Business Research : Data Collection and Questionnaire Construction

Business Research : Data Collection and Questionnaire Construction

1. 11
Data Collection &Data Collection &
Questionnaire ConstructionQuestionnaire Construction
Ravi Shankar RaiRavi Shankar Rai

2. 22
Data CollectionData Collection

Two basic methods of collectionTwo basic methods of collection
–Ask them or Observe themAsk them or Observe them
–Questioning & ObservationQuestioning & Observation

Questioning: Less expensive, Quick, Questioning: Less expensive, Quick,
VersatileVersatile

Observation: Expensive, Difficult to Observation: Expensive, Difficult to
set up, Time consumingset up, Time consuming

Observation Data are more Valid & Observation Data are more Valid &
Reliable than QuestioningReliable than Questioning

3. 33
Why a Questionnaire is Why a Questionnaire is
needed?needed?

To standardize the process of data To standardize the process of data
collection – helps in analysiscollection – helps in analysis

To achieve speed & accuracy in To achieve speed & accuracy in
collection & recordingcollection & recording

To achieve speed & accuracy in To achieve speed & accuracy in
handling data within & between handling data within & between
offices for analysesoffices for analyses

4. 44
Situations suitable for Situations suitable for
QuestioningQuestioning

Knowledge, opinion, motivation, Knowledge, opinion, motivation,
intension etc. are not open for intension etc. are not open for
observationobservation

Past events (like time & quantity of Past events (like time & quantity of
last purchase) can be studied only last purchase) can be studied only
by questioning by questioning

5. 55
Questionnaire Preparation Questionnaire Preparation
not an easy tasknot an easy task

Respondents tend to answer every Respondents tend to answer every
question – whether they have the question – whether they have the
information or notinformation or not

Response depends on the way the Response depends on the way the
question is askedquestion is asked

Two types of biasesTwo types of biases
–Interviewer biasInterviewer bias
–Respondent biasRespondent bias

6. 66
Sources of Error in Sources of Error in
QuestionnairesQuestionnaires

Memory: Respondents cant remember Memory: Respondents cant remember
correctlycorrectly

Motivation: To project a better self image, Motivation: To project a better self image,
respondents may be motivated to respondents may be motivated to
respond to all questions even incorrectlyrespond to all questions even incorrectly

Communication: May not understand the Communication: May not understand the
questionquestion

Knowledge: Simply may not knowKnowledge: Simply may not know
Attitude measurements are more prone to Attitude measurements are more prone to
errorerror

7. 77
Classification of Classification of
Questionnaire StudiesQuestionnaire Studies

Three bases of classifying Three bases of classifying
questionnaire studiesquestionnaire studies
–Degree of ‘Degree of ‘Formalization Formalization ’ or ‘’ or ‘Structured Structured ’’
–Degree to which the objective is Degree to which the objective is
‘‘Disguised Disguised ’ (In Direct) or ‘’ (In Direct) or ‘NonNon Disguised Disguised ’ ’
(Direct)(Direct)
–Communication method used: Phone, Communication method used: Phone,
Mail, Interview, InternetMail, Interview, Internet
First two bases are considered togetherFirst two bases are considered together

8. 88
Classification of Classification of
Questionnaire StudiesQuestionnaire Studies
StructuredStructured Non- StructuredNon- Structured
Non-Non-
DisguisedDisguised
Most Widely usedMost Widely used
Focus GroupFocus Group
Depth InterviewDepth Interview
DisguisedDisguised
Attitude Attitude
MeasurementMeasurement
Projective TechniqueProjective Technique

9. 99
Structured & Non - DisguisedStructured & Non - Disguised

Because StructuredBecause Structured
–Easy to tabulate & analyzeEasy to tabulate & analyze
–More reliable resultsMore reliable results
–Can be handled over phone, mail or personal Can be handled over phone, mail or personal
interviewsinterviews
–Less Interviewer biasLess Interviewer bias

Because Non – DisguisedBecause Non – Disguised
–Involves respondentsInvolves respondents
–More Respondents biasMore Respondents bias

10. 1010
Non – Structured & Non – Structured &
Non - DisguisedNon - Disguised

Because Non – DisguisedBecause Non – Disguised
–Respondents may not reveal baser intentionsRespondents may not reveal baser intentions
–May not have clear idea why last purchase May not have clear idea why last purchase
was madewas made

Because Non – StructuredBecause Non – Structured
–Takes longer timeTakes longer time
–Yields more info: Focus GroupYields more info: Focus Group
–Interviewer biasInterviewer bias
–Difficult to compare resultsDifficult to compare results

