Skip to main content

Table 7 Correct classification rates for extreme unbalanced designs

From: Trend tests for the evaluation of exposure-response relationships in epidemiological exposure studies

Sample sizes

N

Alternative

Power

Correct classif. rate

200/200/200/200

800

.05/.05/.05/.10

.682

.935

540/200/40/20

800

.05/.05/.05/.10

.251

.758

200/200/200/200

800

.05/.05/.10/.10

.792

.831

540/200/40/20

800

.05/.05/.10/.10

.425

.687

200/200/200/200

800

.05/.10/.10/.10

.603

.783

540/200/40/20

800

.05/.10/.10/.10

.755

.854

400/400/400/400

1600

.05/.05/.05/.10

.915

.971

1340/200/40/20

1600

.05/.05/.05/.10

.266

.749

400/400/400/400

1600

.05/.05/.10/.10

.968

.916

1340/200/40/20

1600

.05/.05/.10/.10

.438

.667

400/400/400/400

1600

.05/.10/.10/.10

.903

.904

1340/200/40/20

1600

.05/.10/.10/.10

.832

.883

9740/200/40/20

10000

.05/.05/.05/.10

.252

.702