Data Sources and Computer Research
Data Sources and Computer Research
COMPUTER
RESEARCH
TYPES OF DATA
PRIMARY
SECONDARY
TERTIARY
PRIMARY DATA
Primarydata are generally information
gathered or generated by the researcher for
the purpose of the project immediately at
hand.
Primary data collection
method
•Observation
•Focus Group
•Personal interview
•Telephone interview
•Self – administered (mail or internet) surveys
Observation
Cheaper and effective
Data collected by simply observing
No need of asking the respondents
EXAMPLE
Which is the popular Laptops among college students
What is the trend of clothes in Chennai
What time it takes a vehicle to clear the check post
Which movie is the block bluster
Focus Group
The people who are interested in your
discussion topic are identified and
subsequently called for seminar/meeting
EXAMPLE
Increase in the taxes on a particular group
say transporters
Ban of tobacco in a state
New government policy against the small
scale industry
Personal interview
One to One interview is been taken
In- depth information being collected
Interview can be audio or video recorded
EXAMPLE
Interview of politicians for the forecast of the election
polls
Interview of Industry personal for the new change in
the government laws for imports
Knowing the famous TV channel among teenagers
Which is most used products ex: perfumes, shoes,
cosmetics and why?
Telephone interview
Quickest way of collecting data
Telephonic interview can be recorded for evidence
Some time not reliable
Government restrictions
EXAMPLE
Asking customers of HUTCH are they happy with
the billing pattern of their company
Grievances if any after the use of any washing
machine, Grinder, television ……
Self administered (mail or
internet) survey
Inthis type of collection of data mails are
send across users of a free mail service who
are not registered to a particular group
EXAMPLE
Polls on who will the next CM of Tamil nadu
Which is favorite breakfast Topramen
noodles or Maggie noodles
Surveys
Survey is a process for gathering information,
without detailed verification, on the activity being
examined. The main purposes are to:
Understand the activity under review
QUALITY OF DATA
INCIDENCE RATE
Telephone
Quality of Data
Survey’s to produce valid data
Quality of data affected by:
Questionnaire design
Sampling methods
Scaling techniques
Interviewer qualification & Training
INCIDENCE RATE
Percent of total population that fit the qualification of
people interviewed
Being costly face to face method
RESPONDENT
ADMINISTRATIVE ERROR
ERROR
RESPONSE BIAS
NON RESPONSE
ERROR
SAMPLE SELECTION
ERROR
DELIBERATE FALSIFICATION UNCONSCIOUS
MISREPRESENTATION
SAMPLE FRAM
ERROR
ACQUIESCENCE BIAS
POPULATION
SPECIFICATION
ERROR
EXTREMITY BIAS
DATA PROCESSING
ERROR
INTERVIEWER BIAS
INTERVIEWER ERROR
Avoided by
Response bias
Non-Response error
Inclusion of the sample that failed to response is
non-response error
Occur due to
Person not available during preliminary call &
Subsequent call
Caller Id & Answering machine
Busy with other work
Concealing privacy and sensitive issue
Respondent not involved in the survey Example:
Airlines bad experience
Response Bias
A survey error that results from the inclination
of people to answer a question falsely, either
through deliberate misrepresentation is called
Response Bias
Response Bias has two basic form
1)Deliberate Falsification 2)Unconscious
Misrepresentation
Deliberate Falsification
Knowingly giving wrong answers when respondent
are not certain about the facts
Reasons for Deliberate Falsification
Appear intelligent
Conceal information
Time pressure
Courtesy bias
EXAMPLE:
How many time he visited a supermarket, cinema hall.
Conceal true information towards efficiency of their units.
He is afraid from the supervisor
Unconscious
Misrepresentation
Itis a situation where respondent gives
wrong or estimated information dues to
forgetfulness and ignorance even though he
has no intension of doing it.
Example
Misunderstand question
Prior inexperience
Prestige issue
Unexpected question
Different type of biases
Acquiescence Bias
Respondent inclination to be same/opposite to the interviewer so answers
are ‘Yes’ or ‘No’
Extremity Bias
Respondent either use extreme or neutral answers
Interviewer Bias
Interviewer facial expression, age, gender, tone, etc
Social Desirability Bias
Answers to be socially desirable or acceptable to create favorable image.
