WEBVTT

1
00:00:02.190 --> 00:00:03.390
<v Instructor>Hello students and welcome</v>

2
00:00:03.390 --> 00:00:06.910
to Biostat ER chapter eight, example five.

3
00:00:06.910 --> 00:00:10.800
In this example, we will learn how to calculate sample size

4
00:00:10.800 --> 00:00:14.100
for two-independent samples, continuous outcome,

5
00:00:14.100 --> 00:00:17.010
to determine confidence interval.

6
00:00:17.010 --> 00:00:18.813
This example is from our textbook.

7
00:00:19.680 --> 00:00:21.180
So this is problem 11

8
00:00:21.180 --> 00:00:23.257
and first I'm going to read the problem.

9
00:00:23.257 --> 00:00:25.950
"We wish to design a study to assess risk factors

10
00:00:25.950 --> 00:00:28.080
for stroke and dementia.

11
00:00:28.080 --> 00:00:29.896
There are two primary outcome measures.

12
00:00:29.896 --> 00:00:33.415
The first is a measure of neurologic function.

13
00:00:33.415 --> 00:00:36.810
Patients with normal function typically score 70

14
00:00:36.810 --> 00:00:39.270
with a standard deviation of 15.

15
00:00:39.270 --> 00:00:42.030
The second outcome is incident stroke.

16
00:00:42.030 --> 00:00:44.250
In patients over the age of 65,

17
00:00:44.250 --> 00:00:48.720
approximately 40% developed stroke over 25 years.

18
00:00:48.720 --> 00:00:51.480
How many participants over the age of 65

19
00:00:51.480 --> 00:00:53.943
must be enrolled to ensure the following?

20
00:00:55.230 --> 00:00:58.980
That the margin of error in a 95% confidence interval

21
00:00:58.980 --> 00:01:02.580
for the difference in mean neurologic function scores

22
00:01:02.580 --> 00:01:06.957
between men and women does not exceed two units."

23
00:01:08.735 --> 00:01:11.110
So first here we have to

24
00:01:13.380 --> 00:01:15.240
select the correct formula, and I have

25
00:01:15.240 --> 00:01:17.223
provided you with the formula here.

26
00:01:18.120 --> 00:01:21.810
And after that we have to insert the values

27
00:01:21.810 --> 00:01:26.810
into our formula, which are provided in our problem.

28
00:01:27.300 --> 00:01:30.630
So the Z value here is going to be 1.96 again

29
00:01:30.630 --> 00:01:33.060
because it is 95% confidence interval

30
00:01:33.060 --> 00:01:36.510
and we have used the Z value right now multiple times.

31
00:01:36.510 --> 00:01:38.610
The sigma here is going to be 15

32
00:01:38.610 --> 00:01:42.570
as it is provided in our problem.

33
00:01:42.570 --> 00:01:44.730
And I've highlighted in yellow.

34
00:01:44.730 --> 00:01:48.720
And then the denominator here is going to be two again,

35
00:01:48.720 --> 00:01:50.700
which is provided to us in our problem.

36
00:01:50.700 --> 00:01:55.140
And again, I have highlighted that value in yellow.

37
00:01:55.140 --> 00:01:56.940
So once we insert all these values

38
00:01:56.940 --> 00:02:01.228
and we perform the algebra, we get 432.2.

39
00:02:01.228 --> 00:02:05.340
So the sample size here is going to be 433 for men,

40
00:02:05.340 --> 00:02:10.340
and 433 for women and that is what we will need.

41
00:02:10.530 --> 00:02:14.850
And again, we always round up for sample size calculation.

42
00:02:14.850 --> 00:02:19.720
That's why from 432.2, we will round it up to 433

43
00:02:20.800 --> 00:02:23.820
and we will require an equal sample size for both men

44
00:02:23.820 --> 00:02:27.630
and women to perform the calculation

45
00:02:27.630 --> 00:02:29.133
for the confidence interval.

46
00:02:31.546 --> 00:02:34.296
(mouse clicking)