.  The authors also conducted two itemlevel one-sample t-tests, one on the item-level aggregate of the negative emotional experiences, and another on the itemlevel aggregate of the positive emotional experiences.
Prevalence Estimation Error: Item-Level Independent Sample t-Test Comparing Positive and Negative.  They ran an item-level independent sample t-test of the average estimation error for the negative versus the positive items.
Prevalence Estimation Error: Extension Participant-Level Analyses. We felt that the replication analysis above could be improved and therefore supplemented the analysis with a participant-level analysis, where we first computed the average for the negative events for each participant, and then the average for the positive events. We then conducted two participant-level one-sample t-tests, one on the mean of the negative emotional experiences, and another on the mean of the positive emotional experiences. Finally, we ran a participant-level paired sample t-test of the average estimation error for the negative versus the positive items.
Associations Between Prevalence Estimates of Others and Well-Being: Replication Analyses. We followed the target’s analyses by computing each participant’s average peer-prevalence estimates for negative and positive emotional experiences. Thereafter, we conducted a linear regression analysis to examine the associations between emotional estimation error and the well-being indicators such as depressive symptoms, rumination, life satisfaction, loneliness, subjective happiness, and the number of confidants.
Extension Measures Intensity Estimation Error. We added intensity measures
and conducted all the analyses for prevalence above also for intensity, for both item-level and participant-level. We also conducted similar analyses for associations with well-being measures.
Social Comparison Orientation.  We added the social comparison orientation (SCO) measure to all the correlational and regression analyses detailed above for both prevalence and intensity, with all the other well-being measures. We also conducted regression interaction analyses to examine

Ip and Feldman

9

Table 5.  Classification of the Replication Based on LeBel et al. (2018).

Design facet

Replication

Details of deviation

Reasons for change

IV construct DV construct

Same, with an extension
Similar, with an added extension

IV operationalization Similar DV operationalization Similar

Population

Similar

Procedural details

Different

Physical settings

Similar to Study 3

Contextual variables Similar

Statistical analyses
Replication classification

Similar with additional analyses that were more suitable and tested robustness
Close to far replication

 

We separated prevalence from intensity, by adjusting the self-report measure and adding a separate measure estimating emotional intensity.
We added a social comparison orientation individual differences measure.
Original: Study 1b conflated emotional prevalence and intensity. Study 3 stripped intensity.
Replication: Emotional prevalence and emotional intensity in self were combined into the same question. For others’ they were rated as two separate measures—frequency and intensity.
Original: “rejected by a boy or girl” includes gender. Replication: adjusted to “rejected by someone” for
inclusiveness. We added measures evaluating views on test items
that were vague, such as “received a bad grade.” The population in our studies was also students, but
with a larger and more diverse sample from across the United States. Original: N = 80 (Study 1b); N = 104 (Study 3) Replication: N = 594 We added comprehension checks before the rating task to ensure participants understand (a) target (self or others), and (b) dimension (prevalence or intensity). We randomized the self-other rating order and the order of scale measures. We added three clarification questions that inquired about how they perceived subjective test items. Original: Study 1b: In person; Study 3: Online. Replication: Online.
Original studies: Participants were recruited from a medium-sized West Coast university in the United States.
Current replication: Our replication was conducted in 2023 broadly targeting college students across the United States.
We added participant-level analyses to supplement the item-level analyses in the original article.
We also conducted moderation analyses.
Mostly followed the target, yet the self-rating adjustment to measuring both prevalence and intensity warranted a more conservative categorization.

We proposed that social comparison orientation may interact with the misperception and contribute to differential well-being outcomes.
   
 
Ensuring participants read and understood the instructions and to rate accordingly.
Address order effects Address possibility of diverging
perceptions of our test items.
Target’s Studies 1b and 3 had similar designs that can be combined and extended.
 
Misestimations were determined based on estimate-actual comparisons.
To examine the effect of social comparison orientation.
 

whether SCO interacts with estimation error in predicting well-being measures.
Outliers and Exclusions.  We followed the preregistered plan to only include responses from participants who completed the

entire questionnaire with no further exclusions. We also preregistered that in case we failed to find support for the estimation hypotheses in our replication of the target article, we would supplement our analyses by rerunning the analyses with exclusion and stricter alpha to account for multiple

10

Personality and Social Psychology Bulletin 00(0)

analyses (alpha = .005). We found support for misestimation effects (see below) using the full sample (N = 594) and therefore did not proceed to conduct or report the analyses with exclusions.
Results
Prevalence Estimate Errors (Replication)
We summarized descriptives in Tables 6 (prevalence) and 7 (intensity), and correlations in Table 8.
Consistent with Hypothesis 2a and the original findings, we found support for an underestimation of others’ negative emotional experiences. Mirroring the analyses conducted in the original article, we computed the estimation error for each item of positive and negative experiences by comparing the self-rating with ratings for their peers. We conducted an item-level one-sample t-test on the estimation errors to examine if they differed from 0. We found support for the expected underestimation of negative (t(5) = −4.30, p = .008, d = −1.76, 95% CI [−3.06, −0.41]), yet also support for an unexpected underestimation of positive experiences (t(5) = −3.67, p = .014, d = −1.50, [−2.67, −0.26]).
We further extended our analyses to a series of participant-level one-sample t-tests on the estimation error of emotional experiences, summarized in Table 6. Overall, we found support for participants’ underestimation of all six negative emotional experiences and of five of the six positive emotional experiences. The participant-level analyses mirrored that of the item-level analyses, with support for the expected underestimation of negative emotional experiences (M = 23.56; t(593) = −35.02, p < .001, d = −1.44, 95% CI [−1.55, −1.32]), yet again with the unexpected underestimation of positive emotional experiences (M = 15.97; t(593) = −24.17, p < .001, d = −0.99, [−1.09, −0.89]).
Partially consistent with Hypothesis 2b, the underestimation of negative experiences was stronger than positive experiences, yet only for the participant-level analysis. Specifically, for the item-level analyses, we ran an independent samples t-test comparing negative and positive emotional experiences and did not find support for Hypothesis 2b in estimation error (t(10) = 0.31, p = .31, d = −0.62, 95% CI [−1.79, 0.60]). We anticipated this in advance, given that the item-level analysis had too few items and therefore power to detect such differences, which is why we planned and preregistered to also run participant-level analyses, which we felt were more appropriate and accurate. The participant-level paired-samples t-test analysis allowed us to find support for a larger estimation error for negative compared to positive experiences (t(593) = −11.90, p < .001, d = −0.49, [−0.57, −0.40]).
Intensity Estimate Errors (Extension)
To test whether underestimation of emotional experiences also extends to emotional intensity, we conducted similar

