ci lsrl & sp + ci for B

Question 1

1 sample t interval for B and 1 sample t interval for Udiff

Answer 1

for B, df= n-2
for Udiff, df= n=1

Question 2

b +/- t*SEb

Answer 2

Udiff= xbar diff +/- sdiff/sqroot of ndiff

Question 3

Random?

Answer 3

SRS/RA OR RANDOMIZED EXPERIMENT SELECTED

Question 4

Linear?

Answer 4

check SCATTERPLOT BETWEEN X AND Y for a LINEAR RELATIONSHIP

Question 5

Normal?

Answer 5

for Udiff, n >/= 30 or if too small, the ORIGINALDOTplot doesn’t show strong skewness or outliers!!! NORM AND RANDOM SHOULD STAY THE SAME!!

for B, the RESIDUAL dotplot does not have strong skewness or outliers (b only) (random, and then ONLY if the residual dotplot shows no strong skew or outliers)

Question 6

Equal St.D?

Answer 6

Check residual scatterplot for a RANDOM SCATTER (b only)

Question 7

Interpret SEb

Answer 7

If the study was repeated many times and we got many LSRLS, the SEB ESTIMATES!!! the sampling variability of the slope of the LSRL to predict y!!!!! from x !!!!!! (USE UNITS FOR BOTH) (ex. to predict height of soapsuds (mm) from amount of detergent used (grams))

Question 8

Interpret B

Answer 8

For each additional unit of x, we can predict an increase of y (include units!!)
ex. for each additional gram of detergent used, we can predict a 0.13 mm increase in height of soap suds ON AVERAGE

Question 9

Interpret S
this statistic measures the typical amount of vairablity in the vertical distances between the observed height of the soap suds in mm and the lsrl

Answer 9

S= standard deviation of the residuals
This statistic measures the typical amount of variability in the vertical distance from y to the LSRL

ex. S= 1.7- this statistic measures the typical amount of variability in the VERTICAL DISTANCE between the height of soapsuds (mm) to the LSRL

Question 10

LSRL EQUATION

Answer 10

yhat= coefficient +B(x)

Question 11

STATE B

Answer 11

we want to estimate B= the true slope of the LSRL RELATING yhat= PREDICTED height of soapsuds (mm) to the amount of detergent used (grams) used for ALL PANS SIMILAR IN THE STUDY

Question 12

for xdiff, do after-before and put into L1

Question 13

At most you’ll only do 2 conditionals, random, normal if dot plot, equal sd if scatterplot, and linear if original graph is given with a linear relationship

Question 14

2 groups of ___ selected in a randomized experiment

Question 15

1 sample t test for B

Answer 15

t= b-B/SEb

Question 16

in a 1 sample t test for B, df= n-2

Question 17

!!!!!!!!!!!B= slope of the population regression line relating x to y

Question 18

Conditions for slope
LINEAR

Answer 18

check scatterplot for a relationship between x and y (NOT RESIDUAL)

Question 19

Conditions for
NORMALITY

Answer 19

Check residual dot plot for no strong skew or outliers

Question 20

Conditions for
EQUAL SD

Answer 20

check residual scatterplot for a random scatter

Question 21

Conditions for
RANDOM

Answer 21

random assignment/randomized experiment
ex. random sample of 12 flowers and independeny random sample of 8 flowers

Question 22

Conditions for INDEPENDENCE

Answer 22

check if x and y are independence (unless it’s a randomized experiment, then do not check)

Question 23

differences of means
x1-x2-(U1-U2)/sqr. S1sq/n + S2 sq/n

Question 24

differences of means in differences of means –> take one off the smaller n value (n-1)

Question 25

SEb- this is the standard error of the sample slopes!!!!!!! if we repeated the study many times and got many lsrls, the SEb measures the sampling variability in the sample slope of the LSRL predicting y (context) from x (context)

Answer 25

SD -this is the std. of the residuals!! The st. d measures the average amount of variability of the vertical distances from the OBSERVED y and from the regression lines

Question 26

state- if B not equal to 0, it's two sided, leave p value as is, if not, it's one sided so halve p value

Question 27

Interpret b

Answer 27

if b = 0.204 b predicts that for every additional increase of 1 (x), the EXPECTED (y) increases by 0.204 on average

Question 28

U1-U2>0 U1= the mean difference (pretest-posttest) for all students at the original school U2= the mean difference (pretest-posttest) for all students at the magnet school

Question 29

10%- it is reasonable to assume there are at least....

Answer 29

Unless stated, then Problem states there are more than 20,000 students at this University.

Question 30

with concluding p value include ts because that's in his key

Question 31

state lsrl

Answer 31

yhat= a + bx x= # of _____(unit) yhat= predicted # of ____ (unit)

Question 32

a- if no x was added, the y would be _____

Answer 32

SEb- standard error of the sample slopes SD- standard deviation of the residuals