Upload
fabiana-fullam
View
50
Download
3
Embed Size (px)
DESCRIPTION
Questionnaire Design, Survey Methods, and Sampling. Market Intelligence Julie Edell Britton Session 5 September 4, 2009. Today’s Agenda. Announcements Comparative Advertising, Measurement Scales & Data Analysis Introduction to Survey Research Sampling Procedures Causal Research - maybe. - PowerPoint PPT Presentation
Citation preview
1
Questionnaire Design, Survey Methods, and Sampling
Market IntelligenceJulie Edell Britton
Session 5September 4, 2009
2
Today’s Agenda
Announcements
Comparative Advertising, Measurement Scales & Data Analysis
Introduction to Survey Research
Sampling Procedures
Causal Research - maybe
3
Announcements
For Sat prepare Milan Food case – download data (Milan.sav) from the platform, please post your responses by 8 pm tonight – no slides needed.
For Sat prepare WSJ/ Harris Survey – no slides
4
Comparative Advertising Measurement Scales & Data Analysis
Page 52 packet
What do you conclude?
Remember – Percentage change or difference can only be calculated with Ratio scales
This is an interval scale (at best).
5
Descriptive Survey Research Surveys usually used for descriptive research
Provide a snapshot at a point in time Most analyses univariate or bivariate (but can do
elaboration model with control variables) Would you recommend National to a friend interested in
insurance services? Yes 1 No 2
Bivariate allows for hypothesis testing Hypothesis: Less educated people more likely to recommend
Descriptive, not causal Recommendation could be driven by some 3rd factor
correlated with education such as income
6
Sources of Survey Errors Population definition Representativeness of the sample frame Sampling Procedure Used Respondent Participation:
Willing to participate (Do Not Call) Comprehend questions Have knowledge, opinions Willing & able to respond (language or memory)
Interviewer understands & records accurately
7
Raising Willingness to Participate
A good response rate requires persuasion Survey Introduction
Phone or send letter in advance Introduce self, give affiliation unless this would
bias Describe purpose briefly, w/o making survey
sound threatening or demanding Make respondent feel that s/he is getting chance
to provide opinions that will influence market offerings & that her/his cooperation is extremely important
8
Comprehends Questions? Advice on Question Wording
Be simple and precise Give clear instructions Check for question applicability
respondent screening question branching based on prior
answers
Avoid leading & double barrel questions
9
What’s the Problem?
“Laws should be passed to eliminate the possibilities of special interests giving huge sums of money to candidates”
“Laws should be passed to prohibit interest groups from contributing to campaigns, as groups do not have the right to contribute to candidates they support?”
10
Comprehends Questions? Literacy, translation considerations
Conversational Norms How demanding was Term 3? How demanding was
Core Finance?
How demanding was Core Finance? How demanding was Term 3?
How demanding was Managerial Accounting? How demanding was Core Finance? How demanding was Global Economic Environment of the Firm? How demanding was Term 3?
“Question order effects in measuring service quality,” by DeMoranville and Bienstock, in International Journal of Research in Marketing, September, 2003
11
Do Respondents Have Knowledge?
Retrieve answer from memory vs. construct it on spot
Constructed answers are more likely to be influenced by question wording & prior questions.
When answering later questions or engaging in later behavior, likelihood of using earlier answer input A: positively related to accessibility of A
positively related to diagnosticity (relevance) of A
negatively related to accessibility, diagnosticity of alternative inputs B, C, etc. (Feldman & Lynch)
e.g., when political poll respondents asked: issue opinion A, presidential voting intention, issue opinion B,
answers to A predict intention, but only for those who did not vote for either candidate in primary
1212
Survey Best Practices: Survey Content, Question Order
Survey Questions First figure out what questions are needed! Then order Lead with interesting, nonthreatening, easy questions
Do you like to play golf? Have you ever travelled with your clubs? Can you remember the last time you traveled with your clubs?
Put difficult or sensitive questions well into the interview How many times did you have to see your doctor for your
reconstructive surgery? What is the size of your company (revenue)?
Usually use funnel order (general to specific) Use product category? Brand X? Do you like Brand X? Why?
13
Question Order (Cont.)
Survey Questions (cont.) Inverted funnel (specific to general) for complex topics.
Is your company considering offering training courses on word processing over the Internet?
Database? Spreadsheets? In general, how big is the untapped market for your software
training courses if offered over the Internet?
