In: Statistics and Probability
.
A yield improvement study at a semiconductor manufacturing facility provided defect data for a sample of 450 wafers. The following table presents a summary of the responses to two questions” Were particles found on the die that produced the wafer?” and “is the wafer good or bad?”
CONDITION OF DIE |
|||
Quality of wafer |
No particles |
Particles |
Totals |
Good |
320 |
14 |
334 |
Bad |
80 |
36 |
116 |
Totals |
400 |
50 |
450 |
Show working
(a)
P(Particles/ Bad) = P(Particles AND Bad)/ P(Bad)
= 36/116
= 0.3103
So,
Answer is:
0.3103
(b)
P(Particles/ Good) = P(Particles AND Good)/ P(Good)
= 14/334
= 0.0419
So,
Answer is:
0.0419
(c)
P(Good) = 334/450 = 0.7422
P(No particles) = 400/450 = 0.8889
So,
P(Good) X P(No particles) = 0.7422 X 0.8889 = 0.6597
But
P(Good AND No particles) = 320/450 = 0.7111
Since P(Good) X P(No particles) = 0.6597 P(Good AND No particles) = 0.7111, the two events a good wafer and a die with no particles, are not statistically independent