ABSTRACT
In this chapter we will examine a statistical design called “repeated measures.” This design is used when the same participants are measured more than once. The repeated measurement may be separated by time, as these situations are:
US • past tense forms provided in conversation after five years and ten years of using
the language
All of these situations would require a repeated-measures design, because they test the same people more than once. Repeated-measures designs are also often found in our field when there is only one testing point in time, but you have related tests that you view as an independent variable, such as:
• scores separated by categories, as when you divide the scores of a test instrument into:
noun and verb parts phonemic distinctions syntax and morphology components
In general, research designs which incorporate repeated measures are quite desirable, as they increase the statistical power of a test. In fact, the more closely correlated the measures are, the higher the statistical power will be with fewer participants (Murphy & Myors, 2004). This is an important factor to consider in our field, where sample sizes are generally small.
