SOLVED PAPER OF SAMPLING TECHNIQUES 1 :
Section A
Q1. Choose the correct answer
I.
A group of individuals or entities about which you wish to know something and from which
a sample will be taken is called:
(a) Population
II.
A single member of a population is known as:
(c) Both (a) and (b)
(Element and Sampling unit)
III.
Any numerical value calculated from sample is called:
(a) Statistic
IV.
Any numerical value calculated from population is called:
(b) Parameter
V.
The procedure by which a sample is taken from the population is called:
(b) Sampling
VI.
A list of all the sampling units of the population is called:
(b) Sampling frame
VII.
A program layout including all the procedures to take a sample is known as:
(b) Sampling design
Section A – Q2
i. Probability Sampling
A sampling method in which each unit of the population has a known and non-zero
chance of selection.
Example: Simple random sampling, stratified sampling.
ii. Non-Probability Sampling
A sampling method in which selection probabilities are unknown.
Example: Convenience sampling, quota sampling.
iii. Finite Population Correction (FPC) Factor
When sampling without replacement from a finite population, variance is adjusted by:
𝑁−𝑛
FPC = √
𝑁−1
iv. Simple Random Sampling (SRS)
A sampling method where every unit has an equal chance of selection and every
possible sample of size 𝑛 is equally likely.
v. Difference between Sampling Error and Non-Sampling Error
Sampling Error Non-Sampling Error
Due to sample selection Due to data collection
Reduced by larger sample Not reduced by size
Measurable Not measurable
vi. Sampling Frame
A complete list of all population units from which the sample is drawn.
vii. Quota Sampling
A non-probability sampling method where fixed quotas are assigned to different
population groups and units are selected conveniently.
Section B
Q3.
Prove that in SRS, variance of sampling proportion is
𝑁 − 𝑛 𝑃𝑄
𝑉(𝑝) =
𝑁−1 𝑛
Proof:
Let:
• 𝑃 = population proportion
• 𝑄 = 1−𝑃
• 𝑝 = sample proportion
• 𝑛 = sample size
• 𝑁 = population size
Variance of sample mean under SRSWOR:
𝑁 − 𝑛 𝑆2
𝑉(𝑦‾) =
𝑁−1 𝑛
For binary variable:
𝑆 2 = 𝑃𝑄
Hence,
𝑁 − 𝑛 𝑃𝑄
𝑉(𝑝) =
𝑁−1 𝑛
Proved
Q4. Allocation of Sample Size in Stratified Random Sampling
Definition
Allocation refers to how total sample size 𝑛 is distributed among strata.
(a) Equal Allocation
𝑛
𝑛ℎ =
𝐿
Where:
• 𝐿 = number of strata
Advantage: Simple
Disadvantage: Ignores stratum size and variability
(b) Neyman Allocation
𝑊ℎ 𝑆ℎ
𝑛ℎ = 𝑛
∑𝑊ℎ 𝑆ℎ
Where:
𝑁ℎ
• 𝑊ℎ = 𝑁
• 𝑆ℎ = standard deviation of stratum
Advantage: Minimum variance
Best allocation method
Q5.
Prove that 𝑉(𝑦‾𝑠𝑡 ) is minimum when 𝑛ℎ ∝ 𝑊ℎ 𝑆ℎ
Proof:
Variance of stratified mean:
𝑊ℎ2 𝑆ℎ2
𝑉(𝑦‾𝑠𝑡 ) = ∑
𝑛ℎ
Using Lagrange’s method under constraint:
∑𝑛ℎ = 𝑛
Differentiation gives:
𝑛ℎ ∝ 𝑊ℎ 𝑆ℎ
Thus,
𝑊ℎ 𝑆ℎ
𝑛ℎ = 𝑛
∑𝑊ℎ 𝑆ℎ
Hence proved that Neyman allocation minimizes variance
Q6.
(a) Ratio Estimator
Ratio estimator is used when study variable 𝑌 is highly correlated with auxiliary variable
𝑋.
𝑦‾
𝑅̂ =
𝑥‾
Population total estimator:
𝑌̂𝑅 = 𝑅̂ ⋅ 𝑋‾
Importance
• Improves precision
• Reduces variance
• Useful in surveys with auxiliary data
(b) Bias of Ratio Estimator
Approximate bias of ratio estimator:
1 𝑆𝑥2 𝑆𝑥𝑦
𝐵𝑖𝑎𝑠(𝑅̂ ) ≈ ( − )
𝑛 𝑋‾ 2 𝑋‾𝑌‾
Upper bound of bias relative to standard error:
𝐵𝑖𝑎𝑠(𝑅̂ ) 1
| |≤
𝑆𝐸(𝑅̂ ) √𝑛
Thus, bias decreases as sample size increases.