Test reliability

The reliability of a test is a measure of the consistency of a student’s test scores over repeated testing, assuming conditions remain the same – that is, there was no fatigue, learning effect or lack of motivation. Tests with poor reliability might result in very different scores for a student across two test administrations. The reliability of the test was estimated using the Cronbach’s Alpha formula which produces values ranging from 0 to 1. Values above 0.80 are very good. The reliability of NGRT was very high at 0.90.

For interpreting the score of an individual student, the Standard Error of Measurement (SEM) is a more useful statistic than a reliability coefficient. It indicates how large, on average, the fluctuations in standard scores may be and indicates the 68% chance or confidence band. However, most tests show the 90% chance or confidence bands. The SEM for NGRT is 4.7 and for an average-performing student with a Standard Age Score (SAS) of 100, there is a 90% chance that the student’s true SAS will be in the range +/- 8, i.e. from 92 to 108.

The three parallel versions of NGRT are used to monitor student reading ability over time. A study of around 59,000 students who took different versions of the NGRT tests on average 6 months apart found the correlation was 0.83. The correlation for students who took the tests on average one year apart was 0.82 based on around 44,000 students.