A method for separating the atutomatic and controlled components in affect misattribution procedure (AMP)

An ROC curve approach

Abstract

A method for separating the automatic and controlled components in the affect misattribution procedure (AMP) was proposed. This method assumes that the automatic component follows a signal detection model whereas the controlled component can be described as a high-threshold model, and each parameter is estimated by fitting the model to the ROC curve obtained from the data. A series of experiments showed that the model fits the data well and the estimates obtained under experimental manipulations changed in a theoretically predictable way, suggesting the validity of the proposed estimation method.