## Transcribed Text

1. A study is being planned measuring the amount of time clinicians spend navigating an Electronic Medical
Record, in particular, measuring the difference in time between MDs and their administrators. They hope to
gather 100 physicians and 100 administrators separately, and they think that a difference of 10 seconds between
the groups would be significant. We will assume that the times are reasonably normally distributed, and that the
standard deviation for times spent navigating the system is 40 seconds.
(a) Calculate the power of this study.
(b) What would be a sufficient sample size? Assume a power of 80%.
(c) If the sample size is fixed, what would you change about the study to increase the power? Be specific.
2. A study is being conducted to determine if there is an effect of gender on the rate of breech birth amongst infants.
They expect half the babies to be male, and hope to be able to gather up to 1600 patients. They have no firm
data on the rate of breech births, but they believe that the rate is around 10%, and they believe that a difference
of 5% would be a meaningful clinical difference.
(a) Calculate the power of this study.
(b) What would be a sufficient sample size? Assume a power of 80%.
(c) When the study is completed they found that the breech rate was actually 45% and 50% for males and females
respectively. Why would this be a problem?
3. The following table produced an OR of 0.3822 and a chi-square statistic of 9.15. Interpret the OR and its
significance.
Cancer No Cancer Total
Treatment Yes 15 185 200
No 35 165 200
TOTAL 50 350 400
4. For the following table, calculate RD and NNT (and interpret them). The chi-square statistic is 4.31, is the effect
of drug A significant?
Cancer No Cancer Total
Drug A 28 72 100
B 42 58 100
TOTAL 70 130 200
Paper Questions
5. The Short 2015 paper presented RR as a measure of the effect of their intervention. What other metrics could
be used to summarize the effect? Provide an evaluation of their paper using two other effect measures.
R Questions
Use only the first encounter for each subject in the Framingham Dataset for the R questions.
6. Calculate a 90% confidence interval for the DIABP variable in the Framingham dataset.
7. Test the relationship between smoking (CURSMOKE) and angina (ANGINA) using the Framingham data
(a) State the null and alternative hypothesis
(b) Present the test statistic
(c) Provide a p-value and draw your conclusions
(d) Present the RR and 95% confidence interval for the risk of having ANGINA for smokers compared to nonsmokers
8. Test the ordinal relationship between smoking and BMI category
(a) State the null and alternative hypotheses
(b) Provide the p-value and draw your conclusions

These solutions may offer step-by-step problem-solving 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 skill-building and practice.
Unethical use is strictly forbidden.

# 7

tbl <- table(CURSMOKE=d$CURSMOKE,ANGINA=d$ANGINA) # CURSMOKE will be rows, ANGINA will be columns

tbl

# ANGINA

# CURSMOKE 0 1

# 0 1862 391

# 1 1847 334

# (b) and (c)

chisq.test(tbl, correct=F)

# Pearson's Chi-squared test

#

# data: tbl

# X-squared = 3.3738, df = 1, p-value = 0.06624

# (d)

install.packages("fmsb")

library(fmsb)

riskratio(Y=391,X=334,,m2=2253,m1=2181)

# Disease Nondisease Total

# Exposed 334 1847 2181

# Nonexposed 391 1862 2253

#

# Risk ratio estimate and its significance probability

#

# data: 334 391 2181 2253

# p-value = 0.06627

# 95 percent confidence interval:

# 0.7720333 1.0085895

# sample estimates:

# [1] 0.8824198...