This study investigated questions related to the length of video segments in interactive video programs: (1) the preferred segment length if learners decide how much information they want presented before stopping to answer question; (2) the relationship between segment length and direct recall of factual information when segment length is self-chosen and when segment length is fixed; (3) the relationship between segment length and delayed recall of factual information when segment length is self-chosen and when segment length is fixed; and (4) effects of fatigue on self-chosen segment length. Subjects were 235 freshmen at a university in the Netherlands. Five experimental conditions were used: variable, in which subjects determined the length of each segment; crossed, the same as variable with information elements presented in a different order; linear, which allowed no stopping; short-long, fixed segment length with short segments followed by long; and long-short, fixed segment length with long segments followed by short. Data on field dependence independence, verbal ability, imagery-based strategies for storing information, and retention were gathered by testing; logs of interactive video sessions provided data on relevant time intervals, starting and stopping points, and questions answered correctly. Wide variances in the results of data analyses led to the conclusion that the main message may be that quality communication can be accomplished in many forms, whereby different audiovisual formats and segment lengths may appear feasible as long as learners start their tasks with a realistic expectancy of the demand characteristics and be motivated to tune their mental effort accordingly. (8 figures, 17 references) (MES)