Professional Documents
Culture Documents
Session 4: Descriptive Research: Survey/Observation Causal Research: Experiment
Session 4: Descriptive Research: Survey/Observation Causal Research: Experiment
Descriptive Causal
Quantitative Descriptive
Research
SURVEY OBSERVATION
Information Obtained Information Obtained
by Questioning Respondents by Observing Behavior or
Phenomena
Survey Method
• Obtaining information based on questioning respondents
• Involves sampling a large number of people and asking the
m a series of fixed-response questions
• Questions regarding to behavior, intentions, attitudes, awar
eness, motivations, demographics and lifestyle characteristi
cs can be asked
• Advantages: reliability, simplicity, easy to administer
• Disadvantages: unwilling or unable to response, cannot me
asure subconscious nature of respondent’s motive
Criteria for Evaluating
Survey Methods
• 1. Task factors
– Complexity of the questionnaire
– Use of physical stimuli (e.g., product, ads)
– Quantity of data or length of questionnaire
– Response rate (i.e., % of attempted interviews that are co
mpleted)
– Sample control: the ability to reach the desirable sample e
ffectively and efficiently
5
Criteria for Evaluating Survey Methods
• 2. Situational factors
– Control of the data collection environment
– Potential for interviewer bias (i.e., the error due to the interviewer n
ot following the correct interviewing procedures; interviewer’s facial
expressions, intonations, the way they ask the questions)
– Control of field-force (e.g. whether the interviewers can be supervis
ed from a central location)
– Speed: time taken for administering the survey
– Cost (of administering the survey and collecting the data)
• 3. Respondent factors
– Perceived anonymity of the respondents’ identities
– Social desirability: the tendency of the respondents to give answers
that may not be accurate but that may be desirable from a social sta
ndpoint.
6
Classification of Survey Methods
Survey Methods
Prior Other
Incentives Follow-up
Notification Facilitators
Monetary Nonmonetary
Prepaid Promised
Methods of Obtaining Quantitative Data in
Descriptive Research
Figure 7.3 Methods of O
btaining Quantitative D
ata in Descriptive Resea
rch
Quantitative Descriptive
Research
SURVEY OBSERVATION
Information Obtained Information Obtained
by Questioning Respondents by Observing Behavior or
Phenomena
Observation Methods
• The recording of behavioral patterns of people, objects,
and events in a systematic manner to obtain information
about the phenomenon of interest.
• The observer does not question or communicate with th
e people being observed.
How to obverse?
– Mechanical observation
• Use of electronic device to record behavior
– Personal observation
• A trained observer collects the data by
recording behavior exactly as it occurs
• No attempt to control or manipulate the
phenomenon being observed
Data Collection for Observation Methods
Watching
Watching
Listening
Listening
Touching
Touching
Smelling
Smelling
Reading
Reading
What can be observed?
• Physical movements
• Verbal behavior
• Expressive behavior and physiological reaction
s
• Temporal patterns
• Physical objects
• And more…
Examples of what can be observed
In-store At home On the road On the Internet
Arrival time in store TV viewing patterns Use of public Comments made on
transport review sites
Shopper’s movement Family purchasing Driving routes taken Brands mentioned in
pattern in store behavior blogs and on
Facebook
Time spent Brands of products Behavior while driving Links made to
shopping/queuing owned (e.g., use of car seat corporate or celebrity
belts, mobile phones, sites
etc.)
Behavior/expressions Children’s behavior Products used in
when queuing with a new toy videos
Interaction with Clothes worn in videos
service personnel
Products purchased Opinions expressed
Payment methods Trends discussed
used
Chat:
What can NOT be observed?
• The report summariz
es the experience of
a mystery shopper w
ho pretended to be a
potential bank custo
mer
• Source: Marketing Re
search Kit for Dummi
es
Structured Observational Report for retailer
Record #:________
Female Male
First soft drink can picked up for examination:
Pepsi Coca-Cola 7-up Sprite Other: ______
Total # cans picked up for examination, any brand: _________
Descriptive Causal
• Examples
The combination of all THREE conditions does not guarantee a causal relationship. But, an
accumulation of consistent evidence increases the confidence that a causal relationship exists.
