Table 3.
Example of 2 binary diagnoses in a 2-binary examination response table
| Disorder present |
Total | ||
|---|---|---|---|
| Examination | absent | ||
| Fail | a | b | |
| Pass | c | d | |
| Total | C2 | N | |
The table shows binary data of disorder present/absent in the columns and binary data of wrong/correct answers in the rows. The number of cases corresponding to each cell are represented by a, b, c, d. The totals of rows one and two are r1 and r2, respectively, and those of columns one and 2 are c1 and c2, respectively. The phi coefficient value approximates 1 when a and d are higher, approximates −1 when b and c are higher, and approximates 0 when there is no bias in a–d.