Epi: Biostat Tests Flashcards
Logistic Regression
Nominal Data
Prediction/Association/Regression
Buzzwords: Outcome, prediction, odds ratio
Contingency Coefficient
Nominal
Correlation
Buzzwords: Correlation, Association, Relationship
Log - Rank
Nominal
The proportion with event/Survival
Buzzword: Event to occurrence, time to event, occurrence over time
McNemar
Nominal
2 Related Groups
Buzzwords: Frequencies, counts proportions
Related buzzwords: Before vs after, baseline vs end, pre- vs post-
Cochran
Nominal
3+ Related Groups
Buzzwords: Frequencies, counts proportions
Related Buzzwords: Before vs after, baseline vs end, pre- vs post-
Pearson-Chi Square
Nominal
2 Independent Groups
Buzzwords: Frequencies, counts proportions
Uses reliability when there are less than 5 people in a 2x2 box
Fisher’s Exact
Nominal
2 Independent Groups or 3+ Independent Gropus
Used when there is a group within the 2x2 table that is less than 5
Buzzwords: Frequencies, counts proportions
Chi-Squared test of independence
Nominal
3+ Independent groups
Buzzwords: Frequencies, counts proportions
Multinominal Logistic Regression
Ordinal
Prediction/Association/Regression
Buzzwords: Outcome, prediction, odds ratio
Spearman Correlation
Ordinal
Correlation
Buzzwords: Correlation, Association, Relationship
Cox Proportional Hazards
Ordinal
Proportion with event/Survival
Buzzwords: Event to occurrence, time to event, occurrence over time
Wilcoxon Signed-Rank
Ordinal
2 Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Friedman
Ordinal
3+ Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Mann-Whitney
Ordinal
2 Independent Groups
Buzzwords: Frequencies, counts proportions
Kruskal-Wallis
Ordinal
3+ Independent Groups
Buzzwords: Frequencies, counts proportions
Linear Regression
Interval/Ratio
Prediction or Association/Regression
Buzzwords: Outcome, Prediction, Odds Ratio
Pearson Correlation
Interval/Ratio
Correlation
Buzzwords: Correlation, Association/Relationshiop
Kaplan-Meier
Interval/Ratio
Proportion with event/Survival
Buzzwords: Event to occurrence, time to event, occurrence over time
Paired t-test
Interval/Ratio
2 Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Repeated Measures ANOVA
Repeated Measures AN(C)OVA
Interval/Ratio
3+ Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Student t-test
Interval/Ratio
2 groups independent
Buzzwords: Frequencies, counts proportions
ANOVA
AN(C)OVA
Interval/Ratio
3+ Independent Groups
Frequencies, counts proportions
Nominal Data
Prediction/Association/Regression
Buzzwords: Outcome, prediction, odds ratio
Logistic Regression
Nominal
Correlation
Buzzwords: Correlation, Association, Relationship
Contingency Coefficient
Nominal
The proportion with event/Survival
Buzzword: Event to occurrence, time to evet, occurrence over time
Log - Rank
Nominal
2 Related Groups
Buzzwords: Frequencies, counts proportions
Related buzzwords: Before vs after, baseline vs end, pre- vs post-
McNemar
Nominal
3+ Related Groups
Buzzwords: Frequencies, counts proportions
Related Buzzwords: Before vs after, baseline vs end, pre- vs post-
Cochran
Nominal
2 Independent Groups
Buzzwords: Frequencies, counts proportions
Loses reliability when there are less than 5 people in a 2x2 box
Pearson-Chi Square
Nominal
2 Independent Groups or 3+ Independent Gropus
Used when there is a group within the 2x2 table that is less than 5
Buzzwords: Frequencies, counts proportions
Fisher’s Exact
Nominal
3+ Independent groups
Buzzwords: Frequencies, counts proportions
Chi-Squared test of independence
Ordinal
Prediction/Association/Regression
Buzzwords: Outcome, prediction, odds ratio
Multinominal Logistic Regression
Ordinal
Correlation
Buzzwords: Correlation, Association, Relationship
Spearman Correlation
Ordinal
Proportion with event/Survival
Buzzwords: Event to occurrence, time to event, occurrence over time
Cox Proportional Hazards
Ordinal
2 Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Wilcoxon Signed-Rank
Ordinal
3+ Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Friedman
Ordinal
2 Independent Groups
Buzzwords: Frequencies, counts proportions
Mann-Whitney
Ordinal
