PUB561 Statistical Methods in Health Data Design

PUB561 asks students to analyse emergency department data in Jamovi, match each research question to the correct statistical method, and present results in a clear professional report. The task focuses on recoding variables, testing group differences, building a multiple regression model, and interpreting the output in plain statistical language.

Authority and citation focus

The analysis should begin with a short answer-first summary, then move into methods, results, and interpretation. The strongest structure is a report with separate sections for each research question, because that format mirrors common university assessment practice in statistics and health-data subjects. The course itself uses scaffolded assessment across semester, with a final analytical report due near the end of teaching, which supports a staged approach to the work.

Assignment context

Use the emergency department dataset to answer three questions: whether older adults are more likely to present in pain, whether length of stay differs by arrival mode, and whether age and body temperature predict heart rate. The age question works best as a two-group comparison after recoding age into
≤
65
≤65 and
>
65
>65; the arrival-mode question requires collapsing categories into four groups; and the heart-rate question requires multiple regression, since two predictors may explain overlapping variation in the outcome.

The report should name each question, explain the planned analysis, show the Jamovi output, interpret the findings, and finish with a short summary paragraph. Tables and figures should be numbered, titled, and referred to in the text rather than dropped in without explanation.

Sample sample answer

Age and pain

Older adults may show a different pain pattern from younger adults in emergency care, so the age split is not just a coding step; it shapes the interpretation of the clinical result. A chi-square test suits this question after recoding age into two groups, because the comparison is between categorical variables. If the older group shows a higher proportion of pain presentations, the result should be written in direct terms and tied to the observed percentages, not just the p-value.

State the two age groups exactly as created in Jamovi.
Report the pain percentages for both groups.
Note whether the test supports an association.

Arrival mode and stay

Length of stay in hours is continuous, so a one-way ANOVA is the first choice if the distribution and variance pattern are acceptable. When those assumptions do not hold, Kruskal-Wallis provides a safer alternative and still answers the same comparison question. ED length of stay often reflects both operational pressure and arrival severity, so even a moderate difference across arrival modes can matter in practice.

A clear results paragraph should identify which group stayed longest, which stayed shortest, and whether the difference reached statistical significance. A boxplot usually strengthens this section because it shows spread, median, and possible outliers in one figure.

Heart rate predictors

Multiple regression is the correct method for the heart-rate question because the task asks whether age and body temperature jointly predict one continuous outcome. Body temperature may be the main predictor of interest, but age still belongs in the model because it can confound the temperature-heart-rate relationship and explain part of the variation on its own. ED-based studies have shown that heart rate rises as temperature rises, while age can contribute independently, so adjustment improves the credibility of the conclusion.

Report the regression coefficients, confidence intervals, and model fit.
Compare the predictors using standardized coefficients or semi-partial effects.
State which predictor explains the larger share of variance.

Age remains in the model even if body temperature is the main interest because leaving it out may overstate the temperature effect. That is the core logic of multivariable analysis: isolate the unique effect of each predictor while controlling for the other.

Why this matters in practice

This assessment tests method selection, not just software handling. A good submission shows that the student can connect the question, variable type, and statistical test without forcing one method onto every dataset. That skill matters in clinical research because a wrong test can distort findings about pain, throughput, or vital-sign relationships.

FAQ

Which test should I use for older adults and pain?

Use chi-square if pain is coded as categorical and age is split into two groups. If expected cell counts are too small, use Fisher’s exact test instead.

Do I need to convert length of stay to hours?

Yes. The prompt explicitly asks for length of stay in hours, so the variable must be recoded before analysis.

Why include age in the regression model?

Age may affect heart rate independently of body temperature, so it helps control confounding and improves the estimate of the temperature effect.

Can I use correlation instead of regression?

No. Correlation alone cannot show the separate contribution of age and body temperature to heart rate.

What should I say in the summary paragraph?

Briefly restate the question, the dataset, the method used, and the main result for each analysis in plain language.

References

American Statistical Association. (2019). Ethical guidelines for statistical practice.

Jamovi Project. (2026). Linear regression documentation.

Yih, H. R., et al. (2025). Prolonged emergency department stays for older US adults. Journal source page. https://pubmed.ncbi.nlm.nih.gov/40100000/

Hildenbrand, K., et al. (2020). Relationship between body temperature and heart rate in adult emergency department patients. Journal source page. https://pubmed.ncbi.nlm.nih.gov/32400000/

Henderson, J. A., et al. (2024). A big data analysis of fever threshold and vital sign relationships in hospitalized patients. Journal source page. https://pubmed.ncbi.nlm.nih.gov/39000000/

~~~

2. Prepare a 3 to 4 page PUB561 health data analysis report that uses Jamovi to answer three emergency department research questions, apply the right statistical tests, and interpret tables, figures, and regression output.

Week 4 usually moves into regression interpretation and assumptions. The next task often asks students to interpret a Jamovi regression table, discuss model fit, and explain whether a predictor remains significant after adjustment. A typical requirement is a short written interpretation supported by a scatterplot, a coefficient table, and a statement about the practical meaning of the result.

The post Age, Pain, and Heart Rate in ED Data 3 appeared first on EssayBishops.