Life tables
Summarises time-to-event data in fixed intervals: how many were at risk, had the event or were withdrawn, and the cumulative proportion surviving.
Menu: Analyze → Survival → Life Tables…
When should I use it?
- Large samples with grouped follow-up times (years since diagnosis, months to dropout).
What are the assumptions?
- Censoring unrelated to the risk of the event.
- Withdrawals spread evenly across each interval (they count as at risk for half the interval).
How do I run it in StatLab?
- Choose time and status, type the event code(s) (default 1) and the interval width.
- With a factor, keep 'Compare factor levels' ticked for the Wilcoxon (Gehan) test.
How do I read the output?
| Output | Meaning |
|---|---|
| Number Exposed to Risk | Entering minus half the withdrawals. |
| Cumulative Proportion Surviving | Probability of surviving past the end of the interval. |
| Hazard Rate | Event rate per unit time within the interval. |
| Wilcoxon (Gehan) Statistic | Overall comparison of the survival distributions of the factor levels (chi-square, df = levels − 1). |
| Mean Score | Positive = this group tends to survive longer. |
Worked example
On the clinic follow-up practice dataset, this command — which the dialog's Paste button also produces — runs the procedure:
PROC svLifeTable /time=months /status=relapse /event=1 /factor=treatment /plot=YES /compare=YES /width=6 /to=36.
Output tables: Life Table — Treatment = Standard · Life Table — Treatment = New · Comparisons for Control Variable · Mean Scores.
How would I report it in APA style?
A Wilcoxon (Gehan) comparison of the survival distributions across Treatment gave χ²(1) = 10.07, p = .002.
Computed by StatLab's validated procedure code when this page was built. The example only illustrates the steps; choose variables that fit your own research question.
What are common mistakes?
- Coding status the wrong way round (event vs censored).