Scribd
Upload
447 views

Unit 4: Qualities of A Good Test: Validity, Reliability, and Usability

The document discusses key qualities of a good test: validity, reliability, and usability. It defines four types of validity (face, content, criterion-related including concurrent and predictive, and construct validity) and how each ensures a test accurately measures the desired objectives. Reliability refers to a test's consistency and is categorized into scorer, content, and temporal reliability. Usability considers practical factors like a test's availability, cost, administration, scoring, and interpreting results.

Uploaded by

mekit bekele
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
447 views

Unit 4: Qualities of A Good Test: Validity, Reliability, and Usability

The document discusses key qualities of a good test: validity, reliability, and usability. It defines four types of validity (face, content, criterion-related including concurrent and predictive, and construct validity) and how each ensures a test accurately measures the desired objectives. Reliability refers to a test's consistency and is categorized into scorer, content, and temporal reliability. Usability considers practical factors like a test's availability, cost, administration, scoring, and interpreting results.

Uploaded by

mekit bekele
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 18

Unit 4: Qualities of a good Test: Validity,

reliability, and Usability


I/ Validity
Meaning
• Validity refers to the capacity of a test to generate
accurate and relevant information (data) that are
emphasized in the objectives of interest.
• Shortly validity is the accuracy of a test in
measuring the desired objectives (learning
outcomes in the context of teaching learning
processes).
• For example, a physics test should measure
accurately the desired knowledge, skills or
competence of physics rather than unrelated
abilities such as language abilities.
• There are four types of validity, namely face
validity, content validity, criterion related
validity, and construct validity.
Face Validity
• This refers to the reasonableness of items (questions)
of a test with regard to the background of the test
takers it is meant for.
• Here reasonableness is explained by relevance of the
items in the test to the background of the test takers ,
the adequacy of the items in generating the desired
information, and the coverage of the test.
• A test may have face validity without having actual
validity.
Content Validity
• Content is concerned with the extent to
which items in a test are related to the
contents of the course, chapter or unit
covered in the teaching-learning processes.
• There shouldn't be items from the contents
areas which are not touched upon in
instruction.
Criterion Related Validity
• This type of validity is concerned with checking the
extent to which a test correlates with another test.
• Here there are two tests.
• One is called predictor and the other is criterion.
• There are two types of criterion related validity,
namely concurrent validity and predictive validity.
Concurrent Validity
• This refers to the extent to which a newly
developed test in certain area of interest is
related to the existing (old) test in the same
area. In this case the existing test is criterion.
• This type of validity is determined when there
is a need to replace the old test with a new
one because the difficulties associated with
the existing test in using it, having high validity
and reliability.
Predictive Validity
• This is concerned with the capacity of a test to
predict the future success of an individual in
certain area of interest.
• For example, in Ethiopian context, the results
on ESLCE may predict how the students scoring
high in it will be successful in their academic
career in the colleges and in the universities.
• Here the ESLCE is predictor whereas a test
given later in the universities to assess the
academic performance of the students is
criterion.
• The higher the degree of correlation between
the scores of the predictor and the scores of
the future criterion test the higher the
predictive validity of the former.
• Here the predictive validity is determined after
a long period of time.
Construct Validity
• This type of validity is concerned with the
degree to which a test result tells us
something meaningful about person's trait
(relatively permanent behavior) such as
achievement motivation, degree of sociability,
introversion etc.
• Construct validity is determined by correlating
measures of the construct with known
measures of observable criterion that the
construct is highly related to.
• For example, if a researcher is interested in
construct validation of a test of achievement
motivation, then he/she might correlate the
scores from such a test with the scores from a
test for scholastic aptitude since achievement
motivation is known to be highly related to
academic achievement.
II/ Reliability
• Reliability refers to the capacity of test to
produce consistent (invariable) results when
taken by different norm groups or when it is
taken by the same group at different times.
• There are three type of reliability: scorer
reliability, content reliability, and temporal
reliability
• Scorer Reliability: This type of reliability is
concerned with the agreement between two
markers of the same answer of a test (also
called inter-scorers' reliability) or consistency
of score of the same answer scored by the
same scorer at different occasions (also called
intra-scorers' reliability).
• Content Reliability: This refers to capacity of
items in a test to measure related or similar
area of behavior.
• Temporal Reliability: This is concerned with
consistency (stability) of test results over time.
• N.B. Reliability of a test could be determined
by one or a combination of the following five
common methods of reliability determination.
These are test-retest method, parallel form
method, split-half method, Kuder-Richardson
formula 20 and 21 methods, and Crombach's
alpha method.
III/ Usability of a Test
• This refers to all the practical considerations
that go into our decisions to use a particular
test or another.
• Such practical considerations include
availability of the test, its cost, mode of
administration, scoring procedures, and test
score interpretations.

You might also like