Scribd
Upload
5 views1 page

STAT 443: Sampling Theory Assignment

The document is an assignment for the course STAT 443: Theory of Sampling at the University of Ghana, due by November 17, 2023. It includes tasks such as calculating population statistics, performing simple random sampling, estimating population parameters, and demonstrating specific variance properties. The assignment is structured to enhance understanding of sampling theory through practical application.

Uploaded by

Bless Tetteh
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
5 views1 page

STAT 443: Sampling Theory Assignment

The document is an assignment for the course STAT 443: Theory of Sampling at the University of Ghana, due by November 17, 2023. It includes tasks such as calculating population statistics, performing simple random sampling, estimating population parameters, and demonstrating specific variance properties. The assignment is structured to enhance understanding of sampling theory through practical application.

Uploaded by

Bless Tetteh
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

UNIVERSITY OF GHANA

(All rights reserved)


DEPARTMENT OF STATISTICS AND ACTUARIAL SCIENCE
STAT 443: THEORY OF SAMPLING (3 credits)

ASSIGNMENT (To be submitted latest by 4 pm, November 17, 2023)

1. Consider the artificial population P = {160, 220, 850, 510, 110}.

i. Calculate the population total, population mean and population variance.

ii. Using simple random sampling without replacement, list all possible samples of size
n = 4 with their corresponding probabilities of occurrence.

iii. With the aid of these samples, estimate the population mean and population total
and compare your outcomes with those in (i) above.

iv. Show that:


h i
a. V ar(ȳ) = E V ar (ȳs ) where ȳts is the mean of sample s.
d
h i
b. V ar(Ŷt ) = E Vd
ar (ŷs ) where ŷts is the estimated total of sample s.

Dr. Onsongo W.M Page 1 of 1

Common questions

Powered by AI

Using simple random sampling without replacement, we list all combinations of 4 items from 5, which results in 5 possible samples: (160, 220, 850, 510), (160, 220, 850, 110), (160, 220, 510, 110), (160, 850, 510, 110), (220, 850, 510, 110). The probability of selecting any specific sample is 1 divided by the total number of samples, which is 5C4 = 5. So, each sample has a probability of 0.20 or 20%.

The variance of estimated totals, Var(ˆYt), is verified by calculating it as the expected variance of the estimated totals from samples, E [Var(ˆys)]. Similar to variance of sample means, this involves calculating the total estimates' variance within each sample and taking the average across all samples to show the equivalency. Such verification ensures the predictive validity of estimation procedures grounded in simple random sampling.

Variance of sample means, Var(¯y), represents the variability of the means of all possible samples. It is demostrated by showing Var(¯y) equals the expected value of the variances of the sample means, E [Var(¯ys)], by calculating the variance within each sample and averaging these variances across all samples. This confirms theoretically that the observed variance among sample means can predict the true population variance if sampling is random and unbiased.

The population total is found by summing all the elements of the population. Total = 160 + 220 + 850 + 510 + 110 = 1,850. The population mean is this total divided by the number of elements, which is 5. Mean = 1,850 / 5 = 370. The population variance is calculated by finding the average of the squared differences from the Mean. Variance = [(160 - 370)^2 + (220 - 370)^2 + (850 - 370)^2 + (510 - 370)^2 + (110 - 370)^2] / 5 = 74,080/5 = 14,816.

For each of the 5 samples, calculate the sample mean and total. For example, (160, 220, 850, 510): mean = (160 + 220 + 850 + 510) / 4 = 435, total = 1640. Compute the overall sample mean by averaging the means of all samples, and the overall sample total by averaging the totals. Then compare with the actual population mean of 370 and the total of 1,850. The sample estimates will vary but should be unbiased, providing an average close to the actual mean and total.

You might also like