Lecture Notes on Data Sampling and Bias

Sampling Methods

1. Simple Random Sampling

   • Divide into proportional groups.
   • Example: Drawing names from a hat.

2. Systematic Sampling

   • Select every kth individual from a population list.
   • Example: Choosing every 10th person in a list of participants.

3. Stratified Sampling

   • Divide population into strata/groups and sample from each group.
   • Example: Selecting students from different classrooms.

4. Cluster Sampling

   • Choose clusters randomly and sample everyone within the cluster.
   • Example: Selecting entire classrooms from a school.

5. Convenience Sampling

   • Sampling from a group that is easy to reach.
   • Example: Surveying people at a shopping mall.

6. Voluntary Response Sampling

   • Allowing individuals to volunteer for the sample.
   • Example: Online surveys where participants choose to respond.

Variables and Measurement

• Qualitative Variables
  • Flavour of food
  • Favourite colour
• Quantitative Variables
  • Speed of a car (Continuous)
  • Number of pets (Discrete)
  • A person's weight (Continuous)
  • Buildings in a neighbourhood

Types of Bias and Avoidance Strategies

1. Selection Bias

   • Ensure random selection to cover the entire population.

2. Response Bias

   • Wording of questions should not influence responses.

3. Non-response Bias

   • Follow up with non-respondents to get comprehensive data.

4. Voluntary Response Bias

   • Avoid surveys where only those interested will respond.

Sampling Example

• Stratified Random Sampling Example
  • Example Group: A1, A3, A4, D2, E3
• Cluster Random Sampling Example
  • Example Group: Choose one group and include all members.

Systematic Random Sample Example

• Class of Thirty Students (Labels S01 to S30)
  • Die Roll: 2
  • Sample: S02, S08, S14, S20, S26

Fast-Food Restaurant Survey Bias

• Sources of Bias

  • Survey Location Bias: Survey conducted at the restaurant might skew results to favor the restaurant's menu.
  • Question Wording Bias: Question may be phrased to lead respondents towards a particular answer.

• Effect on Percentage

  • These biases may inflate the percentage of people choosing the Big Burger Combo as their favorite meal.