Students' potential for learning contrasted under tutorial and group approaches to instruction

~30 students per arm (conventional / mastery / tutorial), grades 4, 5 and 8 — figure taken from Bloom's summary, not verified against the document

US grades 4, 5 and 8; 3-week units chosen for zero prior knowledge.

ACCESS FAILURE. One of the two University of Chicago dissertations behind Bloom's 2 sigma claim, an unpublished University of Chicago doctoral dissertation. No DOI is registered and nothing is indexed by Crossref, OpenAlex or Semantic Scholar; the title and details recorded here come from Bloom's own citation rather than from the document. Everything the archive knows about the design and the numbers is second-hand via Bloom (1984) and von Hippel's re-analysis, which is exactly why it cannot serve as independent evidence.

Not established by this archive. Reported by Bloom as one of the two sub-studies producing the tutoring-plus-mastery ~2 sigma advantage over conventional instruction on researcher-built unit tests. Kept rather than dropped because the most-quoted effect size in education rests on it and we could not read it.

Unassessable — the study was not read.

Unretrievable. Recorded so the provenance of the 2 sigma figure stays visible instead of being silently absorbed into Bloom's essay.