MAR3613 - Exam 2

What are the three types of consumer characteristics that are of interest when doing a questionnaire?

1. Behavior

2. State of Being

3. State of Mind

Behavior

The actual actions consumers make that can be measured and directly seen.

State of Being

Characteristics of consumers that are physically verifiable:

• Demographics

• Age

• Gender

State of Mind

Status of consumers’ thinking, perception of their own intuition, and the function of their brains.

• Attitude

• Value

• Perception

• Image

Scaling

The process of adopting/creating a scale with descriptors (either numbers or labels) on it.

• Example: How satisfied are you with your service (Rate 1-5)

Natural Scales

Scales that already exist in real life.

• Readily available and can be adopted by the mass population.

• Used to measure (State of Being and Behavior)

• Example: How many times did you go to Starbucks this week? (Behavior)

• Example: How old are you? (State of Being)

Questions using Natural Scales are called?

Natural Metric Questions

Synthetic Scales

Scales you have to create by yourself.

• Created to measure State of Mind.

Questions using Synthetic Scales are called?

Synthetic Metric Questions

Measurement

Assigning descriptors (either numbers or labels) to consumer characteristics to analyze them.

• Assignment process must be one-to-one (isomorphic), meaning each answer should only correlate to one number or label.

• Rules should be standardized and used uniformly, aka all data must be measured the same way.

One-way Labeled Scale

A LABEL SCALE that is used to measure all types of consumers’ state of mind from one extreme to another extreme.

• Example: Cleanliness of a restaurant (Important to Not important at all)

N-Point (Rating) Scale

A NUMBER SCALE scale that is used to measure all types of consumers’ state of mind from one extreme to another extreme.

• Usually, only positive numbers are used. (No negatives)

• Example: How confident are you right now that your answer is correct? (Very Unconfident = 1, Very Confident= 5)

Anchored N-Point Scale

Numbers have an associated meaning attached to them.

• Example: How confident are you right now that your answer is correct? (Very Unconfident = 1, Very Confident= 5)

Unanchored N-Point Scale

Numbers have no meaning associated with them/

• Example: How confident are you right now that your answer is correct? (1-5)

One-way Labeled Scale (+ & -)

• Advantage: Question is straightforward, does not require much effort on the respondent’s side.

• Disadvantage: Not numeric, and takes time for researchers to come up with labels.

N-Point Scales (+ & -)

• Advantage: Numbers allow for statistical analysis.

• Disadvantage: Requires respondents to translate state of mind into numbers, leading to more errors.

Semantic Differential Scale

Use bipolar adjectives for various features of brands/products/stores, and respondents choose based on their judgments and/or state of mind.

• Used to identify brands/products/stores’ image.

• DOES NOT PROVIDE INPUT INTO THE IMPORTANCE OF ATTRIBUTES.

• Example: How would you describe Kmart, Walmart, and Target on the following scale (Clean 1, 2, 3, 4, 5 Dirty)

Stapel Scale

Unipolar rating scale that measures respondents’ feelings towards various features of brands/stores/products from a negative to a positive end.

• Used to identify an image.

• Example: How would you rate this brands customer service ( 3, 2, 1, 0 , -1, -2, 3)

Semantic Differential Scale (+ & -)

• Advantage: No negative numbers involved, easy to compare.

• Disadvantage: Difficult to come up with bipolar adjectives.

Stapel Scale (+ & -)

• Advantage: One polar adjective is sufficient, easy to construct.

• Disadvantage: Negative numbers are involved.

Likert Scale

Asks respondents to indicate their level of (dis)agreement on a series of statements.

• Captures respondents’ lifestyle, values, views, and more based on their level of (dis)agreement.

• Example: I generally buy the same brands I always buy (Strongly disagree - Strongly Agree)

How many categories should you have in answer options?

• 5 when all responses are to be summed up into one score.

• 9 when comparing attributes between different options.

What is a good amount of time a questionnaire must take to fill out?

The questionnaire should be relatively short, simple, and clear.

• 3-5 Minutes to complete.

• Long surveys risk respondents not completing them.

• Too short surveys risk not gathering enough information.

What are the three types of logic on Qualtrics?

1. Skip Logic

2. Display Logic

3. Branch Logic

Synthetic Scales allow researchers to obtain both the direction and intensity of the state of mind.

True.

When using scales, how much levels should you create?

4-7 Levels.

When creating a survey you should not use industry-specific jargon (Example: CTR, SWOT, PPC)

True.

