STATS 119 SDSU FINAL

CONFIDENCE INTERVAL: For a POPULATION MEAN what kinda of test do you use?

Use a T test

HYPOTHESIES TEST: When standard deviation of the population is known

Use a Z-test

HYPOTHESIES TEST: When standard deviation of population is unknown and sample standard deviation given?

Use a T-test, with the degrees of freedom

Describing a sample distribution that is NORMAL

N( mean, SD/n^.5)

Describing a sample distribution that is SKEWED

AN

HYPOTHESIES TEST: With a T-test on a calculator

Tcdf (low number, test statistic, DF(n-1))

When you see the word total in a question using samples

convert the number within the terms of the sample

Unknown mean and standard deviation in a confidence interval

Use T-table with degrees of freedom and the confidence interval

Population Proportion mean and sd

mean= np SD= (npq)^.5

Describing a confidence interval

We can be % confident that the POPULATION PROPORTION

When n is small and uses: at least...What function do you use?

1 - Binomcdf( n, p, x-1)

Compare t-distribution to to standard normal distribution

As the degrees of freedom increases, the t-distribution approaches the standard normal curve

Average math score for all students who took the ACT was a 21..... what do you call 21

A Value of a Parameter

When Ho is in the interval what do you do?

Fail to reject

When Ho is outside the interval what do you do?

Reject Ho

Level of significance for a hypothesis test

1 - (confidence level)

IQR

Q3-Q1

Lower bound of the quartile

Q1- (1.5 x IQR)

Upper bound of the quartile

Q3 + (1.5 x IQR)

Testing for an outlier

See if x is within the upper or lower bounds of the confidence interval

asking for "Percent of variability" in regression means you do what?

R^2

The correlation

R

When the residual is negative

its an over estimate

When residual is positive

its an under estimate

Type II Error

When you accept something but its actually not true

Cluster Sampling

Divide population into groups then select some clusters and all the people within the cluster

Stratified Sampling

Divide population into groups (non overlapping) and then random sample from each group

Systematic Sampling

Population can be put as a list of numbered individuals. Divide the list into segments and at sample a the same random point for each segment

Calculate the residual

y-y-hat , that is the plug in value if percent given use percent

Finding P-value for a test statistic that has a not equal to sign

Use the negative of that statistic to find p-value and multiply by 2

When you see wining the percentile

INVNORM

How to test independence

Independent: P(A I B) = P(A)

if not then its disjoint

If you see GIVEN in a probability it means

CONDITIONAL PROBABILITY!

If the p-value is LESS/SMALLER than the level of significance, what do you do?

REJECT Ho

If the p-value is LARGER/MORE than the level of significance, what do you do?

FAIL TO REJECT Ho

Randomized block experiment

subjects are divided into specific GROUPS then randomly assigned to be in the treatment or control group

When finding a confidence interval for a POPULATION proportion, what do we use?

1-propZinterval

To find table value (critical value), what do you do?

cross check degree of freedom with CL on table

what do you use to find the second part of a p-hat distribution?

square root of p*q/n

what equation do we use to find sample size?

(z-score*SD/ME) squared

If a data set has a strong skew to the right, what happens to the mean?

mean moves in direction of skew (greater/less than) - mean will be larger than the median

What are the hypothesis test assumptions?

approximately normal pop & simple random samp