source SS
df MS
F
P
Groups
31.610 5
freq
14.111 2
7.056 8.471
P < 0.01
size
9.389 1 9.389
11.271
P < 0.01
fr x size
8.110 2 4.055
4.868
P < 0.05
Resid
10.001 12
0.833
Total
41.61 17
Ho1:
no effect on hygiene due to frequency
F(calc):
MS(fr) / MS(res) = 8.471
F(0.01)(2,12): 6.69 Reject Ho1
(P<0.01)
Ho2:
no effect on hygiene due to size
F(calc):
MS(size) / MS(res) = 11.271
F(0.01)(1,12): 9.33 Reject
Ho2 (P<0.01)
H03:
no interaction between frequency and size in terms of hygiene
F(calc):
MS(frXsize) / MS(res) = 4.868
F(0.05)(2,12): 4.75 Reject Ho3
(P<0.05)
Interaction is present. Effects of the two main factors will
each be described in detail for each of the levels of the other factor.
Frequency: Overall, this was found to have a statistically
significant (P<0.01) effect upon hygiene score. Low freq had lowest
hygiene scpore (mean: 6.165) and high freq the highest/best hyg score (mean:
8.34). Within each the size categories, however, the freq effect was
different. In small restaurants, the best score was in med freq
premises (7.0), with higher and lower freq premises showing lower scores
(6.0 and 6.3 respectively). In Large size restaurants the high freq
premises achieved the best score of 10, med freq scored 7,7 and low freq the
worst score at 6.7
Size:
Overall, the effect was statistically significant (P<0.01).
Small premises showed a mean of 6.67 and large premises a mean of 8.0.
Within the freq categories, however, the size effect varied. Low freq:
large had very slightly greater score than small (6.3. v 6.0). Med
freq: large had slightly higher than small (7.7 v 7.0). But in high freq
premises, the Large prem achieved the highest score (10.0) whereas small
prem recorded only 6.7)
Clearly larger
premises appear to benefit from increased inspection frequency, but this
effect was not as evident among small premises.