Measurement and Scaling Techniques

Posted on Mar 11, 2025 in Other university degrees and diplomas

Measurement: Assigning numbers or symbols to characteristics of objects based on specific rules. Scaling: Continuum creation for measured objects.

Primary Scales of Measurement

  • Nominal: The number of your ID.
  • Ordinal: Ranking of your favorite movies.
  • Interval: A satisfaction survey (from 1 to 5).
  • Ratio: The money you spend in a week.

Scaling Techniques

Comparative Scaling

  • Paired Comparison: The respondent is asked to choose between two options presented simultaneously.
  • Rank Order: The respondent ranks several objects according to their preferences.
  • Constant Sum: The respondent distributes a fixed number of points among the attributes of an object.
    • Advantages: Detects small differences. It allows for comparing options that are very similar to each other.
    • Disadvantages: Only generates ordinal data.

Noncomparative Scaling

Each object is evaluated independently.

  • Continuous Rating: Mark on a line from one extreme to another.
  • Itemized Rating
Non-Comparative Scaling Examples
  • Likert Scale: Measures agreement with statements.
  • Semantic Differential: Bipolar adjectives (e.g., reliable/unreliable) on a scale.
  • Stapel Scale: Single adjective with a unipolar scale (-5 to +5).

Scale Evaluation

  • Reliability: Consistency of measurement (e.g., test-retest reliability, internal consistency using Cronbach’s Alpha). Alpha values: ≥0.9 excellent; <0.6 poor.
  • Validity: Accuracy in measuring the intended trait without error.

Questionnaire Design

Definition of a Questionnaire: A formalized set of questions designed to collect information from respondents. It can be verbal or written.

Objectives of a Questionnaire

  • Translate the research problem into specific, answerable questions.
  • Motivate respondents to cooperate and provide complete answers.
  • Minimize response errors.

Question Structure

Is the Question Necessary? Are Several Questions Needed Instead of One?

  • Unstructured questions are open-ended questions that respondents answer in their own words. This is costly and time-consuming.
  • Structured questions specify the set of response alternatives and the response format. A structured question may be multiple-choice, dichotomous, or a scale.

Use Ordinary Words / Use Unambiguous Words / Dual Statements – Positive and Negative: Broad general questions to narrow specific questions (Funnel approach)

Sampling Techniques

Key Definitions

  • Population: All elements sharing specific characteristics relevant to the research problem.
  • Census: A complete count of all elements in a population.
  • Sample: A subset of the population selected for the study.
  • Parameter: The true value of a characteristic in the population (e.g., average income).
  • Statistic: An estimate of the parameter, calculated from the sample.

Sampling and Non-Sampling Errors

  • Sampling Errors: Occur due to using a sample instead of a census.
  • Non-Sampling Errors: Result from issues such as questionnaire design or data processing errors.

Comparison of Sampling Techniques

  • Stratified Sampling: Subpopulations (strata) are homogeneous within but heterogeneous between. Ensures higher precision without increasing cost.
  • Cluster Sampling: Subpopulations (clusters) are heterogeneous within but homogeneous between. Reduces cost but may lower precision.

Sampling Design Process

  1. Define the Target Population
  2. Determine the Sampling Frame
  3. Select a Sampling Technique
  4. Determine Sample Size
  5. Execute the Sampling Process

Types of Sampling Techniques

Nonprobability Sampling

  • Convenience Sampling: Based on ease of access (e.g., mall interviews).
  • Judgmental Sampling: Selection based on researcher’s judgment (e.g., expert witnesses).
  • Quota Sampling: Predefined quotas based on population characteristics (e.g., 50% male, 50% female).
  • Snowball Sampling: Referrals from initial respondents (useful for niche populations).

Probability Sampling

  • Simple Random Sampling (SRS): Equal chance of selection for all elements.
  • Systematic Sampling: Selects every nth element after a random start.
  • Stratified Sampling: Divides the population into homogeneous strata; random samples are drawn from each.
  • Cluster Sampling: Randomly selects clusters; includes all or parts of the elements within clusters.