analyses on the emotional intensity measures. We began with item-level one-sample t-tests and found support for Hypothesis 2d with an overestimation of positive experiences (t(5) = 2.97, p = .031, d = 1.21, 95% CI [0.099, 2.26]), yet contrary to our expectations in Hypothesis 2c, we also found an overestimation for negative experiences (t(5) = 3.51, p = .017, d = 1.43, [0.23, 2.58]).
The participant-level analyses showed a similar trend, as we found support for the expected overestimation of positive emotional experiences (M = 5.60; t(593) = 8.45, p < .001, d = 0.35, 95% CI [0.26, 0.43]), and again the unexpected overestimation of negative emotional experiences (M = 12.54; t(593) = 19.42, p < .001, d = 0.80, [0.70, 0.89]). We ran one-sample t-tests for each of the experiences and found that participants overestimated five out of the six negative emotional experiences and five out of the six positive emotional experiences.
Mirroring the target’s analyses for item-level prevalence comparing positive and negative, we ran the same analysis for intensity estimates and found no signal for difference in overestimation error between negative and positive emotional experiences, though large effect size (t(10) = 1.72, p = .117, d = 0.99, 95% CI [−0.24, 2.18]). Again, this is most likely due to the small number of items a power analysis of which shows an unreasonably large effect in order to be detectable with null hypothesis significance testing. We supplemented the item-level analysis by conducting a participant-level paired sample t-test, which was far better powered, and indeed found support for stronger overestimation for negative experiences than for positive experiences (t(593) = 10.63, p < .001, d = 0.44, [0.35, 0.52]).
Prevalence Estimates Associations with WellBeing (Replication)
We summarized the Pearson’s correlations in Table 8, and the regression model findings predicting well-being from negative and positive prevalence estimates in Table 9 (comparing with the target’s).
Inconsistent with and opposite to Hypothesis 4, prevalence estimations of negative emotional experiences were negatively associated with well-being, as indicated by a negative association with life satisfaction (β = −.23, t(591) = −4.87, p < .001), and subjective happiness (β = −.26, t(591) = −5.59, p < .001), and a positive association with loneliness (β = .33, t(591) = 7.04, p < .001), brooding (β = .37, t(591) = 8.08, p < .001), depressive symptoms (β = .41, t(591) = 9.01, p < .001). On the other hand, positive prevalence estimates were negatively associated with loneliness (β = −.25, t(591) = −5.40, p < .001), and depressive symptoms (β = −.22, t(591) = −4.69, p < .001), and positively associated with life satisfaction (β = .25, t(591) = 5.35, p < .001), subjective happiness (β = .32, t(591) = 6.73, p < .001), and the number of confidants (β = .19, t(591) = 3.98, p < .001).

Table 6.  Prevalence of Emotional Experiences (Replication and Extension): One-Sample t-Tests of Estimation Error.

Participant-level analyses

Item-level analysis

Experiences
Negative experiences    Had fight/argument    Thought about
distant friends/ family    Thought about enormous workload    Was rejected by boy/girl    Received low grade    Thought about bad personal health habits Overall negative

Actual Prevalence

prevalence estimate

average

mean

Error mean

71.04 78.79

47.41 45.76

−23.63 −33.02

86.36

62.14 −24.22

46.13
50.17 85.35

36.04
42.60 42.51

−10.09
−7.57 −42.85

69.64

46.08 −23.56

Error SD
23.54 22.61
22.29
22.11 23.60 23.02
16.40

t-stat

p

−24.47 <.001 −35.60 <.001

−26.49 <.001

−11.12 <.001
−7.82 <.001 −45.36 <.001

−35.02 <.001

Cohen’s d and CI
−1.00 [−1.10, −0.90] −1.46 [−1.58, −1.34]
−1.09 [−1.19, −0.99]
−0.46 [−0.54, −0.37] −0.32 [−0.40, 0.24] −1.86 [−1.99, −1.73]
−1.44 [−1.55, −1.32]

Interpretation
Signal; same direction Signal; same direction
Signal; same direction
Signal; same direction Signal; same direction Signal; same direction
Signal; same direction

t-stat df

p

/

/

/

/

/

/

/

/

/

/

/

/

/

/

/

/

/

/

−4.30 5 .008

Cohen’s d and CI
/ /
/
/ / /
−1.76 [−3.06, −0.41]

Overall negative: target article (Study 3)

Positive experiences   Received high grade

75.08

54.84

  Attended fun party   Participated in
athletics   Went out with
friends   Talked to distant
friends/ family    Had great meal

51.35 54.04 77.27 72.90 94.44

52.36 39.45 65.61 53.68 63.31

Overall positive

70.85

54.87

−21.4

8.7

—

—

—

—

−5.99 5 <.01

−20.35 24.35 −20.27 <.001 −0.83 [0.92, 0.74]

Unexpected signal; same /

/

/

direction

1.01 24.04

1.03 .305

0.04 [−0.04, 0.12]

Unexpected no signal;

/

/

/

−14.59 22.64 −15.70 <.001 −0.64 [−0.73, 0.56]

Signal; opposite

/

/

/

direction

−11.66 19.63 −14.48 <.001 −0.59 [−0.68, −0.51] Signal; opposite

/

/

/

direction

−19.22 22.47 −20.84 <.001 −0.86 [−0.95, 0.76]