3+ Independent Groups
Buzzwords: Frequencies, counts proportions
Kruskal-Wallis
Interval/Ratio
Prediction or Association/Regression
Buzzwords: Outcome, Prediction, Odds Ratio
Linear Regression
Interval/Ratio
Correlation
Buzzwords: Correlation, Association/Relationshiop
Pearson Correlation
Interval/Ratio
Proportion with event/Survival
Buzzwords: Event to occurrence, time to event, occurrence over time
Kaplan-Meier
Interval/Ratio
2 Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Paired t-test
Interval/Ratio
3+ Related Groups
Buzzwords: Frequencies, counts proportions
Related: Before vs after, baseline vs end, pre- vs post-
Repeated Measures ANOVA
Repeated Measures AN(C)OVA
Interval/Ratio
2 groups independent
Buzzwords: Frequencies, counts proportions
Student t-test
Interval/Ratio
3+ Independent Groups
Frequencies, counts proportions
ANOVA
AN(C)OVA
CONSORT
- Randomized interventional studies
- Clinical Trials
- Consolidated Standards of reporting trials
PIRSMA
Systematic Reviews and Meta-Analyses (of multiple Randomized Trials)
Preferred Reporting Items for Systematic Reviews and Meta-Analyses
STROBE
Observational Studies
Strengthening the reporting of observational studies in epidemiology
TREND
Non-Randomized Designs of behavioral and public health interventions
Transparent Reporting of Evaluations with Non-randomized Designs
REMARK
Tumor Marker prognostic Studies
Reporting Recommendations for Tumor Marker Prognostic Studies
GRIPS
Genetic Risk Prediction Studies
STARD
Diagnostic Studies
Standards of the Reporting of Diagnostic Accuracy Studies
QUADAS-2
Systematic reviews of multiple diagnostic accuracy studies
Quality assessment of studies of Diagnostic Accuracy in Systematic Reviews
- Randomized interventional studies
- Clinical Trials
- Consolidated Standards of reporting trials
CONSORT
Systematic Reviews and Meta-Analyses (of multiple Randomized Trials)
Preferred Reporting Items for Systematic Reviews and Meta-Analyses
PIRSMA
Observational Studies
Strengthening the reporting of observational studies in epidemiology
STROBE
Non-Randomized Designs of behavioral and public health interventions
Transparent Reporting of Evaluations with Non-randomized Designs
TREND
Tumor Marker prognostic Studies
Reporting Recommendations for Tumor Marker Prognostic Studies
REMARK
Genetic Risk Prediction Studies
GRIPS
Diagnostic Studies
Standards of the Reporting of Diagnostic Accuracy Studies
STARD
Systematic reviews of multiple diagnostic accuracy studies
Quality assessment of studies of Diagnostic Accuracy in Systematic Reviews
QUADAS-2
Bonferroni Correction
Post-Hoc
Adjusts the p-Value
Can be used by any test: Only test for Nominal 3+ Groups
Student-Newman Keul
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
All groups must be equal size
Dunnett Test
Post - Hoc
3+ Groups
Compares all pairwise comparisons against a single control
All groups must be equal size
Dunn Test
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
Useful when all groups are not of equal size
Tukey Test
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
All groups must be equal
Slightly more conservative than Student - Newman - Keul
Scheffe Test
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
All groups must be equal
Less affected by violations in normality and homogeneity of variances: Most Conservative
Post-Hoc
Adjusts the p-Value
Can be used by any test: Only test for Nominal 3+ Groups
Bonferroni Correction
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
All groups must be equal size
Student-Newman Keul
Post - Hoc
3+ Groups
Compares all pairwise comparisons against a single control
All groups must be equal size
Dunnett Test
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
Useful when all groups are not of equal size
Dunn Test
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
All groups must be equal
Slightly more conservative than Student - Newman - Keul
Tukey Test
Post - Hoc
3+ Groups
Compares all pairwise comparisons possible
All groups must be equal
Less affected by violations in normality and homogeneity of variances: Most Conservative
Scheffe Test