In 2009 the H1N1 flu virus became a great concern to public health officials around the world. Because it is easily passed from human to human, H1N1 quickly spread from its origins in Mexico to millions of people. SeeFlu Tracker Map.

What added to the concern of health officials is that H1N1 is closely related to the A/H1N1 virus responsible for the 1918 flu pandemic that killed 20 to 40 million people worldwide (see above photo of temporary hospital at Ft. Riley, Kansas).

Pandemics have occurred regularly throughout this century (1918, 1957 and 1968), but the 1918 pandemic was particularly destructive. The fatality rate was ten times higher than in all other flu pandemics and was unusually high in young adults. Epidemiologists and other health professionals are studying all aspects of the 1918 pandemic to be better prepared since pandemic influenza spreads rapidly, has high infection rates, and can kill millions of people worldwide.

Influenza can strike quickly and spread with frightening speed. Data for 6 US cities is in this text file (right click the file and choose "open in new window" or "open in new tab" to open the file):H1N1 Flu. Shown is the excess (that is, above the normally expected) weekly death rates (annual basis) per 100,000 from influenza and pneumonia in Buffalo, Chicago, Nashville, New Orleans, New York City, and Pittsburgh from Sept. 14, 1918 (week 1) to Oct. 26, 1918 (week 7) during the early stages of the 1918 flu pandemic. You can use statcrunch to plot this data on a Google map as follows:

• Right clickGoogle maps at statcrunch.comand choose "open in new window" or "open in new tab" to open the Google mapping facility at statcrunch.com.
• Below the map in theLocation informationwindow copy and paste the location information from the above text file.
• In theLegend titlewindow copy and paste the legend information from the above text file.
• In theOptional legend informationwindow copy and paste the optional legend information from the above text file.
• Click thePlot locationsbutton below the Optional legend information window.
• Your data will be plotted on the Google map.
• The map will have "baloon" markers showing the location of the cities. Point to a marker to see the name and excess weekly death rates for that city; click on the marker to open a window with the information. Alternatively, click a city name in the list to the right of the map to see the location and information for that city.

Question 1. The weekly excess death rate for each city is the sample mean of the data from several reporting stations (hospitals, clinics, etc.) within a 50 mile radius of the city. Determine the 95% CI for the mean week 7 excess death rate in the city with the highest reported week 7 excess death rate. Assume this city receives data from n = 38 reporting stations and that the sample standard deviation s = 463.

lower bound

upper bound

Are the New York City and Pittsburgh mean week 7 excess death rates significantly different? Perform a 2-tail hypothesis test for mu(NYC)- mu(PITT). Assume n(NYC) = 50, n(Pitt) = 35, s(NYC) = 239, s(Pitt) = 214. DO NOT use the approximation minimum(n1-1, n2-1) for the degrees of freedom. Question 2a. What is the value of the test statistic for this hypothesis test?

Question 2b. Select the correct choice concerning the P-value for this hypothesis test.

P-value0.00010.10P-value0.150.02P-value0.040.001P-value.01P-value0.05

Question 2c. What is the correct conclusion for this hypothesis test?

Reject the null hypothesis and conclude that the mean week 7 excess death rates are different. There is evidence that Pittsburgh's mean week 7 death rate is higher than New York's.Do not reject the null hypothesis; there is no significant difference in the mean week 7 excess death rates.Accept the null hypothesis; the mean week 7 excess death rates are equal.Reject the null hypothesis and conclude that the mean week 7 excess death rates are not significantly different.