Reconsider the context of investigating whether giving chest-compression-only (CC) instructions rather than standard cardiopulmonary resuscitation (CPR) instructions to the witness of a heart attack will improve the victim’s chance of surviving. One of the randomized experiments that the researchers Hupfl et al. (Th e Lancet, 2010) looked at was by Hallstrom et al. (New England Journal of Medicine, 2000) where there were 35 survivors in the CC group of 240 and 29 survivors in the CPR group of 278.
a. Find the observed difference in proportion of survivors comparing those who received CC to those who received CPR.
b. Recall that in the Hupfl et al. study, where data from three studies were combined, there were 211 survivors in the CC group of 1,500 and 178 survivors in the CPR group of 1,531. Note that such a study is called a meta-analysis.

Find the observed diff erence in proportion of survivors comparing those who received CC to those who received CPR in Hupfl et al.’s meta-analysis.
c. Compare the answers from part (a) and part (b). Which is bigger?
d. Use an appropriate randomization-based applet to find a p-value for Hallstrom et al.’s data. Report and interpret the p-value in the context of the study.

e. Based on this p-value, how much evidence do you have against the null hypothesis?
f. Compare the p-value from part (d) to that from Hupfl et al.’s study. Which provides stronger evidence against the null hypothesis? How are you deciding?
g. You should have found that even though the observed difference in proportion of survivors is greater for Hallstrom et al.’s study compared to Hupfl et al.’s, the p-value is smaller for Hupfl et al. Explain why that makes sense.
h. Use the 2SD method to find a 95% confidence interval forthe parameter of interest based on Hallstrom et al.’s data.  Report this interval and compare the midpoint and width of this interval to that based on Hupfl et al.’s data. Do the differences between the midpoints and widths of the two intervals make sense? Why?