Module 3 — Describing your data

Question: who is in this study? Every analysis starts here — know your distributions before you test anything.

Open the sample (Ctrl+Shift+O). All three dialogs are under Analyze → Descriptive Statistics.

Frequencies — categorical variables

Frequencies…, variables Sex, SmokingStatus, StudyZone, Run.

Each variable gets a table of counts, percents and cumulative percents. Read SmokingStatus: roughly half never smoked. Percent vs Valid Percent differ only when values are missing — remember that for module 9.

Descriptives — scale variables

Descriptives…, variables Age, BMI, SystolicBP, HeartRate, Run.

For each: valid N, minimum, maximum, mean, standard deviation. Sanity-check against the codebook: ages should span 18–80, BMI should centre in the mid-20s. This kind of "does the data look like it should" check catches import mistakes before they become wrong conclusions.

Explore — a distribution under a microscope, by group

Explore…, variable SystolicBP, factor Sex, Run.

Explore gives each group the full treatment: mean with confidence interval, median, quartiles, skewness — the numbers behind the boxplot you will draw in module 4.

Exercise

1. Which StudyZone contributed the most participants, and what percent?
2. From Descriptives: is HeartRate's standard deviation large or small relative to its mean, compared with SystolicBP's? (This foreshadows the codebook's warning that HeartRate is pure noise.)
3. Explore BMI by SmokingStatus. Do the group means look different, or do their confidence intervals overlap heavily? Keep your answer for module 6, where you will test it properly.

The syntax trail

frequencies variables=[Sex, SmokingStatus, StudyZone]
descriptives variables=[Age, BMI, SystolicBP, HeartRate]
explore variables=[SystolicBP], factor=Sex