Statistical Analysis

Descriptive Techniques

The measurement level used in a case study determines the choice of data analytic technique. In case studies employing a nominal level of measurement, which uses numbers simply to classify respondents into groups (e.g., male or female), frequency and cross-tabulation tables are most common. Ordinal measurements reflect inherent rank ordering of observed phenomena (e.g., work performance of five employees) and use the mean, median, and mode to indicate rank differences. The mean denotes the average value of a distribution, whereas the median and mode represent the middlemost and typical values in the same distribution, respectively. When multiple observations on two or more variables are available, Spearman rank correlation coefficients can also be computed. A value of 1 indicates perfect agreement between ranks of two variables, such as teacher effectiveness and student performance; a value of −1 indicates that the ranks of one variable are in exactly the opposite order as the ranks of the other; and a value of near zero indicates that the two variables are independent.

With interval-level measurement the numbers take on new meaning. Typically, ratings on a scale, such as “below average” (1), “average” (2), and “above average” (3), are employed here. In this instance, order and quality of magnitude of the differences in the data can be computed along with measures of central tendency such as the mean, median, and mode. However, interval scales do not have an absolute zero point; hence, we cannot say that a teacher who earned a score of 2 is twice as good as another who earned a score of 1.

An interval scale with an absolute zero becomes a ratio scale and permits the researcher to perform various mathematical operations (add, subtract, multiply, and divide).