asking questions

Learning Objectives

1. List four scales (levels) of measurement.

2. Identify widely used attitude-scaling techniques in marketing research.

3. Explain validity and reliability concerning measuring instruments.

Defining Scales of Measurement

• Measurement:

  • Assignment of numbers to objects representing quantities of attributes.

  • Benefits: Easier communication, summarization, and statistical analysis.

• Scaling:

  • Generation of continuum to locate measured objects.

  • Types: Nominal, Ordinal, Interval, Ratio.

Example: Measurement and Scale

• Measuring understanding of marketing research using a quiz score (0-20).

• Assign points based on correct answers (1=correct, 0=incorrect).

• Example Representation:

  • Abigail: Score = 10

  • Quantity of attribute: understanding

  • Measurement rules applied.

Scale of Measurement Hierarchy

• Levels:

  • Highest: Ratio

  • Interval

  • Ordinal

  • Nominal

• Data collected at higher levels can represent lower levels, but the reverse isn't true.

Central Tendency in Measurement

• Mean: Average value of data.

• Median: Middle value when data is ordered.

• Mode: Most frequently occurring value.

Types of Measurement Scales

1. Nominal Scale:

   • Assigning numbers solely for identification.

   • Example: Favorite cosmetics.

   • Central tendency measured: Mode.

2. Ordinal Scale:

   • Assigning numbers based on order.

   • Example: Rank brand preferences.

   • Central tendency measured: Median, Mode.

3. Interval Scale:

   • Numbers allow for comparison of size differences.

   • Example: Attitudes towards brands (1=Negative, 5=Positive).

   • Central tendency measured: Mean, Median, Mode.

4. Ratio Scale:

   • Has absolute zero; allows for comparing absolute magnitudes.

   • Example: Total units sold.

   • Central tendency: Geometric mean, Harmonic mean, Mean, Median, Mode.

Interval vs. Ratio Scales

• Interval Scale: Zero as a position, does not represent absence.

• Ratio Scale: Zero has absolute meaning, indicating absence of property.

Comparison of Scales of Measurement

Measuring Attitudes & Other Unobservables

• Attitudes, perceptions, and preferences are often gauged through self-reports.

• Scaling Techniques:

  • Non-comparative Scales:

    1. Itemized-Ratings (Likert) Scale

    2. Semantic-Differential Scale

  • Comparative Scales:

    1. Continuous Ratings

    2. Constant-Sum Scale

Comparative vs. Non-comparative Scales

• Comparative Scales: Evaluate two or more attributes together.

• Non-comparative Scales: Evaluate one attribute at a time.

• Itemized: Fixed response categories vs. Continuous: Many response categories, flexible.

Likert Scale**: A widely-used scale to measure attitudes. Features:

• Total sum represents overall attitude.

• Symmetric and equidistant for interval-level measurement.

Semantic-Differential Scale

• Allows evaluation on a bipolar scale.

• Advantages: Versatility for brand image assessment.

• Disadvantages: Difficulty in finding appropriate bipolar adjectives.

Graphic-Ratings Scale

• Respondents evaluate by placing marks along a continuum.

• Pros: Provides fine distinctions; Cons: Scoring can be challenging.

Constant-Sum Scale**: Assign a fixed number of points among attributes, assessing relative importance.

• An effective method to prevent halo effect influence.

In-class Assignment

• Design a survey including:

  1. A Likert question

  2. A semantic differential question

  3. A ranking task

  4. A point allocation task

• Example Topic: ASU Student Preferences.

Measurement Error

• Measurement error: Difference between observed and true score.

• Sources of Error:

  • Systematic Error: Constant impact.

  • Random Error: Variable impact across measurements.

Reliability vs. Validity

• Reliability: Consistency of measurement results over time (related to random error).

• Validity: Reflects true differences among objects rather than error.

• Perfect reliability does not ensure perfect validity.

Conclusion & Review

1. List the four scales of measurement.

2. Identify widely used attitude-scaling techniques.

3. Explain validity and reliability in measuring instruments.