Lecture 8

Question 1

What are the two types of continuous outcomes?

Answer 1

independent and matched

Question 2

Why are confidence intervals the preffered method of making a conclusion about a population?

Answer 2

• shows the precison of the estimates
• can be applied clinically to the whole population

Question 3

What is an independent variable?

Answer 3

categorical variable with two levels that are independent (exposure variable)

Question 4

What is dependent variable?

Answer 4

an interval or ratio in which a mean can be calculated (outcome variable)

Question 5

What are the requirements of an independent sample t-test?

Answer 5

• 2 independent samples
• normally distributed (n >= 30)
• both variances are equal

Question 6

Which equation is used for which variacne scenariio?

Answer 6

equal variance- pooled variance formula
unequal variance- separate variance formula

Question 7

What is the test statistic value equation if variances are unequal?

Answer 7

z or t = x-bar 1 - x-bar 2 / root (sd1^2/n1 + sd2^2/n2)

Question 8

What must be done to the p-value in a two tailed test?

Answer 8

multiply by 2

Question 9

Which value is used to determine the p-value?

Answer 9

the TEST STATISTIC calculated is applied to the z-table to determine the p-value

Question 10

What must be done to the p-value if the test statistic is < critical value?

Answer 10

be subtracted from 1 (the z-table only gives values to the left of the distribution, not the right- if the critical value is on the right side, positive)

Question 11

Based on p-values, when is the null rejected?

Answer 11

when p-value < 0.05

Question 12

If the null (0) IS included in the CI, is H0 rejected?

Answer 12

No, we fail to reject the null

Question 13

When variances are equal, what is the test statistic equation?

Answer 13

z or t = x-bar 1 - x-bar 2 / Sp root(1/n1 + 1/n2)

Question 14

What is Sp?

Answer 14

the pooled standard deviation

Question 15

If the test statistic calculated is too large for the table to accomadate which value is used?

Answer 15

the highest possible value is used for the p-value

Question 16

When comparing variation in two samples, what is their margin of error?

Answer 16

z a/2Sp root *1/n1+1/n2)

Question 17

What is the SD equation for matched samples?

Answer 17

root[ (sum of difference^2 - (sum of differences)^2 / n] / n-1

Question 17

What is the focus of matched t-tests?

Answer 17

the difference in means

Question 18

What is the t value equation for matched samples?

Answer 18

X-bar d - mew d / sd/root n

Question 19

Why can mew d in the t value equation be zero?

Answer 19

there is no population since samples are being compared to each other

Question 20

In a t-test when is the null rejected?

Answer 20

if the t-statistic falls IN the rejection region determined by df and alpha)

Question 21

How is the p-value determined in t-tests?

Answer 21

the t-statisitic is lined up where it would hypothetically fall in the table and a range or </> of alpha values is determined

Question 22

What is the test statistic equation for proportions?

Answer 22

z = p-hat 1-p-hat 2 / root (p-bar(1-phat)(1/n1+1/n2)

Question 23

What is the p-bar equation?

Answer 23

x1 + x2 / n1 + n2

Question 24

What is the purpose of the test of independence?

Answer 24

testing that there is NO association between the two variables?

Question 25

What is the reuqirement of the expected frequency for a X^2 independence test to be ran?

Answer 25

E must be >= 5

Question 26

What is the test statistic equaiton for independence tests?

Answer 26

x^2 = sum of (O-E)^2 / E

Question 27

What is the equation for the E values?

Answer 27

E = row total * column total / grand total
- for each row/column combination

Question 28

What is the degree of freedom for an independence test?

Answer 28

df = # rows - 1 * # columns - 1

Question 29

When is the null rejected?

Answer 29

since the critical value table is x>X, if the test statistic is GREATER than the critical value, the null is rejected