Here's a breakdown of how scoring generally works:
* Instructions: The test presents a series of increasingly complex instructions involving manipulating tokens (usually colored shapes). These instructions test various aspects of language comprehension, including:
* Object identification: "Point to the red circle."
* Attribute identification: "Point to the large square."
* Relationship identification: "Point to the shape that is smaller than the blue triangle."
* Sequential commands: "Put the yellow triangle on top of the green square, then place the red circle next to them."
* Response Scoring: Each instruction is typically scored as either correct or incorrect. A correct response involves the individual accurately following the instructions. An incorrect response includes any failure to completely and correctly follow the instructions, even if parts of the instruction were followed. Partial credit is rarely given.
* Total Score: The total score is the sum of correctly executed instructions. This raw score is often then converted to a standardized score (e.g., percentile rank, scaled score) based on normative data provided in the test manual specific to the version used, age and other relevant demographic information.
* Qualitative Data: Beyond the quantitative score, the examiner should also note qualitative observations. This includes things like:
* Errors: What types of errors are being made? (e.g., difficulty with spatial relationships, understanding attributes, following sequential commands)
* Strategies: What strategies does the individual use to solve the problems?
* Behavioral Observations: Was the individual attentive? Frustrated? Did they require repetition of instructions?
In short: You don't just count correct responses; you also analyze the *type* of errors to understand the nature of the language processing difficulties. To accurately score a Token Test, you absolutely *must* use the specific scoring manual that accompanies the particular version of the test you are administering. The manual will provide detailed instructions, scoring criteria, and normative data for interpreting the results.