Test retest reliability;
A test retest reliability coefficient is compute to determine whether a test measures consistency from one time to another.
This coefficient is found by correlating the scores obtained by a group of people on one administration with their scores on a second administration of a test.
This coefficient is also known as a coefficient of stability.
In test retest method, the score of error is time sampling.
This type of analysis is of value only when we measure traits or characteristics that do not change over time.
Test that measure some constantly changing characteristics are not appropriate for test-retest evaluation.
It takes into account error resulting from difference in conditions associated with the two occasions on which test are administered.
Advantage:
Test retest reliability estimates are used to evaluate the error associated with administering a test at lest two different times.
Test retest reliability is relatively easy to evaluate.
Test retest reliability tends to be higher when the interval between initial test and retest is short.
Limitation:
Because the some test is administered on both occasions, errors due to different samples of test items are not reflected in a test coefficient.
Test retest reliability tends to be lower when the interval is longer.
A carryover effect or practice effect occurs when the first testing session influence scores from the second session.
Conclusion:
Carryover problems are of concert only when the changes over time are random. In cases where the changes are systematic, carryover effects do not harm the reliability.
Because a carryover and practice effects, the time interval between testing sessions must be selected and evaluated carefully.
Sometimes poor test-retest correlations do not mean that a test is unreliable, instead, the suggest that the characteristic under study has changed.
Alternate Form Reliability
One way of avoiding the difficulties encountered in the test retest reliability is through the use of alternate from of the test, one consisting of similar items (with some difficulty level) but not the same items.
Alternate forms reliability compares two equivalent forms of a test that measure the same attributes. It is also known as parallel from or equivalent forms reliability.
Properties of Alternate-Forms
By administering an alternate form after a suitable interval following administration of the first form, a reliability coefficient reflecting errors of measurement due can be determined.
To control for the confounding effect or test form with administration time, Form A should be administered first to half the group and Form B to the other half; then on the second administration, the first group is given from B and the second group is given Form A.
The resulting correlation between scores on the two forms, referred to as a coefficient of stability and equivalence, takes in to account errors due to different administration times or different items.
Similarities with Test-Retest method
Both strategies involved two administrations to the same subjects with an intervening time interval.
Intervening changes in motivation and individual difference in amount of improvement would produce fluctuation in test scores and thereby somewhat reduce reliability estimates in both methods.
Difference with Test-Retest Method
The alternate forms methodology introduces item sampling difference as an additional score of error variance with time sampling.
Even though the two forms may be equally difficult on average, some subjects may find one form quite a bit harder than the other because supposedly alternate items are not equally familiar to every person.
Advantage
The method of alternate forms provides one of the most rigorous assessments of reliability commonly in use.
It provides a useful measure for evaluating many tests.
It is useful in follow-up studies or in investigations of the effect of some intervening experimental factor on test performance.
The use of several alternate forms also provides a means of reducing the possibility of coaching or cheating.
Limitations
There are lots of limitations such as-
It is often very difficult to develop two versions of a test that are truly parallel.
Alternate forms of a test are also quite expensive-nearly doubling the cost of publishing a test and putting it on the market.
For the above reasons, fewer and fewer tests are being released in this format.
এতে সদস্যতা:
মন্তব্যগুলি পোস্ট করুন (Atom)
কোন মন্তব্য নেই:
একটি মন্তব্য পোস্ট করুন