Questionnaire Design

A systematic process in which the researcher:

• Contemplate various question formats.

• Word the various questions carefully.

• Organizes the layout of the questionnaire.

Types of question formats:

• Open-Ended

• Categorical

• Metric

Open-Ended Questions

Instructs respondents to respond in their own words without providing any answer choices.

• Used when there are too many potential answer choices (Unaided).

• Used when the researcher wants to ask for additional information or clarification (Aided).

• NOT STANDARDIZED.

Order of response formats:

Unaided questions are usually asked first, then followed by a aided question.

Open-Ended Response Formats

• Aided

• Unaided

Aided

Example: What ads do you recall?

Unaided

Example: Do you recall any other ads you saw on TV?

Categorical Questions Response Formats

• Dual-Choice

• Multiple Choice

Dual-Choice

Example: Yes or No

Multiple Choice

Example: From the following list, which brand…

Metric Question Response Formats

• Natural

• Synthetic (Numeric or Label)

Open-Ended Questions (+ & -)

• Advantage: Avoids bias and gives respondents some degree of control.

• Disadvantage: Coding questions are time-consuming, and it is hard to understand in what way people interpret some words (Example: Sometimes).

Categorical Questions

Asks respondents to choose the category that best describes them.

• NO SCALE IS NEEDED.

• Provide nominal data that is useful for elaborating on behavior and state of being.

• Example: What is your gender?

• Example: Why did you not purchase this item? (Select all)

Metric Question

Call for respondents to measure on a scale.

• Provide ordinal, interval, and ratio data.

• Example: What is your age? (a. 18-24, b. 25-30)

• Example: What price range of sports shoes do you typically purchase?

Close-Ended Questions (+ & -)

• Advantages: Quicker for the respondent, simple numerical analysis, allows for uniformity in research design.

• Disadvantage: The researcher must know what the possible answer choices are beforehand, respondent is less involved.

What are the Four Do’s in Questionnaire Wording?

Questions should be:

1. Focused (Only focus on a single issue or topic)

2. Simple (Grammatically simple)

3. Brief (Remove redundant and unnecessary words)

4. Clear (Avoid words that will lead to misinterpretation)

What are the Six Do’s in Questionnaire Wording?

1. Avoid ambiguous words in questions.

2. Avoid leading questions that give respondents a strong (expectation) or subtle cue as to how to answer.

3. Avoid overstated questions that place undue emphasis on some aspects related to the topic of the questions.

4. Avoid loaded questions that refer to unverified assumptions, presumptions, facts, or universal beliefs. And generate emotional/cognitive impetus towards a specific answer.

5. Avoid using double-barred questions that have two questions in one.

6. Avoid ill-defined questions

In a questionnaire, do not use words of extreme absolutes, meaning that respondents are placed in an extreme position where they must either completely agree or disagree.

True.

Extreme Absolute words to avoid:

• All - Example: Did you consider ALL options before purchasing?

• Always - Example: Do you ALWAYS buy brand A from firm B?

• Any - Example: Did you have ANY concerns about the price?

• Others: Anybody, best, ever, every, most, never, worst.

Avoid grading adverbs or gradable adjectives in categorical and metric questions?

True. Avoid these words, because they’re not exact and are rather subjective.

• Grading adverbs: Extremely, fairly, hugely, very, usually.

• Gradable adjectives: Angry, big, busy, rich, strong, tall.

Avoid leading questions that give respondents a strong (expectation) or subtle cue as to how to answer.

True. As it can lead to biased answers.

• Emotional verbs such as: Force and prohibit can lead to biased responses.

  • Example: Do you agree the government should FORCE you to pay higher taxes?

  • Emotional adjectives also lead to bias. Using words such as: Legendary and amazing.

• Don’t use apparent logic in the question:

  • Example: You like fast food, don’t you?

• Avoid socially desirable outcomes:

  • Have you tried this new restaurant, everyone is talking about?

• Do not use preceding questions to set up assumptions in the minds of respondents.

  • Since you agreed that fast-food consumption leads to eating disorders, should fast-food restaurants be responsible for correcting them?

Avoid overstated questions that place undue emphasis on some aspects related to the topic of the questions.

True.

• Do not state facts and research in an unbalanced way.

  • Example: As studies show, overweight kids buy fast food. Should kids’ meals be healthier?

• Do not overstate wrong facts or assumptions.

  • Example: As everyone buys fast food, should everyone be concerned about nutrition?

• Do not overstate answers to preceding questions.