Signal; opposite

/

/

/

direction

−31.22 22.00 −34.49 <.001 −1.42 [−1.53, −1.30] Unexpected signal; same /

/

/

direction

−15.97 16.10 −24.17 <.001 −0.99 [−1.09, −0.89] Unexpected signal

−3.67 5

.014

−2.68 [−4.65, −0.68]
/ / / / / / −1.50 [−2.67, −0.26]

Overall positive: target article (Study 3)

Overall positive and negative combined

70.24

50.48

3.8 −19.77

8.7 14.27

—

—

−33.75 <.001

— −1.38 [−1.50, −1.27]

—

1.06 5 n.s

0.48 [−0.48, 1.38]

−5.61 11 <.001 −1.62 [−2.48, −.73]

Note. One-sample t-tests, N = 594, participant level df = 593, 95% CI (confidence intervals). The interpretation of the outcome is based on LeBel et al. (2019). We summarized whether we found a signal (p < alpha), whether the signal or lack of was expected, and whether it was in the same direction as that of the target’s per the singular item. We did not summarize effect size consistency (target’s effect within replication confidence intervals) given that we conducted analyses on participant level whereas the target conducted analysis on item-level.

11

12

Table 7.  Intensity of Emotional Experiences [Extension]: One-Sample t-Tests of Estimation Error.

Participant-level analyses

Experiences

Actual Intensity

intensity estimate Error Error

average mean mean SD t-stat

p

Negative experiences Had fight/argument Thought about distant friends/ family Thought about enormous workload Was rejected by boy/girl Received low grade
   Thought about bad personal health habits
 Overall negative Positive experiences    Received high grade    Attended fun party    Participated in athletics    Went out with friends    Talked to distant friends/family    Had great meal  Overall positive Overall: positive and negative combined

42.63 42.65 50.81 33.00 37.05 44.07
41.70
56.07 50.63 44.95 59.11 52.03 58.80 53.60 47.65

59.64 48.48 60.44 58.16 53.65 45.07
54.24
60.56 62.65 49.35 67.65 57.96 57.03 59.20 56.72

17.01 5.84 9.63 25.17 16.60 1.01

22.27 21.48 22.13 23.94 22.46 21.93

18.62 6.62 10.60 25.62 18.01 1.12

12.54 15.73 19.42

4.50 12.03
4.40 8.54 5.93 −1.77 5.60 9.07

22.39 21.35 22.16 19.47 20.29 23.40 16.16 13.83

4.90 13.73
4.84 10.69
7.12 −1.85
8.45 15.99

<.001 <.001 <.001 <.001 <.001 =.264
<.001
<.001 <.001 <.001 <.001 <.001 =.065 <.001 <.001

Note. One-sample t-tests, N = 594, participant level df = 593, 95% CI (confidence intervals).

Cohen’s d and CI
0.76 [0.67, 0.86] 0.27 [0.19, 0.35] 0.43 [0.35, 0.52] 1.05 [0.95, 1.15] 0.74 [0.65, 0.83] 0.05 [−0.03, 0.13]
0.80 [0.70, 0.89]
0.20 [0.12, 0.28] 0.56 [0.48, 0.65] 0.20 [0.12, 0.28] 0.44 [0.35, 0.52] 0.29 [0.21, 0.37] −0.08 [−0.16, 0.004] 0.35 [0.26, 0.43] 0.66 [0.57, 0.74]

Item-level analyses

t-stat df

p

Cohen’s d and CI

/

/

/

/

/

/

/

/

/

/

/

/

3.51 5

/

/

/

/

/

/

/

/

/

/

/

/

2.97 5

4.14 11

/

/

/

/

/

/

/

/

/

/

/

/

.017 1.43 [0.23, 2.58]

/

/

/

/

/

/

/

/

/

/

/

/

.031 1.21 [0.099, 2.26]

.002 1.20 [0.43, 1.90]

Table 8.  Correlations Between Prevalence and Intensity Estimates with Well-Being and Social Comparison Measures.

Variable

M SD Alpha

1

2

3

4

5

6

7

8

9

10

11

1—Prevalence estimates for negative emotions 2—Prevalence estimates for positive emotions 3—Prevalence estimates overall
4—Intensity estimates of negative emotions (extension) 5—Intensity estimates of positive emotions (extension) 6—Intensity estimation error overall (extension) 7—SCO (extension)
8—Loneliness
9—Brooding
10—Depression
11—Life satisfaction
12—Subjective happiness

46.1 16.4 — 54.9 16.1 — 50.47 14.27 — 54.2 15.7 — 59.2 16.2 — 9.07 13.83 — 35.30 6.27 .67 18.36 5.97 .88 11.44 3.95 .87 10.28 6.60 .88 20.84 7.75 .92 17.15 5.55 .88

 

0.54*** [0.48, 0.60]
0.88*** [0.86, 0.90]
0.58*** [0.53, 0.63]
0.32*** [0.25, 0.39]
0.52*** [0.46, 0.58]
0.21*** [0.13, 0.28]
0.19*** [−0.11, 0.27]
0.30*** [.22, .37] 0.30*** [0.22, 0.37] −0.09* [−0.17, −0.13] −0.09* [−0.17, −0.11]

0.88*** [0.86, 0.89]
0.35*** [0.28, 0.42]
0.63*** [0.58, 0.67]
0.57*** [0.51, 0.62]
0.04 [−0.04, 0.12]
−0.07 [−0.15, 0.01]
0.06 [−0.02, 0.14]
0.01 [−0.07, 0.09]
0.13* [0.05, 0.21]
0.17*** [0.09, 0.25]

0.53*** [0.47, 0.59]
0.54*** [0.48, 0.59]
0.62*** [0.47, 0.67]
0.14*** [0.06, 0.22]
0.07 [−0.01, 0.15]
0.20*** [0.13, 0.28]
0.18*** [0.10, 0.25]
0.02 [−0.06, 0.10]
0.05 [−0.04, 0.13]

