**ERIC Number:**ED185092

**Record Type:**RIE

**Publication Date:**1979

**Pages:**21

**Abstractor:**N/A

**Reference Count:**0

**ISBN:**N/A

**ISSN:**N/A

On Latent Structure Models for Measuring Achievement on Hierarchically Related Skills.

Wilcox, Rand R.; Yeh, Jennie P.

The purpose of this paper is to derive explicit estimates of the parameters of a latent structure model when the skills represented by the test items are hierarchically related. Two special cases are described, which may be used as an approximation to the Dayton and Macready model and also provide initial estimates in an iterative estimation procedure when the more general model is being used. In the first case, the probability of a correct response from a randomly chosen examinee who guesses is greater than zero, and the probability of an incorrect response given that the examinee knows is zero. In the second special case, the probability of a correct response from a randomly chosen examinee who guesses is zero, and the probability of an incorrect response given that the examinee knows is greater than zero. A numerical illustration is used to demonstrate the required procedures in each case. (BW)

**Publication Type:**Reports - Research

**Education Level:**N/A

**Audience:**N/A

**Language:**English

**Sponsor:**National Inst. of Education (DHEW), Washington, DC.

**Authoring Institution:**California Univ., Los Angeles. Center for the Study of Evaluation.

**Identifiers:**Item Hierarchies; Latent Structure Analysis