DPH 860 Module 6 assignment: doctoral paper on visualizing data for readers, a full sample

Show the Data: Visualizing Blood Pressure Results for Readers and Decision Makers

Student Name

Doctor of Public Health Program, Aspen University

DPH 860: Advanced Biostatistics

Instructor Name

Month Day, Year

Show the Data: Visualizing Blood Pressure Results for Readers and Decision Makers

Figures often shape what readers remember from a study more than tables or text. A good figure reveals patterns and uncertainty; a poor one hides them or misleads. This paper plans the figures for a DPH project in which 700 people enrolled in a hypertension outreach effort led by community health workers are compared with 700 similar residents who did not enroll, and it explains the choices behind each.

The Purpose of Each Figure

Every figure should answer one question. Before choosing a chart type, the project wrote the message each figure should convey, such as participants' systolic pressure fell more than comparison adults', and chose the chart that best shows that message with its uncertainty.

Problems With Bar Charts of Means

Bar charts showing only means and error bars are common but hide the distribution. Weissgerber et al. (2015) showed that very different datasets, with outliers, skew or subgroups, can produce identical bar charts, and they urged researchers to show individual data points or distributions, especially for continuous outcomes.

Planned Figures

The table lists the planned figures.

Showing Individual Data

For systolic change, a dot plot shows every participant's value, with a box plot overlay marking the median and interquartile range. With 700 points per group, dots are made small and slightly transparent so overlapping values remain visible. Readers can see that many comparison adults also improved and that distributions overlap considerably.

Showing Estimates and Uncertainty

For blood pressure control, a figure shows the percentage controlled in each group with 95% confidence intervals, and a second panel shows the difference with its interval. Plotting the difference directly helps readers judge the effect size and its uncertainty rather than comparing two intervals by eye. A horizontal reference line at zero difference helps readers see at a glance that the interval excludes no effect.

Forest Plot for Subgroups

A forest plot displays the difference in control within subgroups by age, sex, insurance and baseline pressure, each with its interval, alongside the overall estimate. Wide intervals for small subgroups remind readers that apparent differences may reflect chance.

Flow Diagram

A flow diagram shows how many adults were eligible, enrolled, matched and included in analysis, with reasons for exclusion and missing outcomes. Reporting guidance for observational studies recommends such a diagram so readers can judge selection and completeness (von Elm et al., 2007). Numbers at each step should add up exactly, so readers can trace every adult from eligibility to analysis.

Accessible Design

Figures use color palettes readable by people with color vision deficiency, add shapes or labels in addition to color, use font sizes legible when printed and include descriptive titles and alternative text. Axis labels include units, such as mmHg. Alternative text describes the pattern, not only the chart type.

Avoiding Misleading Figures

Truncated axes can exaggerate differences, and inconsistent scales across panels invite false comparisons. The project starts axes at meaningful values, states when they do not start at zero and keeps scales consistent across panels showing the same measure.

Figures for Different Audiences

The doctoral committee needs detailed figures showing distributions and uncertainty. County commissioners need one or two simple figures, such as the difference in control with its interval and a plain-language caption. The project will prepare both versions from the same data.

Checklist for Final Figures

Before submission, each figure will be checked: does it answer one question, show data or uncertainty appropriately, have labeled axes with units, use accessible colors, include sample sizes and stand alone with its caption? Reporting guidelines stress that figures should be interpretable without the text (Lang & Altman, 2015).

Tables or Figures

Some information is better in tables. Exact values for baseline characteristics belong in a table, while patterns, distributions and comparisons of effect sizes are clearer in figures. The project uses a table for baseline characteristics and figures for outcomes, so that each format does what it does best.

Scatter Plots and Smoothing

A scatter plot with a smoothed line relates each participant's visit count to systolic change. The line suggests larger reductions with more visits up to about 12, then little further change. Because participants with more visits may differ in other ways, the caption notes that the pattern is descriptive, not causal.

Captions

Captions should state what the figure shows, the sample size, what points and lines represent and what the intervals mean. A well-written caption lets a reader who skips the text still understand the figure. The project drafts captions alongside figures rather than adding them at the end.

Software and Reproducibility

Figures are created with code rather than by hand, so they can be regenerated if data change. Code for each figure is stored with the analysis scripts, and figures are exported at high resolution in formats suitable for both print and screen.

Common Mistakes to Avoid

Common mistakes include three-dimensional effects that distort values, pie charts with many slices, dual axes that suggest false relationships and too many colors. The project uses simple two-dimensional charts, limited palettes and a single axis per panel.

Testing Figures With Readers

Draft figures were shown to two committee members and two health department staff. Staff found the forest plot confusing at first, so a short plain-language caption and a reference line at zero were added. Testing figures with intended audiences often reveals problems authors cannot see.

Figures in the Defense

The oral defense will use simplified versions of the manuscript figures, with larger fonts, fewer panels and one message per slide. The committee can ask to see detailed versions when needed. Planning defense figures now saves time and ensures consistency between the written and spoken presentation.

Linking Figures and Text

The text cites each figure where its message comes up, and the text states the key number the figure shows. Readers should never have to guess why a figure is there or what they are meant to notice.

Conclusion

Thoughtful visualization lets readers see what the DPH project found and how certain it is. Replacing bar charts of means with dot and box plots, plotting differences with intervals, using forest plots and a flow diagram, designing for accessibility and tailoring figures to audiences will make the project's results clear and honest.

References

Lang, T. A., & Altman, D. G. (2015). Basic statistical reporting for articles published in biomedical journals: The "Statistical Analyses and Methods in the Published Literature" or the SAMPL guidelines. International Journal of Nursing Studies, 52(1), 5-9. https://doi.org/10.1016/j.ijnurstu.2014.09.006

von Elm, E., Altman, D. G., Egger, M., Pocock, S. J., Gøtzsche, P. C., & Vandenbroucke, J. P. (2007). The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: Guidelines for reporting observational studies. The Lancet, 370(9596), 1453-1457. https://doi.org/10.1016/S0140-6736(07)61602-X

Weissgerber, T. L., Milic, N. M., Winham, S. J., & Garovic, V. D. (2015). Beyond bar and line graphs: Time for a new data presentation paradigm. PLOS Biology, 13(4), Article e1002128. https://doi.org/10.1371/journal.pbio.1002128