0.50*** [0.44, 0.56]
0.86*** [0.84, 0.88]
0.14*** [0.06, 0.22]
0.18*** [0.10, 0.26]
0.17*** [0.09, 0.25]
0.19*** [0.11, 0.27]
−0.05 [−0.13, 0.03]
−0.10* [−0.18, −0.02]

0.87*** [0.85, 0.89]
0.08 [−0.01, 0.16]
−0.07 [−0.15, 0.01]
0.02 [−0.07, 0.10]
−0.05 [−0.13, 0.03]
0.15*** [0.07, 0.23]
0.20*** [0.12, 0.28]

0.12** [0.04, 0.20]
0.06 [−0.02, 0.14]
0.11** [0.03, 0.19]
0.08* [0.00, 0.16]
0.06 [−0.02, 0.14]
0.06 [−0.02, 0.14]

0.28*** [0.20, 0.35]
0.45*** [0.38, 0.51]
0.32*** [0.25, 0.39]
−0.17*** [−0.25, −0.09]
−0.18*** [−0.26, −0.11]

0.56*** [0.51, 0.62]
0.72*** [0.68, 0.75]
−0.54*** [−0.59, −0.48]
−0.65*** [−0.69, −0.60]

0.69*** [0.64, 0.73]
−0.45*** [−0.52, −0.39]
−0.47*** [−0.53, −0.40]

−0.55*** [−0.60, −0.49]
−0.65*** [−0.69, −0.60]

                    0.65*** [0.61, 0.70]

Note. df = 592. SCO = social comparison orientation. *p < .05. **p < .01. ***p < .001.

13

14

Personality and Social Psychology Bulletin 00(0)

Table 9.  Prevalence Estimates Regression Coefficients with Outcomes: Comparison Between Jordan et al. (2011)’s Study 3 Versus Replication.

Jordan et al. (2011)

Replication

Measure

β for negative β for positive

β for negative β for positive

prevalence prevalence

prevalence prevalence

Cronbach’s α estimate

estimate R2

estimate

estimate R2

Interpretation

Negative

 Loneliness

.81

 Rumination/brooding

.72

 Depressive symptoms

.73

Positive

 Satisfaction with life

.84

 Subjective Happiness

.80

Number of confidants

/

Social orientation scale

/

(extension)

−.30** −.28**
.00
.23* .19 N/A
/

.08

.08

.33***

.17

.08

.37***

−.02

.00

.41***

−.37*** .15

−.16

.05

N/A

N/A

/

/

−.23***
−.26***
−.11* .26***

−.25*** −.15** −.22***
.25*** .32*** .19*** −.10*

.08 Signal; inconsistent opposite
.10 Signal; inconsistent opposite
.12 Signal; inconsistent opposite
.05 Signal; inconsistent opposite
.08 Signal; inconsistent opposite
.03   .05 Supported

Note. Linear regression, N = 594. N/A = not reported. Source. The “original” column was adopted from Jordan et al. (2011, p. 130). The interpretation of the outcome is based on LeBel et al. (2019). *p < .05. **p < .01. ***p < .001.

We conclude that our findings are in the opposite direction to that reported in the target article, for both negative and positive emotional experiences prevalence estimates.
Intensity Estimates Associations with Well-Being (Extension)
We conducted the same analyses on intensity estimates as the ones reported above for prevalence, and summarized those in Tables 8 and 10. We found a similar pattern, with support for a positive association between negative emotional experiences intensity estimates and loneliness (β = .29, t(591) = 6.21, p < .001), brooding (β = .22, t(591) = 4.74, p < .001), and depressive symptoms (β = .29, t(591) = 6.31, p < .001), and negative association with life satisfaction (β = −.18, t(591) = −3.79, p < .001), and subjective happiness (β = −.28, t(591) = −6.07, p < .001).
On the other hand, we found that positive emotional experiences intensity had a negative association with loneliness (β = −.21, t(591) = −4.63, p < .001) and depressive symptoms (β = −.20, t(591) = −4.23, p < .001).
Complementary Analysis: Self-Reports Associations with Well-Being (Exploratory Extension)
We explored the associations between self-reports of emotional experiences and well-being measures. We found that reporting more of the negative emotional experiences was positively correlated with loneliness (r = .16 [0.08, 0.23]),

brooding (r = .25 [0.17, 0.32]), and depression (r = .26 [0.19, 0.34]), and that reporting more the listed positive emotional experiences was positively correlated with life satisfaction (r = .22 [0.14, 0.29]) and happiness (r = .27 [0.20, 0.35]) and negatively associated with loneliness (r = −.16 [−0.23, −0.07]; all p < .001).
Complementary Analysis: Interaction Between Self and Others in Predicting Well-Being (Exploratory Extension)
We also explored interactions between one’s own negative experiences and estimates of others’ negative experiences in predicting well-being, and found support for an interaction for both loneliness, brooding, and depression, such that the positive association between prevalence estimates and negative factors of well-being was stronger the less negative experiences one had. We also found support for an interaction with happiness and well-being, such that the negative association between prevalence estimates and positive factors of well-being was stronger, the less negative experiences one had. We did not find such interactions for the positive experiences.
External Analysis: Suppression Using Target’s Analyses
When consulting with external expert reviewers to examine the possible explanation for our associations being in the opposite direction from that of the target article, a reviewer

Ip and Feldman

15

Table 10.  Intensity Estimates Regression Coefficients with Outcomes (Extension).