Example
Education information overstated
Salary overstated
Fields of Application
Sociology
Economics
Behavioural Science
Political Science
Public health
WHAT IS AN INTERVIEW?
AN INTERVIEW IS A PURPOSEFUL DISCUSSION
BETWEEN TWO PEOPLE OR MORE.
HELPS IN GATHERING VALID AND RELIABLE
DATA.
BY WAY OF FORMULATING QUESTIONS IN
QUESTIONNAIRE.
KEEPING IN MIND REASERCHERS QUESTION
AND OBJECTIVES.
CHARACTERIZATION OF
INTERVIEW
CONSISTS OF FOUR ENTITIES
RESEARCHER
INTERVIEWER
INTERVIEWEE
INTERVIEW ENVIRONMENT
TYPES OF INTERVIEW
STRUCTURED
SEMI-STRUCTURED
UNSTRUCTURED
STANDARDIZED
NON-STANDARDIZED
RESPONDENT
INFORMANT
INTERVIEW METHOD
NON-DIRECTIVE
DIRECTIVE
FOCUSED
IN-DEPTH
COMMON METHODS IN
INTERVIEW
PERSONAL INTERVIEW
TELEPHONIC INTERVIEW
SELF-ADMINISTERED INTERVIEW
MAIL SURVEY
Personal interview
One to One interview is been taken
In- depth information being collected
Interview can be audio or video recorded
EXAMPLE
Interview of politicians for the forecast of the
election polls
Interview of Industry personal for the new
change in the government laws for imports
Knowing the famous TV channel among
teenagers
Which is most used products ex: perfumes,
shoes, cosmetics and why?
Telephone interview
Quickest way of collecting data
Telephonic interview can be recorded for evidence
Some time not reliable
Government restrictions
EXAMPLE
Asking customers of HUTCH are they happy with
the billing pattern of their company
Grievances if any after the use of any washing
machine, Grinder, television ……
CLASSIFICATION OF
PERSONAL INTERVIEW
DOOR TO DOOR INTERVIEW
EXECUTIVE INTERVIEW
LOGICALSTARTING POINT
PERSONALISED APPROACH
DISAVANTAGES OF PRIMARY
DATA
EXPENSIVE
BIASED
DIFFICULT
TO FORM A SAMPLE
REPRESENTATIVE
NON-COOPERATION
ADVANTAGES OF DOOR-TO-
DOOR INTERVIEW
INSTANT FEEDBACK
IMPROVED QUALITY
CLARITY
REPRESENTATIVE POPULATION
DISTANCE
UNEXPECTED HINDERANCE
ACCESSIBILITY
ADVANTAGES AND DISADVANTAGES
OF EXECUTIVE INTERVIEW
BETTERCONTROL
RESPONDENTS REPLY AT THEIR
CONVENIENCE
DISADVANTAGE
NON-RESPONSE RATE
QUESTIONNAIRE
IT IS THE LIST OF QUESTIONS
QUESTIONNAIRS
CHOICEOF QUESTIONNAIR
COSTRUCTIO OF QUESTIONNAIRS
INFLUENCING FACTORS
CLOSED QUESTIONS
MULTIPLE
LIST CATAGORY RANKING
QUESTION
CONSTRUCTION OF A
QESTONNAIRE
Defining the research problem
Target respondents
Give a clear perception to the respondents
Should be practicable ad realistic
Should be simple and should have less jargons
Should be concerned with a single idea
Should follow the FUNNEL METHOD
Necessary items should be incorporated
QUESTIONNAIRE SEQUENCE
Lead in question
Qualifying question
Warm up questions
Specific question
Demographic question
PRETESTING
REVSION
FINAL DRAFTING
COMPUTERISED QUESTIOAIRE
The questions should be in a format
Data to be coded
before entry
Coding scheme
TYPES OF CODING
OPTION 2 5 4 3 2 1
PURPOSE OF THE QUESTIONNAIRE
PILOT TESTING :
1. Helps to obtain some assessment of the questions ‘ validity,
reliability of the data collected.
2. Helps to ensure that the data collected enables to answer the
questions
Administering The
Questionnaire
After the questionnaire is designed ,pilot-tested and
amended and samples is selected , it can be used to
collect data.
This final stage is called ADMINISTERING THE
QUESTONNAIRE.
