Estimation of Controlled Direct Effects in Longitudinal Mediation Analyses with Latent Variables in Randomized Studies

Abstract

In a randomized study with longitudinal data on a mediator and outcome, estimating the direct effect of treatment on the outcome at a particular time requires adjusting for confounding of the association between the outcome and all preceding instances of the mediator. When the confounders are themselves affected by treatment, standard regression adjustment is prone to severe bias. In contrast, G-estimation requires less stringent assumptions than path analysis using SEM to unbiasedly estimate the direct effect even in linear settings. In this article, we propose a G-estimation method to estimate the controlled direct effect of treatment on the outcome, by adapting existing G-estimation methods for time-varying treatments without mediators. The proposed method can accommodate continuous and noncontinuous mediators, and requires no models for the confounders. Unbiased estimation only requires correctly specifying a mean model for either the mediator or the outcome. The method is further extended to settings where the mediator or outcome, or both, are latent, and generalizes existing methods for single measurement occasions of the mediator and outcome to longitudinal data on the mediator and outcome. The methods are utilized to assess the effects of an intervention on physical activity that is possibly mediated by motivation to exercise in a randomized study.     

Development of the Uni Virtual Clinic: an online programme for improving the mental health of university students

ABSTRACT

There is growing recognition of the importance of addressing the mental health needs of young people attending university. Anonymous, scalable, and evidence-based online interventions can help to reduce burden on university services and increase access to care for marginalised or disconnected students. This paper reports the participatory design methods used to develop the Uni Virtual Clinic (UVC), a comprehensive online programme that was designed to prevent and treat mental health problems and related issues in university students. Data evaluating the participatory design process is also presented. The potential for the implementation of the UVC within university-based counselling services is strong, and has potential to reduce the prevalence of mental disorders in a high-risk group of young people.     

Psychological adjustment in patients post-curative treatment for oesophageal cancer: a longitudinal interview study

Objective: Oesophageal cancer is an increasingly prevalent disease with a demanding post-curative treatment recovery period and sustained longer-term effects. Although post-curative treatment is a key transitionary period, the process of psychological adjustment for the individual is under-researched which limits the evidence base to inform supportive care. The aim of this study was to understand the process of adjustment for oesophageal cancer patients post-curative treatment, in particular the beliefs participants hold regarding their condition and how these are appraised against their experience.

Design: Serial interviews were undertaken with six oesophageal cancer patients who have recently completed curative treatment, at baseline and at 6-month follow-up using interpretative phenomenological analysis.

Results: The findings demonstrate an effortful process of adjustment, including recognising and accepting a changed self, fostering control beliefs over the course of the illness and physical sequelae, searching for meaning, developing illness coherence and moving away from self-blame.

Conclusions: This study is the first to utilise a longitudinal qualitative design in oesophageal cancer, and provides an understanding of post-treatment adjustment over time for this clinical population through which to inform clinical practice and service development.     

ACADEMIC HARASSMENT

Although much has been written concerning the sexual harassment of university students, no research has yet directly examined the behaviors of university professors themselves. The present study describes the responses of 235 male faculty members of a prestigious, research-oriented university who responded to a survey inquiring about social and sexual interaction among faculty and students. Although the majority of the responses focused on mentoring and social interactions, a sizable minority (26%) reported sexual involvement with women students. In addition to item frequencies, a structural analysis of the phenomenon of academic harassment is presented, and discussed in the context of the subjects' responses to an open-ended invitation to comment on the study.     

The Multiple Meanings of Age for Television Content Preferences

Two studies examined how adult age and time of measurement influence media preferences. Study 1 (using TV ratings data from 1970s to 1980s) found that the popularity of genres varied over time, but even after controlling for year, age groups differed in (a) genre choices (consistent with socioemotional selectivity theory) and (b) age of characters viewed (consistent with age identity theory). The average age of characters in each program partially mediated age differences in ratings of comedies and news, and suppressed a negative relationship between viewer age and ratings of violent programs. Study 2 (using 2005 survey data) found similar genre and character preferences. In addition, subjective age predicted age of favorite characters beyond the effect of chronological age.     

The Incidence and Outcomes of Sexual Harassment Among Hispanic and Non–Hispanic White Women: A Comparison Across Levels Of Cultural Affiliation

An extensive body of literature has documented the widespread and damaging nature of sexual harassment among Anglo women, but little attention has focused on the experiences of women of color. The current paper begins to address this gap by comparing sexual harassment incidence rates and psychological and work–related outcomes of Hispanic and non–Hispanic White women across varying levels of cultural affiliation. Chi–square tests of association suggest that risk of harassment increases with affiliation to the mainstream U.S. culture, and hierarchical regression analyses suggest that cultural affiliation moderates the relation between sexual harassment and occupational functioning. These provocative findings are discussed in the context of social and economic vulnerability, socio–sexual norms, and cultural background.