QMO Summary for Final

By Dylano Paans

Lecture 1: Introduction to Quantitative Methods in Entrepreneurship

• Decision-Making:

• Many business decisions are based on intuition rather than formal analysis.

• Most decisions in life rely on heuristics (simplifying strategies or rules of thumb)

rather than a rational model.

• Heuristics: Provide a simple way for time-pressured professionals to deal with

complex situations but can lead to systematic errors.

• Two Models of Thinking:

• System 1: Intuitive thinking; automatic, unconscious, quick, implicit, emotional,

and effortless.

• System 2: Reflective thinking; controlled, conscious, slow, explicit, logical, and

effortful.

• Data-Based Decision-Making:

• Involves developing and using computer-decision models for analyzing,

planning, and implementing marketing strategies.

• Layers and Areas of Analyses:

• Defining the relevant market (industry boundaries) is essential for

analyzing competitive forces.

• Overview of the stages of a market research project.

Detailed Summary: This lecture introduces the basics of quantitative methods in

entrepreneurship, focusing on the difference between intuitive (System 1) and reflective
(System 2) thinking. It highlights the role of heuristics in decision-making, the potential for
errors, and the importance of data-based decision-making. The lecture also covers the initial
steps of market research, emphasizing the need to define the relevant market to analyze
competitive forces effectively.

Lecture 2: Essentials of Data Collection and Measurements

• Secondary Data:

• Data already exists within the company or is collected by third parties for
purposes other than solving the problem at hand.

• Possible Uses:

• Provides information at a sufficient level of detail and quality for solving a

problem.

• Acts as a preliminary stage before solving a problem with primary data.

 • Potential Limitations:

• Data is incomplete because it was generally collected for a different

purpose.

• Units of measure and level of detail do not correspond to requirements.

• No control over the data collection process.

• Data might be outdated.

• Primary Data:

• Collected specifically for the problem at hand.

• Different Types of Questioning:

• Qualitative data analysis: efficient in the early, exploratory stage of

addressing a problem but has limitations.

• Limitations:

• No representative character.

• Statements must be interpreted by the interviewer, lacking

objective measurement.

• Difficult to aggregate opinions.

• Limited options for efficient, computer-based processing.

• Quantitative Survey Methods:

• Observations:

• Involves observing and recording behavior or other data points.

• Scale and Measurement:

• Scale: A discrete or continuous space onto which objects are located

according to measurement rules.

• Measurement: Rules for assigning symbols to objects to numerically

represent the amount of a characteristic or categorize objects.

• Types of Scales: Single-item vs. multi-item scales.

• Measurement Errors:

• Context matters, including contrast effects and language differences.

• Common errors include overreporting, interviewer bias, bias due to question

order, halo-effect, tendency to mark the middle position, and non-anonymity.

• Validity and Reliability:

• Ensuring the accuracy (validity) and consistency (reliability) of

measurements.
 • Sampling:

• Basic Idea: Analyzing a selection of elements in a population to draw

conclusions about the entire population.

• Decisions in the Sampling Process:

• Define the population.

• Determine the sampling frame.

• Select the sampling procedure.

• Determine the sample size.

• Sampling Methods:

• Probability Sampling: Every member of the population has a known,

non-zero chance of being selected (e.g., random sampling).

• Non-Probability Sampling: Some members of the population have a

zero chance of being selected (e.g., convenience sampling).

Detailed Summary: This lecture dives into the details of data collection and measurement. It
distinguishes between secondary and primary data, discussing their uses, limitations, and the
methods of obtaining them. The lecture explains different types of questioning, focusing on
qualitative and quantitative data collection. It addresses common measurement errors and
highlights the importance of validity and reliability. Additionally, it provides a comprehensive
overview of the sampling process, including steps and methods used to ensure representative
samples.

Lecture 3: Data Editing, Matching, and Causal Analyses

• Data Editing:

• Observed data generally contains errors and missing values, requiring

preliminary preparation before analysis.

• Statistical Data Editing:

• Error Localization: Identifying erroneous values.

• Correction: Correcting missing and erroneous data in the best possible

way.

• Consistency: Ensuring that all edits are satisfied.

• Why Editing is Needed:

• Interviewer errors, omissions, ambiguities, inconsistencies, lack of

cooperation, and ineligible respondents.

