Evaluating convergence between two data visualization literacy assessments

Area of Science:

• STEM Education
• Data Visualization Literacy

Background:

• Data visualizations are vital for understanding quantitative data.
• Data visualization literacy is a key educational objective.
• Current assessments may not fully capture the construct of data visualization literacy.

Purpose of the Study:

• To investigate the extent to which two common graph comprehension assessments measure the same underlying constructs of data visualization literacy.
• To analyze individual variability and error patterns in data visualization literacy assessments.

Main Methods:

• Administered two graph comprehension assessments to a university sample and a demographically representative US adult sample (N=1,113).
• Analyzed overall scores and individual error patterns to compare assessment constructs.

Main Results:

• Overall scores on both assessments were correlated and linked to prior mathematics coursework.
• Analysis of error patterns indicated that the assessments probe distinct components of data visualization literacy.
• No clear correspondence was found between assessment components and the tests' design categories (e.g., value retrieval vs. comparison).

Conclusions:

• While overall scores align, current assessments may not fully capture the multifaceted nature of data visualization literacy.
• Findings suggest a need for developing more comprehensive assessments that better reflect behavioral patterns in data visualization comprehension.
• Future assessments should be organized by components that more accurately represent distinct skills within data visualization literacy.