Psychological bulletin最新文献

Researchers have become increasingly aware that data-analysis decisions affect results. Here, we examine this issue systematically for multinomial processing tree (MPT) models, a popular class of cognitive models for categorical data. Specifically, we examine the robustness of MPT model parameter estimates that arise from two important decisions: the level of data aggregation (complete-pooling, no-pooling, or partial-pooling) and the statistical framework (frequentist or Bayesian). These decisions span a multiverse of estimation methods. We synthesized the data from 13,956 participants (164 published data sets) with a meta-analytic strategy and analyzed the magnitude of divergence between estimation methods for the parameters of nine popular MPT models in psychology (e.g., process-dissociation, source monitoring). We further examined moderators as potential sources of divergence. We found that the absolute divergence between estimation methods was small on average (<.04; with MPT parameters ranging between 0 and 1); in some cases, however, divergence amounted to nearly the maximum possible range (.97). Divergence was partly explained by few moderators (e.g., the specific MPT model parameter, uncertainty in parameter estimation), but not by other plausible candidate moderators (e.g., parameter trade-offs, parameter correlations) or their interactions. Partial-pooling methods showed the smallest divergence within and across levels of pooling and thus seem to be an appropriate default method. Using MPT models as an example, we show how transparency and robustness can be increased in the field of cognitive modeling. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Our meta-analysis on gender differences in sex drive found a stronger sex drive in men compared to women (Frankenbach et al., 2022). Conley and Yang (2024) criticized how we interpreted the findings and provided suggestions regarding the origins of these gender differences, an undertaking that we had refrained from doing in our original work. We concur with several important points made by Conley and Yang (2024): (a) women's sexual experiences are generally more negative than men's, which could partly explain why men report more sex drive; (b) lack of statistical moderation by some sociocultural variables does not imply that the sex drives of men and women are generally unaffected by the social environment; and (c) gender differences in sexuality are likely smaller than they are often portrayed in research, and that the practical impact of this difference is largely unknown. Still, we reject other assertions made by Conley and Yang (2024): (a) we did not frame our findings in support of the view that gender differences in sex drive are determined by biology, (b) we did not conflate response bias with sociocultural biases more broadly, and (c) we did not fail to incorporate and consider gendered cultural messages about sexuality in our methods and discussion. We make several suggestions about future research on these matters. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Sensitive caregiving behavior, which involves the ability to notice, interpret, and quickly respond to a child's signals of need and/or interest, is a central determinant of secure child-caregiver attachment. Yet, significant heterogeneity in effect sizes exists across the literature, and sources of heterogeneity have yet to be explained. For all child-caregiver dyads, there was a significant and positive pooled association between caregiver sensitivity and parent-child attachment (r = .25, 95% CI [.22, .28], k = 174, 230 effect sizes, N = 22,914). We also found a positive association between maternal sensitivity and child attachment security (r = .26, 95% CI [.22, .29], k = 159, 202 effect sizes, N = 21,483), which was equivalent in magnitude to paternal sensitivity and child attachment security (r = .21, 95% CI [.14, 27], k = 22, 23 effect sizes, N = 1,626). Maternal sensitivity was also negatively associated with all three classifications of insecure attachment (avoidant: k = 43, r = -.24 [-.34, -.13]; resistant: k = 43, r = -.12 [-.19, -.06]; disorganized: k = 24, r = -.19 [-.27, -.11]). For maternal sensitivity, associations were larger in studies that used the Attachment Q-Sort (vs. the Strange Situation), used the Maternal Behavior Q-Sort (vs. Ainsworth or Emotional Availability Scales), had strong (vs. poor) interrater measurement reliability, had a longer observation of sensitivity, and had less time elapse between assessments. For paternal sensitivity, associations were larger in older (vs. younger) fathers and children. These findings confirm the importance of both maternal and paternal sensitivity for the development of child attachment security and add understanding of the methodological and substantive factors that allow this effect to be observed. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Although health-promotion interventions that recommend changes across multiple behavioral domains are a newer alternative to single-behavior interventions, their general efficacy and their mechanisms of change have not been fully ascertained. This comprehensive meta-analysis (6,878 effect sizes from 803 independent samples from 364 research reports, N = 186,729 participants) examined the association between the number of behavioral recommendations in multiple-behavior interventions and behavioral and clinical change across eight domains (i.e., diet, smoking, exercise, HIV [Human Immunodeficiency Virus] prevention, HIV testing, HIV treatment, alcohol use, and substance use). Results showed a positive, linear effect of the number of behavioral recommendations associated with behavioral and clinical change across all domains, although approximately 87% of the samples included between 0 and 4 behavioral recommendations. This linear relation was mediated by improvements in the psychological well-being of intervention recipients and, in several domains (i.e., HIV, alcohol use, and drug use), suggested behavioral cuing. However, changes in information, motivation, and behavioral skills did not mediate the impact of the number of recommendations on behavioral and clinical change. The implications of these findings for theory and future intervention design are discussed. