# Thread: Please help - Assumptions with 2 tail t-tests

1. ## Please help - Assumptions with 2 tail t-tests

Hey new to this forum and looking for some help with this stats question as it is slowly sending me round the bend?

any help would be welcomed!

How effective is washing your hands in eliminating bacteria? A researcher tests four
dierent methods - washing with water only, washing with regular soap, washing with
antibacterial soap, and spraying hands with an antibacterial spray. For 32 mornings she
washed her hands according to one of the four methods (8 mornings with each method).
She then placed her hand on a sterile media plate designed to encourage bacteria growth.
Each plate was incubated for 2 days at 36C, after which the number of bacteria colonies
were counted. The data is shown in the table below.
Alcohol Spray, Antibacterial Soap, Soap, Water
69, 66, 16, 134
18, 112, 106, 145
20, 62, 138, 82
32, 67, 66, 113
59, 95, 84, 71
66, 75, 75, 165
47, 111, 81, 97
41, 24, 120, 158

Using JMP, how do you check whether the assumptions for this test (2 tail t-test) are met???

2. This looks more like a test of equality between 4 treatments.

$\displaystyle H_0:\mu_1=\mu_2=\mu_3=\mu_4$

vs

$\displaystyle H_0:\mu_i\ne\mu_j$ for some $\displaystyle i\ne j$

I would not do a t-test, nor a bonferroni either.
IF you're asking for test of constant variance (that's cochran's test)
or normality (that the kolmogorov-smirnov test)