Back to Annual Meeting Page		American Public Health Association 133rd Annual Meeting & Exposition December 10-14, 2005 Philadelphia, PA

Abstract #119040

Jenine K. Harris, MA and Ryan C. Burke, BS. Center for Tobacco Policy Research, Saint Louis University School of Public Health, 3545 Lafayette Ave, Suite 300, St. Louis, MO 63104-1399, 314-977-8210, harrisjk@slu.edu

Qualitative research is gaining popularity in fields like business, nursing, and public health. Even so, public health has yet to realize many of the advantages of qualitative research due in part to the perception of subjectivity associated with qualitative research. However, subjectivity can be addressed through assessing reliability. Other challenges, such as large amounts of data and time-consuming analysis, can be addressed through a team approach. This research examines how a team can collaborate to code qualitative data with reliable results. In addition, this research will explain how to adapt reliability measures for unusual data characteristics such as multiple coders assigning multiple codes per item. In 2004 the Center for Tobacco Policy Research conducted interviews with 146 tobacco control professionals. The interviews were transcribed, a codebook was developed, and three coders coded the transcripts. Inter-coder reliability was calculated to determine the extent to which coders assigned the same code(s) to the same text units. Kappa is a preferred method of assessing inter-coder reliability. Typically kappa is calculated for two coders assigning single codes to items; this data was unusual in that three coders assigned multiple codes to each text unit. An adaptation of kappa accounting for these characteristics through the use of intraclass correlation coefficients was calculated for a sample of text units. The resulting high kappa value (0.85) showed that a team approach to coding qualitative data can produce reliable results. This adaptation of kappa may be useful to other researchers, allowing public health to benefit from qualitative methods.

Learning Objectives: