00:01
Okay, so in part a, we're asked to find the percentage in each group that suffered a severe adverse outcome.
00:06
So for the drug group, we'll call it pd, and that was 198 out of 1099, which turns out to be as a percentage that's 18 .02%.
00:22
Obviously to get from here to a percentage you have to times by 100.
00:26
And then for the placebo group, it was 82 out of 546.
00:37
And that turns out to be 15 .02%.
00:42
Part b then says, fill out a two way table with the counts in each category.
00:48
So if we say we've got the drug group and the placebo group, and then we say the severe adverse effects or not, then we know that the drug group had 198 out of a total of 1099, meaning that 901 didn't have severe adverse effects.
01:17
For the placebo, 82 out of 546 had severe effects.
01:22
So that means the remaining 464 did not.
01:25
And then we can just add up the totals for the remaining boxes.
01:30
So down here, we get a total of 280, here 1365, and then here 1645.
01:41
So that's our table.
01:43
And then we want to do a hypothesis test by the looks of things.
01:47
So the null hypothesis is going to be that the effects were independent of the treatment.
01:58
And the alternative hypothesis is that the effects are going to be dependent on the treatment.
02:07
And the way we're going to compute this is, well, it says to check that all the required conditions are met.
02:17
We've got all the obvious conditions met for a chi -squared test.
02:20
We're doing a chi -squared test of independence at the 5 % significance level.
02:28
And so the main condition we need to check is that the expected values are all bigger than 5.
02:32
Now, the expected values are given by the row total times the column total divided by the total, the overall total...