Transcribed Text
For this assignment, use .05 for all inference.
1. White blood cell count (WBC) is measured on sample of 183 women. Some descriptive
statistics, as well as a95% for the mean, are given below.
Meam
7128.03 1724.37
i) We want to test fthe mean WBC differs from 7500. Complete the following steps
(you'll use the edited printout below for some fthese).
a) Write the null and alternative hypotheses.
b) State the rejection region.
c) Calculate the test statistic, to.
d) State your decision to reject or fail to reject the null hypothesis.
e) Interpret the result in the context of the problem.
ii) Explain how you could use the CI above to reach the decision of the test part (i).
iii) Suppose WBC not Normally distributed. Are you comfortable using the ttest in this
setting, despite this? Why why not?
2. Using the "babies' data set from the lessons, we want to test the mean birth weight of
female babies born to mothers who got poor prenatal care lessthan 3000 grams.
i) Write the nul and alternative hypotheses.
ii) The data were analyzed using PROC TTEST in SAS: edited results are below. State the test
statistic and pvalue, and interpret results in the context of the problem.
Variable birthat
=

Sad Dew
Sad Err
Minimura
Maximum
435.8
92.7524
2043.0
3598.0
0.1349
3. This problem is related to Review Exercise p. 279 of the text but the numbers have
changed. Cotinine levels (nmol/l) are measured hours and 24 hours after smoking ona
sample of 12 people. We want to see the mean cotinine level differs for the two time periods.
i) Are these paired or independent samples? Explain why you chose your answer.
ii) Write the null and alternative hypotheses. Clearly explain any notation you use.
iii) In the context of this problem, define what Type and Type errors are.
iv) Edited results of test are given below. Interpret test results the context of the problem.
Difference: cotminel2
; Std Dev war Minimurn
5.9362 11.0000 28.0000
DE Pr>t
0.0003
4. Again using the "babies' data set from the lessons, we want to see mean length is smaller
for female babies than for male babies, looking only at mothers who had good prenatal care.
i) Are these paired or independent samples? Explain why you chose your answer
ii) Write the null and alternative hypotheses.
An edited portion of SAS printout is given below. Use this to answer parts (iii) (iv).
Variable: length
Mean Std Dev Sad Err Minimum Maximum
de 20.3571 18.5000 24.0000
0.6005
1.3674
0.2477
Variances
120
Unequal
119.96
2.43
0.0083
Equality dr Varlances
Num DF
ii) When doing the test of interest, do you feel comfortable with the assumption of equal
variances? Give some numeric justification for your answer.
iv) Give the value for the test of interest, and interpret results in the context of the problem.
5. We are planning study to compare mean reaction time between people with no alcohol
consumption those who have had the equivalent of 24 ounces beer. We speculate that
the mean reaction time will be 300 milliseconds theno alcohol group, and 360 milisecond
group who had alcohol, and we therefore will doa one sided test. We also believe the
standard deviation among participants within each group will be 60.
i) Write the nul and alternative hypotheses being tested here.
The code (provided i case you'r interested) needed to do power analysis is below, with
corresponding edited printout on the next page. Use these to answer the following questions.
ii) What sample size do we need to attain power of .6?
iii) What sample size do we need to attain power of .9?
iv) How would the answers to parts (ii) and (iii) change we changed the standard deviation
from 60t 80?
v) If recruiting each patient costs $150. how much would the study cost we want power
of .8? Remember, the printout gives the sample size per group.
vi) Rather than setting an priori power level like we did in parts (ii) (iv) suppose we want
to use the results provided to decide on the sample size. State what sample size you would
recommend, and justify why you chose this value.
stddev
Two Sample Test for Mean Difference
03
o
These solutions may offer stepbystep problemsolving explanations or good writing examples that include modern styles of formatting and construction
of bibliographies out of text citations and references. Students may use these solutions for personal skillbuilding and practice.
Unethical use is strictly forbidden.