Question 139808: Stan you answered this question but my instructor came back with another question. Here is the orginal question. The number of admissions for Carolinas Medical Center hospital emergency room over the course of a lunar cycle was 129 patients per 36 days per full moon as compared to 1,315 patients per 330 days per nonfull moon days
Perform a chi-square test on the data. Use the test to prove (or disprove) that hospital emergency room admissions are dependent (or independent) on (of) the full moon lunar phase.
Full moons were defined as three-day periods in the 29.531-day lunar cycle, with the middle day being described in the world almanac as the full moon.
: The number of admissions for Carolinas Medical Center hospital emergency room over the course of a lunar cycle was 129 patients per 36 days per full moon as compared to 1,315 patients per 330 days per nonfull moon days.
Here is the question my instrutor asked....How did you get the test statistics, I'm sure your classmates are curious of how the .0287 generated as well as the p-value of .5920. Help.
Answer by stanbon(75887)   (Show Source):
You can put this solution on YOUR website! Ho: full-moon and # of patients are independent
Ha: they are dependent
--------------------------
I set up 2 rows of DATA in a matrix on a TI caluculator:
full-moon.....129......36
no full-m....1315......330
----------------------------
I ran a Chi-Sq Test on the matrix and got:
test statistic: Chi-Sq = 0.87089...
p-value = 0.59209..
df = (2-1)(2-1)
------------------------------
You could do this by hand:
The data listed above is the "Observed" data.
You could generate the "Expected data"; that is the numbers you would
expect to see if the # of patients was independent of the full-moon.
That matrix would be:
full-moon.....131.64......33.365
no full-m.....1312.4......332.64
----------------
Then you would have to calculate the Chi-Sq statistic
using:
Chi-Sq = Sum[ (observed-expected)^2/expected] for each of the 4 entry
positions.
Then you would have to use your Chi-Sq chart to calculate the area
to the right of your test statistic to get the p-value.
---------------------------
I don't know how much evidence your instructor requires. The
calculator is handy as it does all the arithmetic. However,
you should try to understand what is behind the numbers the
calculator generates.
--------------------------
Cheers,
Stan H.
|
|
|