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Although trauma-focused psychotherapy (T-F psychotherapy) is the treatment of choice for posttraumatic stress disorder (PTSD), up to one half of patients do not respond to this treatment. Attempts to improve response to T-F psychotherapy have focused on augmenting fear extinction-based factors. Here, a systematic and meta-analytic review of predictors of T-F psychotherapy outcome was conducted with the goal of using an aggregate data-driven approach to elucidate baseline factors associated with treatment outcome. There were 114 studies that met inclusion criteria (N = 61, 970; M_age = 40.1 years; 40.1% female). There were 237 effect sizes across 24 meta-analytic categories. Poorer treatment response is associated with lower pretreatment levels of activation of fear-related brain regions, psychophysiological reactivity to fear provocation, trauma-related cognitions, anger, depression, high-risk alleles of genes linked to fear, lower levels of executive control, and social support. A range of other factors also predicted poorer responses including being male, non-Caucasian, older in age, early trauma occurrence, more trauma experience, history of combat trauma, as well as comorbid sleep, pain, poor quality life, and alcohol abuse difficulties. This review provides one potential explanation for the limited success of T-F psychotherapy augmentation strategies that have focused only on fear circuity mechanisms at the exclusion of other factors. Here, poor response relating to predictors of early trauma onset and comorbidity are consistent with clinical presentations of complex PTSD, which may suggest T-F psychotherapy is less effective for this condition. This collective evidence suggests that clinicians should consider a tailored approach that targets potential barriers to successful treatment response. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Boundary dissolution has broadly been defined as the breakdown of boundaries and loss of psychological distinctiveness in the parent-child subsystem. Qualitative reviews have highlighted the developmental and clinical value of examining boundary dissolution as a multidimensional construct. Though prior work suggests patterns share minimal variance, research has yet to quantitatively synthesize the weighted effect of distinct patterns. The primary aim of this meta-analysis was to aggregate empirical research on associations between boundary dissolution patterns and children's internalizing and externalizing symptoms. Four patterns of boundary dissolution were identified across developmental, clinical, and family systems literatures: (a) enmeshment-entanglement and blurring of the intergenerational parent-child boundary through psychologically controlling and intrusive behaviors, (b) disorganization-chaotic parent-child boundary (e.g., inexplicable, contradictory behaviors, and responses) reflecting no coherent pattern of relating, (c) caregiving-child functions as a caregiver providing parents with instrumental and emotional support and guidance, and (d) coerciveness-child operates as a disciplinarian or authoritarian to intimidate and control parents. The meta-analysis reviewed 478 studies. Although each boundary dissolution pattern was associated with internalizing and externalizing symptoms, weighted effects across patterns significantly varied in magnitude. Regarding externalizing symptoms, the weighted effect of enmeshment was stronger relative to the weighted effect of caregiving. Turning to internalizing symptoms, the weighted effect of enmeshment was stronger than the weighted effect of caregiving and coerciveness. Additionally, the weighted effect of disorganization was stronger than the weighted effect of caregiving. The robustness of weighted effects depended on child, contextual, and methodological characteristics as well as time lag. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

In recent years, our increasing use of technology has resulted in the production of vast amounts of data. Consequently, many researchers have analyzed digital data in attempt to understand its relationship with individuals' personalities. Such endeavors have inspired efforts from divergent fields, resulting in widely dispersed findings that are seldom synthesized. In this two-part study, we draw from two distinct areas of personality prediction across psychology and computer science to explore the convergent validity of self-reports with human perception and machine learning algorithms, the identifiability of the Big Five traits, and the predictability of different types of data. In Study 1, five meta-analyses of human perception studies integrating findings from 24,124 individuals rated across 30 independent samples demonstrated moderate convergent validity across all traits (ranging from ρ = 0.38 for Neuroticism, to ρ = 0.57 for Openness). In Study 2, a multilevel meta-analysis of computer prediction studies reporting 534 effect sizes across 42 studies also demonstrated moderate convergent validity (ρ = 0.30). Multivariate analyses of the significant moderators highlighted that X, Facebook, Sina Weibo, videos, and smartphones had a negative impact on the variance identified. Finally, in synthesizing the extant literature, we discuss the measures used to assess personality and the analytical approaches adopted. We identify the strengths and limitations across each field and explain how interdisciplinary methodologies could advance the testing and development of psychological theory. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

