HI560 Unit 3 Descriptive Statistics Report Example

Course	HI560 Health Care Data Analytics
Unit	Unit 3
Paper type	Descriptive statistics report
Length	About 880 words, 6 pages double-spaced
Format	APA 7 student paper
School	Purdue University Global
Program	MS in Health Informatics
Updated	October 2026

Free sample paper for HI560 Unit 3

Shape Before Testing: A Unit 3 Descriptive Statistics Report on 17,840 Outpatient MRI Exams at Two Sites

[Student Name]

Purdue University Global

HI560: Health Care Data Analytics

Unit 3 Descriptive Statistics Report

[Instructor Name]

[Date]

Thistlewood Health, its sites and every statistic are composite; the sources cited are real.

What this part is doingDescribing every variable by site without any p-values signals that this unit summarizes and the later units compare, which keeps the two tasks from blurring.

Approach

Each variable is described in the form its shape demands. Continuous measures receive a center, a spread and a picture; categorical measures receive counts and percentages. Every summary is split by site, and missing values are reported in the last line of each table rather than dropped. No tests appear here, because comparing the sites is the work of Units 5 and 6.

Table One: Patients and Exams by Site

Hospital suite, 9,850 exams; Fennimore center, 7,990 exams.

1. Age in years. Hospital median 54 (interquartile range 41 to 66), mean 53.1 (standard deviation 16.2). Fennimore median 52 (40 to 63), mean 51.4 (15.8). Both distributions are roughly symmetric with a slight left tail from young adults, so mean and standard deviation are shown beside the median.

2. Sex recorded female. Hospital 5,713 (58.0 percent); Fennimore 4,874 (61.0 percent).

3. Exam group. Head-first neuro exams (brain, cervical spine, orbits, internal auditory canals): hospital 2,800 (28.4 percent), Fennimore 3,330 (41.7 percent). Torso exams (lumbar and thoracic spine, abdomen, pelvis, cardiac): hospital 4,930 (50.1 percent), Fennimore 3,430 (42.9 percent). Feet-first extremity exams: hospital 2,120 (21.5 percent), Fennimore 1,230 (15.4 percent).

4. Contrast ordered. Hospital 3,743 (38.0 percent); Fennimore 2,477 (31.0 percent).

5. Scheduled slot of 30 minutes or less. Hospital 2,167 (22.0 percent); Fennimore 5,114 (64.0 percent). Median slot length 45 minutes at the hospital and 30 at Fennimore.

6. Booking question on claustrophobia answered yes. Hospital 92 (0.9 percent); Fennimore 438 (5.5 percent). Not asked: hospital 1,379 (14.0 percent); Fennimore 240 (3.0 percent).

7. Missing values: weight 887 hospital rows (9.0 percent), 1,358 Fennimore rows (17.0 percent); all other fields in this table complete.

Fennimore does proportionally half again as many head-first neuro exams as the hospital, and those are the exams most likely to stop early.

Table Two: Time on the Table

For exams that finished, time from the first to the last acquired series:

1. Hospital: median 31 minutes (interquartile range 24 to 40), 5th percentile 15, 95th percentile 62.

2. Fennimore: median 27 minutes (21 to 35), 5th percentile 13, 95th percentile 55.

Both distributions are skewed to the right, with a long tail of complex contrast studies running beyond an hour. The median and quartiles lead for that reason; a mean would be pulled upward by a few long exams and would describe almost no actual patient. A box plot by site shows the shared shape and Fennimore's shorter center.

What this part is doingExplaining in one sentence why the mean would mislead for table time shows the grader that the choice of summary followed the data, not habit.

Table Three: Completion Status

Of exams that started, as the technologists coded them:

1. Hospital: 9,519 complete (96.6 percent), 331 incomplete (3.4 percent).

2. Fennimore: 7,578 complete (94.8 percent), 412 incomplete (5.2 percent).

These counts take the completion field at face value. Unit 4 tests that field against the images actually acquired before any comparison relies on it.

Among incomplete exams, the technologist's reason (where recorded) at the hospital was claustrophobia 41 percent, motion 27 percent, pain 16 percent, equipment 6 percent and other 10 percent, with 22 percent of incomplete exams carrying no reason. At Fennimore the shares were claustrophobia 52 percent, motion 21 percent, pain 14 percent, equipment 4 percent and other 9 percent, with 9 percent blank. Because the reason field is incomplete and inconsistently applied, later units use completion status as the outcome and report reasons only descriptively.

Pictures

Histograms of age by site share one axis and one bin width, so the slightly younger Fennimore population can be seen directly. A stacked display was considered for exam group and rejected in favor of side-by-side bars on a common baseline, since readers compare positions on a common axis more accurately than segment lengths (Cleveland & McGill, 1984).

Three Differences That Will Matter

1. Exam group. Fennimore's heavier share of head-first neuro exams matters because claustrophobic reactions are more frequent when the head enters the bore first (Dewey et al., 2007).

2. Slot length. Nearly two thirds of Fennimore exams are booked in slots of 30 minutes or less, against about a fifth at the hospital. Slot length is a scheduling decision, not a patient characteristic, which will raise a question in Unit 6 about whether to adjust for it.

