You have a project where you need to predict the expected number of defects that will be found after your system is in operation for 3 months.You have data from 15 other similar systems, so you decide to analyze it to see if it tells you anything. The data is below.

Project	Lines of Code (KLOC)	Defects found
1	2	6
2	5	8
3	10	22
4	15	25
5	15	20
6	22	35
7	24	34
8	38	88
9	66	122
10	77	200
11	100	350
12	88	200
13	24	50
14	100	200
15	189	300

Questions:

You expect your system to be 50K LOC in size.

1) Analyze the data above to understand the relationship (or not) between defects and size. Use scatter diagrams, correlation, Box and whisker, and trendline techniques.

2) Assuming there is, what do you predict as the expected number of defects for your 50K program?

3) You now define a metric called defect density (defects per KLOC).

1. Plot using a histogram.
2. Determine the mean, percentiles and standard deviation.
3. Plot it.

4) If you assume a normal distribution for defect density (defects per KLOC), what would be the expected range of number of defects for your 50K system, 95% of the time?

5) Do you think a normal distribution is reasonable or not for the data given?