This meta-analytic review investigated the development of narcissism across the life span, by synthesizing the available longitudinal data on mean-level change and rank-order stability. Three factors of narcissism were examined: agentic, antagonistic, and neurotic narcissism. Analyses were based on data from 51 samples, including 37,247 participants. As effect size measures, we used the standardized mean change d per year and test-retest correlations that were corrected for attenuation due to measurement error. The results suggested that narcissism typically decreases from age 8 to 77 years (i.e., the observed age range), with aggregated changes of d = -0.28 for agentic narcissism, d = -0.41 for antagonistic narcissism, and d = -0.55 for neurotic narcissism. Rank-order stability of narcissism was high, with average values of .73 (agentic), .68 (antagonistic), and .60 (neurotic), based on an average time lag of 11.42 years. Rank-order stability did not vary as a function of age. However, rank-order stability declined as a function of time lag, asymptotically approaching values of .62 (agentic), .52 (antagonistic), and .33 (neurotic) across long time lags. Moderator analyses indicated that the findings on mean-level change and rank-order stability held across gender and birth cohort. The meta-analytic data set included mostly Western and White/European samples, pointing to the need of conducting more research with non-Western and ethnically diverse samples. In sum, the findings suggest that agentic, antagonistic, and neurotic narcissism show normative declines across the life span and that individual differences in these factors are moderately (neurotic) to highly (agentic, antagonistic) stable over time. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Experimental research on sequential moral behavior (SMB) has found that engaging in an initial moral (or immoral) behavior can sometimes lead to moral balancing (i.e., switching between positive and negative behavior) and sometimes to moral consistency (i.e., maintaining a consistent pattern of positive or negative behavior). In two meta-analyses, we present the first comprehensive syntheses of SMB studies and test moderators to identify the conditions under which moral balancing and moral consistency are most likely to occur. Meta-Analysis 1 (k = 217 effect sizes, N = 31,242) revealed that engaging in an initial positive behavior only reliably resulted in moral licensing (i.e., balancing) in studies that measured engagement in negative target behaviors (Hedges' g = 0.25, 95% CI [0.16, 0.44]) and only resulted in positive consistency in foot-in-the-door studies using prosocial requests (Hedges' g = -0.44, 95% CI [-0.59, -0.29]). Meta-Analysis 2 (k = 132 effect sizes, N = 14,443) revealed that engaging in an initial negative behavior only reliably resulted in moral compensation (i.e., balancing) in studies that measured engagement in positive target behaviors (Hedges' g = 0.27, 95% CI [0.18, 0.37]). We found no evidence for reliable negative consistency effects in any conditions. These results cannot be readily explained by current theories of SMB effects, and so further research is needed to better understand the mechanisms that drive moral balancing and consistency under the conditions observed. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

This meta-analysis examines the association between exposure to community violence and parenting behaviors (i.e., positive parenting, harsh/neglectful parenting, parent-child relationship quality, and behavior control). A systematic search yielded 437 articles that measured community violence exposure before or at the time of parenting, assessed parenting, and were available in English. There were 342 effect sizes across parenting constructs: positive (k = 101; 68 studies), harsh/neglectful (k = 95; 60 studies), relationship quality (k = 68; 41 studies), and behavior control (k = 78; 51 studies), from 160 reports representing 147 distinct studies. Results of the three-level meta-analyses found small but significant effects between community violence and positive parenting (r = -.059, 95% CI [-.086, -.032]; 95% PI [-.268, .151]), harsh/neglectful parenting (r = .133, 95% CI [.100, .166]; 95% PI [-.107, .372]), parent-child relationship quality (r = -.106, 95% CI [-.145, -.067]; 95% PI [-.394, .182]), and behavior control (r = -.047, 95% CI [-.089, -.005]; 95% PI [-.331, .237]). The association between exposure to community violence and harsh/neglectful parenting and behavior control was moderated by the type of exposure to community violence, informant or source of community violence and parenting data, child age, sex, and race/ethnicity. Given the substantial degree of heterogeneity in overall effect sizes, implications for policy and intervention are tentatively considered while emphasizing that more empirical research on the association between community violence and parenting is essential for advancing the field. (PsycInfo Database Record (c) 2024 APA, all rights reserved).