Five studies of holistic writing assessment procedures examined interactive relationships of the participants, processes, and products of writing assessment episodes. The first study examined practices in designing writing test prompts. The second study investigated the effects of variation in the specification of audience in a writing test prompt upon holistic score. A second part of this study examined rater response to the anchor papers used in training three groups of raters for scoring. The third study investigated writing task and response variables and validated an analysis system for studying the relationships between assessment prompts, written products, and holistic scores. The second part of the third study summarized data collected on the performance elicited on eight prompts in a county-wide writing assessment. The fourth study investigated differences in expectations among participants in a writing test episode, and included an analysis of characteristics of a prompt which influence variations in interpretation. The fifth study investigated holistic assessment as a longitudinal measure of student growth, focussing on aspects of performance variation over time. A discourse typology was developed and applied in an analysis of performance variation on annual assessment topics over a period of five years. (Author/PN)