The 1970s and 1980s were times when communication behavior was a primary interest of many communication scholars. The aim of this book is to reignite some interest in and passion about how human communication behavior should be studied. It presents the best advice, techniques, cautions, and controversies from the 1970s and 1980s and then updates them. Several chapters also introduce statistical methods and procedures to allow readers to analyze behavioral data.
This book is a useful resource for communication scholars and graduate students to guide their study of communication behavior.
Many times, the several categories of a multinomial response measure are not ordered, but rather are nominal in their level of measurement. There is no ordinal hierarchy for grouping the categories and, hence, creating a cumulative logit is neither possible nor meaningful. In such situations, one selects a single category to be the reference category whose probability is compared to the probabilities of all of the other J – 1 categories by calculating their odds. When the natural log of these odds is taken, one has created a baseline-category logit, frequently called the generalized logit.
When an investigator has no substantive or theoretical rationale for selecting a particular category as the baseline, the common practice is to select the last category. In contrast, when one has a substantive rationale (e.g., a particular category is the central focus of the analysis), then it is both acceptable and desirable to select that category as the baseline or reference.
Assume that four categories of behavior were observed and recorded from two theoretically distinct groups of people. Designating the fourth category as the baseline, we have:
where i is the index for the two groups.