skip to main content skip to footer

The Effects of Different Types of Anchor Tests on Observed Score Equating RMSE

Author(s):
Liu, Jinghua; Sinharay, Sandip; Holland, Paul W.; Feigenbaum, Miriam; Curley, W. Edward
Publication Year:
2009
Report Number:
RR-09-41
Source:
ETS Research Report
Document Type:
Report
Page Count:
46
Subject/Key Words:
Anchor Test Equating, Middle Difficult Anchor, Mini Anchor, Bias, Root Mean Square Error (RMSE)

Abstract

This study explores the use of a different type of anchor, a midi anchor, that has a smaller spread of item difficulties than the tests to be equated, and then contrasts its use with the use of a mini anchor. The impact of different anchors on observed score equating were evaluated and compared with respect to systematic error (bias), random equating error (standard errors of equating), and total equating error (RMSE) using empirical data. The results show that the midi anchor generally produces more accurate equating results than the mini anchor—with a few exceptions. Our findings suggest that a midi anchor test would be preferred to a mini anchor test if equating accuracy at the top and at the bottom of the score scale is not a primary concern.

Read More