Graphical presentation of trends in rates

Area of Science:

• Epidemiology
• Biostatistics
• Scientific Visualization

Background:

• Graphical methods are widely used to analyze temporal trends in disease rates.
• Previous presentations of data have sometimes led to misinterpretations.
• Authors identified areas for improvement in their own graphical methods.

Purpose of the Study:

• To improve graphical presentations of temporal trends in disease rates.
• To investigate how different graphical choices affect the perception of change.
• To provide guidelines for clear and accurate data visualization in epidemiology.

Main Methods:

• Analysis of graphical techniques for displaying disease rates over time.
• Illustration using 40-year disease rate data.
• Examination of the impact of arithmetic vs. logarithmic scales.
• Evaluation of axis scaling ratios and uniform scaling.

Main Results:

• The choice of arithmetic or logarithmic scale significantly impacts the portrayal of trends.
• Appropriate scaling ratios are essential for detecting specific rates of change.
• Uniform scaling facilitates comparisons across different datasets.
• Misleading plots can result from incorrect axis choices, especially with odds ratios or relative risks.

Conclusions:

• Graphs should be designed carefully to accurately reflect data and convey information.
• Justifiable uniformity in graphical presentation offers advantages for analyzing disease trends.
• Attention to plotting techniques is vital in epidemiologic and medical literature.
• Clear visualization is key to preventing overemphasis and ensuring accurate interpretation.