00:01
It is given that for 1 ,000 companies based on their ratings, they are grouped as a, b, c, d and e.
00:13
A as the top 20 percentage of companies and b as the next 20 percentage of companies.
00:22
Similarly, e as the last 20 percentage of companies and a sample of 60 % of companies.
00:30
Largest companies are taken and the number of companies based on ratings are 5 ,7, 15, 20 and 13.
00:45
The alpha level is given as 5 percentage or 0 .05.
00:52
Now we have to find that if the sampled largest companies differ, by their performance.
01:05
The nell hypothesis is such that pfa equal to pfb equal to pfc equal to pfc equal to pfd equal to pfe of e and the alternative hypothesis is such that at least one of the largest companies which are sampled differ by their performance and kai square goodness of it is used to test this null hypothesis.
01:35
Let us represent observed values as oi and the expected values as ei.
01:41
It is given that each ratings has 20 percentage of companies.
01:46
This implies that for a sample of 60, the 20 percentage which is 60 into 0 .2, that is 12.
01:55
Now the expected value for each of the ratings are 12 companies.
02:01
The test statistics for the kai square goodness of it is summation oi minus ei, the old square divided by ei.
02:14
It is known as for the ratings of a, b, c, d and e.
02:23
The observed values are 5, 7, 15, 20 and 13...