Psychological Safety: A Meta‐Analytic Review and Extension

Although psychological safety research has flourished in recent years, and despite the empirical support for the important role of psychological safety in the workplace, several critical questions remain. In order to address these questions, we aggregate theoretical and empirical works, and draw on 136 independent samples representing over 22,000 individuals and nearly 5,000 groups, to conduct a comprehensive meta‐analysis on the antecedents and outcomes of psychological safety. We not only present the nomological network of psychological safety but also extend this research in 4 important ways. First, we compare effect sizes to determine the relative effectiveness of antecedents to psychological safety. Second, we examine the extent to which psychological safety influences both task performance and organizational citizenship behaviors over and beyond related concepts such as positive leader relations and work engagement. Third, we examine whether research design characteristics and national culture alter validities within the nomological network, thus promoting a more accurate and contextualized understanding of psychological safety. Finally, we test the homology assumption by comparing the effect sizes of the antecedents and outcomes of psychological safety across individual and group levels of analysis. We conclude with a discussion of the areas in need of future examination.     

Resilience: A Meta‐Analytic Approach

This study investigated the relationship between psychological resilience and its relevant variables by using a meta‐analytic method. The results indicated that the largest effect on resilience was found to stem from the protective factors, a medium effect from risk factors, and the smallest effect from demographic factors.     

A Meta‐Analytic Review of Child‐Centered Play Therapy Approaches

The authors explored the overall effectiveness of child‐centered play therapy (CCPT) approaches through a meta‐analytic review of 52 controlled outcome studies between 1995 and 2010. Hierarchical linear modeling techniques estimated a statistically significant moderate treatment effect size (.47) for CCPT, as well as statistically significant relationships between effect size and study characteristics, including child's age, child's ethnicity, caregiver involvement, treatment integrity, publication status, and presenting issue.     

Narcissism and Leadership: A Meta‐Analytic Review of Linear and Nonlinear Relationships

Past empirical studies relating narcissism to leadership have offered mixed results. This study integrates prior research findings via meta‐analysis to make 4 contributions to theory on narcissism and leadership, by (a) distinguishing between leadership emergence and leadership effectiveness, to reveal that narcissism displays a positive relationship with leadership emergence, but no relationship with leadership effectiveness; (b) showing narcissism's positive effect on leadership emergence can be explained by leader extraversion; (c) demonstrating that whereas observer‐reported leadership effectiveness ratings (e.g., supervisor‐report, subordinate‐report, and peer‐report) are not related to narcissism, self‐reported leadership effectiveness ratings are positively related to narcissism; and (d) illustrating that the nil linear relationship between narcissism and leadership effectiveness masks an underlying curvilinear trend, advancing the idea that there exists an optimal, midrange level of leader narcissism.     

The Therapeutic Reactance Scale: A Measure of Psychological Reactance

This article describes the development of the Therapeutic Reactance Scale, which was developed to measure psychological reactance as defined by Brehm (1966). The scale was factor-analyzed into verbal and behavioral reactance subscales. Reliability and validity data are also presented.     

Leader–Member Exchange (LMX) and Performance: A Meta‐Analytic Review

This paper reports a meta‐analysis that examines the relationship between leader–member exchange (LMX) relationship quality and a multidimensional model of work performance (task, citizenship, and counterproductive performance). The results show a positive relationship between LMX and task performance (146 samples, ρ = .30) as well as citizenship performance (97 samples, ρ = .34), and negatively with counterproductive performance (19 samples, ρ = ?.24). Of note, there was a positive relationship between LMX and objective task performance (20 samples, ρ = .24). Trust, motivation, empowerment, and job satisfaction mediated the relationship between LMX and task and citizenship performance with trust in the leader having the largest effect. There was no difference due to LMX measurement instrument (e.g., LMX7, LMX‐MDM). Overall, the relationship between LMX and performance was weaker when (a) measures were obtained from a different source or method and (b) LMX was measured by the follower than the leader (with common source‐ and method‐biased effects stronger for leader‐rated LMX quality). Finally, there was evidence for LMX leading to task performance but not for reverse or reciprocal directions of effects.     

School Counseling Outcome: A Meta‐Analytic Examination of Interventions

The effectiveness of school counseling interventions is important in this era of evidence‐based practices. In this study, Meta‐Analysis 1 involved treatment‐control comparisons and Meta‐Analysis 2 involved pretest‐posttest differences. The overall average weighted effect size for school counseling interventions was .30. The study examined whether pertinent moderator variables influenced effect sizes. The pretest‐posttest effect size was not significant, so moderator analyses were conducted on treatment‐control comparisons. Analyses of moderator variables indicated school counseling program activities or interventions varied in effectiveness.     

