TOWARD A MULTIPLE-METHOD VIEW OF MOOD INDUCTION - THE APPROPRIATENESSOF A MODIFIED VELTEN MOOD INDUCTION TECHNIQUE AND THE PROBLEMS OF PROCEDURES WITH GROUP ASSIGNMENT TO CONDITIONS

Two methodological issues involving the Velten (1968) mood induction t echnique are examined. The first issue concerns threats to statistical conclusion validity that result from the use of some group mood induc tion techniques: In research on the effects of mood, groups of subject s are often assigned to mood conditions, whereas individuals are treat ed as the unit of analysis. It is argued that this procedure threatens the validity of inferences about mood effects, in particular by infla ting Type I error. The use of a group mood induction, with only one mo od manipulated in a particular group, also increases the likelihood of experimenter bias, local history effects, or both. This article illus trates an alternative method of method mood induction involving four m odifications to the Velten (1968) mood induction procedure. Using this revised procedure, subjects are randomly assigned to different mood c onditions, yet the data are collected simultaneously. Results indicate that this method is successful in inducing moods. This technique avoi ds the problems associated with assigning groups to conditions, but re tains the economy of group sessions. The second issue concerns the tem poral duration of the Velten procedure. We demonstrate that our modifi ed procedure, with the inclusion of an incubation period, leads to moo d change lasting up to 35 min. Further, our modified Velten procedure resulted in mood differences, even after various intervening tasks. Th e technique was discussed in terms of demand characteristics, experime nter bias, local history, and ethics.