11. 1111
Structured & DisguisedStructured & Disguised

Because structuredBecause structured
–Eliminates interviewer biasEliminates interviewer bias
–Interpretation is uniform & easyInterpretation is uniform & easy
–Less time & Lower costLess time & Lower cost

Because disguisedBecause disguised
–Removes respondent biasRemoves respondent bias
–Desires latent in unconscious mind can Desires latent in unconscious mind can
be uncoveredbe uncovered

12. 1212
Non – Structured & DisguisedNon – Structured & Disguised

Because non – structuredBecause non – structured
–Yields more informationYields more information
–Interviewer biasInterviewer bias
–Takes longer time Takes longer time
–Results can not be comparedResults can not be compared

Because disguisedBecause disguised
–Helps uncovering subconscious feelings & Helps uncovering subconscious feelings &
socially unacceptable attitudes & motivessocially unacceptable attitudes & motives
–Removes respondent biasRemoves respondent bias

13. 1313
ObservationObservation

Generally not employed for one time studyGenerally not employed for one time study

For data collection at regular interval For data collection at regular interval
researchers use observation: Television researchers use observation: Television
viewing, Purchases made in shopping mallsviewing, Purchases made in shopping malls

Advantages:Advantages:
–Willingness & Accuracy not relevant. Observers Willingness & Accuracy not relevant. Observers
watch actual actionswatch actual actions
–Interviewer bias eliminatedInterviewer bias eliminated

DisadvantagesDisadvantages
–Can not observe attitude, Opinion, Belief, Can not observe attitude, Opinion, Belief,
Knowledge, MotivationKnowledge, Motivation
–Higher cost than questionnaire method Higher cost than questionnaire method

14. 1414
Methods of ObservationMethods of Observation

Classified on five basesClassified on five bases
1.1.Situation is Natural or ContrivedSituation is Natural or Contrived
2.2.Observation is Obtrusive or UnobtrusiveObservation is Obtrusive or Unobtrusive
3.3.Topic of interest observed Directly or Topic of interest observed Directly or
IndirectlyIndirectly
4.4.Observation is Structured or Un-Observation is Structured or Un-
StructuredStructured
5.5.Mechanical means employed or notMechanical means employed or not

15. 1515
Classification of Classification of
Questionnaire StudiesQuestionnaire Studies

By methods of communicationBy methods of communication
–Telephone, Mail, Interview at home or Telephone, Mail, Interview at home or
mall/office, Internetmall/office, Internet

Select a method based on cost of Select a method based on cost of
obtaining obtaining accurateaccurate information information

Factors affecting accuracyFactors affecting accuracy
–Character of populationCharacter of population
–Subject of interestSubject of interest
–Questionnaire length & complexityQuestionnaire length & complexity
–Geographic spreadGeographic spread
–Time & Budget availableTime & Budget available
–Extent of validity checksExtent of validity checks

16. 1616
(1)
Survey
Objectives
(2)
Data
Collection
Methods
(3)
Determine
Content
(4)
Types of
Question
(5)
Decide
Wording
(6)
Decide
Sequence
(7)
Pre-Code
& Obtain
Approval
(8)
Decide
Layout
(9)
Pre Test
(10)
Revise &
Implement
Questionnaire Design StepsQuestionnaire Design Steps

17. 1717
10 Steps of Questionnaire 10 Steps of Questionnaire
ConstructionConstruction
1.1.Specify objectives: List of needed Specify objectives: List of needed
informationinformation
2.2.Decide method of data collectionDecide method of data collection
3.3.Determine content of individual Determine content of individual
questionsquestions
4.4.Determine type of questions to useDetermine type of questions to use
5.5.Decide wording of the questionsDecide wording of the questions

18. 1818
6.6.Decide question sequenceDecide question sequence
7.7.Pre-Code & Obtain approvalPre-Code & Obtain approval
8.8.Decide layoutDecide layout
9.9.Pre Test: Pilot surveyPre Test: Pilot survey
10.10.Revise & ImplementRevise & Implement
10 Steps of Questionnaire 10 Steps of Questionnaire
ConstructionConstruction

19. 1919
Steps 1 & 2: Specify objectives & Steps 1 & 2: Specify objectives &
Decide method of data collectionDecide method of data collection

Specify survey objectivesSpecify survey objectives
–Focused, Outcomes measurable & Focused, Outcomes measurable &
Within available resourcesWithin available resources
–Objectives will determine the list of Objectives will determine the list of
needed informationneeded information
–Must be comprehensiveMust be comprehensive

Methods of Data Collection: Personal Methods of Data Collection: Personal
Interview, Mail, Phone, InternetInterview, Mail, Phone, Internet

