Teachers have been in the era of accountability for some time. There is an increased demand for accountability and the use of non-referenced testing with President Bush's "No Child Left Behind" initiatives. However, there is a growing demand for less reliance on standardized tests. Admission decisions to colleges and universities are being made with less emphasis on using standardized test scores and more on other criteria such as Grade Point Averages (GPAs). GPA is a standard of accountability. However, when you compare GPA and standardized test scores there are frequently differences among students GPA and scores on a standardized test, sometimes very large differences. From the literature we know standardized tests are valid. The question needs to he asked if GPAs are a valid measures of student achievement. GPAs are based in large measure on teacher made tests. If teacher made tests are not valid, how can a students GPA be valid? This paper looks at teacher made tests and validity. The use of a Table of Specifications can provide teacher made tests validity. This paper provides the why a Table should be used and how to construct a Table for their assessment purposes.
The literature is full of articles on accountability issues in education (Eisenberg & Serim, 2002). Others agree. Mehrens and Lehman referred to the "age of accountability ..." as far back as 1973. More recently, Falk (2002) Nathan (2000) and Newell (2002) have spoken to the growing demand for accountability given the massive use of norm referenced testing in today's schools. The literature is full of articles on norm-referenced achievement testing. The literature is full of articles on admission policies and the selection decisions being made on the basis of standardized test scores and grade point averages (Imber, 2002; Jenkins, 1992; Marshall, 1997; Micceri, 2001; Patton, 1998; and Perfetto, 2002).
However, the literature is not full of accountability issues regarding teacher made classroom tests. Teacher made tests have flirted with, had affairs with and been engaged to accountability, but a permanent relationship has not materialized. Now, it is the time for a marriage to take place. The reasoning is simple--the GPA. The grade point average (GPA) is a standard of accountability based on course grades resulting from teacher made, or teacher chosen, content specific tests. And, although the GPA may be considered in selection processes, norm referenced test results may carry more weight. This happens because there is too often a poor relationship between the GPA and scores on norm-referenced achievement tests.
Lei, Bassiri and Schultz, (2001) found that a college GPA was an unreliable predictor of student achievement. Since we assume that norm referenced tests are valid measures, the tendency is to put more weight on those results concerning student achievement. Opponents of standardized achievement testing would argue otherwise. For example, Bennett, Wesley and Dana-Wesley (1999) suggested that a college admission model should be developed to encompass GPA, rank in class and a district performance index or a similar predictor as an alternative to standardized test scores. A formula index based on these predictors would afford some protection in selectivity issues. But, since a GPA may not significantly correlate with norm referenced test results, which measure is the more valid measure? The belief in the validity of norm referenced achievement tests today is strong. However, can we tell if the GPA is a valid measure? Where is the validity data for the teacher made (or chosen) tests on which GPA is calculated? If the teacher made/chosen test is NOT valid, the GPA will not be valid either.
Therefore a marriage between teacher made tests and accountability should take place to insure validity of its offspring, the GPA. For a wedding you need something old, something new, something borrowed and something blue. …