Biomedical Engineering Reference
In-Depth Information
Table 9.14: Data Sets for Figure 9.24
Data
Set 1
Set 2
Set 3
9.00
21.65
10.69
9.01
7.05
9.01
9.02
9.86
10.71
8.99
7.43
8.99
8.98
10.63
10.67
9.00
13.18
9.00
9.00
0.17
10.69
9.00
0.15
9.00
8.99
0.41
0.20
9.99
20.24
11.86
9.00
21.65
10.69
9.01
7.05
9.01
9.02
9.86
10.71
8.99
7.43
8.99
8.98
10.63
10.67
9.00
13.18
9.00
9.00
0.17
10.69
9.00
0.15
9.00
8.99
0.41
0.20
9.99
20.24
11.86
No. of points
20
20
20
Average
9.1
9.1
9.1
Std Deviation
0.305
7.565
3.195
Error (Eq. 9.5)
0.137
3.383
1.429
MAX
9.40
16.64
12.28
MIN
8.79
1.51
5.89
most error). One would normally assume that the difference between the average of the two
data sets is zero (i.e., they are the same). One then needs to select some choices, for example
equal or unequal variances? If they are the same then the variance should be the same, but to
perform both calculations is easy so it is possible to present both. The final selection is
two-tailed or one-tailed. This means is the average likely to be lower only or higher only
(one-tailed), or could it be both (two-tailed)? The magic item we are looking for is the
p
-
value. If the groups are the same (and we are using 95% confidence) then
p
>0.05; if the
groups are different then
p
<=0.05. The more dissimilar they are the smaller the value of
p
.
Using Microsoft Excel® and using the t-test option in data analysis, we have already seen
that the standard deviations are different, hence the variances are unequal. Performing this
analysis on data sets 1 and 2 yields the results shown in
Table 9.15
.
As can be seen,
p
for the one-tailed assumption is 0.49; clearly the groups are identical.
Hence there is no
significant difference
between the two groups.
Table 9.16
analyzes a
different data set to illustrate significant difference.
Search WWH ::
Custom Search