Racial Discrimination and Health Outcomes Among Racial/Ethnic Minorities: A Meta‐Analytic Review

In this meta‐analysis, the authors reviewed 105 studies on the relationships between racial discrimination and health outcomes among racial/ethnic minority Americans. The authors tested for moderator effects of measurement strategies, cultural factors, substance use, gender, and racial group differences. Findings indicate a statistically significant effect size between racial discrimination and health, with the largest effect for mental health and strongest for studies with multi‐item measures. En este metaanálisis, los autores revisaron 105 estudios sobre las relaciones entre la discriminación racial y los resultados para la salud entre americanos pertenecientes a minorías raciales o étnicas. Los autores examinaron los efectos moderadores de las estrategias de medición, los factores culturales, el uso de sustancias, el género y las diferencias entre grupos raciales. Los hallazgos indican un efecto de tamaño estadísticamente significativo entre la discriminación racial y la salud, con el máximo efecto observado en la salud mental y el más fuerte en estudios con mediciones múltiples.     

The Interpersonal Adaptiveness of Dispositional Guilt and Shame: A Meta‐Analytic Investigation

Despite decades of empirical research, conclusions regarding the adaptiveness of dispositional guilt and shame are mixed. We use meta‐analysis to summarize the empirical literature and clarify these ambiguities. Specifically, we evaluate how guilt and shame are uniquely related to pro‐social orientation and, in doing so, highlight the substantial yet under‐acknowledged impact of researchers’ methodological choices. A series of meta‐analyses was conducted investigating the relationship between dispositional guilt (or shame) and pro‐social orientation. Two main methodological moderators of interest were tested: test format (scenario vs. checklist) and statistical analysis (semi‐partial vs. zero‐order correlations). Among studies employing zero‐order correlations, dispositional guilt was positively correlated with pro‐social orientation (k = 63, Mr = .13, p < .001), whereas dispositional shame was negatively correlated, (k = 47, Mr = –.05, p = .07). Test format was a significant moderator for guilt studies only, with scenario measures producing significantly stronger effects. Semi‐partial correlations resulted in significantly stronger effects among guilt and shame studies. Although dispositional guilt and shame are differentially related to pro‐social orientation, such relationships depend largely on the methodological choices of the researcher, particularly in the case of guilt. Implications for the study of these traits are discussed.     

Outcome Severity and Judgments of “Responsibility”: A Meta‐Analytic Review1

Research has provided mixed support for the hypothesis that when an incident results in a more severe outcome, more responsibility will be attributed to a potentially responsible actor. This paper uses the techniques of meta‐analysis to examine this literature. The results support the contention that people attribute greater responsibility for the outcome of a negative incident when that outcome is more severe than when the outcome is minor. The direction of this relationship is consistent across methodologies. However, the strength of the correlation varies depending on which type of judgment participants are asked to make. Because many of these variables are tied to legal concepts, the results are discussed in the context of the expectations of the legal system regarding the impact of outcome severity on each variable.     

Psychological Reactance in College Students: Family‐of‐Origin Predictors

This study examined the relationship between psychological reactance and dimensions of functioning in the family of origin. Psychological reactance is the tendency to exhibit resistance in relation to one's freedoms being restricted. Three hundred par ticipants completed the Therapeutic Reactance Scale (E. T. Dowd, C. R. Milne, & S. L. Wise, 1991), the Family Environment Scale (R. H. Moos & B. S. Moos, 1986), and demographic questions. Results showed that 5 family dimensions (i.e., cohesion, conflict, moral‐religious emphasis, independence, and achievement orientation) significantly predicted psychological reactance. Results also showed that college students from divorced families were more psychologically reactant than students from intact families. Implications of these results for counselors and for future researchers are provided.     

Mitigating Psychological Reactance: The Role of Message‐Induced Empathy in Persuasion

This article examines the role of message‐induced state empathy in persuasion. Message‐induced empathy is conceptualized as a perception–action process that consists of affective, cognitive, and associative components. Twenty professionally produced public service announcements (PSAs) were used as stimuli messages in a 2 (high vs. low empathy) × 2 (antismoking vs. drunk driving) × 5 (messages) mixed‐design quasi‐experimental study. The 289 participants were randomly assigned to each cell and watched five PSAs presented in a random sequence. Results showed that state empathy has unique contribution to predicting persuasion outcomes above and beyond the individual's affective and cognitive responses to the messages. In addition, state empathy also has an indirect effect on persuasion via mitigating psychological reactance.     

