Skip to main content

Table 4 Mean d’ for classifying images with the tumor region removed (“background”)

From: Perceptual learning in the identification of lung cancer in chest radiographs

Experiment and train schedule

Testing image

Pre-test

Post-test 1

Post-test 2

Post-test 3

Final test

Pre-test vs. final test

1A (N = 12)

30-image

Trained

0.12

0.79

0.70

0.83

0.87

P < .002

Untrained

0.32

0.37

0.35

0.48

0.33

P = .98

1B (N = 12)

60-image

Trained

0.18

0.24

0.26

0.73

0.73

P < .03

Untrained

0.26

0.36

0.58

0.30

0.16

P = .63

3 (N = 24)

30-image

Trained

−0.10

Not applicable

0.44

P < .001

Untrained

0.03

Not applicable

0.14

P = .29

  1. P values show results from t tests that compared pre-test with the final post-test. The overall d’ during pre-test was not different from chance, p = .10