Data visualization is an effective strategy to achieve clear scientific data communications. Gerontology has produced a tremendous amount of cross-disciplinary data based on surveys, experiments, and qualitative methods. Types of data and analytical techniques have become more diverse than ever before. However, little is known about how gerontologists use data visualization to communicate their research. The purposes of this study are to document data visualization practice in the premier gerontology journal and to provide basic guidelines for more effective data communication. To capture historical trends, all data graphics published in 1984, 1994, 2004 and 2014 issues in Journal of Gerontology Series B (JGSS) were examined and classified into common types. Results showed that the numbers of articles with data graphics increased from 29 data graphics in 16 (out of a total of 63) articles in 1984 to 163 graphic elements in 67 (out of 109) articles in 2014 in JGSS. Interestingly, the most frequently used data graphics were fairly simplistic: dynamite plots --- a modified bar-chart that displays means and standard deviations --- (45% of articles with data graphics), line-graph (18%) and box-plot (9%). Based on the review of data graphics and current literature in data visualization, it is recommended that data graphics should be diversified and tailored for target audiences, better able to stand alone, and more informative than simple summary statistics. Finally, the advantages and disadvantages of commonly used data graphics for displaying data distributions are illustrated with data from different simulated distributions (e.g., normal, skewed, bimodal).
展开▼
