Consider whether data distribution skewness exists in some of the numeric variables, and if it does, determine how to transform the data into a less skewed distribution. Perform data sub-setting and use simple summary measures such as averages and frequency counts to find out if any differences exist across subsets.

Tower	Model	Damage
1	A	Minor
2	A	None
3	A	None
4	A	None
5	A	None
6	B	Minor
7	B	None
8	A	Partial
9	A	Minor
10	A	None
11	B	Minor
12	B	None
13	A	Minor
14	A	None
15	A	Severe
16	A	Minor
17	A	None
18	A	Minor
19	A	Partial
20	A	None
21	A	Partial
22	A	Minor
23	A	None
24	B	None
25	A	None
26	A	None
27	A	Minor
28	A	None
29	B	Partial