Ocelot STATISTICAL-ANALYSIS-AND-RESULTS Class Documentation

Class STATISTICAL-ANALYSIS-AND-RESULTS

Subclasses:

Template Slots

Slot name Documentation Type Allowed Values/Classes Cardinality Default

justification-for-one-tailed-test "If the statistical test is one-tailed, it's use must be explained." STRING 0:1

pairwise? "Whether pairwise comparisons of outcomes were analyzed for trials with three or more arms (e.g., in Arm A vs. B, in B vs. C, and in A vs. C)." SYMBOL Yes No 0:1

program-name "Name of the computer program used to run this analysis, e.g., STATA, SAS, SPSS, Minitab." STRING 0:1

number-of-tails "Two-tailed tests consider and test for the possibility of the effect being beneficial or harmful. One-tailed tests consider only one possibility. The use of one-tailed tests must be justified." SYMBOL One-tailed Two-tailed 0:1

statistical-test-name "The name of the statistical test for analysis of this outcome. e.g., paired-t-test, Kruskal-Wallis test, Wilcoxon rank-sum, etc. These test names are partially modelled in UMLS (mostly in PSY97). The list below is taken from Lang, Secic 'How to Report Statistics in Medicine' p. 71." SYMBOL ANOVA ANOVA-Friedman ANOVA-one-way-Friedman ANOVA-repeated-measures Chi-square Cox Proportional Hazard F-test Kruskal-Wallis Log-rank Mantel-Haenszel Mann-Whitney-U McNemar's Paired-t Students-t Wilcoxon-rank-sum Wilcoxon-signed-rank Other 0:1

other-notes STRING

parametric? "Whether the test is a parametric or non-parametric test." SYMBOL parametric non-parametric 0:1

confidence-interval-percent SYMBOL 90% c.i. 95% c.i. 99% c.i. 0:1 "95% c.i."

justification-for-statistical-test-used "Justification for why this statistical test was used." STRING 0:1

censoring-approach "The statistical approach taken to dealing with censored observations." STRING 0:1

whether-test-assumptions-hold "Description of whether the assumptions necessary for the test (e.g., normal distribution), were seen in the actual data analyzed." STRING 0:1

justification-for-transformation "Justification for why the data is parameterized or transformed (e.g., log-transformed)." STRING 0:1

repeated-measures-adjustment SYMBOL Bonferroni Other None

results "Points to collection of classes, grouped by all comparisons at a particular timepoint (e.g., 6 months). For study outcomes, the value of this slot are instances of ALL-COMPARISONS-AT-TIME-X (one instance for each timepoint). For BASELINEs, the value is BASELINE-COMPARISON." Instance All-comparisons-at-time-x 1:*

summary-statistic-type "Type of summary statistic. In STATISTICAL-ANALYSIS-AND-RESULTS and REGRESSSION-ANALYSIS-AND-RESULTS, specifies the standard summary statistic used. In SUMMARY-DATAPOINT, other summary values (e.g., ARR in addition to OR) can also be provided." SYMBOL Odds Ratio Relative Risk Relative Risk Reduction Risk Ratio Hazard Ratio Rate Ratio Difference Score 0:1

p-level-threshold "The p-level at which statistical 'significance' is declared." FLOAT 0:1

difference-score? "Whether analysis of intrasubject differences was performed, e.g., difference within a subject on repeated measures of a test." SYMBOL Yes No 0:1

difference-score-results Instance Single-time-x-comparison 0:1

protocol-changes Instance Protocol-change

roles "Tells which class and slot will take instances of this class depending on the role of the instance. The format is (role OWNER-CLASS |owner-slot|) e.g., if this class is DATE and the instance's role is recruitment-start-date, (recruitment-start-date RECRUITMENT |start-date|) tells us that this instance of date should be the value of the slot |start-date| in an instance of class RECRUITMENT." SYMBOL 0:1

