GZW3024 || Statistics & Methodology || 2018
Inhoudsopgave
Frequent associations....................................................................................................................................3
Measurement levels......................................................................................................................................3
Continuous:...............................................................................................................................................3
Categorical:...............................................................................................................................................3
Measure assocations.....................................................................................................................................3
Designs.........................................................................................................................................................3
Measurement instruments.............................................................................................................................4
Validity (one measurement)......................................................................................................................5
Reliability.................................................................................................................................................5
Responsiveness (more measurements)......................................................................................................6
Measurement purposes..............................................................................................................................6
No measurement properties......................................................................................................................6
External validity...........................................................................................................................................7
Threats of the external validity.................................................................................................................7
Solutions...................................................................................................................................................7
Internal validity & types bias........................................................................................................................7
Quantitative bias.......................................................................................................................................8
Qualitative bias.........................................................................................................................................8
Information bias........................................................................................................................................8
Selection bias............................................................................................................................................9
Confounding.............................................................................................................................................9
Solutions...................................................................................................................................................9
Effectmodification & interaction..................................................................................................................9
Research - set up........................................................................................................................................10
Types of causality.......................................................................................................................................10
Viewpoints on Causation – Bradford Hill...............................................................................................11
General rules..............................................................................................................................................11
Linear regression........................................................................................................................................12
Compare 2 groups...................................................................................................................................12
Multiple linear regression.......................................................................................................................12
Polytomous predictors............................................................................................................................12
Logistic regression.....................................................................................................................................13
One binary X...........................................................................................................................................13
1
, Two binary X’s.......................................................................................................................................13
One continuous X...................................................................................................................................13
One continuous X and one binary X.......................................................................................................13
Test for effect..........................................................................................................................................14
Marginal models.........................................................................................................................................14
Unstructured...........................................................................................................................................14
Heterogeneous........................................................................................................................................14
Homogeneous.........................................................................................................................................15
Holm procedure......................................................................................................................................15
Techniques in validity................................................................................................................................15
Correction for attenuation.......................................................................................................................15
Intraclass correlation...............................................................................................................................16
Techniques in reliability.............................................................................................................................16
Spearman-Brown = overall formula of explained techniques.................................................................16
Test-retest...............................................................................................................................................16
Parallel-test reliability.............................................................................................................................16
Split-half reliability.................................................................................................................................16
Cronbach’s alpha....................................................................................................................................16
Standardized .......................................................................................................................................17
Methods to determine …............................................................................................................................17
Sample size (N).......................................................................................................................................17
Effect size...............................................................................................................................................18
Power......................................................................................................................................................18
Type I error.............................................................................................................................................18
Additional...................................................................................................................................................18
Missing values........................................................................................................................................18
Response tendencies...............................................................................................................................18
Methodology subjects GZW3024
2
, Frequent associations
Prevalence
- Point prevalence: part of population that is diseased at a
certain point in time
- Period prevalence: part of population that is diseased
within a time period
- Life-time prevalence: part of population that was diseased
during life
Incidence relates to the number of new cases
- Cumulative incidence (CI)
- Incidene Density (ID) / Incidence Rate (IR)
Measurement levels
Continuous:
Nominal: 2 or more categories of types (type house flat/appartement/villa)
Ordinal: 2 or more categories which can be ranked (type school HBO/WO)
Dichotomonous = binary: 2 categories or levels (yes/no)
Polychroom: more than 2 categories or levels (bloodtype A/B/AB/O)
Categorical:
Interval: numeric (Fahrenheit)
Ratio: same as interval but with zero-point (degrees)
Measure assocations
Risk Ratio (RR): (a/(a+b)) / (c/(c+d))
- If the risk ratio is 1 (or close to 1), it suggests no difference or little difference in risk (incidence in
each group is the same) = not significant
- A risk ratio > 1 suggests an increased risk of that outcome in the exposed group.
- A risk ratio < 1 suggests a reduced risk in the exposed group.
Attributive (Proportion) Ratio (AR/AP): RR – 1 / RR
Used when you want to calculate the proportion of disease in the total population that can be attributed to
the population.
Attributive (Proportion) Ratio among the exposed (APe): Exposed – not exposed / exposed
Used when you want to calculate the proportion of disease in the exposed group that can be attributed to
the exposure.
Odds Ratio (OR): (a/b) / (c/d) of (a*d) / (b*c)
Odds Ratio is an estimate of the Risk Ratio when for example the disease is uncommon. So mostly you
use it when it is about rare outcomes. OR mostly overestimates the RR.
