Statistical Mediation Analysis

MacKinnon, David P.; Tofighi, Davood

doi:10.1002/9781118133880.hop202025

Cited by 164 publications

(221 citation statements)

References 99 publications

Supporting

Mentioning

210

Contrasting

Unclassified

Order By: Relevance

“…We designed models with (1) illicit drug use predicting the outcome and the two potential mediators, (2) the two potential mediators predicting the outcome, and (3) all control variables predicting the outcome and the two potential mediators. We computed indirect associations using product of coefficients, paired with bootstrapping (2500 draws) to estimate standard errors (MacKinnon, 2008). Bootstrapping was unavailable for models involving numerical integration (logistic and Poisson models).…”

Section: Discussionmentioning

confidence: 99%

Adolescent illicit drug use and subsequent academic and psychosocial adjustment: An examination of socially-mediated pathways

Brière

Fallu

Morizot

et al. 2014

Drug and Alcohol Dependence

View full text Add to dashboard Cite

Background. Questions remain regarding the consequences of illicit drug use on adolescent adjustment and the nature of mechanisms that may explain these consequences. In this study, we examined whether early-onset illicit drug use predicts subsequent academic and psychosocial adjustment and whether associations are socially-mediated by decreased school engagement and increased peer deviancy.Method. 4885 adolescents were followed throughout secondary school. We used regressions to determine whether illicit drug use in grade 7 predicted academic achievement, school dropout, depressive symptoms, and conduct problems in grades 10-11, adjusting for potential confounders.We used path analysis to test whether significant associations were mediated by school engagement and peer deviancy in grade 8.Results. Illicit drug use predicted conduct problems and school dropout, but not academic achievement and depressive symptoms. The association between illicit drug use and conduct problems was fully mediated by increased peer deviancy. The association between illicit drug use and school dropout was partially mediated by increased peer deviancy, but remained mostly direct. No indirect association via decreased school engagement was found. Examination of reverse pathways revealed that conduct problems and academic achievement in grade 7 predicted drug use in grades 10-11. These associations were mediated by peer deviancy and school engagement (conduct problems only).Conclusion. Adolescent illicit drug use influences the risk of school dropout and conduct problems in part by contributing to deviant peer affiliation. Reciprocal social mediation characterizes the association between drug use and conduct problems. A reverse mechanism best explains the association with academic achievement.

show abstract

Section: Discussionmentioning

confidence: 99%

Adolescent illicit drug use and subsequent academic and psychosocial adjustment: An examination of socially-mediated pathways

Brière

Fallu

Morizot

et al. 2014

Drug and Alcohol Dependence

View full text Add to dashboard Cite

show abstract

“…The hypotheses in the present study are tested using two measurement points to establish temporal precedence [29]. Thus, the independent variable was measured at T1, and the mediators and dependent variable at T2.…”

Section: Methodsmentioning

confidence: 99%

“…Furthermore, to fully understand temporal effects among body satisfaction, healthy eating or unhealthy dieting and body mass index (BMI), self-reported variables should be measured at separate time points [29]. Unfortunately, most of research so far has mostly included one measurement point [30][31][32].…”

Section: Introductionmentioning

confidence: 99%

Body satisfaction and body weight in under- and healthy-weight adolescents: mediating effects of restrictive dieting, healthy and unhealthy food intake

et al. 2018

View full text Add to dashboard Cite

Purpose Theoretical models, such as the transdiagnostic model of eating disorders highlight the role of cognitive factors (e.g., the way people perceive their bodies) and their associations with maladaptive weight management behaviors resulting in underweight. This paper aims at testing the indirect association of adolescent's body satisfaction and body mass index (BMI) through restrictive dieting, healthy eating or unhealthy eating as well as moderating role of adolescent's weight status. Methods The study was conducted in 16 public middle and high schools in Central and Eastern Poland. A sample of 1042 under-and healthy-weight white adolescents aged 13-20 (BMI: 12.63-24.89) completed two self-reported questionnaires (fruit, vegetable, and energy-dense food intake) with a 11-month interval. Weight and height were measured objectively. Multiple mediation analysis and moderated multiple mediation analysis were conducted to test the study hypotheses. Results Adolescents less satisfied with their bodies were more likely to diet restrictively and at the same time ate more unhealthy energy-dense food rather than healthy food, which in turn predicted lower BMI. No moderating effects of weight status were found. Conclusions Low body satisfaction is a risk for restrictive diet and unhealthy food intake. Prevention programs may target under-and healthy-weight adolescents who are highly dissatisfied with their bodies, have a high intake of energy-dense food and apply a restrictive diet at the same time. Level of evidence Level III: longitudinal cohort study.

show abstract

“…This value represents the proportion of the total effect that can be explained by the mediator. Approximate randomization tests (MacKinnon, 2008) were exploited to assess the difference in mediation effect between the EXP and BE group. This resampling technique was further used to check the robustness of the results derived from the Sobel test.…”

Section: Mechanisms Of Change In Exposure 12mentioning

confidence: 99%

Pathways to change in one-session exposure with and without cognitive intervention: An exploratory study in spider phobia

Raes

Koster

Loeys

et al. 2011

Journal of Anxiety Disorders

View full text Add to dashboard Cite

Statistical Mediation Analysis

Cited by 164 publications

References 99 publications

Adolescent illicit drug use and subsequent academic and psychosocial adjustment: An examination of socially-mediated pathways

Adolescent illicit drug use and subsequent academic and psychosocial adjustment: An examination of socially-mediated pathways

Body satisfaction and body weight in under- and healthy-weight adolescents: mediating effects of restrictive dieting, healthy and unhealthy food intake

Pathways to change in one-session exposure with and without cognitive intervention: An exploratory study in spider phobia

Contact Info

Product

Resources

About