FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2912505407> ?p ?o ?g. }

• W2912505407 endingPage "395" @default.
• W2912505407 startingPage "395" @default.
• W2912505407 abstract "In their retrospective case series, Parmar et al1Parmar D.N. Awwad S.T. Petroll W.M. et al.Tandem scanning confocal corneal microscopy in the diagnosis of suspected Acanthamoeba keratitis.Ophthalmology. 2006; 133: 538-547Abstract Full Text Full Text PDF Scopus (119) Google Scholar confirm the efficacy of corneal confocal microscopy in making the often difficult diagnosis of Acanthamoeba keratitis. However, their conclusions need to be reinterpreted regarding the study’s 10 separate t tests (three each in Figs 1, 2, and 3 and one in Fig 4). A single t test with a cutoff α level of 0.05 carries with it a 5% (1/20) chance of demonstrating a statistically significant difference when, in fact, none exists (i.e., the demonstrated difference is due to coincidence, resulting in a type I error). When multiple t tests are employed in a single study, the overall chance of type I error increases if the α level remains at 0.05 (to 40% when 10 separate comparisons are made, even if all 10 null hypotheses are true).2Motulsky H. Intuitive Biostatistics. Oxford University Press, New York1995: 118-126Google Scholar In this situation, the Bonferroni adjustment is an easy way to adjust the α cutoff level appropriately—the method is simply to divide the original α level by the number of comparisons.3Shaffer J.P. Multiplicity, directional (type III) errors, and the null hypothesis.Psychol Methods. 2002; 7: 356-369Crossref PubMed Scopus (37) Google Scholar For the 10 t-test comparisons in the study of Parmer et al, to reduce the possibility of making a type I error to the presumably intended 5% the α level needs to be lowered to 0.005 (0.05/10).2Motulsky H. Intuitive Biostatistics. Oxford University Press, New York1995: 118-126Google Scholar, 3Shaffer J.P. Multiplicity, directional (type III) errors, and the null hypothesis.Psychol Methods. 2002; 7: 356-369Crossref PubMed Scopus (37) Google Scholar With this new adjusted cutoff level, the change in mean logarithm of the minimum angle of resolution (logMAR) visual acuity (VA) (Fig 2 in the article) is not statistically significant between any of the initial VA groups—not between groups 1 and 2 (P≤0.045), between groups 2 and 3 (P≤0.006), or between groups 1 and 3 (P≤0.041). The authors’ conclusions regarding statistically significant observed differences for the mean final logMAR VA among the 3 initial VA groups (Fig 1 in the article) remain valid, as the reported P values are below the properly adjusted α level of 0.005 (P≤0.002 between groups 1 and 2; P≤0.001 between groups 1 and 3). In general, in any study that employs multiple comparisons all statistical analyses should be planned before the data are collected (not decided upon after perusal of the data), and all planned analyses should be completed and reported (not only the “significant” ones). Studies that do not follow these rules cannot be interpreted properly because the reader cannot evaluate the likelihood of coincidence being responsible for results reported as statistically significant. I congratulate Parmar et al for following these rules so that proper statistical interpretation of their well-done study is possible. Author replyOphthalmologyVol. 114Issue 2PreviewWe thank our colleagues for their cogent comments, and find that we agree with their central point that, in several instances, the data analysis is better served by applying a reduced α level correction. In general, we also agree that studies requiring multiple comparison analysis are better served by prospective planning; however, Acanthamoeba keratitis is an uncommon condition, and the series reported required a decade to accumulate sufficient cases with complete data. Full-Text PDF" @default.
• W2912505407 created "2019-02-21" @default.
• W2912505407 creator A5001256348 @default.
• W2912505407 date "2007-02-01" @default.
• W2912505407 modified "2023-09-25" @default.
• W2912505407 title "Acanthamoeba Keratitis" @default.
• W2912505407 cites W2056445260 @default.
• W2912505407 cites W2104082422 @default.
• W2912505407 doi "https://doi.org/10.1016/j.ophtha.2006.08.023" @default.
• W2912505407 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/17270694" @default.
• W2912505407 hasPublicationYear "2007" @default.
• W2912505407 type Work @default.
• W2912505407 sameAs 2912505407 @default.
• W2912505407 citedByCount "4" @default.
• W2912505407 countsByYear W29125054072013 @default.
• W2912505407 countsByYear W29125054072021 @default.
• W2912505407 crossrefType "journal-article" @default.
• W2912505407 hasAuthorship W2912505407A5001256348 @default.
• W2912505407 hasConcept C105795698 @default.
• W2912505407 hasConcept C114614502 @default.
• W2912505407 hasConcept C118487528 @default.
• W2912505407 hasConcept C121332964 @default.
• W2912505407 hasConcept C191988596 @default.
• W2912505407 hasConcept C2778217198 @default.
• W2912505407 hasConcept C2778560704 @default.
• W2912505407 hasConcept C2778868110 @default.
• W2912505407 hasConcept C2780576314 @default.
• W2912505407 hasConcept C33923547 @default.
• W2912505407 hasConcept C62520636 @default.
• W2912505407 hasConcept C71924100 @default.
• W2912505407 hasConcept C86803240 @default.
• W2912505407 hasConcept C89423630 @default.
• W2912505407 hasConceptScore W2912505407C105795698 @default.
• W2912505407 hasConceptScore W2912505407C114614502 @default.
• W2912505407 hasConceptScore W2912505407C118487528 @default.
• W2912505407 hasConceptScore W2912505407C121332964 @default.
• W2912505407 hasConceptScore W2912505407C191988596 @default.
• W2912505407 hasConceptScore W2912505407C2778217198 @default.
• W2912505407 hasConceptScore W2912505407C2778560704 @default.
• W2912505407 hasConceptScore W2912505407C2778868110 @default.
• W2912505407 hasConceptScore W2912505407C2780576314 @default.
• W2912505407 hasConceptScore W2912505407C33923547 @default.
• W2912505407 hasConceptScore W2912505407C62520636 @default.
• W2912505407 hasConceptScore W2912505407C71924100 @default.
• W2912505407 hasConceptScore W2912505407C86803240 @default.
• W2912505407 hasConceptScore W2912505407C89423630 @default.
• W2912505407 hasIssue "2" @default.
• W2912505407 hasLocation W29125054071 @default.
• W2912505407 hasLocation W29125054072 @default.
• W2912505407 hasOpenAccess W2912505407 @default.
• W2912505407 hasPrimaryLocation W29125054071 @default.
• W2912505407 hasRelatedWork W1972972501 @default.
• W2912505407 hasRelatedWork W2006924288 @default.
• W2912505407 hasRelatedWork W2026853921 @default.
• W2912505407 hasRelatedWork W2088824920 @default.
• W2912505407 hasRelatedWork W2805356426 @default.
• W2912505407 hasRelatedWork W3120204657 @default.
• W2912505407 hasRelatedWork W4286498151 @default.
• W2912505407 hasRelatedWork W4313517058 @default.
• W2912505407 hasRelatedWork W2335909519 @default.
• W2912505407 hasRelatedWork W2415699495 @default.
• W2912505407 hasVolume "114" @default.
• W2912505407 isParatext "false" @default.
• W2912505407 isRetracted "false" @default.
• W2912505407 magId "2912505407" @default.
• W2912505407 workType "article" @default.