Please enable JavaScript.
Coggle requires JavaScript to display documents.
Module 4: test preparation and specification (step 1: specify the type of…
Module 4: test preparation and specification
step 1: specify the type of measure
measures of maximal performance
examples of this include aptitude tests or achievement tests, including classroom exams and personal selection tests
measures of typical performance
step 2: define the domain of the test
clear specification of the domain to be assessed determines the limits the test is intended to have and is essential for evaluating the content validity of the test
step 2a: what is the intended context of the test?
by limiting the domain to the context in which we are specifically interested, we are likely to reduce the amount of error associated with the test
as an added benefit, test takers might perceive the more specific measures as possessing greater face validity than does a more general instrument measuring a broader domain
step 2b: how is the construct to be measured different than the related constructs?
here we are specifying the differences between our constructs and related constructs
step 2c: what is the dimensionality of the construct?
one might want to develop sub-scales in order to measure the broad construct as accurately as possible
step 2d: how much emphasis should be placed on each dimension or facet of the construct?
should certain facets or dimensions of the test have different weights or measures of importance
step 3: determine whether a closed-ended or open-ended item format will be used
closed ended items limit the amount of explaining a participant can do, if at all. where open ended need an administrator with vast expertise to measure the answers
step 4: determine the item format
for likert scale items anchor scales for strongly agree or strongly disagree are needed to get an accurate understanding of what the participant meant
step 5: determine whether the exam will be administered individually or in a group setting
time and resources play an issue here
when we are allowed to administer individually we are able to get a much better understanding of each individuals responses
step 6: determine the appropriate test length
a shorter test is best to avoid test taker fatigue
step 7: determine the appropriate difficulty level of items that will compose the test
the difficulty of items is dependent on the ability of the sample population tested under the classical test theory assumptions.
questions written above the test takers educational level will most likely lead to errors in variance and reliability