| fbh_test {auctestr} | R Documentation |
Apply z-test for difference between auc_1 and auc_2 using FBH method.
Description
Apply z-test for difference between auc_1 and auc_2 using FBH method.
Usage
fbh_test(auc_1, auc_2, n_p, n_n)
Arguments
auc_1 |
value of A' statistic (or AUC, or Area Under the Receiver operating characteristic curve) for the first group (numeric). |
auc_2 |
value of A' statistic (or AUC, or Area Under the Receiver operating characteristic curve) for the second group (numeric). |
n_p |
number of positive observations (needed for calculation of standard error of Wilcoxon statistic) (numeric). |
n_n |
number of negative observations (needed for calculation of standard error of Wilcoxon statistic) (numeric). |
Value
numeric, single aggregated z-score of comparison A'_1 - A'_2.
References
Fogarty, Baker and Hudson, Case Studies in the use of ROC Curve Analysis for Sensor-Based Estimates in Human Computer Interaction, Proceedings of Graphics Interface (2005) pp. 129-136.
See Also
Other fbh method: auc_compare,
se_auc
Examples
## Two models with identical AUC return z-score of zero
fbh_test(0.56, 0.56, 1000, 2500)
## Compare two models; note that changing order changes sign of z-statistic
fbh_test(0.56, 0.59, 1000, 2500)
fbh_test(0.59, 0.56, 1000, 2500)