Marketing Research: Research Design – Comprehensive Notes

Research Design

• Definition: A master plan that specifies the methods and procedures for collecting and analyzing the information needed for addressing the marketing research problem.

Types of Research Design

• Three types: Exploratory, Descriptive, Causal.
  - Exploratory research is often unstructured and informal, used to gain background information and clarify problems.
  - Descriptive research aims to describe marketing phenomena by answering questions of who, what, where, when, and how.
  - Causal research is designed to determine cause-and-effect relationships, enabling the formation of "if-then" statements.

• Choice of design depends largely on the objectives of the research and how much is known about the problem and research objectives.

• Key concepts: Constructs, variables, independent variable (IV), dependent variable (DV).

Basic Research Objectives and Research Design

• To gain background information → Exploratory design.

• To describe and measure marketing phenomena at a point in time → Descriptive design.

• To determine causality and make if-then statements → Causal design.

• Note: Causal research focuses on conditional statements of the form "If x, then y"; x is the IV and y is the DV. \text{If } x, \text{ then } y

Exploratory Research

• Often unstructured and informal; used to gain background information about the research problem.

• No formal set of objectives, sample plan, or questionnaire in some cases.

• Usually conducted at the outset of research projects.

Exploratory Research: Uses

• Gain background information.

• Define terms.

• Clarify problems and hypotheses (refine research objectives).

• Establish research priorities.

• Involves many questions and many sources; useful for defining the problem and getting a "feel" for the issue.

Exploratory Research: Methods

• Secondary data analysis.

• Experience surveys.

• Case analysis.

• Focus groups.

• Projective techniques.

Exploratory Research: Case Examples

• Patagonia results:

  • Interviews: Patagonia is demographically oriented toward the Northeast and the mountains of the West; implication: consider increasing advertising on social media and television.

  • Focus group: People wear Patagonia for social reasons due to quality and strong brand name (to those aware) and due to expensive price tag; to compete with North Face, increase brand awareness via ads and word of mouth.

• Focus Group/Plagiarism on Twitter:

  • Nine participants, one moderator, two note-takers; demographics: 3 seniors (2 female, 1 male), 6 freshmen (6 male).

  • Moderator introduced theme: plagiarism and Twitter; asked participants to contribute words associated with plagiarism.

  • Responses included: "Cheating, Getting expelled, Copyright, Essay, Copying someone else’s work, English paper, Don’t do it, Worrying about it in English class, Getting in trouble."

  • Mention of The Fat Jewish Instagram account and repercussions for plagiarizing; participants were unaware of repercussions.

  • Hashtag: #MK370.

• Experience survey:

  • 8 participants (4 male, 4 female) interviewed individually.

  • Shown examples of plagiarism; goal: experience plagiarism in professional and social media settings.

  • Examples included: The Fat Jewish, plagiarism headlines from CNN and U.S. sources; hashtag: #MK370.

Descriptive Research

• Objective: Describe answers to questions of who, what, where, when, and how; describe phenomena/events.

• Desirable when projecting findings to a larger population through representative samples.

Classification of Descriptive Research Studies

• Two basic classifications: Cross-sectional studies and Longitudinal studies.

Cross-Sectional Studies

• Measure units from a sample at a single point in time.

• Sample surveys are cross-sectional and representative of a specific population.

• Provide a "snapshot" of the population at one point in time.

Longitudinal Studies

• Repeatedly measure the same sample units over time.

• Often use a panel of respondents who answer questions at periodic intervals.

• Many large research firms maintain consumer panels.

Cross-sectional vs. Longitudinal: Example (Restaurant Visit)

• Data across two surveys (Week 1 vs Week 2) for Wendy’s, McDonald’s, Taco Bell:

  • Week 1: Wendy’s 100, McDonald’s 200, Taco Bell 200; Total 500.

  • Week 2: Wendy’s 75, McDonald’s 200, Taco Bell 225; Total 500.

• Observations:

  • Wendy’s sales decreased from 100 to 75; market share dropped from 20% to 15%.

  • McDonald’s held at 200 (40%).

  • Taco Bell rose from 200 to 225 (40% to 45%).

• Possible conclusion: Wendy’s lost market share to Taco Bell; McDonald’s share remained unchanged.

• Another cross-tab example (Week 1 vs Week 2 totals): illustrates how shares can move and how to interpret changes across time.

Implications of Cross-sectional vs Longitudinal

• Cross-sectional provides a snapshot; cannot track changes over time.

• Longitudinal tracks changes and can reveal true dynamics but is more resource-intensive.

Causal Research

• Demonstrates what an experiment is and how to establish causality.

• Examples: mk370/mk201 class scenarios, Pepsi ad vs no ad, jelly beans experiment (referenced YouTube link).

Causal Research: Causality and IV/DV

• Causality is described by conditional statements: "If x, then y".

• IV: independent variable (the cause) — what the researcher controls/manipulates.

• DV: dependent variable (the effect) — what is measured.

• Examples:

  • If price is lowered, then sales will be higher.

  • If shelf display is simplified, then customers will find it easier to make selections.

