Ask how a developmental level was scored before you act on it

Before you act on a developmental score, ask what descriptions it was matched against and how consistently scorers agree on it.

Por qué funciona

In the human scoring system that the Developmental Testing Service validity report describes, analysts scored responses against developmental-level descriptions that the developer kept revising. The report gives inter-rater agreement, meaning how often two scorers give the same rating, including 85 percent or more within a quarter of a level as of 2007. The CLAS page says the computerized system implemented on February 7th, 2026 is its “final standardized version,” which “has fully replaced the human version.” This page infers that the agreement figures describe the earlier system, and the sources read publish no reliability figures for the computerized one. Asking how a score was produced still helps you judge how much weight it can bear. These are the developer’s own summaries, not an independent audit.

Cómo hacerlo

  1. Ask whether the score came from the computerized system used from February 7th, 2026, or from the human scoring before it. The CLAS page says the change moved what were levels 11 and 12 into 11–13, and attributes most score changes to new research on levels 12 and 13.
  2. Ask for the level descriptions the score was matched to, or the public version of them.
  3. For human-scored results, ask whether a second person scored the same response, and how often the two agreed. For computerized results, ask what reliability figures the developer has published.
  4. Treat the level as a rough guide until those answers are clear.

Evidencia

The Developmental Testing Service validity report gives inter-rater agreement figures for the human scoring system, undated. The CLAS page says the computerized system implemented on February 7th, 2026 is its “final standardized version,” and says CLAS “generates highly reliable” growth records without publishing reliability figures. Dawson and Stein (2008) designed a scoring rubric for teachers and hypothesized that teachers could use it with minimal instruction; the abstract states that as a hypothesis. Dawson and Wilson (2004) is cited only for its title, a computerized scoring system for developmental assessments; its text was not read. (anecdotal)

The agreement figures describe the human scoring system and are undated. No reliability figures for the computerized system were found in the sources read. The search behind this page covered only Crossref, so no independent reliability study of the Lectical Assessment System was located by it.

Fuentes

Error común

Reading a single level as a fixed stage of a person, rather than as a reading of one response to one dilemma.

Practica esto con IX Coach

Empieza con IX Coach

Más prácticas para The Lectical Scale and the Virtuous Cycle of Learning

Conceptos relacionados