WEBVTT

1
00:00:02.940 --> 00:00:04.830
<v Instructor>Hello students, welcome to Biostat ER,</v>

2
00:00:04.830 --> 00:00:08.130
chapter eight, example seven.

3
00:00:08.130 --> 00:00:11.070
In this example, we will learn how to calculate sample size

4
00:00:11.070 --> 00:00:12.990
for one-sample dichotomous data

5
00:00:12.990 --> 00:00:14.853
to determine confidence interval.

6
00:00:16.620 --> 00:00:19.110
The problem here is not from our textbook,

7
00:00:19.110 --> 00:00:21.960
and the issue here is that we have a new policy

8
00:00:21.960 --> 00:00:24.240
implemented in a private practice

9
00:00:24.240 --> 00:00:26.640
and we would like to know what proportion of patients

10
00:00:26.640 --> 00:00:28.950
are in favor of this new policy.

11
00:00:28.950 --> 00:00:31.440
With 95% confidence, we want to state

12
00:00:31.440 --> 00:00:33.960
that our estimate is within 3%

13
00:00:33.960 --> 00:00:36.000
of the true proportion of patients

14
00:00:36.000 --> 00:00:38.520
who are in favor of the new policy.

15
00:00:38.520 --> 00:00:40.290
So, how many subjects are required?

16
00:00:40.290 --> 00:00:42.990
As always, I have summarized the information here

17
00:00:42.990 --> 00:00:45.123
and I have also stated the formula.

18
00:00:48.090 --> 00:00:49.560
Now, before we can move forward,

19
00:00:49.560 --> 00:00:52.470
we need to determine the value of the P here.

20
00:00:52.470 --> 00:00:55.950
Sometimes we can determine this P value from the literature

21
00:00:55.950 --> 00:00:58.020
or have an established value.

22
00:00:58.020 --> 00:01:02.070
However, here that is not feasible as this is a new policy.

23
00:01:02.070 --> 00:01:04.470
Because we always want to be conservative

24
00:01:04.470 --> 00:01:06.180
in sample size calculation,

25
00:01:06.180 --> 00:01:07.770
we want to utilize a P value

26
00:01:07.770 --> 00:01:10.890
that will give us the largest sample size

27
00:01:10.890 --> 00:01:13.323
within the margin of error we want to remain,

28
00:01:14.190 --> 00:01:16.443
and that value is 0.50.

29
00:01:17.850 --> 00:01:20.370
So, for the initial part of the calculation,

30
00:01:20.370 --> 00:01:25.370
as we insert 0.50, we obtain a value of 0.25.

31
00:01:27.540 --> 00:01:31.050
I encourage you to experiment with different P values

32
00:01:31.050 --> 00:01:36.050
to determine if you can obtain a value larger than 0.25.

33
00:01:36.510 --> 00:01:39.780
The Z value here will be 1.96,

34
00:01:39.780 --> 00:01:44.020
and once we insert all these values and perform the algebra,

35
00:01:45.870 --> 00:01:50.870
we will get 1066.02, and as always, we will round up.

36
00:01:52.050 --> 00:01:57.050
So we will need 1067 subjects.