20. 2020
Step 3: Determine Contents of Step 3: Determine Contents of
Individual QuestionsIndividual Questions

Is the question necessary: Check with Is the question necessary: Check with
objectiveobjective

Does respondent have the informationDoes respondent have the information

Is the Question asked within respondent’s Is the Question asked within respondent’s
experienceexperience

Can he remember: we may use aided recallCan he remember: we may use aided recall

Will the respondent give the infoWill the respondent give the info

Are several questions needed to replace oneAre several questions needed to replace one

21. 2121

Types of questionsTypes of questions
–Open questionsOpen questions
–Multiple choice questionsMultiple choice questions
–Dichotomous questionsDichotomous questions
Step 4: Determine the type of Step 4: Determine the type of
QuestionsQuestions

22. 2222
Step 4: Determine the type of Step 4: Determine the type of
QuestionsQuestions

Open QuestionsOpen Questions
–AdvantagesAdvantages

Good as first questionGood as first question

Involves respondentsInvolves respondents

Every one can answerEvery one can answer
–DisadvantagesDisadvantages

Long answerLong answer

Difficult to analyze: Brings too much Difficult to analyze: Brings too much
varietyvariety

Introduces Interviewer biasIntroduces Interviewer bias

23. 2323

Multiple choiceMultiple choice
–AdvantagesAdvantages

Eliminates interviewer biasEliminates interviewer bias

Quick and easy to answerQuick and easy to answer

Easy to tabulate & analyzeEasy to tabulate & analyze
–DisadvantagesDisadvantages

Introduces ordering biasIntroduces ordering bias
(Explain Split Ballot)(Explain Split Ballot)
Step 4: Determine the type of Step 4: Determine the type of
QuestionsQuestions

24. 2424

Dichotomous questions: questions Dichotomous questions: questions
with only two possible answerswith only two possible answers
–AdvantagesAdvantages

Easy to answerEasy to answer

No interviewer biasNo interviewer bias

Easy to tabulate & analyzeEasy to tabulate & analyze
–DisadvantagesDisadvantages

Not all questions lend themselves to such a Not all questions lend themselves to such a
formatformat
Step 4: Determine the type of Step 4: Determine the type of
QuestionsQuestions

25. 2525
Step 5: Decide wordingStep 5: Decide wording

Define the issue: Who, What, When, Why, Define the issue: Who, What, When, Why,
Where & How?Where & How?

Subjective versus objectiveSubjective versus objective

Positive or negativePositive or negative

Use simple wordsUse simple words

Avoid ambiguous words e.g., Avoid ambiguous words e.g., Which brand Which brand
of cigarette do you smoke regularly?of cigarette do you smoke regularly?

Questions about personal information Questions about personal information
need to be buffered with an explanation need to be buffered with an explanation

26. 2626
Step 6: Decide question Step 6: Decide question
sequencesequence

Opening questions must win Opening questions must win
respondent’s interestrespondent’s interest

Easy to difficultEasy to difficult

Logical orderLogical order

No cross referencingNo cross referencing

Demographic questions at the endDemographic questions at the end

27. 2727
Step 7: Pre code & Obtain Step 7: Pre code & Obtain
approvalapproval

Each questionnaire must contain a code Each questionnaire must contain a code
to help in tracing up to the sourceto help in tracing up to the source

To facilitate data entry & analysis, assign To facilitate data entry & analysis, assign
codes to each question & its possible codes to each question & its possible
answersanswers

Develop a systematic approach about Develop a systematic approach about
analysesanalyses

Analyses can be done even with cross Analyses can be done even with cross
referencing (AND & OR concepts of Wenn referencing (AND & OR concepts of Wenn
Diagram) of questionsDiagram) of questions

28. 2828
Step 8: Decide layout Step 8: Decide layout

Layout should be such thatLayout should be such that
–Secures acceptanceSecures acceptance
–Easy to controlEasy to control
–Easy to handleEasy to handle

Font size used should be easy to readFont size used should be easy to read

Easy to fill with minimum effort on Easy to fill with minimum effort on
respondents’ partrespondents’ part

Should not appear to be bulkyShould not appear to be bulky

Should withstand the wear & tear of Should withstand the wear & tear of
handling by multiple peoplehandling by multiple people

29. 2929
Steps 9 & 10: Pre test, Revise Steps 9 & 10: Pre test, Revise
& Implement& Implement

Pre-testing is a mustPre-testing is a must
–Faults in wording, sequence, layout etc. Faults in wording, sequence, layout etc.
can be identified & eliminatedcan be identified & eliminated

Revise Revise

Implement Implement