Vocabulary Flashcards
odds ratio=1
equal chance
odds ratio<1
exposure protective
odds ratio >1
exposure increased risk
info bias
missclasification
recall bias
skewed data further from events
Response bias
question bias/ interview bias
- stereotyping
- halo
- nonverbal
halo bias
+ or - overshadow everything
nonverbal bias
ex. tattoos piercings
berkson’s bias
compare hospital patients to reg. population
hawthorn effect
change behavior when watched
length time bias
different diseases
slow growing cancers dx sooner
lead time bias
screened sooner seems like live longer
same disease
nominal data
castegories with no order
use mode only
ordinal
categories with order
use median or mode
interval
each step same size
mean median or mode
ratio
interval data with true zero
continuous data
age, height, definate numbers
categorical data
tall, short, old, young
20-30, 5’-5’6”
histogram
continuous
use if in order
bar graph
categorical data
no order
confidence interval
how precise estimate of true value is
standard deviation
variability of pop around the mean
no change in samploe size
standard error
variability of estimated mean
statistically significant
p=.05
Type 1 error
false positive
type 2 error
false negative
T- test
compares 2 groups
ANOVA
Analysis of variable
2 groups: b/w and within
