what is incidence of disease?
number of new cases of disease/number of people at risk
what is sensitivity in a test?
ability of the test to detect every case of the disease
what is specificity in a test?
ability of the test to exclude healthy non diseased people
what is PPV vs NPV?
PPV is probability of having the disease if the test if positive
NPV probability of being disease free if test if negative
what is selection bias?
issue with selection process of sample
what is measurement bias?
data gathering process can distort information
what is confounding bias?
characteristic associated with factor being studied that cannot be separated from the factor being studied
what is recall bias?
subjects cannot recall events
what is experimenter expectancy bias?
what the investigator expects is communicated and alters the info
what is late look bias?
info collected late so most severe cases are less likely to be found e.g due to death
what is the N in a case report?
N=1
how are results of cross sectional studies analysied?
Chi square
what is a case control study? how is data presented?
diseased compared with disease free
results expressed via an odds ratio
how are results of cohort studies expressed?
relative vs absolute risks
what are experimental study designs?
RCT
double blind RCT
crossover study
how to calculate range?
subtracts smallest from largest
what does a low SD mean?
data points are close to the average
what is SEM?
standard error of the mean
how well the mean estimates the mean of the population
what is a confidence interval?
provides a measure of uncertainty around an estimate based on a sample
what is a p value?
probability of obtaining a test statistic as extreme as the one observed
what does it mean if CI does not include a range with 1.0 in it?
significant difference in risk
standard p value?
0.05 (5%)
what does a meta analysis do?
synthesis of results of different studies to obtain a single p value
what is the best study for assessing an intervention?
RCT