The purpose of this study is to assess the feasibility, acceptability, and effectiveness
associated with Taking Action for college students with serious mental illnesses (SMI).
The study will be a randomized controlled trial. The investigators plan to recruit 300
college students with SMI to participate in the study. A total of 60 students will be
enrolled in the study per semester, 30 of whom will be randomized to Taking Action and 30
to the control condition. Enrollment will occur over a period of 5 semesters, with two
Taking Action groups running concurrently each semester. The investigators will recruit
from within Temple University and other colleges and universities in the United States.
Participants will be randomly assigned to one of two groups, the experimental condition
(Taking Action) or the control condition (information only). Participants in the
experimental condition will attend five 2.5-hour Taking Action sessions that will be
delivered in a small group format online via videoconferencing.
All participants will complete 3 research interviews: baseline, post-intervention and
follow-up, lasting about 1.5 hours each. Data related to the feasibility of research and
intervention procedures, acceptability of the intervention, and clinical and academic
outcomes will be collected and analyzed. The investigators will assess the feasibility of
the intervention procedures by tracking data pertaining to recruitment, retention, and
assessment using CONSORT guidelines, and data related to implementation of the
intervention (i.e., fidelity, number of intervention meetings attended by participant).
Fidelity to Taking Action will be assessed using a facilitator checklist based on
prescribed content, procedures, and materials for each Taking Action session.
Facilitators will also maintain attendance logs in RedCAP for each Taking Action session.
The investigators will assess the acceptability of the intervention via quantitative and
qualitative questions administered at post-intervention about participants' level of
satisfaction and experiences with the intervention. The investigators will assess the
clinical and academic outcomes of the intervention using the Mechanisms of Action Scale,
Hopkins Symptom Checklist, Recovery Assessment Scale, College Persistence Questionnaire,
Perceived Competence Scale, College Self-Efficacy Inventory, Study Habits Inventory, and
Procrastination Assessment Scale - Students. Some exploratory measures will also be
included in the interviews.
Feasibility and quantitative acceptability data will be reported using descriptive
statistics. Qualitative acceptability data will be analyzed using thematic analysis. To
examine the impact of Taking Action on mental health and academic outcomes random effect
ANOVAs on repeated measures data in PROC MIXES (SAS) will be used to compare the
experimental and control groups over time on the various measures. Prior to running the
random effects ANOVAs above, the investigators will examine whether the experimental and
control groups were different at baseline on any background characteristics despite
randomization, and if so, control for them in the analyses. In addition, exploratory
analyses will be conducted to provide insight regarding the characteristics of students
who are most likely to benefit from Taking Action.
The investigators will recruit 300 individuals, which, with an estimated attrition rate
of 25% based on previous experience with this population, will yield a sample of 225
participants with complete data. In a repeated measures ANOVA with two groups, three time
points, and the correlation between repeated measures assumed to be 0.5, the
investigators will be able to achieve .80 power to detect a small effect size of f=.085
in a time*group interaction term when the probability of a type I error, alpha, is set at
.05. Even so, the investigators anticipate these power estimates are conservative,
because of the use of PROC MIXED in SAS for analyzing repeated measures data. PROC MIXED
does not discard participants with missing data at some time points, making the impact of
missing data and attrition on power much less severe. GPower was used for power analysis.