Return to index

Generated on Mon Dec 15 11:40:40 2008

Template Slots
Slot name	Documentation	Type	Allowed Values/Classes	Cardinality	Default
justification-for-one-tailed-test	"If the statistical test is one-tailed, it's use must be explained."	STRING		0:1
pairwise?	"Whether pairwise comparisons of outcomes were analyzed for trials with three or more arms (e.g., in Arm A vs. B, in B vs. C, and in A vs. C)."	SYMBOL	Yes No	0:1
program-name	"Name of the computer program used to run this analysis, e.g., STATA, SAS, SPSS, Minitab."	STRING		0:1
number-of-tails	"Two-tailed tests consider and test for the possibility of the effect being beneficial or harmful. One-tailed tests consider only one possibility. The use of one-tailed tests must be justified."	SYMBOL	One-tailed Two-tailed	0:1
statistical-test-name	"The name of the statistical test for analysis of this outcome. e.g., paired-t-test, Kruskal-Wallis test, Wilcoxon rank-sum, etc. These test names are partially modelled in UMLS (mostly in PSY97). The list below is taken from Lang, Secic 'How to Report Statistics in Medicine' p. 71."	SYMBOL	ANOVA ANOVA-Friedman ANOVA-one-way-Friedman ANOVA-repeated-measures Chi-square Cox Proportional Hazard F-test Kruskal-Wallis Log-rank Mantel-Haenszel Mann-Whitney-U McNemar's Paired-t Students-t Wilcoxon-rank-sum Wilcoxon-signed-rank Other	0:1
other-notes		STRING
parametric?	"Whether the test is a parametric or non-parametric test."	SYMBOL	parametric non-parametric	0:1
confidence-interval-percent		SYMBOL	90% c.i. 95% c.i. 99% c.i.	0:1	"95% c.i."
justification-for-statistical-test-used	"Justification for why this statistical test was used."	STRING		0:1
censoring-approach	"The statistical approach taken to dealing with censored observations."	STRING		0:1
whether-test-assumptions-hold	"Description of whether the assumptions necessary for the test (e.g., normal distribution), were seen in the actual data analyzed."	STRING		0:1
justification-for-transformation	"Justification for why the data is parameterized or transformed (e.g., log-transformed)."	STRING		0:1
repeated-measures-adjustment		SYMBOL	Bonferroni Other None
results	"Points to collection of classes, grouped by all comparisons at a particular timepoint (e.g., 6 months). For study outcomes, the value of this slot are instances of ALL-COMPARISONS-AT-TIME-X (one instance for each timepoint). For BASELINEs, the value is BASELINE-COMPARISON."	Instance	All-comparisons-at-time-x	1:*
summary-statistic-type	"Type of summary statistic. In STATISTICAL-ANALYSIS-AND-RESULTS and REGRESSSION-ANALYSIS-AND-RESULTS, specifies the standard summary statistic used. In SUMMARY-DATAPOINT, other summary values (e.g., ARR in addition to OR) can also be provided."	SYMBOL	Odds Ratio Relative Risk Relative Risk Reduction Risk Ratio Hazard Ratio Rate Ratio Difference Score	0:1
p-level-threshold	"The p-level at which statistical 'significance' is declared."	FLOAT		0:1
difference-score?	"Whether analysis of intrasubject differences was performed, e.g., difference within a subject on repeated measures of a test."	SYMBOL	Yes No	0:1
difference-score-results		Instance	Single-time-x-comparison	0:1
protocol-changes		Instance	Protocol-change
roles	"Tells which class and slot will take instances of this class depending on the role of the instance. The format is (role OWNER-CLASS \|owner-slot\|) e.g., if this class is DATE and the instance's role is recruitment-start-date, (recruitment-start-date RECRUITMENT \|start-date\|) tells us that this instance of date should be the value of the slot \|start-date\| in an instance of class RECRUITMENT."	SYMBOL		0:1