Marketing Analytics Midterm Study Guide

Descriptive Statistics

How do you process and understand data?

Inferential Statistics

Drawing conclusions based on a sample population

Linear Regression

Teasing our relationships between variables

Marketing Research

Involves the specification, gathering, analysis, and interpretation of information to help managers understand the environment, identify problems and opportunities, and develop and evaluate courses of marketing action.

Internal Groups

Marketing and sales departments, market research department, product development and R&D departments

External Groups

Independent marketing research companies

What Triggers the Need for Research

Problems/Opportunity/Changes in the World

Defining Problem or Oppurtunity

First and most important step of marketing research

Research Question

a clear, focused, and actionable question that guides the purpose and direction of your study

Problem

Suite of symptoms with unknown causes

Management Decision Problem

What does the decision maker need to do, action oriented

Marketing Research Problem

What does the decision maker need to know/learn to make the decision? How should that information be obtained, information oriented

Idea Grouping

Your actions, your competitors actions, your consumers actions, your environment changes

Exploratory Research

Very open-ended, lots of possibilities

Conclusive Research

Well-defined alternatives, hypotheses to test

Performance-Monitoring Research

Check how actions/decision are performing

Cross Sectional Research

Snapshots of a phenomena at one point in time

Longitudinal Research

Measure phenomena repeatedly over several time periods, how things change over time

Business Performance Metrics

Sales, market share, customer satisfaction, loyalty, etc

Marketing Strategy Metrics

Retail penetration, price points, customer wait times, brand awareness

Marketing Environment

Economic indicators, social trends, competitor actions and performance, etc

Qualitative Research Methods

Focus Groups, interviews, case studies

Descriptive Research

Allows you to uncover patterns

Causal Research

Allows you to see cause/effect

Primary Data Sources

Data collected specifically by researcher for the problem at hand. Survey data, experimental data, focus group data

Secondary Data Sources

Data collected for some other purpose. Syndicated industry data, data from previous research, historical sales

Internal Secondary Data

Generated within the organization

External Secondary Data

Generated by sources outside your organization

Syndicated Data Sources

Companies who collect and sell standardized data to other firms

Quantitative Research

A research methodology that seeks to quantify the data and typically applies some form of statistical analysis

Pitfalls of Qualitative Research

Confirmatory Bias, Researcher Bias, Participant Bias

Focus Group

Groups of 6-12 participants, 1-3 hours, more creative answers

Focus Group Design

Screen Participants, Homogenous target group, different groups, different focus group

Demographics

Age, gender, income, education, occupation, geographic location, etc

Psychographics

Interests, values, lifestyle choices, attitudes

Behavioral Characteristics

Past purchase behavior, brand loyalty, product usage, etc

Experiential Criteria

Participant may need to have specific experiences, such as using a product or service, attending a particular event, or having expertise in a particular area

Inclusion Criteria

Characteristic or experiences that participants must have to be part of the focus group

Exclusion Criteria

Characteristics or experiences that qualify someone from participating

Good Moderator

Combination of empathy and firmness, encourages involvement, can think on their feet, flexible and sensitive, knowledge of subject

In Depth Interview

One on one interview with an individual

Why Choose Interview

Allows for more depth and understanding, ideal for discussing with expert, better for sensitive topics

Why Choose Focus Group

Group dynamics encourage idea sharing and discussion, Efficient, Helps form fuller picture, When needing to gather a wide range of ideas

Laddering

Move from the concrete features and attributes of a product to the emotional needs and values of the consumer

Counterfactual World

World where there is an absence of X

Omitted/Confounding Variable

Variable not accounted for in research

Conditions required for causality

Cause and effect of X and Y must be correlated, Cause must occur before the effect, X should be the only plausible causal explanation

Independent Variable

Manipulated by Researcher

Dependent Variable

Measured outcome

Hypothesis

A testable statement or an educated guess about the relationship between variables

Key Hypothesis Characteristics

Specific, Testable, Based on Research

Treatment

Independent variables that are manipulated

Test Units

Recipients of the treatments

Extraneous Variables

All other variables that potentially affect the dependent outcome/variable

Treatment Group

Test units that are recipients of the IV manipulation

Control Group

Test units that are not recipients of the IV manipulation

Random Assignment

Randomize who gets treatment versus control

Required Sample Size

At least 30 observations per group

Population

Entire group of people about whom information is needed

Sampling

Selection of a subset of the population of interest on which the research will be conducted

Random selection

How you draw the sample of people for your study from a population

A/B Testing

Experimental method used for various versions/aspects/designs

Pre Experimental Design

One-shot case study, one group pretest-posttest, static group, do not use random assignment

Statistical Design

Factorial design, more than one independent variable

True Experimental Design

Pretest-posttest control group, postest-only control group, do use random assignment

One Shot Case Study

Studying one case without much else going on

One Group Pretest Postest Design

One group that measures something pretest and post test

Static-Group Design

Treatment group and control group tested

Pre-test Post-test Control Group Design

Includes pre-test testing and control group, with random assignment

Post-test Only Control Group Design

Avoids priming, no Pre test

Laboratory Experiment

More control and cheaper but less realistic/generalizable

Field Experiments

More expensive and less control but more generalizable

Internal Validity

Extent to which competing explanations for the experimental results observed can be ruled out

External Validity

The extent to which causal relationships measured in an experiment can be generalized to outside persons, setting, and times

Measurement

The process of assigning numbers or labels to persons, objects, or events in accordance with specific rules for representing quantities or qualities or attributes

Construct

Types of concepts that exist at higher levels of abstraction

General Process for Measurement

Conceptualize, Define in concrete terms, Construct measurement, Evaluate, Utilize

Scale

Set of symbols or numbers so constructed so they can be assigned by a rule to individuals, behaviors, or attitudes to whom the scale is applied.

Nominal Measurement

Scales that partition data into mutually exclusive and collectively exhaustive categories

Ordinal Measurement

Scales that maintain the labeling characteristics of nominal scale and can order data

Interval Measurement

Scales that have characteristics of ordinal scales plus equal intervals between points

Ratio Measurement

Scales that have the characteristics of interval, plus a meaningful zero point

Systemic Error

Biases in your measurement, error that effects measurement in a constant way

Measurement Error

Occurs when there is a discrepancy between the actual value and the recorded value in survey responses or research findings

Sampling Error

When the sample of participants is systematic not capturing the sample you intend

Random Error

Unpredictable noise during measurement, larger sample sizes help minimize it

Validity

Are we measuring what we think we are measuring? Describes or quantifies the construct it is designed to measure

Reliability

Is our measurement process free from random errors? Gives us consistent, stable values when measured repeatedly

Question Order Bias

Switch up what order questions are asked in

Basic Survey Components

Introduction, Main Survey Questions, Demographics and identifiers, Thank you for participating

Attention Filter Question

Question used to tell if the person taking your survey is actually paying attention

Open Ended Question Vs Close Ended Question

Both have various pros and cons, should use both types of questions on surveys

Synthetic Survey Respondents

AI which can be helpful for Filling data gaps and more cost effective, however it can misrepresent populations if poorly generated.