1 lesson 3 measurement & attitude scales in marketing research

28
1 Lesson 3 Measurement & Attitude Scales in Marketing Research

Upload: eric-garrett

Post on 13-Jan-2016

227 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

1

Lesson 3Measurement & Attitude Scales in Marketing Research

Page 2: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

2

Lesson outline:Considerations in Choosing a Question-Response Format

Some Basic Concepts in Measurement

The Four Types of Scales used by Marketing Researchers

Reliability & Validity of Measurements

Page 3: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

3

Basic Question-Response Formats

Designing a questionnaire can be compared with Backstreet Boys composing a song, or Vincent Van Gogh painting a landscape… it requires creativity!There are however, some basic aspects to questionnaire should be taken into considerationThere are basically three basic question-response formats for a researcher to choose

Page 4: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

4

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

(1) Open-Ended Response Format QuestionsUnprobedProbed Format

(2) Closed-Ended Response Format QuestionsDichotomousMultiple Category

(3) Scaled-Response Format QuestionsLabeledUnlabeled

Page 5: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

5

(1) Open-Ended Response Format QuestionsUnprobed

Probed Format

There is no response options for the respondent

Respondent is instructed to respond in his/her own words

The response naturally depends on the topic

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

Page 6: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

6

(1) Open-Ended Response Format Questions

Unprobed OR Probed Format

An unprobed format seeks no additional information from the respondent

When the researcher wants comments or statements or simply wants the respondent to indicate the name of a brand or the name of a movie or store – he uses a probed format

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

Page 7: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

7

(1) Open-Ended Response Format Questions

Unprobed

Q1 ‘What do you think of the latest Nokia mobile phone?’

‘I don’t like it!.’

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

Page 8: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

8

BASIC QUESTION-RESPONSE BASIC QUESTION-RESPONSE FORMATSFORMATS

(1) Open-Ended Response Format Questions

Probed Format

‘Did you watch F4 in concert last week?

‘Yes’

What do you think of the concert?’

‘It was terrible!’

Why was it terrible?

‘They danced funny and their dressing were so obiang!’

Page 9: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

9

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

(2) (2) Closed-Ended Response Format QuestionsProvides response options to questions

Can be Dichotomous OR Multiple Category

A dichotomous close ended question has only two response options, e.g. ‘yes’ or ‘no’, ‘agree’ or ‘disagree’If there are more than two options for the response, then the researcher is using a multiple category closed-ended question

Page 10: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

10

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

(2) Examples:

DichotomousDo you agree or disagree with the statement ‘Nokia handphones are more durable than Motorola handphones?

Multiple CategoryIf you were to buy a handphone tomorrow, which brand would you be most likely to purchase? Would it be: (a) Nokia (b) Motorola c) Ericsson (d) Siemens (e) National Panasonic (f) other brands?

Page 11: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

11

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

(3) Scaled-Response Format QuestionsLabeled OR Unlabeled

A Scaled-Response question uses a scale developed by the researcher to measure the attributes

The response options are identified on the questionnaire

Page 12: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

12

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

(3) Scaled-Response Format QuestionsLabeled scaled-response format uses a

scale in which all of the scale positions are identified with some descriptors

Eg. Do you strongly disagree, disagree, agree or strongly agree with the statement “Nokia handphones are of better value than Motorola handphones?”

Page 13: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

13

BASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATSBASIC QUESTION-BASIC QUESTION-RESPONSE FORMATSRESPONSE FORMATS

(3) (3) Scaled-Response Format QuestionsFor Unlabeled scaled-response format,

the scale may be purely numerical and only end-points of the scale are identified

E.g. On the scale of 1 to 5, how do you rate the Nokia handphone on the ease of operation?

(1 = very easy to 5 = very difficult)

Page 14: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

14

Labeled scaled-response format

Marketing Research is an interesting subject

Strongly DisagreeAgreeStrongly Agree Disagree

1 2 3 4

Unlabeled scaled-response format

In choosing a Poly, it must have modern teaching facilities

Not Important at all V. Important

1 2 3 4 5

MORE EXAMPLES…

Page 15: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

15

Pros & Cons of Alternative Response Formats

Example:Unprobed open-ended question+ allows respondents to use his/her own

words- Difficult to code and interpretProbed open-ended question+ able to get complete answers- Difficult to code and interpret

Please read Chapter 9 page 305 for more details!

Page 16: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

16

CONSIDERATIONS IN CHOOSING A QUESTION RESPONSE FORMAT

CONSIDERATIONS IN CHOOSING A QUESTION RESPONSE FORMAT

Correct Format!Correct Format!