Designs
3
Inhoudsopgave
Frequent associations....................................................................................................................................3
Measurement levels......................................................................................................................................3
Continuous:...............................................................................................................................................3
Categorical:...............................................................................................................................................3
Measure assocations.....................................................................................................................................3
Designs.........................................................................................................................................................3
Measurement instruments.............................................................................................................................4
Validity (one measurement)......................................................................................................................5
Reliability.................................................................................................................................................5
Responsiveness (more measurements)......................................................................................................6
Measurement purposes..............................................................................................................................6
No measurement properties......................................................................................................................6
External validity...........................................................................................................................................7
Threats of the external validity.................................................................................................................7
Solutions...................................................................................................................................................7
Internal validity & types bias........................................................................................................................7
Quantitative bias.......................................................................................................................................8
Qualitative bias.........................................................................................................................................8
Information bias........................................................................................................................................8
Selection bias............................................................................................................................................9
Confounding.............................................................................................................................................9
Solutions...................................................................................................................................................9
Effectmodification & interaction..................................................................................................................9
Research - set up........................................................................................................................................10
Types of causality.......................................................................................................................................10
Viewpoints on Causation – Bradford Hill...............................................................................................11
General rules..............................................................................................................................................11
Linear regression........................................................................................................................................12
Compare 2 groups...................................................................................................................................12
Multiple linear regression.......................................................................................................................12
Polytomous predictors............................................................................................................................12
Logistic regression.....................................................................................................................................13
One binary X...........................................................................................................................................13
1
, Two binary X’s.......................................................................................................................................13
One continuous X...................................................................................................................................13
One continuous X and one binary X.......................................................................................................13
Test for effect..........................................................................................................................................14
Marginal models.........................................................................................................................................14
Unstructured...........................................................................................................................................14
Heterogeneous........................................................................................................................................14
Homogeneous.........................................................................................................................................15
Holm procedure......................................................................................................................................15
Techniques in validity................................................................................................................................15
Correction for attenuation.......................................................................................................................15
Intraclass correlation...............................................................................................................................16
Techniques in reliability.............................................................................................................................16
Spearman-Brown = overall formula of explained techniques.................................................................16
Test-retest...............................................................................................................................................16
Parallel-test reliability.............................................................................................................................16
Split-half reliability.................................................................................................................................16
Cronbach’s alpha....................................................................................................................................16
Standardized .......................................................................................................................................17
Methods to determine …............................................................................................................................17
Sample size (N).......................................................................................................................................17
Effect size...............................................................................................................................................18
Power......................................................................................................................................................18
Type I error.............................................................................................................................................18
Additional...................................................................................................................................................18
Missing values........................................................................................................................................18
Response tendencies...............................................................................................................................18
Methodology subjects GZW3024
2
, Frequent associations
Prevalence
- Point prevalence: part of population that is diseased at a
certain point in time
- Period prevalence: part of population that is diseased
within a time period
- Life-time prevalence: part of population that was diseased
during life
Incidence relates to the number of new cases
- Cumulative incidence (CI)
- Incidene Density (ID) / Incidence Rate (IR)
Measurement levels
Continuous:
Nominal: 2 or more categories of types (type house flat/appartement/villa)
Ordinal: 2 or more categories which can be ranked (type school HBO/WO)
Dichotomonous = binary: 2 categories or levels (yes/no)
Polychroom: more than 2 categories or levels (bloodtype A/B/AB/O)
Categorical:
Interval: numeric (Fahrenheit)
Ratio: same as interval but with zero-point (degrees)
Measure assocations
Risk Ratio (RR): (a/(a+b)) / (c/(c+d))
- If the risk ratio is 1 (or close to 1), it suggests no difference or little difference in risk (incidence in
each group is the same) = not significant
- A risk ratio > 1 suggests an increased risk of that outcome in the exposed group.
- A risk ratio < 1 suggests a reduced risk in the exposed group.
Attributive (Proportion) Ratio (AR/AP): RR – 1 / RR
Used when you want to calculate the proportion of disease in the total population that can be attributed to
the population.
Attributive (Proportion) Ratio among the exposed (APe): Exposed – not exposed / exposed
Used when you want to calculate the proportion of disease in the exposed group that can be attributed to
the exposure.
Odds Ratio (OR): (a/b) / (c/d) of (a*d) / (b*c)
Odds Ratio is an estimate of the Risk Ratio when for example the disease is uncommon. So mostly you
use it when it is about rare outcomes. OR mostly overestimates the RR.
Designs
3
