Research Design(조사설계)

Week 4
Professor Weon Sang YOO
School of Business
Marketing Research Research Design

Agenda

• Research Design: An Overview

• Exploratory Research (탐색적 조사)
• Descriptive Research (기술적 조사)
• Causal Research (인과적 조사)

Prof. Weon Sang YOO 1

Marketing Research Research Design

Research Design

• Framework (틀) or plan (계획) for a study as a

guide to collect and analyze data (Detailed
blueprint (설계도) to guide a study towards its
objectives)

• Key Question: Which is the most appropriate (

적절한) research method(s) to solve the current
marketing problems / research objectives?

Prof. Weon Sang YOO 2

Marketing Research Research Design

Types of Research Design

Exploratory Descriptive Causal

Discover Ideas and Describe Trends or Determine Cause and
Insights Relationships Effect

Our Notebook PC Who buys our Does ad influence

sales decrease. We Notebook PC and who Notebook
don’t know why. buys our PC purchases?
Time competitor’s?

Uncertainty

More specific hypothesis

Qualitative

Prof. Weon Sang YOO 3

Marketing Research Research Design

Exploratory Research
(탐색적 연구)

Prof. Weon Sang YOO 4

Marketing Research Research Design

Objectives (목적) of Exploratory Research

• Formulating Problems More Precisely (보다 정확한

문제 정의)
– Sales down → only us? which region? repeat
purchase vs. trial?
• Developing Hypotheses (가설 개발)
– Sales down → High price? Poor service quality?
• Establishing Priorities for Further Research (향후 조
사의 우선 순위 설정)
– Which attribute is more important?

Prof. Weon Sang YOO 5

Marketing Research Research Design

Objectives (목적) of Exploratory

Research (Cont’d)

• Clarifying Vague Concepts and Definitions (애매한

개념이나 정의를 명확히 함)
– Loyalty? Satisfaction? Fairness?

• Qualitative Research (질적) with Flexibility (유연성):

Feelings, Thoughts, Intentions, Attitudes, etc.

Prof. Weon Sang YOO 6

Marketing Research Research Design

Examples
• A manufacturer observes (관찰) sales of their new
baby frozen food (유아용 냉동 식품) are not as
high as they expected, but is not sure about the
reasons.
• A luxury home builder (고급 주택 건축 업자)
wonder what features customers most want
• A restaurant is wondering if it has to improve its
interior (인테리어) design or re-design its menu.
• A hospital wants to do some survey to measure
the patient satisfaction (환자 만족도), but it is not
clear which items could be included in the
questionnaire.
Prof. Weon Sang YOO 7
Marketing Research Research Design

Types of Exploratory Research

• Without Interactions (상호작용 부재)

– Secondary Data Analysis (2차 자료 분석)
– Case Analysis (사례 분석)
– Observational Methods (관찰법)

• With Interactions (상호작용)

– Individual In-depth Interviews (심층면접)
– Focus Group Interviews (중심집단면접)
– Projective Techniques (투사법)
Prof. Weon Sang YOO 8
Marketing Research Research Design

Secondary Data Analysis

• One of the quickest and cheapest ways (e.g., Google)

• Academic Journals (학술지 조사)
– Journal of Marketing Research
– Journal of Marketing
– Marketing Science
– Journal of Consumer Research
• Trade Publications (전문 매체) (e.g., PC World)
• Published Statistics (통계자료) (e.g., 통계청)
• Library Databases (e.g., Business Source Premier)

Prof. Weon Sang YOO 9

Marketing Research Research Design

Case Study (사례 연구)

• Learn from others

e.g., Factors for success (성공요인) in restaurant
business : compare some successful ones with other
failed ones
• Searching new “Insights (통찰력)” not just an
“Explanation (설명)” → See the Big Picture
• Productive Cases Include
– Changes (변화)
– Extreme Comparisons (극단적 대조)
– Sequence of Events (순차적 사건)

Prof. Weon Sang YOO 10

Marketing Research Research Design

Descriptive Research
(기술적 조사)

Prof. Weon Sang YOO 11

Marketing Research Research Design

Objectives of Descriptive Research

• Describing consumer characteristics (소비자 성격)
e.g. Weight Watchers’ average customer
Woman about 40 years old
Household income of about $50,000
At least some college education
• Estimating the proportion (비율) of people who have certain
behavior (행동) or attitudes (태도)
– How many people support Donald Trump?
• Identifying consumer evaluations (평가) or attitudes
– How do my customers evaluate my product comparing to
competitors’ products?
• Making Specific Predictions (예측)
– What would be the demand level for tablet PC over the next five
years?
Fact finding!! (Who, What, Where, When and How)
Prof. Weon Sang YOO 12
Marketing Research Research Design

Types of Descriptive Research

Continuous
Panel

Longitudinal
(종단조사)

Descriptive
Studies
Cross-
Sectional
(횡단조사)

Sample
Survey
Prof. Weon Sang YOO 13
Marketing Research Research Design

Longitudinal Analysis (종단 분석)

• Measuring same variables over time (같은 변수를 시

간차를 두고 측정)
– More (demographic) information about the subjects
– More accurate by repeated measurements (반복 측정)
– Less bias (편향) due to the interaction b/w an
interviewer and the respondent
– More Insights (통찰력)
· The long-term relationship among variables
· Managerial implications (시사점) from repeatedly-recorded
behavior (e.g., brand switching matrix: 상표변경 행렬표)

Prof. Weon Sang YOO 14

Marketing Research Research Design

Longitudinal Analysis (Cont’d) (종단 분석)

