Marketing Research - Designing the Sample Plan

Business Marketing Research - Chapter 10: Designing the Sample Plan

Learning Outcomes

  • Distinguish between sampling and a census and explain why sampling is preferred.
  • Discuss the steps in the sampling process.
  • Describe probability and non-probability sampling methods.
  • Explain how to calculate sample size for representativeness.
  • Develop a sample plan for a research project.

Introduction

  • Sampling involves deciding whom to interview and how many people.
  • The goal is to draw conclusions about the entire population cost-effectively and efficiently.

Sampling

  • A sample is a subset of a population.
  • A population is the total group of people or entities from whom information is required.
  • Sampling involves identifying a group within the population and contacting them to obtain information.
  • Successful sampling requires:
    • Clearly defining the population.
    • Defining the sample frame.
    • Selecting a sampling strategy.
    • Determining the sample size.
    • Selecting sample elements.

Sampling versus Census

  • A census obtains data from the entire population.
  • Sampling is often preferred because:
    • It can be more accurate.
    • It takes less time.
    • It costs less.
    • It's more practical.

Basic Concepts

Definitions
  • Census Survey: Obtains information about the characteristics of the entire population.
  • Extent: Geographic region defining the population boundary.
  • Generalization: Using statistical inference to draw conclusions about the population from sample data.
  • Nonprobability Samples: Units selected non-randomly; likelihood of inclusion is unknown.
  • Overall Efficiency: Combination of statistical efficiency and cost.
  • Population: Aggregate of all units of analysis (target and survey populations).
  • Precision: Confidence that statistical generalizations match sample evidence.
  • Probability Sample: Each unit has a known positive probability of selection.
  • Random Sampling: Every element has the same probability of being selected, independent of others.
  • Representative Sample: A true reflection of the population in all aspects.
  • Sample: A selection of elements from the population.
  • Sample Elements: Individuals from whom information is required.
  • Sample Error: Difference between population value and sample value.
  • Sample Frame: The actual list of sample units from which the sample is drawn.
  • Sample Size: The number of respondents selected for the sample.
  • Sample Units: Items studied in a specific survey.
  • Sample Variance: Variation in the estimated value of a population.
  • Sampling Methods: Probability or non-probability methods for obtaining representative samples.
  • Sampling Portion: Number of respondents in the sample relative to the population size.
  • Sampling: Process used to draw the sample from the population.
  • Standard Error: Indication of the reliability of the population parameter estimate.
  • Statistical Efficiency: Standard for comparing sampling methods.
  • Survey Variable/Population Parameter: Characteristic of the population being investigated.

The Steps in the Sampling Process

  1. Define the Population
  2. Identify the Sampling Frame
  3. Select the Sampling Methods
  4. Determine the Sample Size
  5. Select the Sample Elements
  6. Gather Data from Designated Elements

Step 1: Define the Population

  • Population: Comprehensive number of individuals, units, or items for observation.
  • Can consist of individuals, households, families, businesses, etc.
  • Must be clearly defined in terms of sample unit, sample element, extent, and time.
    • Sample Unit: Basis for sampling.
    • Sample Element: Unit from which information is required (e.g., individuals).

Step 2: Identify the Sampling Frame

  • A sample frame is a list of all sample units available for selection.
  • A reliable sample frame must:
    • Represent all elements and strata of the population.
    • Be up to date.
    • Have complete and correct details.
    • Have no duplication of entries.
    • Be accessible and well-organized.
    • Ideally contain additional information for stratification.
Shortcomings of Sample Frames
  • Common issues:
    • Missing sample units.
    • Duplicate entries.
    • Foreign elements (units not part of the population).
Various Types of Sample Frames
  • Computerized registers
  • Address directories, buyer’s guides, and yearbooks
  • Membership lists
  • Telephone directories
  • Records of local authorities

