In: Statistics and Probability
there are 200 families of parents heterozygous for albinism, which present normal pigmentation. They all had 4 children, where 60 families with 4 children were found with normal pigmentation, 58 families with 3 children with normal pigmentation, 57 families with 2 children with normal pigmentation, 16 families with 1 child with normal pigmentation and 9 families with 4 albino children. Using the binomial expansion determine the theoretical probability of the possible 5 combinations. Using chi 2, prove that the distribution obtained from people without the condition and with albinism in the 200 families is consistent with what was expected.
Consider X denotes number of children with normal pigmentation.
X | No. of Children (f) | x*f |
0 | 9 | 0 |
1 | 16 | 16 |
2 | 57 | 114 |
3 | 58 | 174 |
4 | 60 | 240 |
total | 200 | 544 |
The random variable X follows binomial distribution with n = 4 and p.
where p =p ( child with normal pigmentation)
for binomial distribution E(X) = mean = np
mean = 544 / 200 = 2.72.
Hence estimated value of p is
phat = 2.72 / 4 = 0.68.
The probability mass function of X is given by
and Expected frequency = N * P(X=x).
Here N = total number of families = 200.
X | P(X=x) | Expected Freq. = N*P(X=x) |
0 | 0.0105 | 2.0972 |
1 | 0.0891 | 17.8258 |
2 | 0.2841 | 56.8197 |
3 | 0.4025 | 80.4946 |
4 | 0.2138 | 42.7628 |
Total | 1.0000 | 200.0000 |
We have to test the hypothesis that
Whether or not distribution obtained from people without the condition and with albinism in the 200 families is consistent?
i.e. The given frequency distribution is good fitted as binomial distribution or not?
We used Chi-square goodness of fit test.
The value of chi-square test statistic is
Where Oi : Observed Frequency , Ei = Expected frequency and k = number of classes = 5.
Oi | Ei | (Oi-Ei)^2/Ei |
9 | 2.0972 | 22.7201 |
16 | 17.8258 | 0.1870 |
57 | 56.8197 | 0.0006 |
58 | 80.4946 | 6.2862 |
60 | 42.7628 | 6.9481 |
200 | 200.000 | 36.1420 |
The value of chi-square test statistic is 36.1420
at 5% level of significance Critical value is
Since calculated value of test statistic is greater than critical value( 36.1420 > 9.488).
We reject Ho at 5% level of significance.
Conclusion : The distribution obtained from people without the condition and with albinism in the 200 families is not consistent.