FINAL LOOK at EBP Flashcards
what is level I evidence
prospective, RCT, high quality evidence/diagnostic testing
what is level II evidence
lower level diagnostic studies prospective or RTC (low power, weak diagnosis criteria, less than 80% follow up, no blinding or poor randomization)
what are level III evidence
case control studies or retrospective studies
what are level IV studies
case series
what are level V studies
expert opinions
the p value looks at
if the difference occurred by chance
what do you need to interpret a P value
an alpha value
if alpha is set at 0.05, then what P value must you have to have statistical significance
less than 0.05
the P value has to be ____ than alpha for statistical significance
lower than
type I error “?”
concluding there was statistical significance when there was not
claiming that there was an effectiveness when there was not
“backing a loser”
backing a loser represents a
Type I error
Type II error “?”
not finding a statistical difference when there really was a statistical significance,
saying something is not effective when it is
“missing a winner”
missing a winner represents what
Type II error
how can you reduce the chance to type II error
by inc the sample size and inc the number of subjects in the study
effect sizes look at…..
interventions
what are large, moderate, small and trivial effect sizes
0.8 and up: large
0.5-0.799: moderate
0.2-0.499: small
less than 0.2 is trivial
what is the difference between interrater and intrarater reliability
inter: between multiple rates (think internet connects multiple people). if multiple people can get the same result
intra: between the same person, same person able to get the same result all of the time.
what looks at reliability, what measure
Cohen’s Kappa (k)
what is the reliability scale from no reliability to perfect reliability
0: no reliability/no better than chance
<0.4: poor
0.4-0.6: fair
0.6-0.75: good
>0.75: excellent
1: perfect reliability
what does a (+) LR do
inc suspicion based on a positive test
what does a (-) LR do
decrease suspicion based on a negative test
what are values of (+) LR for large, moderate, small and no shift in probability for something
> 10 large shift
5-10: moderate shift
<5: small shift
1: no shift
what are values of (-) LR for large, moderate, small and no shift in probability away from something
<0.1: large shift
0.1-0.2: moderate shift
>0.2: small shift
1: no shift