• Data Preparation Techniques:

• Data Coding: Specifying how information should be categorized to facilitate

analysis.
 • Data Matching: Identifying, matching, and merging records corresponding to
the same entities from multiple databases.

• Data Imputation: Estimating and filling in missing data.

• Data Adjusting: Enhancing data quality for analysis.

• Common Procedures for Adjusting Data:

• Weighting: Assigning weights to observations according to pre-specified

rules.

• Variable Respecification: Modifying existing data to create new

variables or reducing the number of variables.

• Scale Transformation: Adjusting scales to ensure comparability with

other scales.

• Tasks to Prepare Text Data for Model Estimation:

• Entity Extraction: Identifying which words people write about.

• Topic Modeling: Categorizing topics people write about.

• Sentiment Analysis: Assessing the positivity or negativity of text.

• Relationship Between Entities: Understanding how words relate to each other.

• Writing Style: Analyzing the style between the words.

• Uses of Text Data:

• Language Reflects: Text reflects intentions, actions, relationships.

• Language Affects: Text affects perceptions and outcomes.

• Descriptive Statistics and Important Distributions:

• Overview of key statistical distributions.

• Correlational Analysis:

• Measures linear association strength between two metrically scaled

variables.

• Advantages:

• Direction and strength of correlation are visible.

• Values are comparable across variables due to restriction to

interval [-1, +1].

• No dependence on sample size.

• Limits:

• Only depicts linear correlations.

• Provides no causal evidence.

 • Potential for spurious associations if background variables are
uncontrolled.

Detailed Summary: This lecture focuses on preparing data for analysis by editing and
correcting errors. It covers the importance of data coding, matching, imputation, and adjusting.
It also delves into the preparation of text data, including tasks like entity extraction, topic
modeling, and sentiment analysis. Additionally, the lecture provides an overview of descriptive
statistics, key distributions, and the benefits and limits of correlational analysis. It emphasizes
the need for accuracy and consistency in data preparation to ensure reliable analysis results.

Lecture 4: Data Testing and Regression Analysis

• Testing Modes:

• Overview of selected statistical testing modes.

• Type 1 Error: False positive (rejecting a true null hypothesis).

• Type 2 Error: False negative (failing to reject a false null hypothesis).

• Factors Influencing Alpha Error:

• Size of the effect (smaller effect leads to more error).

• Dispersion of measurement values (more dispersion leads to more

error).

• Sample size (smaller sample leads to more error).

• Significance: Every effect becomes significant with a sufficiently large

sample, but not every statistically significant effect is practically
significant.

• Objectives of Regression Analysis:

• Quantifying the slope of a regression line.

• Estimating the influence of one variable on another.

• Properties of R²:

• Indicates how well the model explains variance in the dependent

variable.

• No fixed rules for acceptable R² value.

• Does not indicate the importance of an influencing variable.

• Offers no information on model performance outside the sample.

• Influenced by sample properties.

• Assumptions of Linear Regressions:

• Linear relationship between dependent and independent variables.

• Normally distributed error term.

 • No multicollinearity among independent variables.

• Homoscedasticity of error terms.

• Sample size of at least 20.

Detailed Summary: This lecture explores statistical testing modes, emphasizing the
importance of understanding Type 1 and Type 2 errors and factors influencing them. It delves
into regression analysis, detailing its objectives, such as quantifying slopes and estimating
variable influences. The lecture explains the properties and limitations of R² and the critical
assumptions required for linear regression, including linearity, normality, no multicollinearity,
and homoscedasticity. Understanding these concepts is crucial for effectively conducting and
interpreting regression analyses in quantitative research.

Lecture 5: Event Study

• Efficient Market Hypothesis:

• Stock prices reflect all available information instantaneously.

• Premises:

• Many profit-maximizing participants independently analyze and value

securities.

• New information regarding securities enters the market randomly.

• Profit-maximizing investors adjust security prices rapidly to reflect new

information.

• Event Study Design:

• Steps:

• Define the event and selection criteria.

• Search for companies fulfilling the criteria.

• Handle confounding events.

• Select an appropriate model and determine the event window.

• Return Calculation Models:

• Mean Adjusted Return Model:

• Simple but naive, does not consider market movements.