Measure

β for negative β for positive

prevalence

prevalence

estimate

estimate

R2

Negative well-being  Loneliness  Rumination/brooding  Depressive symptoms Positive well-being  Satisfaction with life  Subjective happiness  Number of confidants  Social orientation scale (extension)

.29*** .22*** .29***
−.18*** −.28*** −.07
.13**

−.21***

.07

−.10*

.04

−.20***

.07

.24***

.05

.34***

.10

.15**

.02

.01

.02

Note. Linear regression, N = 594. *p < .05. **p < .01. ***p < .001.

was kind to review the target article’s and our results. In his analyses (shared on https://osf.io/zy5qa/), he pointed out that there might be suppression in the regression analyses (Sharpe & Roberts, 1997; Thompson & Levine, 1997) that were conducted by the target article and that we repeated in our replication. His conclusion was that the regression analyses should at best be interpreted with caution, and that interpretations should be focused on the raw correlation effects, which were weaker yet in the same direction (see Table 8).
We reached out to the original authors of the target article, and received the reply that the raw data, analysis code, and correlations for the target article’s studies are not available. However, the reviewer, the original authors, and we agreed that suppression alone is not likely to explain the complete reversal of the pattern of effects.

SCO (Exploratory Extension)
We conducted Pearson’s correlation tests to examine the associations between SCO and well-being variables, and summarized them in Table 8. We found support for Hypotheses 6a to 6c, in that SCO was positively associated with depressive symptoms (r(592) = .32, 95% CI [0.25, 0.39], p < .001), loneliness (r(592) = .28, [0.20, 0.35], p < .001), and rumination (r(592) = .45, [0.38, 0.51], p < .001). We also found support for Hypotheses 6d and 6e that SCO was negatively associated with subjective happiness (r(592) = −.18, [−0.26, −0.11], p < .001), and life satisfaction (r(592) = −.17, [−0.25, −0.09], p < .001).
We also examined interactions with the misestimation of emotional experiences in predicting psychological wellbeing. We failed to find support for SCO as moderating the associations between misestimations and well-being measures. We summarized the results with plots of the moderation analysis in the Supplemental Material. Given the many analyses conducted and the many findings that have shown

no support, we only summarized interactions that were supported and documented below p < .05, yet we strongly caution against over-interpreting those and recommend focusing on the much stronger and clearer main effects.
Discussion
We conducted a preregistered replication of Studies 1b and 3 from Jordan et al. (2011) and tested their theoretical framework, measurement, and analysis strategy with a larger sample. We also went beyond the target article by adding additional tests that seem fit for the experimental design, and by adding extensions testing for the generalizability to emotional experience intensity and examining SCO.
Our results were mixed, and we summarized a comparison of our findings to that of the target in Tables 6 and 9. We concluded a successful replication only regarding people’s underestimation of the prevalence of others’ negative emotions, yet with an unexpected underestimation of the prevalence of positive emotions. In addition, we only observed differences between the underestimation of positive and negative emotional experiences when conducting the betterpowered participant-level analysis, yet not when repeating that target’s item-level analysis, likely due to the small number of items and the analysis being underpowered.
Most surprising was that we found opposite effects to the target’s theory and findings regarding the associations between prevalence estimates and psychological well-being factors. Based on the target’s findings, we expected higher estimations for prevalence of negative emotions to be positively associated with higher well-being, and instead, we found support for a negative relationship. We discuss possible explanations for the mixed findings, followed by a discussion of limitations and suggestions for future research directions.
Replication: Prevalence
We conclude mixed findings concerning our replication of the systematic misperception of others’ emotional experiences. We found that (a) people underestimated the prevalence of others’ negative and positive emotional experiences, (b) underestimation errors were stronger for negative experiences than for positive experiences, (c) prevalence estimation of others’ negative emotions was positively associated with loneliness, rumination, depressive symptoms, and negatively associated with life satisfaction and subjective happiness, (d) prevalence estimation of others’ positive emotions was positively associated with life satisfaction and subjective happiness and negatively associated with loneliness, rumination, and depressive symptoms.
Jordan et al. (2011) argued that underestimation was mainly about negative emotional events, yet instead, we found that participants also underestimated positive emotional experiences, albeit to a lesser extent than they did for

16

Personality and Social Psychology Bulletin 00(0)

negative events. This supports a needed reframing of their null hypothesis that we suggested in our preregistration (Hypothesis 2b; see Table 2), that instead of null effects for positive emotions, the hypothesis could be that the underestimation of positive events is weaker than that of negative events. We note, however, that there would still be an unexplained inconsistency with the target article’s Study 2, which extended the idea from Study 1, and using a preexisting dataset from Srivastava et al. (2009) showed very large differences in misestimation between negative and positive emotions, such that negative emotions were underestimated (8 out of 9 emotions), compared to an overestimation of positive emotions (7 out of 8 emotions). We therefore see much value in conducting also a follow-up replication of the target’s Study 2 with possible theoretical and empirical extensions that would try and resolve the differences in results.
Prevalence Estimations and Well-Being
We failed to find support for the findings regarding correlates of prevalence estimates with well-being measures of loneliness, rumination, depressive symptoms, life satisfaction, and subjective happiness. Instead, we found support for the opposite effects of those reported in the target article.
It is difficult to resolve the inconsistent findings regarding the associations between prevalence estimations and well-being. We reached out to the target article’s authors to consult with them regarding the diverging findings (May, 2023), and we were unable to identify the reason for the divergence.
We note that in our view there was a misalignment between the theoretical framework and the analyses performed in the target article that may at least partly account for the contradictory replication findings. Jordan et al. (2011) argued that feeling alone in negative emotional experiences may lead to the feeling that negative emotions are less common and subsequently pathologize their experiences, causing negative psychological impact. Consistent with the target’s findings, Whillans et al. (2017) found that first-year arrivals who perceived themselves as less socially connected than their peers reported lower belonging and well-being. However, the methods in the target article are not fully aligned with the testing of their main theory, since the main argument seems to refer to a comparison between self and others’ emotional experiences, which differs from the main effect correlational analyses that they reported to test their hypotheses. The original analyses may have stemmed from an assumption that the rating person had experienced the negative emotional events before and so perceiving others as also experiencing those negative events helped them feel less alone. However, an alternative scenario might be that the rating person had not experienced the negative emotional event before, and so thinking that negative experiences were common contributed to them feeling more alone. In our replication analyses, we first simply followed their methodology