• Also known as treatment or program variables for IV; DV is the effect variable.

Experiments

• An experiment manipulates an independent variable to see its effect on a dependent variable while controlling extraneous variables.

• Key elements:

  • Manipulation of the IV (treatment) and a comparison with a control group where the treatment is not applied.

  • Random assignment to experimental and control groups to minimize extraneous influences.

Symbols of Experimental Design

• O = measurement of a dependent variable.

• X = manipulation or change of an independent variable.

• These symbols are used to denote the timing and presence of measurements and treatments in the design.

Pretest and Posttest

• Pretest: measurement of the dependent variable before changing the IV.

• Posttest: measurement of the dependent variable after changing the IV.

A “True” Experimental Design

• A true experimental design isolates the effects of the IV on the DV while controlling extraneous variables.

Not True Experimental Designs

• After-Only Design: X O1 (no pretest; only post-treatment measurement).

• One-Group Before-After Design: O1 X O2 (no control group; pre- and post-measurement with a treatment in between).

True Experimental Design: Before-After with Control Group

• Experimental group: O1 X O2

• Control group: O3 O4

• Additional examples exist beyond the basic configuration.

Validity in Experiments

• Internal validity: the extent to which the observed change in the DV is actually due to the IV and not to extraneous variables.

• External validity: the extent to which the relationship observed in the experiment generalizes to the real world.

Types of Experiments

• Laboratory experiments: IV manipulated and DV measured in a contrived, highly controlled setting to control extraneous variables.

• Field experiments: IV manipulated and DV measured in a real-world setting.

• Guidance: Be creative yet comprehensive in experimental design.

Laboratory vs Field Experiments (Comparison)

• Environment: Artificial (lab) vs Realistic (field)

• Environment control: High vs Low

• Ability to manipulate factors: High vs Low

• Internal validity: High vs Low

• External validity: Low vs High

• Time: Short vs Long

• Number of units: Small vs Large

• Ease of implementation: High vs Low

• Cost: Low vs High

Test Marketing

• Test marketing is an experiment, study, or test conducted in a field setting.

• Uses:

  • Test sales potential for a new product or service.

  • Test variations in the marketing mix (product, price, place, promotion).

• Applicable in consumer markets, industrial B2B markets, and international marketing.

Types of Test Markets

• Standard Test Market

• Controlled Test Markets

• Electronic Test Markets

• Simulated Test Markets

Criteria for Selecting Test Markets

• Representativeness: do demographics match the total market? do product/media use patterns match? is the level of competition comparable? is the distribution network comparable?

• Degree of isolation: some markets are more isolated than others (e.g., Erie, Tulsa vs Los Angeles, New York).

• Ability to control distribution and promotion: preexisting distribution arrangements; local media tools to test promo messages.

Test Marketing: Pros and Cons

• Pros:

  • Good method for forecasting future sales.

  • Allows pretesting of marketing mix variables.

• Cons:

  • Not infallible.

  • Expensive.

  • Alerts competitors to the product/promotion being tested.

  • Takes time to conduct.

EXPERIMENT EXERCISE

• Our Brand Designs: design experiments for our brands.

• Create control and experimental groups.

• Select IVs, DVs, and CVs.

• Represent experiments with symbols (O for DV, X for IV) to illustrate designs.

Our Brand Designs (Contextual Reference)

• The exercise emphasizes practical application of experimental design to brand cases, including structuring control vs experimental groups and identifying relevant variables.

Notation and Variable Types (Recap)

• Independent Variable (IV): the variable manipulated by the researcher; aka causal/treatment variable.

• Dependent Variable (DV): the variable measured to assess the effect of the IV; aka the effect variable.

• Extraneous Variables (CVs or confounds): variables other than the IV that could influence the DV; must be controlled through design.

• Experimental vs Control Groups: experimental receives treatment; control does not.

• Randomization: essential for minimizing extraneous variable effects and ensuring comparable groups.

Quick Reference: Variable Definitions in the Lecture

• IV: \text{IV} = \text{variables manipulated by the researcher (e.g., price, ad spend, store layout)}

• DV: \text{DV} = \text{outcomes measured (e.g., sales, attitude, satisfaction)}

• O: O = \text{measurement of a dependent variable}

• X: X = \text{manipulation of an independent variable}

Summary of Core Concepts

• Research design provides the blueprint for collecting and analyzing data to address marketing problems.

• Exploratory research is for gaining background and clarifying problems; descriptive research describes phenomena; causal research tests hypotheses about cause-and-effect.

• Descriptive research can be cross-sectional (one-time snapshot) or longitudinal (over time with panel data).

• Causal research relies on experiments to establish causality, using IVs, DVs, randomization, and control of extraneous variables.

• True experimental designs isolate the effect of the IV on the DV and balance internal and external validity considerations.

• Lab experiments offer high internal validity but lower external validity; field experiments offer higher external validity but potentially lower control.

• Test marketing is a real-world method to forecast sales and test marketing mix variations, but it has costs and competitive exposure.

• Practical application involves designing experiments with clear IVs, DVs, and controls, and representing designs with symbolic notation (O and X).