332210
Is the entire story being told? Examining confidence interval and effect size reporting in health education and behavior research
Tuesday, November 3, 2015
Adam Barry, PhD,
Department of Health and Kinesiology, Texas A&M University, College Station, TX
Hong Liu,
Department of Health and Kinesiology, Texas A&M University, College Station, TX
Qian Ji, MS, CHES,
Department of Health & Kinesiology, Texas A&M University, College Station, TX
Amie Klein,
Department of Health and Kinesiology, Texas A&M University, College Station, TX
Ehikowoicho Idoko,
Department of Health and Kinesiology, Texas A&M University, College Station, TX
Jennifer Farmer,
Department of Health and Kinesiology, Texas A&M University, College Station, TX
Louis Dangelo,
Department of Health and Kinesiology, Texas A&M University, College Station, TX
Background: The newly released APA Publication Manual (6th Ed) strongly recommends researchers always report confidence intervals and effect sizes as minimum research expectations. The purpose of the current investigation sought to: (a) determine the frequency of published, peer-reviewed articles in health education and behavior journals reporting confidence intervals and effect sizes, and (b) discuss implications for statistical ‘best practice’ reporting in social science research. Methods: Across a four year span (2010-2013), 1950 peer-reviewed articles were examined from six health education and behavior journals: American Journal of Health Behavior (AJHB), American Journal of Health Promotion (AJHP), Health Education & Behavior (HEB), Health Education Research (HER), Journal of American College Health (JACH), and Journal of School Health (JoSH). Quantitative data from each eligible manuscript was documented using Qualtrics, an online survey software. Results: Of the 1246 quantitative articles in the final sample, approximately 46% (n=581) reported confidence interval use and of those, 519 provided interval estimates, numerically or visually. The journals reporting the highest frequency of confidence interval use were AJHB (21%) and HER (20%). Almost 10% (n=122) of the quantitative articles examined reported an effect size and of those, 44 reported both confidence intervals and effect size. The JoSH (23%) and HER (18%) most frequently reported both measures of statistical best practice. Conclusion: Health education and behavior literature demonstrate inconsistent statistical ‘best practices’ in reporting the minimum expectations recommended by the APA. In efforts to maximize findings for interpretation, evaluation and future replicability scholars should provide comprehensive information regarding research findings.
Learning Areas:
Public health or related research
Social and behavioral sciences
Learning Objectives:
Identify the frequency of confidence interval and effect size reporting in health education and behavior research; Explain the importance of confidence interval and effect size reporting; Discuss implications for statistical ‘best practice’ reporting in social science research
Keyword(s): Statistics, Publication
Presenting author's disclosure statement:Qualified on the content I am responsible for because: I am a second year health education doctoral student at Texas A&M University pursuing an additional certification in advanced research methods. Being the co-principal investigator of this project has provided an in-depth experience in statistical reporting practices among health education and behavior research.
Any relevant financial relationships? No
I agree to comply with the American Public Health Association Conflict of Interest and Commercial Support Guidelines,
and to disclose to the participants any off-label or experimental uses of a commercial product or service discussed
in my presentation.
