1/61
Looks like no tags are added yet.
Population
: all possible value
Sample
: a portion of the population
Statistical inference
generalizing from a sample to a population with calculated degree of certainty
The purpose of hypothesis testing is to
aid an individual (clinician, researcher, or administrator) in decision-making
A hypothesis may be defined simply as (and examples)
a statement about one or more populations.
A hospital administrator may hypothesize that the average length of stay of patients admitted to the hospital is 5 days
a public health nurse may hypothesize that a particular educational program will result in improved communication between nurse and patient;
General steps for hypothesis testing (4)
1. State null hypothesis and alternate hypothesis 2. Compute test statistic 3. Interpret and make (statistical) decision 4. Final decision
Null vs alternitve
Null - means are the same (always the default hypothesis that states that the measurement of interest is zero)
alternative - means are different
A p value is
the probability that the computed value of a test statistic is at least as extreme as a specified value of the test statistic when the null hypothesis is true.
small value - reject the null
large - accept the null hypothesis
t -test
assumes the data is normal
compares means of a sample
one sample t test
for comparing a sample to a certain mean
ie. is the mean of this data = 100 or not
two sample t test
test the difference in mean from two samples
paired t test
for the same subjects may be measured before and after receiving some treatment
z test
discrete variable of interest measured from populations
comparing the proportions of a certain category for the populations
two sample z test
test the difference in two population proportions
F-test
test if the difference that will exist between the sample variances is indicative of a real difference in population variances
number of children born the pregnancy
numberic discrete
gender of child
categorical nominal
race of mother
categorical nominal
weight gained during pregnancy
numeric continous
0 or 1
cateogrical nominal
histrogram
visualiz single numberical varible
barplot
visual summary of catgorical or numberic variable
box plot
relationship between nuberical and categorical
scatterplots
visualize relationship between 2 numeric variables
mean
sum of values / number of values
median
middle value must be in numerical order
if even amount of number then take the average between the two
mode
value w/greatest freq
20th percentile represents
a value is higher than or equal to 20% of the other values in a data set, and lower than the remaining 80%
geometric mean
all obervations have to be positive
used for growth rates
take log of data, then find the mean
10^(mean(log))
variance formula and meaning for a sample
(value - mean)²/n-1
smaller variance means closer to the mean
mutally exlusive events
if A and B cannot occur simulataneoulsy
complementary event
is the event consisting of those simple events or sample points not in A
joint occurrence (intersection) of 2 events
proability that A and B occur at the same time
typically you can mulitply the proability of the first event times the second event
union of 2 events
event that occurs if either A or B or both occur in a single performance
red cards
diamond and heart
black cards
spade and club
addition theorem
A and B are mutually exclusive so the proability that either A or B occures is P(A) + P(B)
Are not mutually exclusive then P(A) + P(B) -P(A and B)
conditional probability =
p(A and B) divided by P(B)
A - find
B- “given”
numeric continous
any vaule, typically meausred
numeric discrete
countable (number or freq of)
caterogocial nominal
no order like major
caterogrical ordinal
ordered lik education level
u or n
u is for addition
n is for or
Normal distribution
for continuous variables
evaluate with Q-Q plots
if normal then the points will closely follow the line (deviations are bad)
z score
oberservation is the number of sd it falls above or below the mean
if 1.547 is z score the observations is 1.547 SDs away from the mean
then gives value of .83 meaning 83% of the data falls at or below that z-score
1- z score to get above that value
Sensitivity
Sensitivity of the test is conditional probability that the test can to find the true postive cases
Specificity
Sensitivity of the test is conditional probability that the test can to find the true negatives cases
positive predictive value
the proportion of positive test results that are true positives, showing the likelihood that a person with a positive test actually has the condition or disease
negative predictive value
the probability or proportion that a person with a negative test result truly does not have the disease, condition, or trait being tested for
Binomial distribution
discrete variables
use permentions or combination
use formula
Bernouli trail
A Bernoulli trial is a random experiment in which there are only two outcomes (success or failure)
Poisson Distribution
models the number of events occurring in a fixed interval of time or space if these events occur independently
use formula
STARD guidlines
Guidelines that you would use to ensure when putting together a report for the FDA tells you how to organize and what/how to report to them
template/checklist for everything needed for FDA report (use for mini project)
ROC curve
a plot of 1 – Specificity (x) versus Sensitivity (y).
illustrates the diagnostic ability of a test
more area under the curve = more reliable test
confidence interval
is a range or interval of values that we can be really confident contains the population parameter of interes
"We can be 95% confident that the true proportion of Americans addicted to the internet is between 0.0534 and 0.0606 (0.057 ± 0.0036)."
ANOVA Test
used test hypothesis related means from more than groups
we learn to test the null hypothesis that three or more means are equ
Dependent variables
represent the output or outcome whose variation is being studied or by the set of measured values that result from an experiment
Independent variables
represent inputs or causes
randomized block design
design in which the observations (say, subjects) to which the treatments are applied are subdivided into homogeneous groups called blocks, so that the number of observations (subjects) in a block is equal to the number
each treatment appears in every block
Two-way ANOVA test
to analyze data from a randomized block design, since an observation is categorized on the basis of two criteria
repeated measures design
which measurements of the same variable are made on each subject on two or more different occasions
responses over time
What does the Error() term mean
we want to look at how the functional score changes as a function of time