PU525 Unit 2 Birth Data Description Exercise Example

Grams, Weeks, Yes or No and a Trimester: A Data Description Exercise Summarizing One Year of a Composite County's Singleton Births With the Statistic and Graph Each Variable's Measurement Level Permits, and Ending on What the Low Birth Weight Figure Can and Cannot Say

[Student Name]

Purdue University Global

PU525: Biostatistics for Public Health

Unit 2 Data Description Exercise

[Instructor Name]

[Date]

The county's births and every figure are composite and bracketed for the writer's own data; the methods cited are real.

Variable Table

Birth weight: continuous, grams. Gestational age group: ordinal, preterm (under 37 weeks), term (37 to 41) or post-term (42 or more). Maternal smoking during pregnancy: nominal, yes or no. Trimester prenatal care began: ordinal, first, second, third or none. Levels of measurement determine which summaries are meaningful: means and standard deviations for continuous data, counts and percentages for categories, and medians or ordered percentages for ordinal data (Stevens, 1946).

Checking the Data Before Describing It

Before any summary, the file was checked for impossible or implausible values. Two birth weights were recorded as over [6,000] grams for infants born at [34] weeks, almost certainly entry errors; both were checked against the hospital record and corrected. Eleven records lacked the trimester prenatal care began and were reported as unknown rather than dropped, so the percentages for that variable use [2,289] known records and say so. Smoking was missing for [19] records. Describing missing data openly matters, because a percentage that silently drops unknowns can look more complete than it is.

Why the Program Asked for This Description

The county's maternal and child health program is preparing its annual report and deciding whether to expand a smoking cessation service for pregnant women. Its staff wanted to know how common low birth weight and late prenatal care were, and whether smoking was common enough to justify the service. A clear description, with the right statistic for each variable, answers those questions without overreaching, and it sets the baseline for later comparisons.

Describing Without Testing

The exercise deliberately stops short of hypothesis tests. Describing first shows the shape of each variable, catches errors and reveals which comparisons are worth testing. Jumping to tests before describing the data risks testing the wrong thing or missing problems in the data themselves.

Birth Weight

Graph: a histogram with [250]-gram bins. The shape is roughly bell-shaped with a slightly longer left tail, from very small preterm infants. Summary: mean [3,290] grams; standard deviation [560]; median [3,330]; interquartile range [2,990] to [3,650]. Because of the slight left skew, the median sits a little above the mean, and both are worth reporting. Low birth weight, under [2,500] grams: [184] infants, [8.0] percent.

Gestational Age Group

Graph: a bar chart in the order of the categories, not sorted by size, because the order carries meaning. Summary: preterm [230], [10.0] percent; term [1,996], [86.8] percent; post-term [74], [3.2] percent. No mean is calculated: averaging category labels would be meaningless.

Maternal Smoking

Graph: a simple bar chart or a single statement; a pie chart adds nothing for two categories. Summary: [253] mothers smoked, [11.0] percent.

Trimester Prenatal Care Began

Graph: an ordered bar chart. Summary: first trimester [1,702], [74.0] percent; second [437], [19.0] percent; third [115], [5.0] percent; none [46], [2.0] percent. The cumulative share beginning care by the second trimester, [93.0] percent, is a useful summary that respects the order. Each variable gets the summary its measurement level allows, which rules out an average trimester or an average smoking status.

One Cross-Tabulation

Low birth weight by maternal smoking: [34] of [253] infants of smokers, [13.4] percent, against [150] of [2,047] infants of nonsmokers, [7.3] percent. The descriptive comparison suggests a difference worth testing in a later unit; this exercise only describes it.

What the Low Birth Weight Figure Can and Cannot Say

Low birth weight combines two different groups: infants born early, who are small because they had less time to grow, and infants born at term who grew poorly in the womb. A single percentage mixes them, and the two have different causes and different responses (Wilcox, 2001). Of the [184] low birth weight infants, [121] were preterm and [63] were born at term. A program focused on preventing preterm birth would address the first group, while a program on smoking cessation and nutrition might affect both. The county's report should therefore present low birth weight alongside preterm birth, not as a stand-alone measure of infant health.

Presentation Choices

Tables should give counts and percentages together, so readers can judge how many infants a percentage represents. Graphs should use consistent scales and avoid three-dimensional effects. Every figure should carry its denominator, here [2,300] singleton births, so readers never have to guess how many infants a percentage refers to.

Data Notes

Birth certificate data are complete for birth weight but less reliable for smoking, which mothers may underreport, and for the start of prenatal care, which depends on how the first visit is recorded. Multiple births were excluded because they differ systematically in weight and timing (Rosner, 2016).

References

Rosner, B. (2016). Fundamentals of biostatistics (8th ed.). Cengage Learning.

Stevens, S. S. (1946). On the theory of scales of measurement. Science, 103(2684), 677-680. https://doi.org/10.1126/science.103.2684.677

Wilcox, A. J. (2001). On the importance, and the unimportance, of birthweight. International Journal of Epidemiology, 30(6), 1233-1241. https://doi.org/10.1093/ije/30.6.1233