• Cost: Surveying entire population is prohibitively expensive
• Time: Would take too long to reach everyone
• Practicality: Logistically impossible for large populations
• Accuracy: Well-designed sample can be very accurate

Goal: Select a sample that is representative of the whole population so findings can be generalized.

General principle: Larger samples = More accurate, but also more expensive and time-consuming

Typical sample sizes:

• Qualitative research: 10-100 participants
• Quantitative research: 100-1,000+ participants
• National surveys: 1,000-2,000 considered statistically valid

Factors affecting sample size:

• Budget available
• Time constraints
• Required accuracy level
• Population diversity (more diverse = larger sample needed)

1. Simple Random Sampling

Definition: Every member of the population has an equal chance of being selected

Method:

• Create complete list of population (sampling frame)
• Use random number generator or lottery method to select
• Each selection is independent

Example: Put all customer names in a database, use software to randomly select 200

Advantages:

• Unbiased and representative
• Simple to understand
• Statistical validity

Disadvantages:

• Requires complete population list
• May miss important subgroups by chance
• Can be time-consuming

2. Systematic Sampling

Definition: Select every nth person from the population list

Method:

• Calculate sampling interval: \( k = \frac{\text{Population size}}{\text{Sample size}} \)
• Randomly select starting point between 1 and k
• Select every kth person thereafter

Example: Need 100 people from population of 1,000. \( k = \frac{1000}{100} = 10 \). Start at random person #7, then select every 10th person: 7, 17, 27, 37...

Advantages:

• Easier than simple random
• Spreads sample across population
• Still statistically valid

Disadvantages:

• Risk of periodicity (if list has hidden pattern)
• Still requires complete list

3. Stratified Sampling

Definition: Divide population into subgroups (strata) based on shared characteristics, then randomly sample from each stratum

Method:

• Identify important characteristics (age, gender, location)
• Divide population into strata
• Sample proportionally from each stratum

Example: Population is 60% female, 40% male. Sample of 200 should be 120 female, 80 male

Advantages:

• Ensures representation of all subgroups
• More accurate than simple random
• Can compare subgroups

Disadvantages:

• Requires detailed population information
• More complex and time-consuming
• Must identify relevant strata

4. Cluster Sampling

Definition: Divide population into clusters (usually geographic), randomly select clusters, then survey everyone in selected clusters

Method:

• Divide population into clusters (e.g., neighborhoods, schools)
• Randomly select some clusters
• Survey all members of selected clusters

Example: Research students in Dubai. Select 10 random schools, survey all students in those schools

Advantages:

• Cost-effective for geographically spread populations
• Faster data collection
• Don't need complete population list

Disadvantages:

• Less accurate than other methods
• Clusters may not represent whole population
• Higher sampling error

1. Convenience Sampling

Definition: Select participants who are easily accessible or available

Examples:

• Survey people walking by in shopping mall
• Ask friends and family
• Use volunteers
• First 100 website visitors

Advantages:

• Quick and cheap
• Easy to implement
• Useful for pilot studies

Disadvantages:

• Highly biased
• Not representative
• Cannot generalize results

2. Quota Sampling

Definition: Select specific number (quota) from each subgroup based on their proportion in population, but selection within quota is non-random

Method:

• Set quotas for each subgroup (e.g., 60 women, 40 men)
• Researcher selects any individuals meeting quota criteria
• Continue until all quotas filled

Example: Market researcher in mall stops people until quotas met: 30 men aged 18-25, 30 women aged 18-25, etc.

Advantages:

• Ensures subgroup representation
• Faster and cheaper than stratified random
• No need for complete population list

Disadvantages:

• Researcher bias in selection
• Not statistically random
• Cannot calculate sampling error

3. Judgment/Purposive Sampling

Definition: Researcher deliberately selects participants based on expertise or specific knowledge

Examples:

• Interviewing industry experts
• Selecting loyal customers for feedback
• Choosing opinion leaders

Best for: Qualitative research, specialized topics

4. Snowball Sampling

Definition: Existing participants recruit future participants from their networks

Method:

• Start with few participants
• Ask them to refer others with similar characteristics
• Sample grows like snowball rolling downhill

Best for: Hard-to-reach populations (e.g., rare medical conditions, specific professions)