Making statistics meaningful : The factorial survey as a teaching tool in social work education

The aim of this working paper is to expand the range of pedagogical tools for teaching statistics to social work students, by introducing an innovative tool that is based on the factorial survey and intended for practicing statistical analysis. The factorial survey, which was pioneered by the American sociologist Peter Rossi around 45 years ago, involves the use of experimentally constructed fictitious descriptions (vignettes) in the study of social judgments. In the present two-day-exercise, each student first judged the severity of the alcohol or drug use described in 100 randomly sampled vignettes. By carrying out univariate, bivariate and multivariate statistical analyses, the students then identified the predictors of their own individual severity judgments. The statistics labs were instructor-led and included time both for interpreting the results from a statistical point of view and for comparing the substantive results of the students’ individual analyses. A brief assessment of the exercise revealed high mean levels of perceived increases in the students’ statistical knowledge. This teaching tool incorporates several key features that are considered essential for making statistics education more engaging and relevant to students. It is also proposed as an ideal method for integrating statistical instruction into substantive subject courses.KEYWORDS: Factorial survey, vignette experiment, social work, teaching statistics, teaching tool