Step 3: Selecting the Sampling Methods

  • Two major categories:
    • Probability Sampling:
      • Each unit has a known positive probability of being selected.
    • Non-Probability Sampling:
      • Probability of selection is unknown.
      • Based on researcher's judgment.
      • Less time-consuming, convenient, and less expensive.
Key Differences
  • Non-probability: Cannot give indication of bias or error margins.
  • Probability: Sample error can be estimated statistically.
  • Non-probability methods do not allow for generalization outside the sample.
Classification of Sampling Methods
  • Sampling Methods
    • Probability Sampling Methods
      • Simple Random Sampling
      • Systematic Sampling
      • Stratified Sampling
      • Cluster Sampling
    • Non-Probability Sampling Methods
      • Convenience Sampling
      • Judgment Sampling
      • Snowball Sampling
      • Quota Sampling
Non-probability Sampling Methods
  • Convenience Sampling: Sample drawn from a readily accessible part of the population.
  • Judgment Sampling: Sample elements selected subjectively by the researcher.
  • Snowball Sampling: Used when samples of special populations are needed.
  • Quota Sampling: Combination of convenience and judgment sampling, using census data to classify the population.
Probability Sampling Methods
  • Simple Random Sampling: Units selected individually and directly by a random process.
  • Systematic Sampling: Sample elements drawn systematically from a complete list.
  • Stratified Sampling: Population divided into homogeneous strata, then a random sample is drawn from each stratum.
  • Cluster Sampling: Used when it is difficult to compile a sampling frame of the elements.

Step 4: Determining the Sample Size

  • Sample size should not be a fixed proportion of the population.
  • Depends on:
    • Population parameter.
    • Behavior variable in the population.
    • Standard error.
    • Confidence level.
Sample Error
  • Sampling is never 100% accurate due to differences within the population.
  • Types of errors:
    • Sampling Errors
    • Systematic (Non-Sampling) Errors (observation or measurement errors)
  • Sample error: Difference between sample statistic and population parameter.
Limiting Sample Error
  • Clearly define the population.
  • Limit non-reacting respondents.
  • Consider variation within the sample and population.
Sample Size Considerations
  • Sample size is often a matter of judgment.
  • Must be large enough for a precise estimate but also be economical and practical.
  • Determined by:
    • Purpose of the study and precision required.
    • Size of the population.
    • Available time, money, and resources.
    • Type of report required.
  • Sample design should minimize standard errors.
Methods of Determining Sample Size
  • Blind Guesses: Arbitrary, based on judgment.
  • Statistical Method: Uses statistical formulae, based on:
    • Required level of confidence.
    • Required precision.
    • Standard deviation of the population.
General Procedure for Statistical Calculation of Sample Size
  1. Determine the tolerable error.
  2. Specify the confidence level.
  3. Determine the Z value.
  4. Estimate the standard deviation of the population.
  5. Use the appropriate statistical formula.
  6. Draw the desired sample.
Steps Explained
  • Step 1: Determine the Tolerable Error: Acceptable difference between sample estimate and population parameter.
  • Step 2: Specify the Confidence Level: Desired degree of certainty in the estimate.
  • Step 3: Determine the Z Value: Use the standard Z Table.
  • Step 4: Estimate the Standard Deviation of the Population:
    • Pilot survey
    • Similar previous studies
    • Guess (rule of thumb: standard deviation is one-sixth of the range).
  • Step 5: Use the Appropriate Statistical Formula:
    • Formula for calculating the sample size is: calculating sample size n=(Z2∗s2)/E2n = (Z^2 * s^2) / E^2 where:
      nn is the sample size.
      ZZ is the Z-value associated with the desired confidence level.
      ss is the estimated standard deviation of the population.
      EE is the tolerable error (margin of error).
  • Step 6: Draw the Desired Sample.

Step 5: Selecting the Sample Elements

  • Clear guidelines for selecting respondents should be included.
  • Involves selecting:
    • Sample units (e.g., households).
    • Sample elements (e.g., head of household).
  • Clear sampling procedure is more important for probability sampling.

Step 6: Gathering Data from Designated Elements

  • Gather data from designated respondents.
  • Implement measures to minimize potential problems.

Summary

  • Sampling isolates a subgroup from the population.
  • Aims to draw conclusions about the subgroup and the population.
  • Provides a practical research method.
  • Advantages over census:
    • Financial considerations.
    • Speed and time.
    • Easier implementation.
    • Better quality and accuracy.
  • Accuracy depends on sample size, selection method, and estimation methods.