NWEA s approach to test-retest reliability poses a more rigorous test of reliability. Guidelines for acceptable correlations are difficult here, though r =.85 seems reasonable especially if the parallel test is given at about the same time and are assumed to be equivalent. The question being answered with this type of reliability is, To what extent do two equivalent forms of the test yield the same results? Answers to this question are also stated in terms of a Pearson correlation coefficient (r). That is, the two tests would have the same number and types of items in the same structure, with the same difficulty levels, measuring the same content within a domain. Here two tests are considered to be equivalent in every way except that their items differ. Reliability across forms is typically referred to as parallel forms reliability. In fact, common practice keeps the time short span between tests, acknowledging that time can act to erode the resulting correlation coefficients. Traditionally, a span of two to three weeks is used to separate the two test administrations. The minimum acceptable correlation is considered to be.80 1.00 is a perfect correlation. The question being answered with this type of reliability is, To what extent does the test administered to the same students twice yield the same results from one administration to the next? Answers to this question are stated in terms of a Pearson product-moment correlation coefficient (r).

Reliability across time is often referred to as test-retest reliability or temporal stability. This consistency typically refers to performance of the test across time, across forms or across its items or parts. Reliability Reliability is essentially an index, or more precisely, a set of indices of a test s consistency. Below is a modest attempt to answer these questions as they pertain to the NWEA Achievement Level Tests (ALT) and Measures of Academic Progress (MAP).

Yet, even though questions such as, Is the test valid and reliable? are commonly asked about a test, the answers to this question are not always clearly understood. Literally hundreds of journal articles and book chapters have been written about these interrelated topics. 1 Reliability and Validity Estimates NWEA Achievement Level Tests and Measures of Academic Progress Northwest Evaluation Association 5885 SW Meadows Road, Suite 200 Lake Oswego, Oregon (fax) March 2004Ģ A Few Notes About Reliability and Validity as They Are Reported in NWEA Reliability and Validity Estimates: Achievement Level Tests and Measures of Academic Progress Reliability and validity are two of the words most commonly associated with tests.

Nwea map test validity and reliability