UNS Conference Portal, The 1st International Conference on Science, Mathematics, Environment and Education 2017

Font Size: 
Equating Test Development Based Anchor Item For Nonequivalent Anchor Test Design (NEAT)
Nonoh Siti Aminah

Last modified: 2017-07-26

Abstract


The research is an early part of the research on equating test that aims: 1) Develop anchor-based Equating Test items for Non Equivalent Anchor Test Design (NEAT). 2) Find the test characteristics on Equating Test results. Equalization steps consist of: 1) Determining the purpose of equivalence. 2) Build more than one test device to measure the same ability. 3) Selecting the data collector design, equivalence requires data that can provide statistical information. 4) Defining operationally equivalence, equivalence requires a model of the relationship between the test to be estimated. 5) Select the statistical estimation method to use. 6) evaluating the results of the equivalence that has been done, evaluation of equalization results include test making, test execution, statistical procedures and the nature of the equation. Equating test procedure used horizontal equating. In NEAT design, there are two tests consisting of (X + Z) and (Y + Z). Many anchor tests are at least 20% of the test length (Brennan, 2006: 20). NEAT design or anchor test design is relatively much chosen, because the design is safe from test leaks and is effectively reviewed in terms of test execution. If the test is followed by a relatively large testee and a relatively long test then the NEAT design is a design worthy of choice, as it is more practical. Reasons to do Horizontal equalization methods include horizontal equalization if you want to add or replace an existing test into a new test that has not done the scale scale. The results of the study have been prepared instruments for the anchor test of the Instrument of Creativity and Critical Thinking Ability in high school physics.