CCH/HSRC Methods Seminar: Can We Agree On Agreement?

Loading Map....

Date/Time
Date(s) - Feb 12, 2013
12:00 AM - 12:00 AM

Location
Center for Community Health, UCLA Wilshire Center

Category(ies) No Categories


CCH/HSRC Methods Seminar – UCLA-Semel Institute Center for Community Health – UCLA-Semel Institute Health Services Research Center

Can We Agree On Agreement? 

Validity, reliability, and agreement between daily mobile phone self-reports and retrospective recall web-surveys for sex, drugs, and QoL with people living with HIV.

Presented by:

 

Dallas Swendeman, Ph.D.

Assistant Professor-in-Residence, UCLA Department of Psychiatry and Biobehavioral Sciences; Core Scientist, CHIPTS Methods Core; Core Scientist, CHIPTS Development Core

Scott Comulada, DR.P.H.

Assistant Professor-in-Residence, UCLA Department of Psychiatry and Biobehavioral Sciences; Research Scientist, UCLA Center for Community Health

 

Tuesday, February 12, 2013
2pm – 3pm

Center for Community Health, UCLA Wilshire Center
10920 Wilshire Blvd., Suite 350, Conference Room

*** LIGHT REFRESHMENTS WILL BE SERVED ***
We will begin promptly at 2:00 p.m.

Biographies:

 

Dr. Dallas Swendeman, PhD, MPH, is Assistant Professor in the Department of Psychiatry and Biobehavioral Sciences at UCLA DGSOM, a research scientist at the Global Center for Children and Families (GCCF) and the Center for HIV Identification, Prevention & Treatment Services (CHIPTS), and a member of the UC Global Health Institute’s Center of Expertise in Women’s Health and Empowerment. His multidisciplinary training spanned anthropology, psychology, sociology, organizational studies, community health sciences, and applied experience in evidence-based intervention design, randomized trials, and dissemination and scale-up.  Over the past four years his work has focused on innovative methods for research and intervention using mobile phones  (i.e., “mHealth”) in collaboration with computer scientists and co-investigators at UCLA, mHealth software development companies, and collaborators and communities in Los Angeles, India, and South Africa. This work focuses on development of open-source and modular mHealth platforms for patients/clients and providers that are tailorable to wide range of contexts and health challenges, specifically, for child and family well-being and for HIV/AIDS. Dr. Swendeman is currently directing development and testing of mHealth platforms for people living with HIV/AIDS in Los Angeles and India focusing on medication adherence, sexual risks, substance use, and mental health (mood, stress). He is also key collaborator on mHealth projects addressing diet, stress, and physical activity for young mothers in Los Angeles, and for pre- and post-natal support through home visiting by community health workers on projects in South Africa and Los Angeles. Dr. Swendeman also directs the scientific development of family coaching and child/youth development programs at the UCLA Family Commons in Santa Monica and at the RFK/Ambassabor School complex in Koreatown.

 

 

Dr. Scott Comulada is a biostatistician who has served on the UCLA School of Medicine faculty since he joined the Department of Psychiatry and Biobehavioral Sciences as an Assistant Professor-in-Residence in 2010.  He has been a Statistician and then a Research Scientist for the Semel Institute Center for Community since 1999.  Dr. Comulada was an Associate Director of the Methods Core for the Center for HIV Prevention, Identification, and Treatment Services (CHIPTS) from 2009 to 2010 and is currently a CHIPTS Methods Core Scientist.  He earned his B.S. in Biophysics at Pacific Union College, Angwin.  Dr. Comulada earned his M.P.H. in Public Health at Loma Linda University, Loma Linda.  He earned his M.S. and Dr.P.H. in Biostatistics at the University of California, Los Angeles.  Dr. Comulada is currently part of a cross-disciplinary team of scientists, including psychologists, sociologists, and computer scientists, who are developing research methods to assess and evaluate behavioral data from mobile phone-based health applications.

For further information or to view slides of past seminars, please visit

http://chipts.ucla.edu/upcoming-events/archived-events/ or contact Jessica Harwood at JHarwood@mednet.ucla.edu

This seminar series is supported by the National Institute of Mental Health (P 30 MH-58107).

Abstract:

Given the rapidly increasing availability of inexpensive, easy to program, and massively scalable mHealth (and web-based) measurement and intervention tools, researchers and interventionists will be well served in understanding the strengths and limitations of these methods.  This presentation will explore agreement (concordance) between mobile EMA/diary and web-based retrospective recall surveys for self-reported substance use, sexual behaviors, and quality of life among people living with HIV (PLH).  EMA methods are broadly considered to be more valid than retrospective self-reports because they are less subject to recall bias.  EMA/diary methods also provide a much finer granularity of data that allows researchers to examine daily variations in events, behaviors, states, and their covariation.  However, daily/EMA measures are much more burdensome to participants and subject to higher levels of missing data than recall, typically reporting “good” data completion (compliance) rates at around 80%. Compliance rates in EMA/diary methods are dependent on many factors such as, duration of study, incentive structures, participant interest and motivation, reminder or prompting systems, question content (type and frequency of behavior), number of questions, opportunities or options to refuse, and participant training. Discrepancies between EMA/diary and recall reports have been found to vary by the domain measured and by individuals within studies, with some domains and individuals having higher or lower EMA/diary reports compared to recall. In this presentation we will briefly review the relatively small literature on EMA/diary and recall inter-method reliability (agreement) studies for sexual behaviors and substance use. We will then review the methods on our recently completed study with PLH and present the inter-method reliability results.  We will conclude with lessons learned and recommendations for future work in this area.