Postal Questionnaire
2 Important things:
1. clear timetable that identifies the task to be done
2. the resources that are required
Advantage:
Considers changes occurring over time
Great co operation from respondents
Warranty Cards
Distribution or Store Audits.
Pantry Audits.
Consumer Panel.
Hard copy
Books, magazines, journals, special collection
Internet
Microsoft internet explorer, Netscape Navigator
CLASSIFICATION BY
DATABASE
Collectionof information in a detailed and
standard format.
Online database : central data bank
Offline database : diskettes, CD-ROMS
Internet
Further classified as
Reference databases
Source databases
SECONDARY DATA
ADVANTAGES:
Identify, clarify and redefine the research problem
Alternate methods that can be used for primary research
Generates information for better creativity
DISADAVANTAGES:
Lack of availability
Lack of relevance
Inaccurate data
Syndicated Data
Data is produced by a market research firm
Not client-specific
Periodic
Retail Scanner Panels Consumer Panels
Panel Volume Tracking
Data Dairy
Periodic
Panel Survey
Shared Surveys
Audits
In-depth analysis of the existing situation in a
firm
Performance
Stores
Product
Retail
Pantry
Uses of Audits
Assessing brand shares and competitive
activity
Identifying inventory problems
2.TIME TREND
CROSS SECTIONAL
COMPARISON
SARASOTA PASCO LEE
Laddering
Hidden issue questioning
ADVANTAGES OF DEPTH
INTERVIEWS
Independent responses
Attitudes and emotions can be explored in
detail
Possible to determine motivation
Mutual interaction between the interviewer
and interviewee.
DISADVANTAGES OF DEPTH
INTERVIEWS
Difficult
and expensive
Lack of structure in questionnaire
Constructive techniques
Expressive techniques
Sociometry
ASSOCIATIVE TECHNIQUES
Word association
! The frequency of the words given in
response
! The time elapsed before the response
! The number of non-respondents
• Used for brand image or product attributes
COMPLETION TECHNIQUES
Sentence completion
Story completion
CONSTRUCTION TECHNIQUES
Construction of a story, description, dialogue, or a
picture.
Thematic Apperception Test (TAT)
Cartoon techniques
EXPRESSIVE TECHNIQUES
Role playing
This technique reveals the personality of the
respondent.
SOCIOMETRY
Devised by Jacob L.Moreno
mathematical study of psychological properties of
populations
People make choices in interpersonal relationships
ADVANTAGES
Useful in “breaking the ice” in focus group
discussions.
Increased amount of data collection
DISADVANTAGES
Complexity of techniques
expensive
OBSERVATION METHODS
Direct observation
Contrived observation
Content analysis
Participant observation
Despite these, your searches may still be unsuccessful. The most frequent
causes:
1. Is the spelling incorrect?
2. Is the language incorrect?
3. Are you using incorrect terminology?
4. Are you using acronyms and abbreviations?
5. Are you using jargon rather than accepted terminology?
6. Are you using a word that is not in the controlled index
language?
Use a combination of key words linked using Boolean logic.
This enables you to combine, limit or widen variety of items found using
link terms. Boolean logic can also be used to construct searches using dates,
journal titles and names of organizations or people.
• Home pages:
Access to a website is through its home page. The home
page, which has links to other pages in the site, is similar to a title or
contents page. Although they often contain publicity for a company
or institution, they are an excellent way of navigating the site.
• Search tools:
Search tools , often referred to search engines, are
probably the most important way of searching the Internet for
reviewing the literature on your research subject, as they enable you
to locate the most current and up-to-date items.
Most search tools search by key words or subject trees.
• Book marking :
Book marking uses the Internet browser to note
the address so that you will be able to access it again directly. When
sufficient sites have been bookmarked it is possible to arrange then
in the hierarchical order that you want.
Mining Internal Sources
Data warehouse:
A data warehouse is an electronic repository for
databases that organizes large volumes of data into categories to
facilitate retrieval, interpretation, and sorting by end-users.
Data mining:
The term “data mining” describes the process of
discovering knowledge from databases stored in data marts or data
warehouses.
The purpose of data mining is to identify valid, novel,
useful, and ultimately understandable patterns in data.
Data mining is a useful tool, an approach that combines
exploration and discovery with confirmatory analysis.
Techniques:
Numerous techniques are used in data mining; often
they are used together.