3. Age. The age difference is small but runs in the same direction as other differences between the sites, and age is associated with tolerance of long scans.

The booking question also differs sharply, 5.5 percent yes at Fennimore against 0.9 percent at the hospital, but the hospital's 14 percent not asked makes the comparison hard to read, and this report records it without interpreting it.

What the Comparison Units Must Account For

Unit 5 will test the crude difference in incomplete rates. Unit 6 must adjust at least for exam group and age, and must decide openly whether slot length belongs in the adjustment. Weight cannot be used, because it is missing on 17 percent of Fennimore rows. Any comparison that skips these steps would be what epidemiologists call crude, a contrast of two populations rather than of like exams (Rothman et al., 2008).

What this part is doingEnding with a short list of what later units must adjust for turns a descriptive report into the plan for the analysis that follows.

References

Cleveland, W. S., & McGill, R. (1984). Graphical perception: Theory, experimentation, and application to the development of graphical methods. Journal of the American Statistical Association, 79(387), 531-554. https://doi.org/10.1080/01621459.1984.10478080

Dewey, M., Schink, T., & Dewey, C. F. (2007). Claustrophobia during magnetic resonance imaging: Cohort study in over 55,000 patients. Journal of Magnetic Resonance Imaging, 26(5), 1322-1327. https://doi.org/10.1002/jmri.21147

Rothman, K. J., Greenland, S., & Lash, T. L. (2008). Modern epidemiology (3rd ed.). Lippincott Williams & Wilkins.

Reading the HI 560 Unit 3 instructions

The third unit usually asks you to produce descriptive statistics for your data set, split by the groups you will later compare. Expect tables of centers, spreads, counts and percentages, a few well-chosen figures and a short interpretation. Many prompts suggest about six pages. Choose the summary that suits each variable's shape: mean and standard deviation for roughly symmetric measures, medians and quartiles for skewed ones, counts and percentages for categories. Report missing values in each table rather than dropping them. Do not run tests yet; this unit describes and later units compare. End by naming the differences between groups that the comparison units will need to account for.

Inside the HI560 Unit 3 example

An approach section sets the rule that each variable is described in the form its shape demands. Table one gives age, sex, exam group, contrast, slot length, the booking question and missing values by site, and the highlighted sentence notes Fennimore's heavier share of head-first neuro exams. Table two reports time on the table with medians and percentiles and explains why a mean would mislead. Table three gives completion status as coded and the technologists' reasons, with a note that the reason field is unreliable and that the completion field is tested in Unit 4. A short section on pictures explains shared axes and side-by-side bars. Three differences that will matter follow, and a final paragraph lists what Units 5 and 6 must adjust for.

Where the points sit in the HI560 Unit 3 rubric

Descriptive reports are generally scored on appropriate choice of statistics, accuracy, presentation and interpretation. Appropriate choice carries substantial weight: using a mean for a skewed variable without comment, or testing groups in a descriptive unit, costs points. Accuracy is judged on whether percentages and counts reconcile. Presentation earns credit for clear tables, labeled figures and missing values shown openly. Interpretation is rewarded when the report identifies the group differences that matter for later analysis rather than describing every number. APA format and organization complete the rubric. Reports that paste software output without explanation tend to lose most interpretation points, and so do tables whose group columns use different denominators without saying so.

HI560 Unit 3 help: mistakes that cost points

A frequent problem in this unit is reporting means for every variable regardless of shape. Another is dropping missing values silently. Some students run significance tests too early. Figures with different axes for each group make comparison misleading. Percentages that do not add to the stated total draw comments. Interpretation sections sometimes repeat the tables instead of naming what later units must handle. A short closing list of differences to adjust for fixes that. If your data set covers a different service, such as emergency visits or lab orders, share your variables and the prompt, and a descriptive report can be prepared from your own file.

Get an HI560 Unit 3 example written to your instructions

Which variables does your Unit 3 prompt name, and does it supply a file? Send both with the rubric. A free first report follows within 24-48h, each variable summarized in the form its shape calls for, the tables split by whatever groups your assignment compares, and missing values shown rather than dropped. The paper above is an original model document written by our desk, not a submitted student paper and not an official Purdue University Global document.

More HI560 papers and MS in Health Informatics samples

HI560 Unit 3 questions, answered

Should descriptive statistics be split by group before testing?

Usually yes, if a later unit compares those groups, because a reader needs to see how the groups differ in who they contain. Keep the split descriptive: counts, centers, spreads and figures, with no tests. The example shows each hospital side by side and names the differences that later units will have to adjust for.

How should missing values appear in a descriptive table?

As their own row, with a count and a percentage, for every variable and every group. That lets a reader see whether gaps are concentrated somewhere. Percentages for the other categories can then be computed with or without the missing rows, and the table should say which denominator it used so nobody has to guess.

Where can I find a free HI560 Unit 3 sample paper?

The full Unit 3 report is posted here, with three tables by site, a note on each variable's shape and the differences later units must adjust for.

When should a descriptive report use the median instead of the mean?

When the variable is skewed, such as time on the scanner table, because a few long values pull the mean away from what most cases look like.

Why describe completion as coded if the field may be wrong?

Because the descriptive unit reports the data as received. Testing the field against another source belongs to the data quality unit, which can then show how much the repair changes.