and data analysis strategy, yet a more suitable data analysis would have been to examine the interaction between one’s own experiences and one’s perception of others and its association with factors like loneliness. We added exploratory analyses that indeed suggest that self-reports and estimates of others’ negative emotional experiences may interact in predicting well-being factors, such that having fewer negative experiences makes viewing more others as having negative experiences predict one’s own lower well-being.
We therefore offer several suggested hypotheses to be confirmed in future studies: (a) feeling alone in emotional experiences is negatively associated with well-being, (b) for those who have experienced negative emotions, a lower prevalence estimate makes them feel as struggling alone and hence be associated with poorer well-being, and (c) for those who have not experienced negative emotions, a higher prevalence estimate of negative emotions is associated with poorer well-being. Given the correlational methods, the causal direction remains unclear and future research may try to further explore causality in the interplay between prevalence estimations and well-being.
Extension: Intensity
We ran extensions examining if the underestimations extend to an underestimation of others’ emotional intensity. Our findings showed that (a) people overestimate the intensity of others’ negative emotional experiences, (b) people overestimate the intensity of others’ positive emotional experiences, but to a lesser extent than for negative emotions. This diverged from our initial predictions that misestimation of intensity would differ for negative versus positive, and should therefore be subjected to further confirmation with additional research.
Challenging and Reframing Misestimation: Prevalence and Intensity, Positive and Negative
Our findings challenge the target article’s findings in several important ways. First, the effects seem to encompass both positive and negative emotional events, yet with stronger prevalence underestimations for negative than for positive emotional events. We also found that misestimation of others’ emotions extends to misestimation of others’ emotional intensity, yet, crucially, in opposite directions. Our findings suggest that people underestimate prevalence yet at the same time overestimate intensity. If that is indeed the case, then the story shifts from people not being sensitive enough to others’ positive emotions to people sensing fewer but stronger intensity instances of both negative and positive emotions.
One direction is focusing on the broadcasting side of interpersonal exchanges, on how people exhibit their emotional experiences. It could be that estimations are accurate and effectively capture what people indeed exhibit—that people suppress or are able to hide low intensity emotional

Ip and Feldman

17

events (Srivastava et al., 2009) yet are less able to suppress their high intensity emotional events.
A different possibility all together is focusing on the receiving side, in that it might not be at whether emotional experiences occur in solitude or how they are exhibited, but rather more about how we sense, code, remember, and recall social information about emotional experiences, in that people tend to focus on, respond to, and remember the strongest emotional events. People rely on vivid memories of past expressions of emotional experiences (Doré et al., 2016), which tend to be events that are emotionally intense (Yonelinas & Ritchey, 2015). Hence, the most salient expressions may color their evaluations of others’ experiences. While they may not recall many instances, leading to an underestimate of prevalence, the ones that are recalled are the most vivid ones, leading to an overestimation of intensity.
Extension: SCO
We found support for associations between SCO, misestimations, and well-being: (a) SCO was positively associated with depressive symptoms level, loneliness, and brooding, (b) SCO was negatively associated with life satisfaction and subjective happiness, and (c) SCO was positively associated with prevalence estimates and intensity estimates overall.
These associations are in line with the wealth of literature suggesting that SCO, which is mostly upward social comparison (Festinger, 1954), created a discrepancy between the ideal self-presented to others and the real self (Yu & Kim, 2020). This induced a sense of inferiority and distress as these comparisons maintain or even exacerbate negative self-evaluations (Vogel et al., 2015). As a result, these negative evaluations about oneself were found to cause detrimental effects on well-being characterized by lower life satisfaction and lower subjective well-being (Verduyn et al., 2015).
We predicted yet found no support for the idea that SCO interacts with the association between misestimation and psychological well-being. This may be suggestive of the story being less about comparisons between self and others and more about how people communicate or receive socialemotional information, though null effects should be interpreted with caution and humility. Future research can further contrast social comparisons with social attention/ awareness and/or empathy as impacting emotional event misestimation.
Limitations and Future Directions
We note several limitations in our current replication, which may partly explain some of our diverging findings. We noted several weaknesses we spotted in the target’s analyses and adjustments we made to address those (see Table 5 for the summary of deviations), and so it is possible that one of

those affected the findings. For example, we followed the target’s method and their items as closely, and yet, some items were specific about an emotion regarding a specific event (e.g., feeling sad because of a low grade), and it is possible that the estimate of the frequency of the emotions was based on estimates of the frequency of the event and that for the described events there were other emotions that were more relevant than those that the authors had in mind (e.g., feeling shame rather than sadness over a low grade). To avoid that, future research may aim to separate context from emotions, to help gain a more accurate understanding of the cause for misestimation.
We note that we adopted the same emotional experiences items that appeared in the original article, trying to stay as close as possible to the setting of the original article. Yet, the procedure in the target article included a study aiming to generate the items, meant to adjust the items to be as relevant as possible to the target sample. Though we were careful to run the study with a sample of students that would find the experiences listed relevant, it could be that not following the whole procedure may have affected the findings somehow. We do not think this to be a major issue, given the strong support for the targets’ findings regarding prevalence underestimation, which suggests these were suitable.
Our sample was similar to the target article’s sample with U.S. American students, yet with a more diverse and more heterogeneous population recruited online on Prolific compared to the target article’s sample which was recruited from a single school. It is possible that this change may have impacted our findings in some way and that there are some differences (sociocultural, economics, etc.), which would on the one hand show that the consistent findings are generalizable, yet that the heterogeneity may have led to some of the inconsistent findings. We believe that research should aim for more diverse and heterogenous samples to go beyond a singular context, sample, or point in time, in order to maximize impact and practical use. Future research is needed to examine the generalizability of the phenomenon and retest our and the target article’s methods in other contexts.
Conclusion
Overall, we found only partial support for the findings of Jordan et al. (2011) research regarding misestimation of the prevalence of others’ emotional experiences. Consistent with the original article, we found support for an underestimation of the prevalence of others’ negative and positive emotional experiences. However, inconsistent with the target article, we found that the estimation of negative emotions prevalence was positively associated with loneliness, rumination, and depressive symptoms, and negatively associated with life satisfaction and subjective happiness. On the other hand, the estimation of positive emotions prevalence was positively associated

