Using the Revised Bloom Taxonomy to Analyze Psychotherapeutic Games

Applying Bloom's Revised Taxonomy, with its four-level Knowledge dimension and six-level Cognitive Process dimension, to psychotherapeutic games proves too subjective to support objective comparison of game content. Haring and colleagues tested the framework on five games: Personal Investigator, Treasure Hunt, Ricky and the Spider, Moodbot, and SuperBetter. Their central aim was to determine whether the taxonomy could serve as a design and analysis tool for translating therapeutic goals into game content.

The authors first built consensus lists of Player Actions with Learning Objectives (PALOs), a verb-and-object structure derived from the taxonomy. This yielded 39 PALOs: 13 for Personal Investigator, 9 for SuperBetter, 7 for Ricky and the Spider, 6 for Treasure Hunt, and 4 for Moodbot. Four coders then independently placed each PALO in the Taxonomy Table. Agreement was strikingly poor. Only one PALO was scored identically by all coders, and 21 had no exact agreement. Krippendorff's alpha was .075 overall, .131 for knowledge level, and .174 for cognitive processing, far below the conventional .60 to .80 threshold. Disagreement was greatest on the knowledge dimension, especially in distinguishing conceptual, procedural, and metacognitive knowledge.

Removing any single coder did not improve reliability, which points to genuine and evenly distributed disagreement rather than one outlier or a methodological flaw. The authors therefore reject the taxonomy as a design checklist, a means of objectively comparing games, or a guide for psychologists selecting games.

Even so, the exercise has value. Formulating PALOs pushes designers to make their intentions explicit and gives people from different disciplines a shared vocabulary for structured discussion. The study is limited to five games and relies on written descriptions rather than actual gameplay, so its judgments reflect only a lower bound of required knowledge, and reliability would likely fall further with observed play.