A Meta‐Analytic Investigation of Job Applicant Faking on Personality Measures

This study investigates the extent to which job applicants fake their responses on personality tests. Thirty‐three studies that compared job applicant and non‐applicant personality scale scores were meta‐analyzed. Across all job types, applicants scored significantly higher than non‐applicants on extraversion (d=.11), emotional stability (d=.44), conscientiousness (d=.45), and openness (d=.13). For certain jobs (e.g., sales), however, the rank ordering of mean differences changed substantially suggesting that job applicants distort responses on personality dimensions that are viewed as particularly job relevant. Smaller mean differences were found in this study than those reported by Viswesvaran and Ones (Educational and Psychological Measurement, 59 (2), 197–210), who compared scores for induced “fake‐good” vs. honest response conditions. Also, direct Big Five measures produced substantially larger differences than did indirect Big Five measures.     

A Meta‐Analytic Structural Model of Dispositonal Affectivity and Emotional Labor

The exponential growth of the service economy has increased the attention that organizational researchers have paid to the concept of emotional labor. Although much progress has been made in the field, few studies have provided an integrated picture of how individual dispositions, perceived display rules, and emotional labor behaviors shape employee outcomes. To clarify and compare results across this growing body of literature, a quantitative review was developed, along with a theoretically derived path diagram of key emotional labor constructs. Evidence from our structural meta‐analytic model based on 116 primary studies demonstrates that examining affective dispositions and emotional labor constructs and the pattern of positive and negative results helps to clarify and add specificity to the literature. Results were consistent with the perspective that surface acting emotion regulation strategies have a pattern of negative relationships with work outcomes of job satisfaction and stress/exhaustion (but not with job performance), whereas deep acting emotion regulation strategies have a pattern of positive relationships with all of these work outcomes.     

Age Bias in Laboratory and Field Settings: A Meta‐Analytic Investigation1

A meta‐analytic review of age‐discrimination research from laboratory and field settings revealed a significant, yet modest overall effect size with younger applicants and workers evaluated more positively than older applicants and workers. The present predictions and findings were compared with those from previous meta‐analytic investigations by Kite and Johnson (1988) and by Finkelstein, Burke, and Raju (1995). A number of significant moderational relationships were revealed, including negative linear relationships between the relative generalizability of the research (in terms of participants, setting, and stimulus materials) and effect size. In addition, the type of design used (between‐subjects vs. within‐subject) and the specific type of dependent measures used to evaluate applicants and workers were found to significantly moderate effect size.     

The Longevity Effects of Religious and Nonreligious Participation: A Meta‐Analysis and Meta‐Regression

Scholars have long argued that the reduced mortality risk associated with frequent participation in religious services derives from two sources: social participation and religious belief efficacy. In contrast, the reduced mortality risk associated with participation in nonreligious groups is thought to derive solely from the social participation component. This study tests the religious efficacy hypothesis by comparing the effects of religious participation with nonreligious participation using meta‐analyses of 312 mortality risk estimates from 74 publications (providing data on more than 300,000 persons). We found no significant difference between the mean hazard ratio (HR) for low religious participation (HR, 1.32; 95% CI, 1.24–1.41) and the mean HR for low nonreligious participation (HR, 1.25; 95% CI, 1.17–1.33). These findings suggest that the positive health effects of religious participation may largely be attributed to the social participation component, rather than to the religious component of the act.     

Prevalence of Nonsuicidal Self‐Injury in Nonclinical Samples: Systematic Review,Meta‐Analysis and Meta‐Regression

Published prevalence estimates of nonsuicidal self‐injury (NSSI) among nonclinical samples are highly heterogeneous, raising concerns about their reliability and hindering attempts to explore the alleged increase in NSSI over time. Accordingly, the objectives of this study were to investigate the influence of methodological factors on heterogeneity in NSSI prevalence estimates, explore changes over time, and estimate overall international NSSI prevalence. Results showed that methodological factors contributed over half (51.6%) of the heterogeneity in prevalence estimates, and, after adjusting for these factors, NSSI prevalence did not increase over time. Overall, pooled NSSI prevalence was 17.2% among adolescents, 13.4% among young adults, and 5.5% among adults. Clearly, development of standardized methodology in NSSI research is crucial if accurate estimates are desired.     

Job Control and Burnout: A Meta‐Analytic Test of the Conservation of Resources Model

This meta‐analytic study of 71 independent samples from 66 studies (N = 48,528) examined the relationship between job control and burnout. Based on the Conservation of Resources model, job control was hypothesised to have a stronger relationship with depersonalisation and personal accomplishment than with emotional exhaustion. Overall, results supported the main hypothesis. Moderator analyses indicated that the relations tended to be different depending on job types, the national power distance scores of the samples, and the response formats of the Maslach Burnout Inventory. The results imply that interventions in job control can reduce depersonalisation and enhance personal accomplishment.