Abstract:

Measurement and evaluation as an inevitable part of teaching learning process is one of the most difficult phases of all educational contexts. Apart from the debate of favoring criterion-referenced evaluation over norm-referenced evaluation or vice versa, the need to conduct norm-referenced evaluation leads educators and practitioners face some statistical problems. However, the reason regarding the preference of criterion-referenced evaluation is usually related to the lack of knowledge regarding the norm-referenced evaluation methods. One of the concerns of this paper is to sate a concise step by step guide for deciding which evaluation method to use with reasons behind. When conducting norm-referenced evaluation becomes a necessity, then determination of its various usages makes the situation even more challenging for educators and practitioners; therefore, different approaches to conduct norm referenced evaluation is another debate topic of this paper. Likewise, many prefer to use z scores and T scores for within group or between groups assessment to place the actual performance of students; however, the determination of how to use these parameters actually roots back to the distribution of scores for a single or multiple groups. As a principle of assessment, Gaussian’s normal distribution has always been considered as a starting point but it is never obvious how to treat the central limit theorem for the sake of conducting norm-referenced evaluation. In addition to that, the place of skewness or kurtosis as parameters of determination to use or not to use norm-referenced evaluation is not clear in the literature. When the complexity of conducting norm-referenced evaluation has been considered, a clear explanation of its correct use for different measurement contexts becomes a necessity. On the other hand, the reasons to conduct norm-referenced evaluation are many. For example, unexpected occurrence of lack of content sampling, testing type, test administration policy and environment, psychological factors and many others are among the reasons to choose norm-referenced evaluation. However, one cannot easily identify these sources of measurement errors and does not tend to correct it for better assessment results. For this very reason, this paper not only aims to outline different approaches to conduct norm-referenced evaluation for different measurement contexts but also provides empirical support for each approach under consideration as well as their advantages and disadvantages to one another.

Keywords:

Norm-Referenced Evaluation, Approaches to Evaluation, Alternative Assessment Methods.