Commonly used machine learning applications seem to relate to big data. This article provides a gentle review of machine learning and shows why machine learning can be applied to small data too. An example of applying machine learning to screen irregularity reports is presented. In the example, the support vector machine and multinomial naïve…

This essay argues that reliability should be reconceptualized in a way that reflects the importance of the theoretical expectations of the test specialist and the learning and problem solving of the test takers. It is time to characterize clearly the substantive theoretical framework supporting reliability studies and the technical evaluation of…

One major reason for the problems of test validation is an overemphasis on the need for empirical validity data, and a failure to recognize the primary importance of explicit verbal definitions of what the test is intended to measure and rational arguments in support of the means chosen for obtaining the measurement. (Author/LC)

In response to Ebel (TM 508 146) Gardner argues that neither intrinsic rational validity associated with ability tests nor a validity coefficient relating a test to performance as the sole information about validity is sufficient. All relevant data about a test and its functioning are essential in describing the validity of the test. (Author/LC)