In: Statistics and Probability
A survey of 100,000 college graduates and 100,000 non-college graduates is conducted to determine how many of them own stock. 49,781 of the non-college graduates own stock vs. 50,219 for the college graduates. Analyze for a relationship between education and owning stock using a chi-square hypothesis test with alpha=1%.
Data Summary
Observed Frequencies (O) | College | Non-college |
Own stock | 50219 | 49781 |
Do not own stock | 49781 | 50219 |
The null and alternative hypotheses are
Ho : Education and Owning stock are
independent
Ha : Education and Owning stock are
dependent
Level of significance α =
0.01
We use Chi Square test of independence
Grand Total of frequencies = 200000
To find Expected Frequencies
Expected Frequencies = (Row Total * Column Total)/Grand
Total
Expected Frequencies (E) | College | Non-college |
Own stock | 50000 | 50000 |
Do not own stock | 50000 | 50000 |
Following table gives the value of (Observed - Expected)² / Expected
College | Non-college | |
Own stock | 0.9592 | 0.9592 |
Do not own stock | 0.9592 | 0.9592 |
Chi Square Value = ∑[(Observed - Expected)² /
Expected]
χ² = 3.8369
Degrees of freedom = df = (number of rows - 1) * (number of columns
- 1)
Degrees of freedom = df =
1
From chi square table or CHISQ.DIST.RT(X, df) function of
Excel
we find the p-value
p-value = CHISQ.DIST.RT( 3.8369, 1) =
0.0501
p-value = 0.0501
Decision :
0.0501 > 0.01
that is p-value > α
Hence we DO NOT REJECT Ho
Conclusion
:
There does not exist enough statistical evidence at α =
0.01 to show that Education and Owning stock are
dependent.
that is
There is no relationship between Education and Owning
stock.