Exam 2 Flashcards
What is validity?
Truthfulness, meaningfulness, usefulness, and/or accuracy of study results
External vs Internal Validity
External Validity: generalizability of results
Internal Validity: controlled by the study design (blinding, instrumentation, attrition)
What is face validity?
Does a specific measure actually measure what it is designed to measure
What is content validity?
Does the measure represent all constructs of the measure (does it take all things into account)
What is concurrent validity?
Comparing your intervention to the Gold Standard
What is predictive validity?
Can it be used to predict a future score/outcome
What is construct validity?
How well the measure captures a defined entity (theoretical construct)
What is convergent validity?
Examines the degree to which the operationalization is similar to other operations that is should be similar to (one head start compared to others)
What is discriminant validity?
Examines the degree to which one thing differs from others (one head start compared to non-head starts)
How is validity typically measured?
Correlations, -1 to 1
What analyses are used for which data types?
Interval and Ratio (Continuous) - Pearson
Ordinal - Spearman Rank
Nominal (Dichotomous) - Phi
What is Reliability?
Consistency of a specific measure
Ability to produce consistent repeated measures of a test
What are the two components of reliability
True Component Error Component (variety of sources)
What type of data is required for Reliability Measures?
Continuous - Ratio or Interval
What are the breakdown scores for ICC (Reliability)?
Good: > .75
Moderate: .51 - .75
Poor: < .50
What type of data is agreement?
Categorical (Nominal)
Kappa statistic takes out the chance aspect
What are the Kappa score breakdowns?
Almost Perfect: .81 - 1.0 Substantial: .61 - .80 Moderate: .41 - .60 Fair: .21 - .40 Slight: .01 - .20 Poor (equal to chance): < 0
What is the Minimal Detectable Change?
Smallest amount of change an instrument can accurately measure
Changes must exceed MDC to be beyond measurement error
Does not provide clinical meaningfulness
What is the Minimal Clinically Important Difference?
Smallest difference that clinicians and patients would care about
Identify change in health status measure associated with improvement that is meaningful
Compares two measures (Pain: VAS, clinician-derived measure: ROM)
What is Ceiling Effect?
Instrument does not register a further increase in score for higher scoring individuals
What is Floor Effect?
Instrument does not register a further decrease in score for lower scoring individuals
What variables are Statistically significant?
p-values
Precision of estimation/confidence intervals
Type 1 and Type 2 errors
Power
What variables are Clinically significant?
Size of the difference
Does change exceed MCID
Effect Size measurements
Specificity, sensitivity, LR, NNT, RR, ARR
p-values
Risk of Type 1 error
Does not indicate importance or clinical relevance