ANOVAs + Stats Flashcards
One-way anova
used to determine whether there are any statistically significant differences between the means of three or more independent groups
Repeated measures ANOVA
used to determined if there are significant differences within the same group over time
MANOVA
ANOVA with several dependent variables
ANCOVA
used when there may be a confounding or interacting variable in your ANOVA
Covariate
another variable that may be affecting resutls
Two-Way ANOVA
studying two or more independent variables at the same time, each with at least two levels/time-points
Mixed Effects ANOVA
combination of a one-way ANOVA and a repeated measures ANOVA
looking at both a between-subjects and within subjects effect
Main Effect
looking just at the effect of 1 independent variable on 1 dependent variable
Interaction effect
looking at whether the effect of one variable depends on the effect of another variable
False Positive
Type 1 Error
Reporting an effect when there is actually no effect
Chance of committing a Type 1 error = alpha
reducing type 1 error means reducing alpha level
False Negatives
Type 2 Error
reporting no effect when there is an effect
chance of committing type 2 error = beta
reducing type 2 error = increasing sample size
Power
equivalent to 1 - Type 2 error
increasing sample size increases the power of your study
.8 power is very good
Power analysis
generally performed before the study
uses sample means and SDs to estimate sample size
more power is needed to detect small differences, more participants increases power
Effect Size
quantitative measure of the magnitude of the experimental effect
a statistically significant result can have a small effect, or vice versa
Minimal Clinically Important Difference
smallest change in a treatment outcome that a patient would identify as important
subjective measurement
Reliability
extend to which a measured value can be obtained consistently during repeated assessment
Validity
the extent to which a test measures what it is intended to measure, center of bullseye
Sources of Measurement error
Individual
Measurement instrument
characteristic being measured is variable
Test-Retest Reliability
consistency of results when you repeat the same test on teh same sample at a different point in time
Intra-Rater Reliability
consistency of the data recorded by one rater over several trials
Inter-rater reliability
consistency used to evaluate the extent to which different judges agree in their assessment decisions
Minimal Detectable Change
amount of change in a variable that must be achieved before we are confident that the change is not simply due to measurement error
Number needed to treat
metric used to calculate the effectiveness of an intervention. Smaller for longer duration studies
prevention-focused: # of pts that would need to be treated to prevent 1 adverse outcome
improvement focused: $ of pts that would need to be treated to achieve 1 beneficial outcome
if NNT = 1, every pt benefits from tx
Number needed to harm
used to calculate the chance of harm for an intervention
larger NNH means pt is less likely experience an adverse event