Methods
By Tom Methven
• Results: Person 4
Person 5
7.93
7.19
4.27
5.50
Person 6 5.73 5.22
Person 7 8.40 4.09
Person 8 5.88 4.46
Person 9 5.60 4.00
Person 10 4.89 4.90
Mean: 6.03 4.61
Randomisation and Ordering Effects
• Performing an experiment gives experience
which could improve subjects’ ‘scores’
• Standard Deviation:
𝑠= 𝑠2
• Standard Error:
𝑠
𝑠𝑒 =
𝑛
Degrees of Freedom
• For sample populations, often ‘N – 1’ is used
rather than N. This is the simplest calculation
of D.O.F., but it can get very complex.
𝐷 = Average difference
𝜇𝐷 = Expected difference (0 for null hypothesis)
𝑆𝐷 = Standard deviation of differences
𝑁 = Sample Size
Worked Example T Result
𝐷 = 1.420756421
𝑆𝐷 = 1.985348881
𝑁 = 10
t-value = 2.26
Interpreting T-Value
• The T-Test gives a t-value, which can be
converted to a p-value
• If the p-value < 0.05 there is less than a 5%
chance the null hypothesis is true
• Therefore we conclude that the change of
condition caused the observed effect
• Our p-value was 0.025, so we can accept our
alternate hypothesis: ‘Wii Piano is easier to
learn’
Effect Size
• Tells us how important the result is in practical
terms. Gives an ‘r-value’.
𝑡2
𝑟=
𝑡 2 + 𝑑𝑓
6.5
5.5 Mean 1
Mean 2
5
4.5
4
0 1 2 3