Nature of the Property being Measured

Previous Research Studies

Data Collection Mode

Ability of the Respondent

Scale Level Desired

Page 17: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

17

Nature of the Property Being Measured

The inherent nature of the response often determines the question-response format

For instance, if the F4 pop group wants to know if the respondents have seen their last MTV, the only answers are ‘yes’, ‘no’ and perhaps ‘do not recall.’But if Kit Kat wants to know how much consumers like their new chocolate bar, then a scaled-response approach is better

Page 18: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

18

Previous Research Studies

Sometimes a survey follows an earlier one, and there may be a desire to compare new findings with the previous survey

Researchers may want to try to use question formats that are tried and used before, proven to be reliable and valid and suits the purpose of the current study, it is then a good practice to adopt or adapt it rather than ‘re-inventing the wheel’

Page 19: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

19

Data Collection ModeSome data collection modes are better suited for certain question formats

It can be difficult to administer certain types of response formats (e.g. scaled response type) over the telephone as some require detailed explanation before the respondent can envision the response-scale items

However, mail or other self-administered questionnaires accommodates scaled-response questions well because the respondent can see the response categories on the questionnaire itself

Page 20: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

20

Ability of the RespondentSome respondents may relate better to one type of response format than another

For instance, if the respondent is not as articulate or may be reluctant to verbalize their opinions, open-ended option is not a wise move

Some respondents may not be used to rating objects on numerical scales (unlabeled format). It is then appropriate to use a label format, or perhaps to move back to a ‘yes’ ‘no’ dichotomous closed-ended question format

Page 21: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

21

Scale Level Desired

If response options are simply ‘yes’ or ‘no’, the researcher can report the percentages - (60% answered ‘yes’, 40% answered ‘no’)

But if the question asks how many times respondents used an ATM machine during the past month, the researcher could calculate an average number of times e.g. 8.5 times a month

An average is different from a percent, dichotomous yes-no response is less informative than a scaled-response option such as ‘0’, ‘1’ or ‘2’…

Page 22: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

22

Basic Concepts in Measurement

In research, what are we really measuring?We are measuring ‘properties’ – ‘attributes’ or ‘qualities’ – of objectsObjects refer to consumers, brands, stores, advertisements or whatever of interests to the researcher in the research projectProperties are the specific features or characteristics of an object that can be used to distinguish it from another objectIn research, measurement is determining HOW MUCH of a property is possessed by an object

Page 23: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

23

MEASUREMENT IN MARKETING RESEARCH

Basic Concepts in MeasurementBasic Concepts in Measurement

• Objective PropertiesObjective Properties

• Subjective PropertiesSubjective Properties

Page 24: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

24

MEASUREMENT IN MEASUREMENT IN MARKETING RESEARCHMARKETING RESEARCH

Objective properties are physically verifiable characteristics such as age, income, gender, race, number of times xxx etc. Easy to measure

Sometimes market researcher often desire to measure subjective properties which cannot be directly observed e.g. a person’s attitudes, beliefs or intentions

Page 25: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

25

MARKETING RESEARCH DONE MARKETING RESEARCH DONE BY A DRINK MANUFACTURERBY A DRINK MANUFACTURER

Object Mr. Kiasu Ms. Heow

PropertiesAge 20 32Annual Income $40,000 $25,000Gender Male FemaleCanned DrinkLast bought Coca-cola PepsiEvaluationOf ‘our’ brand Good FairIntention to buyOur brand Strong Weak

Page 26: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

26

Levels of Measurement of Scales

4 typesNominal Scale – using only labels/characteristics of descriptionsOrdinal Scale – ability to rank-orderInterval Scale – distance between descriptor is knownRatio Scale – ones in which a true zero origin exists

Page 27: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

27

Nominal Scaled Question

Please indicate your housing type 1) HDB 2) Private Apts

Please indicate your gender 1) Male 2) Female

Ordinal Scaled Question

Please rank your preference of the following mobile phone

brands (1 – most preferred to 3 – least preferred)

Nokia _____

Motorola _____

Siemens _____

In your opinion, can you say the prices of NTUC are

_____ higher than Carrefour

_____ about the same as Carrefour

_____ lower than Carrefour

Page 28: 1 Lesson 3 Measurement & Attitude Scales in Marketing Research

28

Very Poor Very Good

Toyota 1 2 3 4 5

Mitsubishi 1 2 3 4 5

Nissan 1 2 3 4 5

Interval Scaled Questions

Please rate each car brand in terms of its overall performance

Ratio Scaled Questions

1. Please indicate your weight. _______ kg

2. About how many times have you spend more than $10 in

NTUC last week?

0 1 2 3 4 5 6 7 >7 (please specify_____)

3. What’s the probability that you will take a cab if you want

go to Orchard Road?

_______ percent