Marketing Exam 3 Tokyo Drift

scale development

The process of designing questions and descriptors that accurately measure constructs in primary data collection.

Why is scale development important?

It makes sure the data you collect are accurate, meaningful, and can clearly distinguish differences between respondents.

Three components of scale measurement

Criteria for good scales

Intelligible questions

Appropriate descriptors

High discriminatory power

Likert Scale

Uses agree/disagree statements to measure attitudes.

Semantic Differential Scale

Uses bipolar adjective pairs (e.g., clean-dirty) to measure perceptions/image profiles.

Behavioral Intention Scale

Measures likelihood of future behavior (e.g., "definitely would," "probably would not").

Nominal Scale

Labels or categories; no order (e.g., gender, brand name). Which smartphone brand do you currently use?"

Apple

Samsung

Google

Motorola

Ordinal Scale

Ordered categories but unequal intervals (e.g., rank preferences). "Rank these restaurant chains from most preferred (1) to least preferred (4)."

Chipotle

Panera

Chick-fil-A

Wendy's

Interval Scale

no true zero. A scale that demonstrates

absolute differences between

each scale point. Rate Target on the following bipolar attributes:

Expensive ⟵—1—2—3—4—5—⟶ Affordable

Ratio Scale

Most information-rich scale.Most Equal intervals and true zero. "How many times did you visit Starbucks last month?"

Rating vs Ranking

Rating: Judge each item independently.

Ranking: Compare items against each other (forced order).

COMPARATIVE SCALES

Requires comparing one object to another (e.g., rank these brands) .Rank-order scale (ordinal)

Example:"Rank these sneaker brands in order of preference:"1 = Most preferred

Nike

Adidas

New Balance

Converse

Noncomparative Scale

Rate one object at a time (e.g., Likert).

Constant-sum scale

"Divide 100 points among the attributes based on importance in choosing a hotel:"

Price: 40

Location: 30

Cleanliness: 20

Brand reputation: 10

Points must total 100.

Discriminatory Power

A scale's ability to distinguish (discriminate) between respondents who have different attitudes, perceptions, or behaviors.

Single-item Scale

Measures construct with one question.

Multiple-item Scale

Measures construct with several questions.

Construct

A variable that cannot be directly observed (e.g., satisfaction, loyalty).

response order bias

a type of response bias where a respondent may react differently to questions based on the order in which questions appear in a survey or interview. (favoritism between top and bottom answers.)

Common method variance

the spurious correlation between variables caused by using the same measurement method, such as a single survey, to measure them. This can lead to inaccurate conclusions because the correlation may be due to the method itself rather than a genuine relationship between the constructs being studied.