In: Statistics and Probability
Neurofibromatosis Type 1 (NF1) is a human genetic disorder. As well as physical symptoms, affected children often suffer from impaired cognition and learning. A cognitive task that involves solving a puzzle is administered to a group of children. For each child the time taken (in seconds) to solve the task is recorded. It is not known whether there is any suitable parametric model for the times so we will investigate non-parametric methods.
Carry out the Wilcoxon signed-rank test on these data to test the null hypothesis that the mean time to solve the puzzle for children with NF1 is the same as for healthy controls.
d) Calculate the value of the test statistic and give the approximate normal distribution of the test statistic under the null hypothesis.
e) Calculate the p-value for the test assuming a two-sided alternative hypothesis.
f) What do you conclude about the time taken by children with NF1 to solve the puzzle compared to healthy controls?
g) Describe in a few sentences how you would calculate a 95% confidence interval for the mean time without assuming any particular parametric model for the data. You do not need to calculate the interval.
The data below are the observed times ?? taken by 57 children with NF1.
ID | yi |
1 | 51 |
2 | 60 |
3 | 75 |
4 | 43 |
5 | 92 |
6 | 72 |
7 | 49 |
8 | 39 |
9 | 62 |
10 | 127 |
11 | 51 |
12 | 75 |
13 | 69 |
14 | 59 |
15 | 25 |
16 | 58 |
17 | 95 |
18 | 63 |
19 | 91 |
20 | 63 |
21 | 32 |
22 | 50 |
23 | 108 |
24 | 41 |
25 | 93 |
26 | 43 |
27 | 74 |
28 | 50 |
29 | 55 |
30 | 60 |
31 | 62 |
32 | 91 |
33 | 79 |
34 | 71 |
35 | 85 |
36 | 86 |
37 | 78 |
38 | 100 |
39 | 146 |
40 | 62 |
41 | 134 |
42 | 41 |
43 | 40 |
44 | 51 |
45 | 68 |
46 | 59 |
47 | 59 |
48 | 38 |
49 | 66 |
50 | 79 |
51 | 111 |
52 | 69 |
53 | 68 |
54 | 110 |
55 | 69 |
56 | 62 |
57 | 91 |
Solution :
d)
H0: The median difference is zero versus
H1: The median difference is not zero α=0.05
Wilcoxon's signed-rank test statistic is given by
W = W+ - W-
where,
W+ = sum of the observations with positive ranks
W- = sum of the observations with negative ranks
In the given data,
Median = 66
ID | yi | yi-Median |
1 | 51 | -15 |
2 | 60 | -6 |
3 | 75 | 9 |
4 | 43 | -23 |
5 | 92 | 26 |
6 | 72 | 6 |
7 | 49 | -17 |
8 | 39 | -27 |
9 | 62 | -4 |
10 | 127 | 61 |
11 | 51 | -15 |
12 | 75 | 9 |
13 | 69 | 3 |
14 | 59 | -7 |
15 | 25 | -41 |
16 | 58 | -8 |
17 | 95 | 29 |
18 | 63 | -3 |
19 | 91 | 25 |
20 | 63 | -3 |
21 | 32 | -34 |
22 | 50 | -16 |
23 | 108 | 42 |
24 | 41 | -25 |
25 | 93 | 27 |
26 | 43 | -23 |
27 | 74 | 8 |
28 | 50 | -16 |
29 | 55 | -11 |
30 | 60 | -6 |
31 | 62 | -4 |
32 | 91 | 25 |
33 | 79 | 13 |
34 | 71 | 5 |
35 | 85 | 19 |
36 | 86 | 20 |
37 | 78 | 12 |
38 | 100 | 34 |
39 | 146 | 80 |
40 | 62 | -4 |
41 | 134 | 68 |
42 | 41 | -25 |
43 | 40 | -26 |
44 | 51 | -15 |
45 | 68 | 2 |
46 | 59 | -7 |
47 | 59 | -7 |
48 | 38 | -28 |
49 | 66 | 0 |
50 | 79 | 13 |
51 | 111 | 45 |
52 | 69 | 3 |
53 | 68 | 2 |
54 | 110 | 44 |
55 | 69 | 3 |
56 | 62 | -4 |
57 | 91 | 25 |
W+ = 2506
W- = 1428
W = 1078
The approximate normal distribution has an expected value of 0 and a variance of
.
= 63365
e)
p-value = P(|W| > 1078) = 1.848324e-05 ... from normal probability table.
The calculations are as shown below:
f)
As p-value < 0.05, we can reject H0. Hence, the time taken by children with NF1 to solve the puzzle compared to healthy controls is significantly different.
g)
If we have to use no parametric distribution, we can find the 2.5th and the 97.5th quantile to find a 95% confidence interval for the mean time.
Let me know in the comment section if anything is not
clear. I will reply ASAP!
If you like the answer, please give a thumbs-up. This
will be quite encouraging for me.Thank-you!