00:01
So one thing that i need to note here is that with the way that the problem is set up, it is a little bit ambiguous.
00:07
But i'm going to assume that this is a case where you are expected to use a kai squared test.
00:21
In fact, i went through and tested this using a zd score test for difference in population proportions.
00:28
But in that context, the critical value given would not make sense.
00:34
So using the given critical value as a kai square test makes a lot more sense.
00:39
So the approach that we want to take is set up our sort of contingency table.
00:46
So we have, i'll just have the event a and a complement for they received the medication and they didn't.
00:54
And then events, h and h complement for if they hemorrhaged or not.
00:59
So we have 1371 and also i'll have a column total.
01:07
And additionally we have a row total and so we have 84 is the total across the column then we have 3 74 and 77 then our row totals are 16 145 and 161 then we want to find our expected frequencies which we calculate as the let's see here in the formula eij that's total cross row i times the total across column j divided by the grand total for each one of these and it'll just indicate these using i'll just indicate these using values put into sort of brackets next to the original measurements there may need to make this a little bit small but so we'll find our values our expected values are going to be 8 .35 7 .65 75 .65 and 69 .65 so, we then want to find our kai squared value by taking the sum, the difference between the observed value, the expected value, squared, divided by the expected value, where we're summing across all values of i and j...