Statistic Methods- Lecture 15-17-18-19-20

Question 1

How do you calculate standard error when working with one proportion?

Answer 1

When working with one proportion,
○ if constructing a confidence interval, use p= p̂
○ if doing a hypothesis test, p comes from the null hypothesis, p=p0

􏰀

Question 2

How do you calculate the standard error when working with two proportions?

Answer 2

When working with two proportions,
○ if constructing a confidence interval, use p̂ and p̂
􏰀
○ if doing a hypothesis test H0: p1 = p2: use
p̂pool= (#suc1+#suc2)/n1+n2

Question 3

How do you calculate the test statistic for two proportions?

Answer 3

Z=(p̂1-p̂2)/SE(p̂1-p̂2)

Question 4

What is the t-distribution?

Answer 4

When we use the sample standard deviation to estimate the SE of the sample mean using then we replace z* with t* in the CI

x̄+- t*(n-1)xSE

Question 5

What is the shape of the t-distribution compared to the normal distribution?

Answer 5

-same bell shape as normal distribution but “fatter” and “shorter”

Question 6

What is degrees of freedom(df)?

Answer 6

Degrees of freedom describes the shape of the t-distribution. The larger the degrees of freedom, the more closely the t-distribution approximates the normal model.

When modeling x̄ using the t-distribution, use df=n-1

Question 7

How do you traditionally calculate a test statistic?

Answer 7

Z=(point estimate (x̄)- hypothesized mean)/SE(point estimate)

Question 8

What is the rule of thumb to check normality levels for CLT and sample mean?

Answer 8

N<30 data came from a nearly normal distribution

N>30 sampling distribution of x̄ is nearly normal (CLT applies)

Question 9

How do we analyze paired data?

Answer 9

● When two sets of observations have this special correspondence (not independent), they are said to be paired
● To analyze paired data, it is often useful to look at the difference in outcomes of each pair of observations,

Question 10

What is the degrees of freedom for two means?

Answer 10

Df=n1+n2-2