• No additional data needed.

• Market Model:

• Popular for calculating expected returns.

• Tests for Significance:

• T-tests, standardized residual tests, Corrado rank test, generalized sign

test, skewness adjusted t-test.
 • Moderating Analysis:

• Regression analysis to examine conditions under which financial

outcomes change.

Detailed Summary: This lecture introduces the efficient market hypothesis, explaining its
premises and implications for stock prices. It outlines the design of an event study, detailing the
steps involved in defining events, selecting criteria, searching for companies, handling
confounding events, and choosing an appropriate model. The lecture describes return
calculation models, including the mean adjusted return model and the market model, and
discusses various tests for significance. It also covers moderating analysis, emphasizing the
use of regression analysis to understand the conditions affecting financial outcomes.

Lecture 6: Response Model and Elasticities

• Sales Response Models:

• Types of Static Sales Response Models:

• Constant marginal returns (linear model).

• Decreasing marginal returns (multiplicative, semi-logarithmic models).

• Saturation volume (modified exponential model).

• S-shaped (log-reciprocal, logistic models).

• Market Share Models:

• Multiplicative interaction model (MCI).

• Multinomial logit model (MNL).

• Logistic Model:

• Predicts probabilities for binary outcomes.

• Requires robust specification, including nonlinear regression function.

• Special assumptions for the error term distribution.

• Requires adequate estimation methods and fit dimensions.

• Dynamic Effects:

• Customers, retailers, and competitors might need time to react to marketing

activities.

• Advertising impacts are often considered dynamic processes, with possible

advance reactions to marketing activities.

Detailed Summary: This lecture focuses on various sales response models, including static
models like linear, multiplicative, semi-logarithmic, exponential, and logistic models. It also
discusses market share models, such as the multiplicative interaction and multinomial logit
models. The lecture explains the logistic model's role in predicting binary outcomes and
highlights the importance of robust specification and special assumptions for the error term
distribution. Additionally, it covers dynamic effects, emphasizing the time-lagged responses of
customers, retailers, and competitors to marketing activities, particularly advertising.

Lecture 7: Marketing Analytics

• Strategy Definitions:

• Business Strategy:

• A concept, plan, or competitive positioning idea.

• Produces desired outcomes in the marketplace.

• Marketing Strategy:

• Aimed at customer value creation, market orientation, segmentation,

targeting, and positioning.

• Involves decisions on selling products in the market.

• Criteria for Strategy Formulation and Execution:

• Suitability, plausibility, consistency, acceptability, performance impact,

business risk, feasibility, stakeholder compatibility, and internal
readiness.

• Barriers to Strategy Execution:

• People, resources, skills, systems, leadership, and vision barriers.

Detailed Summary: This lecture defines business and marketing strategies, emphasizing their
role in achieving competitive positioning and desired market outcomes. It outlines the criteria
for formulating and executing effective strategies, including suitability, plausibility, consistency,
acceptability, performance impact, business risk, feasibility, stakeholder compatibility, and
internal readiness. The lecture also identifies common barriers to strategy execution, such as
issues related to people, resources, skills, systems, leadership, and vision, and discusses ways
to overcome these barriers to ensure successful strategy implementation.

Lecture 8: Future Trends, Big Data, and Predictions

• Digitalization Benefits:

• Interactivity, individualization/customization, intelligence/measurability,

channel integration, location independence, accountability, testing, flexibility.

• Machine Learning Process:

• Steps include data collection, data preparation, model training, performance

evaluation, and model improvement.

• Five V’s of Big Data:

• Volume, variety, velocity, veracity, value.

• Benefits of Databases:
 • Data independence, integration and sharing of data, consistency, minimal
redundancy, security, accessibility, ease of application development and
program maintenance.

Detailed Summary: This lecture explores the benefits of digitalization, highlighting aspects
such as interactivity, customization, measurability, and channel integration. It outlines the
machine learning process, from data collection and preparation to model training, performance
evaluation, and improvement. The lecture discusses the five V’s of big data—volume, variety,
velocity, veracity, and value—and their implications for data management. Additionally, it
emphasizes the advantages of modern database systems, including data independence,
integration, consistency, minimal redundancy, security, accessibility, and ease of application
development and maintenance.