Group questions in logical order All questions about one subject together, with transitional
phrases in between, “Now I’m going to ask you about agricultural applications of GPS systems...”
14
Survey Best Practices: Question Order (cont.)
Demographics Questions Put last—these are less sensitive to prior questions Seem nosy if put first Rely on standard approaches for assessing
http://www.norc.org/GSS+Website/
The Process of Survey Design Use Backwards Marketing Research to decide what is
“need to know” Draft the survey Pretest for time, clarity, variability in responses Revise and retest Field the survey and keep an eye open for problems
15
16
Survey Best Practices:Choosing a Survey Method
Mail, phone, web, in person? Cost Complexity of inquiries (branching) Need for aids Issue sensitivity Control over sample
1717
Web and Telephone
Web surveys now dominate. To compare web, in person, phone, mail, see http://knowledge-base.supersurvey.com/
18
Free to Fuqua students: Qualtrics
http://www.qualtrics.com/duke#submit Set up an account Build surveys Allows for complex designs
Available to you during this course
19
Multi-Attribute Attitude Model (MAAM)
Liking for a product as a whole = sum of liking for component parts
Attitude toward brand j = (sum from i = 1 to n for salient attributes)
Importance of Attributei * Evaluationij
Importance 0 – 100 (allocate 100 points across attributes) Rating on 1 (unimportant) to 7 (very important) where 0
undefined but implicitly entirely unimportant )
Evaluation of brand j on attribute I -4 = poor to +4 = excellent
20
Land Rover RAV Land Rover RAV
Attribute
Importance 1=unimp, 7= important
Brand Evaluation -4 = poor, +4 =
excellent Imp*Eval Imp*EvalSporty Styling 6 1 2 6 12Handling 5 0 1 0 5Cost 2 -2 3 -4 6Ruggedness 4 4 2 16 8Off-Road Ability 2 2 4 4 8Total Attitude 22 39
MAAM and SUVs
21
Diagnostics of Advantage
22
Measure Types Revisited
Nominal (Unordered Categories) Just need unique number for each category
Ordinal: ranking scale, intervals not assumed equal
Interval: Intervals assumed equal, zero is arbitrary
Ratio: Intervals assumed equal, zero means zero To multiply X * Y, (e.g., importance * evaluation), both X and Y must
be on ratio scales. If X1*Y1 > X2*Y2 (XYbrand 1 >XYbrand 2), it does NOT follow that
(X1+a)*Y1 > (X1+a)*Y2…. e.g., 2*2 > 2*(-2), but (2-4)*2 < (2-4)*(-2) To say % change in Y, Y must be on ratio scales
23
More on Scaling
To multiply importance x evaluation for each attribute, both must be on ratio scales
0 on scale must be 0 of underlying quantity
Importance unipolar (all positive). Completely unimportant = 0 weight
Evaluation bipolar (negative to positive). To multiply, must code “neutral” as zero.
24I got these by subtracting 4 from the values three slides back
Land Rover RAV Land Rover RAV
Attribute
Importance -3 =unimp, +3 = imp
Evaluation -4 = poor, +4 = excellent Imp*Eval Imp*Eval Diff
Sporty Styling 2 1 2 2 4 2Handling 1 0 1 0 1 1Cost -2 -2 3 4 -6 -10Ruggedness 0 4 2 0 0 0Off-Road Ability -2 2 4 -4 -8 -4Total Attitude 2 -9 -11
Improper Rescaling
25
Consumer Attitudes We want to be able to predict consumer behavior
However, instead of examining behavior directly (e.g., choice modeling), we often measure attitudes because… Measuring attitudes is sometimes easier than observing
choice
Attitudes are more diagnostic
Attitudes are sometimes easier to interpret
Attitudes can be reasonable predictors of behavior
Attitudes toward products or brands typically derive from beliefs, actions, and perceptions
26
Types of Attitude Scales Semantic differential Colgate Combo is:
low quality __:__:__:__:__:__:__ high quality
unappealing __:__:__:__:__:__:__ appealing
Constant sum (e.g., Importance)
Purchase intent
Likert scale (Agree-Disagree)
27
Recap
Survey Design: responses constructed on the spot Moving parts of a good survey Population definition,
choosing a survey method, determining what information needed
Order of questions Attitude Measurement & multi-attribute attitude model To multiply or examine percentage differences, data
must be on ratio scales