In: Statistics and Probability
conduct your own example of a related-samples t-test.
IV:
level 1:
level 2:
DV:
Let's conduct a paired-samples or related sample t-test. We need to test whether eating corn flakes lowered the cholesterol levels. 20 individuals were assigned a diet randomly that was either oats of corn flakes. After a period of one week on the initial diet, the levels were measured and the other diet was given. After another week, the levels were measured.
IV: Type of Diet
Level 1: Oats
Level 2: Corn Flakes
DV: Cholesterol levels
The data is:
ID | Corn (mmol/L) | Oats (mmol/L) |
1 | 6.84 | 4.60 |
2 | 7.78 | 4.97 |
3 | 4.78 | 5.69 |
4 | 4.35 | 7.64 |
5 | 7.96 | 3.96 |
6 | 3.85 | 7.85 |
7 | 6.80 | 7.53 |
8 | 6.44 | 6.45 |
9 | 7.16 | 7.51 |
10 | 5.00 | 3.39 |
11 | 3.38 | 6.63 |
12 | 3.26 | 5.76 |
13 | 4.25 | 4.10 |
14 | 3.45 | 3.26 |
15 | 7.83 | 3.85 |
16 | 6.55 | 6.83 |
17 | 5.28 | 4.16 |
18 | 5.67 | 3.33 |
19 | 3.85 | 5.76 |
20 | 3.73 | 6.42 |
The hypothesis is:
Ho: ud = 0
Ha: ud > 0
The calculations are:
ID | Corn (mmol/L) | Oats (mmol/L) | Difference, d = Corn - Oats |
1 | 6.84 | 4.60 | 2.24 |
2 | 7.78 | 4.97 | 2.81 |
3 | 4.78 | 5.69 | -0.90 |
4 | 4.35 | 7.64 | -3.29 |
5 | 7.96 | 3.96 | 3.99 |
6 | 3.85 | 7.85 | -4.00 |
7 | 6.80 | 7.53 | -0.73 |
8 | 6.44 | 6.45 | 0.00 |
9 | 7.16 | 7.51 | -0.35 |
10 | 5.00 | 3.39 | 1.60 |
11 | 3.38 | 6.63 | -3.25 |
12 | 3.26 | 5.76 | -2.50 |
13 | 4.25 | 4.10 | 0.15 |
14 | 3.45 | 3.26 | 0.18 |
15 | 7.83 | 3.85 | 3.98 |
16 | 6.55 | 6.83 | -0.28 |
17 | 5.28 | 4.16 | 1.12 |
18 | 5.67 | 3.33 | 2.33 |
19 | 3.85 | 5.76 | -1.91 |
20 | 3.73 | 6.42 | -2.69 |
Average | -0.07 | ||
Standard Deviation | 2.41 |
t-statistic = (u - u0)/(s/√n)
= (-0.07 - 0)/(2.41/√20)
t = -0.13
p-value (one-tailed) at df = 20 - 1 = 19 is: 0.449
As the p-value = 0.449 > 0.05, we fail to reject Ho.
We conclude that we do not have sufficient evidence to say that eating corn flakes lowered the cholesterol levels significantly