Professional Documents
Culture Documents
AB1202 19202 Quiz2 Prep Special Tutorial (Solns) Jan'21nokmcl
AB1202 19202 Quiz2 Prep Special Tutorial (Solns) Jan'21nokmcl
AB1202 19202 Quiz2 Prep Special Tutorial (Solns) Jan'21nokmcl
1. 2.
4.
3. 0.08196
5. There is an important difference between this question & Q2, for which sigma is known => z
For this question, sigma is NOT known => t => qt => need to specify df (n-1).
So, we are trying to find n using qt, but, we need to specify an n for the qt!
Chicken-and-egg situation! That explains the need for iterations.
The first iteration will be based on the sample of 50.
Hence, for the first iteration, we use n = 50 (df = 49) for qt.
You will get 101, which means you need to “top-
up” your sample size by an additional 51 units.
Iterate using n = 101 ...
Those who are R-savvy, can use the conditional while loop to iterate:
e=1
i=49
while (e>0.001) {e=qt(0.995,i)*0.00375/sqrt(i+1)
print(e)
print(i)
i=i+1}
print(i)
7.
8.
9.
R computes variances and covariances for samples only. When using R to compute
these population parameters, please remember to apply the correction factor.
Note however that this is beyond the scope of AB1202.
FAQ: How is the correction factor obtained?
In the formula for sample variance, the denominator is n-1.
But, in the formula for POPN variance, the denominator is N.
R will always take your data as a sample.
So, when you give it 10 data-points, it will use 9 in the denominator.
But, in my Q7, the 10 data-points in the question constitute the popn.
So, the denominator should be 10 instead of 9!
That’s why the correction factor is 9/10.
In general, the correction factor will be (n-1)/n.
y x
300 200
300 100
400 200
300 200
300 200
300 100
300 200
400 100
300 200
300 200
µ 320 170
σ 40.00 45.83
s 42.16 48.30
σ2 1600 2100
s2 1778 2333
___________________________________________________________________________
ch.Tan@NBS Jan’21