• Concerns (위험성)
– Hard to Recruit “Representative (대표성 있는)”
Panels (응답자 집단)
– Panel Attrition (사라진 패널)
– Cost (비용)
• Continuous Panel (연속적 응답집단) (Time Series
Analysis: 시계열 분석)
– Relies on repeated measurements of the same
variables and the same respondents
Prof. Weon Sang YOO 15
Marketing Research Research Design

Brand-Switching Matrix (상표변경 행렬표)

Brand Purchased 2018 2023
Japan 200 300
U.S. 300 200
Europe 400 300
Others 100 200

1000 1000
v.s.
2023
Brand Purchased Japan U.S. Europe Others
Japan 100 50 50 0 200
U.S. 100 100 50 50 300
2018
Europe 100 40 200 60 400
Others 0 10 0 90 100
300 200 300 200 1000
Prof. Weon Sang YOO 16
Marketing Research Research Design

Cross-Sectional Analysis (횡단 분석)

• Most frequently used (가장 대중적)

• A snapshot (단면) at a single point in time

• Sample selection (표본 선정) is a key! (Sample

Survey)

• Good tabulation (집계표 작성법) is necessary

Prof. Weon Sang YOO 17

Marketing Research Research Design

Causal Research
(인과 조사)

Prof. Weon Sang YOO 18

Marketing Research Research Design

Causal Research (인과 조사)

• If-Then analysis (혹은 What-if 분석)
– If I change my Website design, then would consumer trust on purchasing
online from my online store improve?
– If I lower the price, then what happen to my market share?
• The Scientific Concept of Causality (과학적 인과성의 개념)
– One of many possible causes (하나의 결과에 여러 개의 원인이 존재 가능)
– Probabilistic (확률적)
– Based on inference (추론에 근거한)
• The Evidence of Causality in MR
– Concomitant (accompanying or coexisting) Variation
– Time Order of Occurrence of Variables
– Elimination of Other Possible Causal Factors

We often need “controlled (통제된)” experiments for “Why” questions.

Prof. Weon Sang YOO 19

Marketing Research Research Design

Experimentation (실험)

• A Good Method for Causal Research due to Great

Controllability (통제성)
– Random Sampling (무작위 표본추출)
– Random Assignment (무작위 배정)
– Causal Factor Manipulation (원인 요인 조작)
• Types of Experiments
– Laboratory Experiment (실험실 실험)
• “Created” Situation (인위적 환경)
• 예: Shopping Trip Simulation
– Field Experiment (현장 실험)
• More Realistic Situation (No Special Conditions)
• 예: Recording Weekly Sales of Two Towns Having Different Prices

Prof. Weon Sang YOO 20

Marketing Research Research Design

Experimental Design: Examples

One-Shot Case Study

(일회성 사례 연구) Observation

X O
Manipulation

• Convenience sample (임의 표본)

• Little control
• Exploratory research
• No causal relationship can be tested
Prof. Weon Sang YOO 21
Marketing Research Research Design

Experimental Design: Examples

One-Group Pretest-Posttest
(단일 집단의 전후 비교) Posttest

O 1 X O2
Pretest

• Convenience sample (임의 표본)

• Measuring marketing effectiveness by O2 – O1
• Be cautious in controlling extraneous (외부적)
factors (e.g., other ads)
Prof. Weon Sang YOO 22
Marketing Research Research Design

Experimental Design: Examples

Before-After with Control Group
(통제 집단을 사용한 전후 비교)

(R) O1 X O2 (Experimental Group)

(R) O O
3 4 (Control Group)

• Measuring the impact of experimental stimuli (실험

자극) by (O2 – O1) – (O4 – O3)
• Random assignment (임의 배분) is important
• Eliminating extraneous (외부적) factors
Prof. Weon Sang YOO 23
Marketing Research Research Design

How Valid (타당) Are Experiments?

• Internal validity (내적 타당성): is the result due to the

experimental variables, not other factors? (High in
Lab Experiment)

Factors Undermining Internal Validities

– Extraneous Factors (외부적 요인): External Events such
as Competitors’ reaction, Seasonality, Strike
– Instrument Variation (다양한 실행 도구): Variations in
Administration (e.g., Telephone Survey, Interviews)

Prof. Weon Sang YOO 24

Marketing Research Research Design

How Valid Are Experiments? (Cont’d)

– Statistical Regression (통계적 회귀): Tendency of

“extreme (극단적)” cases to move closer to the average
during an experiment (e.g., Alcohol)
– Selection Bias (편향된 선택): The sample does not well
represent the population → Solved by Matching &
Randomization (무작위화)
– Experimental Mortality (실험 참가 인원의 손실 ): Loss of
Test Units during an Experiment (e.g., Subjects
Attrition)

Always think about “alternative” explanations!!

Prof. Weon Sang YOO 25
Marketing Research Research Design

How Valid Are Experiments? (Cont’d)

• External validity (외적 타당성): Can the result be

generalizable (일반화)? (High in Field Experiment)

External Validity Concerns

– Representative Sample (대표성 있는 표본)
– Realism (현실성)
– Generalizability (일반화의 가능성)

Prof. Weon Sang YOO 26

Marketing Research Research Design

Summary of Research Designs

Exploratory Descriptive Causal

Objective Discovery of ideas and Describe market Determine cause and
insights characteristics or effect relationships
functions
Character Flexible Marked by the prior Manipulation of one or
-istics Versatile formulation of more independent
Often the front end of specific hypothesis variables
total research design Preplanned and
structured
design

Methods Expert surveys Secondary data Experiments

Secondary data Surveys
Case studies Panels
Pilot studies
Qualitative research

Prof. Weon Sang YOO 27

Marketing Research Research Design

Next Class: Data Collection Methods

• Primary Data
• Secondary Data

Prof. Weon Sang YOO 28