Oct 10 - Sample plan

Question 1

Why do we take Samples?

Answer 1

Sampling errors: any errors caused by using a sample
Sample Frame: a master index of the whole population
Sample Frame errors: omission, not updated

Feasible; integrate just a small part of a group to infer assumptions on a population

Sample Frame <=> Sample screening - a master index of the whole population

BUT
Sampling errors
Sample Frame errors: Omissions out of date data that leads to choice of a defective sample

Question 2

Probability Sampling Methods: Objective Technique

Simple random Sampling

Answer 2

Sample size / Population size
- Everybody has the same ads of being picked

E.g random digit calling

Question 3

Simple random Sampling Adv & Dis

Answer 3

Known and equal chance of selection

Dis: Pre-designation and complete listing required

Question 4

Probability Sampling Methods: Objective Technique

-Systematic Sampling:

Answer 4

Efficient as we use a skip interval (if SI = 10, 10th, 20th…)
SI = Pop size / Sample Size

Question 5

Probability Sampling Methods: Objective Technique-

Cluster Sampling:

Answer 5

Dividing the population into identical several groups (clusters) that are representative of a population

One step approach - 1 cluster
Two Step Approach - 2+ clusters

Question 6

Probability Sampling Methods: Objective Technique

-Stratified sampling:

Answer 6

Divide population into 2 or 3 smaller groups

Good for not normally distributed groups

Question 7

Why is stratified sampling more accurate?

Answer 7

Analysis of each stratum

Estimation of the overall sample mean by use of weighted means

Question 8

Estimation of the population mean by using the stratified sampling approach

Answer 8

mean pop= mean a *proportion a + mean b *proportion b

Question 9

Non probability sampling; Subjective techniques

Answer 9

1) Convenience Samples (convenient to researcher, not particularly relevant)
2) Judgement (purposive) samples: educated guess (from biz experience)
3) Referral samples (referred by previous respondents - good for limited sample frames)
4) Quota samples: Specific quota determined by research objective (most used)

Question 10

Steps in the sampling process

Answer 10

1) Define the target population
2) Obtain a list of the population as a sample frame if available
3) Design the sample plan
4) Access the population
5) Draw the sample: Need to have contingency plans
Substitution methods : Oversampling - Resampling - Drop downs
6) Validate the sample
7) Resample, if necessary

Question 11

Probability Sampling Methods: Objective Technique

Answer 11

Simple random sampling
Systematic Sampling
Cluster Sampling
Stratified Sampling