Experimentation
• Terminology of experimentation
– Test Units
• Are individuals, organizations, or other entities whose response to IV or trea
tments is being studied.
– Independent Variable (IV) (or experimental treatment) Symbol: X
• Variables that are manipulated by the researcher and whose effects are mea
sured and compared (i.e. the cause)
– Treatment levels
• The different levels of manipulation of IV
– Dependent Variable (DV) Symbol: O
• Variables that measured the effects of the IV on the test units (i.e. the effect,
Y)
– Extraneous Variable
• All variables other than the IV that affect the response of the test units e.g.,
store size, store location, and competitive effort
Experimentation
• Terminology of experimentation
– Experimental Group (EG)
• The group exposed to the manipulated IV
– Control Group (CG)
• The group which is not exposed to the IV manipulation, thus provi
des a point of comparison when examining the effects of these ma
nipulated on the DV
– Randomization Symbol: R
• The random assignment of test units or groups to separate treatm
ents (use to control extraneous variables)
– Matching
• Comparing test units on a set of key background variables before a
ssigning them to the treatment conditions (use to control extraneo
us variables)
Example
Will consumer buy more with in-store audio advertising?
An field experiment
was conducted to
investigate if in-store
audio advertising will
lead to purchase in a
supermarket.
Validity in Experimentation
• Internal validity (Does X -> Y in this test?)
– Whether the manipulation of the independent variables or
treatments actually caused the observed effects on the de
pendent variables
– Internal validity is threatened when the influence of extran
eous variables are mixed with the independent variables
42
Pre-experimental Designs:
1. One-Shot Case Study
EG: X 01
43
Pre-experimental Designs:
1. One-Shot Case Study
Rating of
Name
brand (O1) EG: X 01
Ann 2
Mary 5
Celebrity effect?
compare O1 with a
John 3
normal score (e.g., 3)
Peter 4
…
…
Mean 3.5
44
Pre-experimental Designs:
2. One-Group Pretest-Posttest Design
EG: 01 X 02
45
Pre-experimental Designs:
2. One-Group Pretest-Posttest Design
Pepsi tests the effectiveness of new ad with celebrities:
• A group of consumers first rated their attitude to the pepsi
• After a month, these consumers watch the new ad
• They rated their attitude to the brand again
Rating of Rating of
Name
brand (O1) brand (O2)
Ann 4 2 EG: 01 X 02
Mary 5 5
John 3 3 extraneous variables
Peter 5 4 are largely
… uncontrolled
…
Mean 4.25 3.5
46
True Experimental Designs:
3. Pretest-Posttest Control Group Design
EG: R 01 X 02
CG: R 03 04
47
True Experimental Designs:
3. Pretest-Posttest Control Group Design
Pepsi tests the effectiveness of new ad with celebrities:
• Consumers are randomly assigned to two groups
• Both groups rate their attitude toward the brand
• After a month, only one group watched the new ad
• Both groups rated again their attitude towards the brand
Rating of Rating of
Name
brand (O1) brand (O2)
Group 1 Ann 2 3
(watch new Mary 4 3
People are ad)
Peter 3 4
EG: R 01 X 02
randomly
assigned CG: R 03 04
to two Name
Rating of Rating of
brand (O3) brand (O4)
groups Group 2
(control John 5 5
group) Kitty 4 5
… 48
True Experimental Designs:
4. Posttest-Only Control Group Design
EG : R X 01
CG : R 02
49
True Experimental Designs:
4. Posttest-Only Control Group Design
Rating of
Name
brand (O1)
Group 1 Ann 2
(watch new Mary 4
ad)
People are Peter 3 EG : R X 01
randomly
assigned to Rating of CG : R 02
Name
two groups Group 2 brand (O2)
(control John 5
group) Kitty 4
… 50
Experimental Setting
• Time consuming
• Expensive
• Difficult to administer (e.g., maybe impossi
ble to control for the effects of extraneous vari
ables)
• The end!