TPC Journal V7, Issue 1-FULL ISSUE

The Professional Counselor | Volume 7, Issue 1 93 system was used, which meant that coders were required to assign each article to one category. The use of a single classification system is recommended when coding variables that represent latent meaning and require greater interpretation by the coder (Insch et al., 1997). The variable included the following categories: approaches to counseling, professional practice, population variables, client variables, counselor variables, measurement, and effectiveness of counseling and preventative interventions (OA = 87.6%; α = .85). Approaches to counseling included articles that presented specific addictions-related counseling techniques, models or treatment programs. The professional practice issues category contained articles that described addiction-related counselor training, credentialing, ethics, diagnosis or trends in the field. Population variables included articles that described characteristics of a non-clinical population; the client variables category contained articles that addressed addictions or addictions counseling within a clinical population. Articles on the characteristics or perceptions of professional counselors were assigned to the counselor variables category. The measurement category included any article with a focus on instrument development, formal assessment or psychometrics. Effectiveness of counseling and preventative interventions represented articles that focused on evaluating an intervention or prevention program or technique. Addictive behavior. This variable represented the types of addictive behavior addressed in the articles. Coders were instructed to record all addictive behaviors and substances described in each article (i.e., multiple classification) using a list of categories that included relevant keywords developed by the researchers for the text search. If a specific type of behavior or substance was not discussed, coders labeled the article as general substance use . Categories representing specific behaviors and substances included: general substance use, alcohol, nicotine, opioids, cannabis, stimulants, ecstasy and behavioral addictions (OA = 95%; α = .90). Type of research. Each article was coded as non-research , qualitative , quantitative or mixed methods (OA = 98.5%; α = .98). Classifications were based on past content analysis research of counseling journals (Moro et al., 2016; Ray et al., 2011), and coders were required to assign each article to one category. Articles that were assigned to the three research categories (i.e., qualitative, quantitative and mixed methods) were used to address research question 4. Data analysis. All research articles were coded in order to determine the types of data analytics used by the authors. The coding variable included 15 categories (i.e., descriptive statistics, regression analysis, theme analysis and coding, chi-square test, multivariate analysis of variance/multivariate analysis of covariance, correlation, analysis of variance/analysis of covariance, structural equation modeling, t -test, confirmatory factor analysis, exploratory factor analysis, other nonparametric test, discriminant analysis, canonical analysis, and cluster analysis), and coders were instructed to assign each article to multiple categories when appropriate (e.g., case where a single research article included multiple types of analysis; multiple classification). The scheme for grouping different types of data analysis was based on a framework by Erford et al. (2011) in their content analysis of articles published in JCD . Percentage of agreement among coders in the present study was 82.4% and α = .79. Because inter-rater reliability is slightly below the recommended minimum of .80 (Krippendorff, 2013), readers are encouraged to interpret these results with caution. Research population. The various populations examined in the research articles were recorded using this variable (OA = 91%; α = .88). Because an article could potentially include multiple populations (e.g., African American, male, college students), coders were instructed to code each article with as many categories as necessary (i.e., multiple classification). When coding, research team members used a preliminary list of possible categories derived from several previous content analysis studies of counseling journals (Byrd, Crockett, & Erford, 2012; Smith, Ng, Brinson, & Mityagin, 2008).