Вы находитесь на странице: 1из 1

Correlation Analysis (Pearson Product Moment Correlation)

Initial wound size (no. of pixels) vs no. of days for the wound to completely heal.
To normalize the data for comparison, all the data were log transformed

Raw Data

3 % Ointment 5 % Ointment Positive Control Negative Control


Size Days Size Days Size Days Size Days
97257 14 81181 16 63650 16 104134 20
149888 14 80349 14 75940 22 114407 30
78997 16 66093 18 69346 16 93069 22
94264 18 96877 16 78840 20 66634 20
80388 16 102874 14 54841 16 124971 20

Log transformed data

3 % Ointment 5 % Ointment Positive Control Negative Control


Size Days Size Days Size Days Size Days
4.99 1.15 4.91 1.20 4.80 1.20 5.02 1.30
5.18 1.15 4.90 1.15 4.88 1.34 5.06 1.48
4.90 1.20 4.82 1.26 4.84 1.20 4.97 1.34
4.97 1.26 4.99 1.20 4.90 1.30 4.82 1.30
4.91 1.20 5.01 1.15 4.74 1.20 5.10 1.30

Analysis Results

Treatment r p-value
3 % Ointment -0.5586 0.328
5% Ointment -0.6551 0.230
Positive Control 0.7723 0.126
Negative Control 0.3251 0.593

None of the treatment exhibited significant correlations. In the experiment, the size of the wound did
not significantly influence the no. days for the wound to completely heal.

Although 3 of r values are > than 0.5 indicating a good correlation, however it is not significant
statistically (at 95% confidence level or alpha=0.05).

r values approaching 1 indicates a direct relationship (e.g. as the size of the wound increases, the no. of
days it will take to completely heal also increases), as r approaches -1 it indicates an inverse relationship
(e.g. as the size of the wound increases, the no. of days it will take to completely heal decreases) and as
it approached 0, it indicates no correlation.

Вам также может понравиться