Normal, Rank Correlation, PMCC Hypothesis Tests Flashcards
Normal with means , what are we testing for exactly
We are given a proposed mean of the population and we want to test if the mean is actually this. We do this by taking a SAMPLE of the population, finding the means and using the fact that sample means are NORMALLY distributed with the same mean and a different vadiance
We test then if the probability of mean we found is likely to happen or not, and then reject or accept the H0 accordingly
What two variables MUST YOU DECLARE for doing means sample
And what sentence you say about H0
Let mew be population mean
Let x be the disturb turn of whatever
If H0 true x follows normal
Thus x bar follows normal
How to do critical region test for normal sample means?
Depending on which tail you’re in, you’re testing for the absolute value that anything abive this will give yiu a value less than significance level
- so if 5% big tail, either use wacky calc or have to do 95 % to find this value insetsd
If the POPULATION VARIANCE IS NOT GIVEN, only the sample, How can we do rest of question?
WHAT CONDITION MUST BE MET
IMPORTANT
If the SAMPLE SIZE IS BIG ENOUGH, we can basically say the sample variance is the same as the ACTUAL VARIANCE
Therefore = can use as normal , s2 /n = model variance for means of x
But this is ONLY if the sample population is big enough, typically bigger than 30
How to do 2 tail tests for p value and critical for normal sample mean
1) also critical both tails and p value 1?
Critical means divide significance by 2
Same for hypothesis, tho I suppose you could multiply p value by 2
This is because significance split both tails
2) yes critical must show both regions but hyptoheks based on tail just that one
What does -1 to 1 mean for
1) measuring association
2) measuring ranks
1) s association , how well it fits that association, 1 means perfect
2) for ranks, if two judges ranked, then it’s how close they ranked the same thing
How to accept or reject null hypothesis based on the CRITICAL VALUE
What is the acceptance range
Modulus answer
Acceptance range is 0 to critical
Beyond critical is the critical region
So if value falls in there, reject!
Else do not reject
What does the null and alternative hypothesis twke for correlation tests
H0= row = 0 ( assume no correlation)
H1 is row bigger or less than 0, or not = to 0 for two tail
Where ROW = THE PARENT PIPULATION CORRELATION COEFFEICMT
What are final statements for rejectiting H 0 and accepting p greater than less than and not equal to 0?
Evdicent to suggest positive correlation
Negative
Some correlation
Difference between correlation and association?
Association just means two thinfs are linked
I think correlation is linear association in particular
Why can’t we use PMCC hypothesis test on a RANK correlation (2)
Thus what must we calculate!
Subjective
+ there is no PARENT POPULATION IN THE FIRST PLACE
2) spearman’s
As there is NO parent population for spearman’s rank, there is no row
So what form does null and alternative hypothesis must take?
IMPORTANT
H0 = there js NO association between x and y (assume not)
H1 = there js some , Positive or negative associaiton
How to format spearman’s rank hyptiheis
H0 words
H1 words
Critical value at @% signify Ian is this
Is x > or whatever, reject
So MAKE SURE TO STATE CRITICAL VALUE
How to use ADJUSTED z score to find critical regions for sample mean testing
Adjusted z score = divided by new sigma
Find the z score by inverting normal of 0.05 and 0.95 whatever
And then using this value to work out the value for critical region by rearranging z = d - d / new sigma for first d
How to use z scores for sample mean?
