Reliability Estimations
A major part of any quality program is the assurance that
the products are in accordance with the performance objectives. Reliability can
be estimated through a number of methods that are grouped into two types:
single-administration and multiple-administration.
Multiple-administration methods require two assessments. In the test-retest method,
reliability is estimated as the correlation of the Pearson product-moment
coefficient between two administrations of the same measure. In the alternate
forms method, reliability is estimated by the correlation of the Pearson
product-moment coefficient between two different forms of a measure
administrated together. The idea behind test/retest is that one should get the
same score on test 1 as well as in the test 2.
The three main components to this method are as follows:
1. Apply the measurement instrument at two separate times
for each subject.
2. Calculate the correlation between the two separate
measurements.
3. Assume there is no change in the underlying condition
between test 1 and 2.
Single-administration methods include the split-half and internal consistency reliability
methods. The split-half method considers the two halves of a measure as
alternate forms. This "halves reliability" estimate is then moved to
the full test length using the Spearman-Brown prediction formula. Internal
consistency estimate groups questions in a questionnaire to estimate
reliability. For example, one could write two sets of four questions that
measure the same concept (example, class performance) and after collecting the
responses, run a correlation between those two groups of four questions to
decide if the instrument is consistently measuring the concept. The most common
internal consistency estimate is Cronbach’s alpha, which is usually interpreted
as the mean of all possible split-half coefficients.
Clarity of expression (for written assessments), increasing the measure
and informal 
No comments:
Post a Comment