Prelim #1 Flashcards
Random Sample
each member of the population has an equal and independent chance of being selected
Volunteer bias
volunteers for a study are likely to be different, on average, from the population
sample of convenience
a collection of individuals that happen to be available at the time
accuracy
an estimate is accurate (unbiased) if the average of estimates is centered on the true population value
precision
a measure of how far apart repeated estimates might be, directly related to sample size
experimental study
researcher randomly assigns individuals to treatment groups
observational
assignment of treatments is not made by researcher
categorical variables
describe membership in category/group. could be dichotomous (binary), ordinal (categories are ordered), or nominal (categories have no natural ordering)
numerical/quantitative variables
quantitative measurements that have magnitude on a numerical scale. either continuous (can be measured) or discrete (can be counted, individual units)
graphing: 1 categorical variable
frequency table, bar graph/bar plot
graphing 1 numerical variable
histogram
graphing 2 categorical variables
grouped bar graph, mosaic plot
graphing 2 numerical variables
scatterplot, line graph (time)
graphing 2 variables: one categorical and one numerical
strip chart, side-by-side box plot, multiple histograms
mean vs median
mean is the center of gravity, median is middle measurement… mean is more affected by outliers/extreme values than median
sample range
max-min…poor measure of distribution width…biased estimator of true range of the population
sample variance (s^2)
the average squared difference from the mean….(standard deviation)^2
standard (s)
positive square root of variance…it is related to the average distance between the mean and each observation…measure of the variability (or spread) of a distribution
what percent of data falls within 1 sd of mean (normal distribution)?
2/3 (66%)
what percent of data falls within 2 sd of mean (normal distribution)?
95%
interquartile range
3rd quartile-1st quartile
skew
measure of asymmetry…refers to pointy tail of a distribution
box plot…where do whiskers extend to?
largest and smallest non-extreme observation
standard error
the standard error of an estimator is the standard deviation of the estimator’s sampling distribution…measures precision
standard error formula
se=s/square root (n)
formula for rough confidence interval and assumptions
sample mean +/ 2SE
normally distributed population and sufficiently large sample size, random sample
does confidence interval contain sample mean or true population mean?
true population mean
standard errors (increase/decrease) with increasing sample size?
decrease (large samples yield more precise estimates)
will a 99% confidence interval be wider or narrower than a 95% confidence interval?
wider
will a 90% confidence interval be wider or narrower than a 95% confidence interval?
narrower
pseudo-replication
the error that occurs when samples are not independent, but are treated as though they are ex: multiple heart rates measured from the same person and treated as different samples
mutually exclusive
two events are mutually exclusive if they cannot both be true… Pr(A and B)=0
addition principle
if two events A and B are mutually exclusive, then Pr(A or B)=Pr(A)+Pr(B)
general addition principle
A and B don’t have to be mutually exclusive.
Pr(A or B)=Pr(A)+Pr(B)-Pr(A and B)
probability distributions: discrete
probability is measured by the height of the bar
probability distributions: continuous
probability is quantified by area under the curve
independence
two events are independent if the occurrence of one gives no information about whether the second will occur
multiplication principle
if two events A and B are independent, then Pr(A and B) =Pr(A) X Pr(B)
Conditional probability
The conditional probability of an event is the probability of that event occurring given that a condition is met
Positive Predictive Value
the probability that subjects with a positive screening test actually have the disease pr(d|+)
