Two by Two Crossover Design Analysis
Two by two crossover design analysis compares two treatments given to each subject in opposite orders over two periods, estimating treatment, period and carryover effects from within- and between-subject contrasts.
2×2 교차설계 분석은 각 대상자가 두 처리를 서로 반대 순서로 두 시기에 받은 자료에서 대상자 내·대상자 간 대비로 처리효과, 시기효과, 이월효과를 추정하는 방법입니다.
In my AB/BA crossover trial, is drug B better than drug A after accounting for period effects, and is there carryover?
AB/BA 교차설계 임상시험에서 시기효과를 고려해도 B 약이 A 약보다 효과가 있는지, 이월효과는 없는지 알고 싶어요.
Halve each subject's period 1 minus period 2 difference and compare these between the two sequences with a two-sample t test to estimate the treatment and period effects. Test carryover by comparing each subject's sum of the two responses between sequences, keeping in mind that this test has low power.
When to use Two by Two Crossover Design Analysis
- Each subject received both treatments in an AB or BA order across two periods
- The condition is stable so subjects return to baseline after a washout
- You want a within-subject treatment comparison adjusted for a period effect
When to use something else
- Subjects received only one treatment; use an independent-samples t test or ANCOVA
- There are more than two periods or treatments; use a linear mixed model for higher-order crossover designs
Assumptions and requirements
- Washout is long enough that carryover is negligible
- Period differences are approximately normal with equal variance across sequences
- Subjects are independent and dropouts are unrelated to treatment
What the analysis reports
- Treatment effect with confidence interval and t test
- Period effect and t test
- Grizzle carryover test on subject totals
- Period 1 only treatment comparison for reference
How to report it
In the 2 × 2 crossover trial (n = [value] AB, n = [value] BA), the treatment effect was [value] (95% CI [[value], [value]]), t([df]) = [value], p = [value]; the period effect was [value], p = [value], and there was no evidence of carryover, t([df]) = [value], p = [value].
Related methods
Run Two by Two Crossover Design Analysis on Quorant
- Data you need
- a table (CSV or Excel) with an outcome variable, a grouping variable, a time variable, a unit (entity) ID and a treatment indicator
- Price
- 250 credits for one analysis up to 10,000 rows (complexity level 2); larger or wider data costs more. Optional methodologist review: 1200 credits.
- Checking
- Checked by property tests in every deployment
- Before you pay
- A free pre-analysis diagnosis checks your data and gives an exact quote; nothing is deducted until you approve it.