Exploring differences in teachers', administrators', and parents' preferences for data display and whether type of graphic display influences accuracy when extracting information

I used a mixed-method design to investigate (a) what preferences members of three educational audiences---teachers, administrators, and parents---have for post-school outcome data displayed graphically, and (b) whether the audiences differ in their accuracy when extracting information from different graphic displays. Focus group participants were selected purposively to discuss their preferences and inform the development of a questionnaire. Results from the focus group indicated teachers, administrators, and parents preferred grouped column graphs to other graphs presented in the study. The majority of the participants disliked stacked columns.;Based on the results from the focus groups, I developed a 38-item self-administered questionnaire to (a) further investigate preferences for graphic displays, and (b) measure how accurately the educational audiences extracted information from three types of graphs---stacked columns, grouped columns, and line graphs. Respondents viewed graphs containing post-school outcomes data (i.e., employment and postsecondary enrollment rates) and responded to multiple choice questions about the data displayed in the graph.;To investigate the between subjects and within subjects differences in the groups, I used nonparametric tests, Kruskal-Wallis and Friedman. Results from the questionnaire indicated administrators and parents differed significantly in their accuracy when extracting information from graphs; differences were not statistically significant between teachers and administrators or teachers and parents.;These findings have practical implications for state education administrators who typically bear the responsibility of conveying educational results and outcomes to parents and other stakeholders. Implications for practice include the need to (a) know the audience, (b) define the task, (c) simplify data complexity, and (d) provide learning opportunities. Implications for further research include (a) validating and replicating the current study, and (b) investigating preferences and accuracy when extracting information from graphs using authentic graphs from states' public reports.