18

Personality and Social Psychology Bulletin 00(0)

with life satisfaction and subjective happiness and negatively associated with loneliness, rumination, and depressive symptoms. We also ran an extension examining estimation of others’ emotional intensity, and unexpectedly found that people tended to overestimate others’ positive and negative emotional intensity. In another extension, we found SCO as associated with misestimation and well-being, yet does not moderate the link between prevalence estimations and well-being.
Acknowledgments We thank Nikolay Petrov (ORCID: 0000-0002-1305-0547) and Hirotaka Imada (ORCID: 0000-0003-3604-4155) for their help with checking and verifying our analyses and thinking about the surprising associations. We are very grateful for their insights.
Authorship Declaration Ho Ching Ip conducted the replication as part of her thesis. Gilad Feldman was the thesis supervisor and led the replication efforts. He supervised each step in the project, conducted the preregistrations, ran data collection, revised and edited the manuscript for final submission.
Author Contributions Ho Ching Ip: Conceptualization; Pre-registration; Formal analysis; Investigation; Methodology; Software; Visualization; Writing – original draft. Gilad Feldman: Conceptualization; Preregistration; Data curation; Funding acquisition; Preregistration peer review/verification; Data analysis peer review/verification; Project administration; Resources; Supervision; Validation; Writing – review and editing.
Declaration of Conflicting Interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the University of Hong Kong seed funding awarded to Gilad Feldman.
ORCID iDs Ho Ching IP https://orcid.org/0000-0002-9549-7544 Gilad Feldman https://orcid.org/0000-0003-2812-6599
Supplemental Material Supplemental material is available online with this article.
Note 1. This is our reframing of the target’s original framing claiming no
misestimation of positive emotions, a null hypothesis, more on that below.

References
Alfasi, Y. (2019). The grass is always greener on my Friends’ profiles: The effect of Facebook social comparison on state self-esteem and depression. Personality and Individual Differences, 147, 111–117. https://doi.org/10.1016/j.paid.2019.04.032
Allport, F. H. (1924). Social psychology. Riverside Press. Burke, M., & Kraut, R. (2013, February). Using Facebook after
losing a job: Differential benefits of strong and weak ties [Conference session]. Proceedings of the 2013 Conference on Computer Supported Cooperative Work, pp. 1419–1430. https://doi.org/10.1145/2441776.2441936 Cole, J. C., Rabin, A. S., Smith, T. L., & Kaufman, A. S. (2004). Development and validation of a Rasch-derived CES-D short form. Psychological Assessment, 16(4), 360. https://doi. org/10.1037/1040-3590.16.4.360 Darley, J. M., & Latané, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. Journal of Personality and Social Psychology, 8(4), 377–383. Diener, E. D., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49(1), 71–75. https://doi.org/10.1207/s15327752jpa4901_13 Doré, B. P., Meksin, R., Mather, M., Hirst, W., & Ochsner, K. N. (2016). Highly accurate prediction of emotions surrounding the attacks of September 11, 2001 over 1-, 2-, and 7-year prediction intervals. Journal of Experimental Psychology: General, 145(6), 788. https://doi.org/10.1037/xge0000168 Ekman, P., & Friesen, W. V. (1969). The repertoire of nonverbal behavior: Categories, origins, usage, and coding. Semiotica, 1(1), 49–98. https://doi.org/10.1515/semi.1969.1.1.49 Feldman, G. (2023). Registered report stage 1 manuscript template. https://doi.org/10.17605/OSF.IO/YQXTP Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140. https://doi.org/10.1177/0018 72675400700202 Gibbons, F. X., & Buunk, B. P. (1999). Individual differences in social comparison: Development of a scale of social comparison orientation. Journal of Personality and Social Psychology, 76(1), 129. https://doi.org/10.1037/0022-3514.76.1.129 Hays, R. D., & DiMatteo, M. R. (1987). A short-form measure of loneliness. Journal of Personality Assessment, 51(1), 69–81. https://doi.org/10.1207/s15327752jpa5101_6 Jané, M., Xiao, Q., Yeung, S., Ben-Shachar, M. S., Caldwell, A., Cousineau, D., Dunleavy, D. J., Elsherif, M., Johnson, B., Moreau, D., Riesthuis, P., Röseler, L., Steele, J., Vieira, F., Zloteanu, M., & Feldman, G. (2024). Guide to effect sizes and confidence intervals. http://doi.org/10.17605/OSF.IO/ D8C4G Jordan, A. H., Monin, B., Dweck, C. S., Lovett, B. J., John, O. P., & Gross, J. J. (2011). Misery has more company than people think: Underestimating the prevalence of others’ negative emotions. Personality and Social Psychology Bulletin, 37(1), 120–135. Larson, R., Csikszentmihalyi, M., & Graef, R. (1982). Time alone in daily experience: Loneliness or renewal. In: L. A. Peplau & D. Perlman (Eds.), Loneliness: A sourcebook of current theory, research and therapy (pp. 40–53). John Wiley & Sons. LeBel, E. P., McCarthy, R. J., Earp, B. D., Elson, M., & Vanpaemel, W. (2018). A unified framework to quantify the credibility

