Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
74 Cards in this Set
- Front
- Back
In statistics what is a population
|
represents the entire group of indivudals in whom we are interested
|
|
B/c it is costly and labourly intensive to study the entire population therefore we collect data on a SAMPLE of indviduals who are
|
representative of this population we are taking a sample we have introduced sampling error
|
|
We are often interested in the value of a parameter in the population (i.e) What is a parameter
|
numerical
mean or proportion |
|
What is statistics
|
an estimate of parameters
|
|
What is a sampling distribution of the mean
|
take several separate smaples and each has its own mean, and plot over time--should get a normal distribution
|
|
What is a standard error the mean (SEM)
|
amount of variation in menas between our sample and other repated samples of a fixed size
|
|
SEM=
|
s.d./square root of n
n=sample size |
|
SEM is always <
|
s.d
|
|
SEM is an indicator of precision, the smaller the SEM the
|
indicates that are estime is more precise
|
|
What is the sample distribution fo the proportion
|
SE(p)= √ P(1-p)/n
|
|
ONce we have taken a sample from our popuation we obtain a point estimate (single data) to make an interval estimate b/c
|
more useful
|
|
Confidence intervals
|
range in which we are 95% sure that the tue value lies between
|
|
95% CI for the mean
|
average +- 1.96(SEM)
|
|
CI for the proportion
|
P +- (SEp)
|
|
We often gather sample data in order to assess how much evidence there is against a specific hypothetsis about the popuatlion this is AKA
|
hypothetisis testing
|
|
What are the 5 stages of hypothesis testing
|
1. Define the Null and alternative hypothesis
2. Collect relevant data 3. Calculate the value of the test statistic specific to Null hypthetsis 4. Compare the values of the test statistic to values from a known probability distruvbtion 5. Interpest P value |
|
What is the null hypothesis or Ho=
|
the is NO difference in the population
|
|
What is the alternative hypothesis
|
difference does exist in the population
|
|
After you collect data from a sample of indivuals then you
|
calculate the correct statistical test
|
|
After you calculate a test statistic, then you then you compare with a known
|
probability distribution
|
|
What is the P value is
|
the probability of obtaining our results if the null hypotheis is true
|
|
THe samller the p-value, the more
|
UNLIKELY the null is true
|
|
What does P <0.05 really mean
|
the observed result was less then 5% likely to have occur due to chance alone
|
|
If P <0.05, the we
|
REJECT the null hypothesis and conclude there is a real difference
|
|
We say we "REJECT the null hypothetisis NOT
|
accept or proven it
|
|
If P is greater or equal to 0.05 then we
|
FAIL to reject the null hypotheris--NO real difference was found
|
|
A small p-value DOES NOT mean that the alternative hypothesis is
|
TRUE
|
|
Does a P value tell us the result are valid
|
NO
|
|
Does the p value tell us the magnitude of effect
|
NO--how small the p value does not mean anything
|
|
Does the P value the probabilty the null hypothesis is true
|
NO
|
|
The p value only tells us
|
that the observered result was less than 5% to have occur due to chance alone
|
|
Why is the two-tailed test the best
|
NO assumptinos on the direction of effect
|
|
Are confidence intervals and hypothesis test closely linked
|
YES
|
|
What is a type 1 error
|
we reject the null hypothesis when it is tru
|
|
What is alpha
|
is the maximum proability of fasely rejecting the null hypothesis that awe are willing to tolerate
|
|
What is alpha=
|
0.05
|
|
What is a type II error
|
we DO NOT reject the null hypothesis when it is false
|
|
What is Beta
|
maximum probability of flaslsely accepting the null hypothesiis that we are willing to tolerate
|
|
1-Beta is the
|
power of the test to detect if adifference turly exists
|
|
B=0.2 what is power
|
80%
|
|
What are factors influencing power
|
Sample size, increase size increases power
Decrease varailbity increases poewr increasing effect size increase power |
|
Does the power increase as Type 1 error increase, and why
|
YES--type II error decreases
|
|
What is minimum power needed
|
70-80% or more
|
|
For numerical data what is parametric test for 1 group
|
one-sample t test
|
|
What is the numberical data for NON-parametic test for 1 group
|
sign test
|
|
What is the non-parametric 1 group test
|
sign test
|
|
When do we use non-parametric data
|
when the data is skewed
|
|
Is the data is skewed then we do not use the average, rather
|
median
|
|
What 2 groups for numerical data what are 2 catergoires
|
pair or independent
|
|
What are Paired group
|
measuring one-group twice---measuring BP twice pre and post
|
|
What is test for measuring Paried numerical group
|
paired t-test
|
|
What is the test for measuring NON-parametic pair groups
|
Wilcoxon signed test ranks
|
|
The most common numerical 2 group is
|
indepenent
|
|
What is an example of numuerical parametric 2 group independent
|
2 groups one given a drug and the other placebo
|
|
What is the test used for 2 group independent
|
unpaired t test
|
|
When do have 2 independent group for one to have drug vs placebo what must you first do
|
ramdonize
|
|
Should the average for both drug and placebo be the same for both groups prior to starting
|
YES
|
|
What statistical test is used for 2 independent groups
|
two sample t-test or (unpaired t-test)
|
|
When is the ANOVA test used
|
3 or more groups
|
|
What test is used for 3 or more groups that data is skewed
|
Kruskal wallis test
|
|
What is the test statistic for ANOVA
|
F ratio
|
|
The Global test means if the P value is greater than or equal to 0.05, than
|
all means are equal
|
|
If P values is < 0.05 for ANOVA, what do we do
|
POST HOC TESt to determine which ones are different
|
|
How does the Post Hoc Test determine which one is differnt
|
using hte Minimum differnt size which is greater than or equal to 5
|
|
Categorical data--meaning each inviducal
|
posses a characterics of intrest (is males, is pregant, died/vs
|
|
Catergoical data does NOT use means, rather
|
proportions
|
|
In Categoical datas we are interested in detmerine wether
|
the TRUE proprotion in the popuation takes a particular value (HAVE DIABETES OR DO NOT HAVE DIABETES)
|
|
What is test for ONE group categorical data
|
z test for proption
|
|
What is test for categorical data with 2 groups (paired)
|
McNemar's test
|
|
What is test for categorical data with 2 groups (independent)
|
Chi-squared test
|
|
What must we have to calculate a test statistic for categoical data with 2 indpendent groups
|
2X2 and the there are mutally exclusive categorical
|
|
If a 2x2 and they are mutally exclusive, then you use
|
Chi-saure test with1 degree of freedom
|
|
When you use the Fisher's exact test
|
sample <5
|
|
What test do you use for Categorical data >2 groups
|
Chi-squared test
|