Overloaded Questions with Unverified Assumptions

Example: Did you stop smoking? What do you love about my product/service?

HOW TO SOLVE THIS: SKIP LOGIN

• Example: Ask Do you smoke? If so, ask Did you stop smoking?

Overloaded Questions (AVOID)

Questions overloaded with facts.

• Example: Do you think KFC should have warning labels for obese people suffering from deadly diseases such as diabetes?

Questions are overloaded with general beliefs.

• Example: Should people be allowed to protect themselves from harm by using taser guns as self-defense?

Ill-Defined Questions Include…

Questions respondents couldn’t recall.

• Example: When was your first time buying aspirin?

Asking respondents to predict their actions in situations they cannot fathom.

• Example: How often would you go out to eat at this new, upscale restaurant that will be built 10 miles from your home?

Questionnaire Organization

Concerns the arrangement of the questionnaire and the sequence of the questions that make up a questionnaire.

What does Questionnaire Organization affect?

• Response rate

• Quality of responses

What are the three aspects of the Questionnaire Organization researchers focus on?

1. Introduction (Included in the email or the opening of the questionnaire)

2. Logic (Screening and filtering questions)

3. Flow (Sequence of questions)

Introduction

What potential respondents read or hear before they begin answering the questions.

• Increases response rate and quality of responses.

What are the six functions of the Introduction?

1. Identify the surveyor.

2. Indicate the purpose of the survey.

3. Explain how the respondent was selected.

4. Request and provide for an incentive for participation.

5. Screening (Example: Do you use Firefox, Internet Explorer, or Opera?)

6. Appreciation

Screening Questions

Used to ferret out unqualified respondents.

• Screening is achieved through Skip Logic = Skip to the end of the survey.

Filtering Questions

Determine what the next question’s respondents see depending on how they answer it.

• Done in Qualtrics by using Display Logic and Branch Logic.

Question Flow

Pertains to the sequencing of questions or blocks of questions, including any instructions, on the questionnaire.

Funnel Approach

1. Screening Questions (Example: Have you shopped at VS in the past month?)

2. Warm-Up Behavioral Questions (Example: How often do you go shopping for casual clothes?)

3. Transitions (Example: Next, I am going to list several statements. Tell me if you agree or disagree)

4. Complicated Synthetic Metric Questions (Example: Rate each of the following 10 stores on a scale 1-7)

5. Demographic Questions (Gender, Age, Income)

What are the four properties of measurement?

1. Assignment

2. Order

3. Difference

4. Origin

Assignment

Classify consumers into categories.

• Example: Consumer segmentation, current consumers vs. potential consumers.

• Example: Every student has an ID.

Order

Allows researchers to rank consumers in terms of having more or less of certain characteristics through measurement.

• Example: Heavy users vs. Light users of cigarettes, consumers who are extremely dissatisfied vs extreme satisfied.

Difference

Quantifies the difference in consumer characteristics in a standardized way.

• Example: Difference is class standing (Freshman vs Seniors = 3 year difference)

Origin

Unique, meaningful starting point (usually it is zero, which has objective meaning) in a set of scale points when measuring certain consumer characteristics.

• Example: Every student has an age, gender, salary, and weight.

• DOES NOT APPLY TO TEMPERATURE, SATISFACTION SCALES & MORE.

What are the four data formats? And their ranking

Highest

1. Ratio

2. Interval

3. Ordinal

4. Nominal

Lowest

Nominal Data

Data that serves as a label or a tag (unordered category, no implied criteria to order) to identify certain customer characteristics.

• Has the assignment property only.

• You can find the percentage, mode, and frequency with it.

• Categorical questions provide nominal data.

• Example: What is your occupation?

Ordinal Data

Type of data where consumer characteristics can be put in natural, ordered categories.

• Has both assignment and order properties.

• The difference cannot be determined with this.

• Can be used to find the percentage, frequency, mode, and median.

• One-way labeled and rank-related questions provide ordinal data.

• Example: How important do you believe it is to vote in the elections? (Extremely important - Not important at all).

Interval Data

Types of data that permit meaningful statements about the differences in certain characteristics between any two consumers.

• Contains assignment, order, and difference properties.

• Evenly-spaced synthetic scales such as the Likert scale, N-Point Scale, Semantic Differential Scale, and Stapel Scale provide interval data.

Ratio Data

Data that has a “true” zero meaningful starting point.

• Has all assignment, order, difference, and origin properties.

• All statistical operations can be performed on ratio data.