stats unit 6 Flashcards
Exploration
Informal, open ended examination of data for patterns
Disjoint (Mutually Exclusive)
can’t happen at the same time
Sample Space:
complete list of disjoint (mutually exclusive) outcomes
Outcome:
result of a single trial
Probability Distribution:
Gives all possible values resulting from a random process and gives the probability of each.
Law of Large Numbers:
In a random sampling, the larger the sampling the closer the Emprical probability come to the classical probability
Fundamental Principle of Counting (Tree Diagram or Two-Way Table):
if there are n1 outcomes for stage 1 and n2 outcomes for stage 2 then the possible number of outcomes for the 2 stages is n1n2
Or
one or the other or both
False positive
when a test result says that the person tested has the disease, but they actually don’t
False negative
when a test result says that the person tested does not have the disease, but they actually do
Sensitivity
the ability of a test to correctly identify people with a disease
Specificity
the ability of a test to correctly identify people without a disease
Independent Events
2 things can occur and they don’t affect each other
cases
subjects of statistical examination
variable
the characteristics of of the cases
summary statistic
condensing” data into a single number to help with data analysis
simulation
a procedure in which you set up a model of a chance process that copies a real situation.
Standard Deviation
The horizontal distance from the line of symmetry to the inflection point
skewed left/right
left: mean is pulled towards the left of the median
right: mean is pulled towards the right of the median
median
the value that divides data equally
quartiles
three numbers that divide data into quarters
interquartile range
distance between Q1 and Q3
percentile
value that separates the lowest percent
recentering
adding or subtract x to all values
rescaling
multiply each value by x; x>0
z-score
how many standard deviations away from mean
interpolation
predicting within a model
extrapolation
predicting outside model
sample
part of a population
census
entire population
units
individual members in a population
population
set of units you study
bias
tends to give samples where some characteristics of the population is under or overrepresented
correlation
measure of strength of a linear relationship
sampling frame
all possible units
convenience sampling
units chosen are easy to include
size bias
size causes misinterpretation
voluntary response bias
sample members self selected
simple random sample
all possible samples of a given fixed size are equally likely
stratified random sample
divide units into strata, take SRS of strata
cluster sample
break population into clusters, take SRS of clusters, obtain data from entire cluster
two staged sampling
break population into clusters, take SRS of clusters, take SRS from clusters
lurking variable
variable in the background that once identified could explain a pattern
experiment
establish cause comparing two or more treatments using a response
observation study
no treatment assigned, conditions are built in
blind
patients don’t know control and treatment group
double blind
patients and doctors don’t know control and treatment group
matched pair
only look at pairs, one with and one without treatment
blocking
selecting pairs with similar units to help control for lurking variables
