Repository navigation
FLoRA rows to correct, found while cross-checking the Reversals page #175
Description
Comparing FLoRA (output/flora.csv at 18e889b) with the Replications & Reversals page turned up the FLoRA rows below. These are the problems with FLoRA itself; the page corrections are in forrtproject/forrtproject.github.io#1139.
Checked against the source and the CSV
| Row | Problem | Suggested fix | Evidence |
|---|---|---|---|
| REPL-001692 (Open Science Collaboration 2015, Schnall et al. 2008) | outcome_quote is Johnson et al. 2014's abstract, the same text as FLORA-000334 |
Use the RP:P report's own wording, or merge with FLORA-000334 if it is the same replication | Both rows carry the identical quote, "We did not find evidence that physical cleanliness reduced the severity of moral judgments…" |
| FLORA-002815 (Besman et al. 2013, cleanliness) | Coded successful on the student template's own claim |
uninformative |
Johnson et al. 2014 (10.1027/1864-9335/a000186): n = 60, 95% CI includes zero, "did not reach conventional levels of significance with a two-tailed test (p = .08)"; their small-telescopes test finds it "not smaller than d33% nor different from 0" |
| FLORA-002282 (O'Donnell et al. 2021, Shah et al. Study 6) | Quote says F(1, 203) = 1.69, p = .19, source "abstract" | The study report says "F(1, 205) = 1.68, p = .20", in the Results of https://osf.io/download/xz6bn/ | Study report; the PNAS abstract does not give this test |
| FLORA-002293 (O'Donnell et al. 2021, Chou et al. Study 3) | outcome_quote is generic text about the whole project ("Although most of our replications did not approach statistical significance…") |
Study 3's own result, "t(601) = 5.27, p < .001, Cohen's d = 0.43" (https://osf.io/download/g6kx3/), noting that p = .08 after covariates | Not yet checked against that report; Sol's reading |
| Many Smiles Collaboration 2022, 10.1038/s41562-022-01458-9 | Missing: no row has this DOI or title | Add as a replication of Strack et al. 1988, coded per task: mimicry and voluntary action positive, pen-in-mouth "less conclusive" | Abstract: n = 3,878 |
Reported by the GPT-6.1 Sol review, not yet checked by hand
- FLORA-002170 (Boyce et al. 2023) and FLORA-002814 (Lee 2013) seem to be the same student study of Schnall et al. (n = 90, F(1, 88) = .20); Boyce's dataset matches Lee's report. They may need deduplicating.
- FLORA-002283 and FLORA-002294 (Boyce et al. 2023) code student replications of other studies in the same papers: Shah et al. Studies 1b–1c, and Chou et al. Study 5, which manipulated control. Their outcomes look right, but the quote or source should name the study, so they are not read as tests of the study the page lists.
- FLORA-002816 (Arbesfeld et al. 2014): the "successful" code rests on the project's own label, and no inferential test has been found. It needs checking against the report.
Activity