Cochran-Armitage Trend Test

The Cochran-Armitage trend test assesses whether the proportion of a binary outcome changes linearly across the ordered levels of a categorical exposure, using assigned scores for the levels.

Cochran-Armitage 추세 검정은 순서가 있는 범주형 노출의 수준에 점수를 매겨 이분 결과의 비율이 선형으로 변하는지 검정한다.

Does the proportion of adverse events increase with the dose level?

용량이 커질수록 이상반응 비율이 높아지나요?

The Cochran-Armitage test regresses the group proportions on scores assigned to the ordered groups and tests whether the slope is zero. It has more power than the general chi-square test when the true pattern is a monotone trend.

When to use Cochran-Armitage Trend Test

• A binary outcome is observed across ordered groups such as dose levels or age bands.
• A monotone increase or decrease in risk is hypothesized.
• Group scores can be specified, such as actual doses.

When to use something else

• The groups have no natural order; use the chi-square test of independence.
• The relationship is non-monotone or covariates need adjustment; use logistic regression.

Assumptions and requirements

• Observations are independent.
• The outcome is binary.
• Groups are ordered and the scores reflect their spacing; numeric group values are used as scores, otherwise levels are scored 1, 2, … in sorted order unless scores are supplied.

What the analysis reports

• z statistic and chi-square for trend
• p value
• Proportions by ordered group
• Slope of proportion on score

How to report it

A Cochran-Armitage test indicated a significant linear trend in the proportion of [outcome] across [ordered groups], z = [z], p = [p].

Related methods

• Chi Square Test
• Binary Logistic Regression
• Jonckheere-Terpstra Trend Test

Run Cochran-Armitage Trend Test on Quorant