Ip and Feldman

19

of scientific findings. Advances in Methods and Practices in Psychological Science, 1, 389–402. LeBel, E. P., Vanpaemel, W., Cheung, I., & Campbell, L. (2019). A brief guide to evaluate replications. Meta-Psychology, 3, 1–9. https://doi.org/10.15626/MP.2018.843 Liu, P., He, J., & Li, A. (2019). Upward social comparison on social network sites and impulse buying: A moderated mediation model of negative affect and rumination. Computers in Human Behavior, 96, 133–140. https://doi.org/10.1016/j.chb.2019.02.003 Lyubomirsky, S., & Lepper, H. S. (1999). A measure of subjective happiness: Preliminary reliability and construct validation. Social Indicators Research, 46, 137–155. https://doi. org/10.1023/A:1006824100041 Miller, D. T., & Prentice, D. A. (1994). Collective errors and errors about the collective. Personality and Social Psychology Bulletin, 20(5), 541–550. Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of depressive episodes. Journal of Abnormal Psychology, 100(4), 569. https://doi.org/10.1037/ 0021-843X.100.4.569 Nosek, B. A., Hardwicke, T. E., Moshontz, H., Allard, A., Corker, K. S., Dreber, A., Fidler, F., Hilgard, J., Struhl, M. K., Nuijten, M. B., Rohrer, J. M., Romero, F., Scheel, A. M., Scherer, L. D., Schönbrodt, F. D., & Vazire, S. (2022). Replicability, robustness, and reproducibility in psychological science. Annual Review of Psychology, 73(1), 719–748. Palan, S., & Schitter, C. (2018). Prolific. Ac—A subject pool for online experiments. Journal of Behavioral and Experimental Finance, 17, 22–27. https://doi.org/10.1016/j. jbef.2017.12.004 Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1(3), 385–401. https://doi.org/10.1177/0146621 67700100306 Russell, D., Peplau, L. A., & Cutrona, C. E. (1980). The revised UCLA Loneliness Scale: concurrent and discriminant validity evidence. Journal of Personality and Social Psychology, 39(3), 472. https://doi.org/10.1037/0022-3514.39.3.472 Sargent, R. H., & Newman, L. S. (2021). Pluralistic ignorance research in psychology: A scoping review of topic and method variation and directions for future research. Review of General Psychology, 25(2), 163–184. https://doi. org/10.1177/1089268021995168 Schanck, R. L. (1932). A study of a community and its groups and institutions conceived of as behaviors of individuals. Psychological Monographs, 43(2), i–133. https://doi. org/10.1037/h0093296 Sharpe, N. R., & Roberts, R. A. (1997). The relationship among sums of squares, correlation coefficients, and suppression. The American Statistician, 51(1), 46–48. Simonsohn, U. (2015). Small telescopes: Detectability and the evaluation of replication results. Psychological Science, 26(5), 559–569. Steers, M. L. N., Wickham, R. E., & Acitelli, L. K. (2014). Seeing everyone else’s highlight reels: How Facebook usage is linked to depressive symptoms. Journal of Social and Clinical Psychology, 33(8), 701–731. https://doi.org/10.1521/ jscp.2014.33.8.701

Srivastava, S., Tamir, M., McGonigal, K. M., John, O. P., & Gross, J. J. (2009). The social costs of emotional suppression: a prospective study of the transition to college. Journal of Personality and Social Psychology, 96(4), 883–897. https:// doi.org/10.1037/a0014755
Tandoc, E. C., Ferrucci, P., & Duffy, M. (2015). Facebook use, envy, and depression among college students: Is Facebooking depressing? Computers in Human Behavior, 43, 139–146. https://doi.org/10.1016/j.chb.2014.10.053
Thompson, F. T., & Levine, D. U. (1997). Examples of easily explainable suppressor variables in multiple regression research. Multiple Linear Regression Viewpoints, 24(1), 11–13.
Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumination reconsidered: A psychometric analysis. Cognitive Therapy and Research, 27, 247–259. https://doi. org/10.1023/A:1023910315561
Verduyn, P., Lee, D. S., Park, J., Shablack, H., Orvell, A., Bayer, J., Ybarra, O., Jonides, J., & Kross, E. (2015). Passive Facebook usage undermines affective wellbeing: Experimental and longitudinal evidence. Journal of Experimental Psychology. General, 144(2), 480–488. https://doi.org/10.1037/xge0000057
Vogel, E. A., Rose, J. P., Roberts, L. R., & Eckles, K. (2014). Social comparison, social media, and self-esteem. Psychology of Popular Media Culture, 3(4), 206.
Vogel, E. A., Rose, J. P., Okdie, B. M., Eckles, K., & Franz, B. (2015). Who compares and despairs? The effect of social comparison orientation on social media use and its outcomes. Personality and Individual Differences, 86, 249–256. https:// doi.org/10.1016/j.paid.2015.06.026
Wang, J. L., Wang, H. Z., Gaskin, J., & Hawk, S. (2017). The mediating roles of upward social comparison and self-esteem and the moderating role of social comparison orientation in the association between social networking site usage and subjective well-being. Frontiers in Psychology, 8, 771. https://doi. org/10.3389/fpsyg.2017.00771
Whillans, A. V., Christie, C. D., Cheung, S., Jordan, A. H., & Chen, F. S. (2017). From misperception to social connection: Correlates and consequences of overestimating others’ social connectedness. Personality and Social Psychology Bulletin, 43(12), 1696–1711. https://doi.org/10.1177/014616 7217727496
Yonelinas, A. P., & Ritchey, M. (2015). The slow forgetting of emotional episodic memories: An emotional binding account. Trends in Cognitive Sciences, 19(5), 259–267. https://doi. org/10.1016/j.tics.2015.02.009
Young, S. D., & Jordan, A. H. (2013). The influence of social networking photos on social norms and sexual health behaviors. Cyberpsychology, Behavior, and Social Networking, 16(4), 243–247. https://doi.org/10.1089/cyber.2012.0080
Yu, E., & Kim, H. (2020). Is she really happy? A dual-path model of narcissistic self-presentation outcomes for female Facebook users. Computers in Human Behavior, 108, 106328. https://doi. org/10.1016/j.chb.2020.106328
Zwaan, R. A., Etz, A., Lucas, R. E., & Donnellan, M. B. (2018). Making replication mainstream. Behavioral and Brain Sciences, 41, e120.

