Abstract
In cross‐over designs for the comparison of two treatments, A and B, each patient receives both treatments in a randomized order. We will consider such design when the response of each patient is binary and propose a new test for treatment effect incorporating the information available from like pairs of response.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 583-587 |
| Number of pages | 5 |
| Journal | Biometrical Journal |
| Volume | 26 |
| Issue number | 5 |
| DOIs | |
| State | Published - 1984 |
Keywords
- Binary response
- Cross‐over designs
- Multinomial distribution
- Order effects