SUMMARY OF EVERYTHING YO U NEED
ALL LECTURES + SELECTED READINGS
+ ALL AVAILABLE WEEKLY QUIZZES
+ EXAMS (2016-17, 2017-18, 2018-19)
E n h a n c e d w i t h a d y n a m ic t a b le o f c o n t e n t s .
, MADS MADLAD |2
Note from MADS MADLAD:
Thank you for buying my summary. I sincerely hope it helps you excel and learn
from this course. When I was writing these I sometimes struggled with this
program, but there were no summaries available.
This is why I decided to write something that is truly complete with a lot of
effort put into it. It helped me and my friends get good grades, but I also
always had you in mind, the future reader. When necessary, I always went the
extra mile to make this summary, more readable, organized and complete.
If you feel like it, leave me a review of how the course is going using this
summary, it will make my day to hear your opinion good or bad!
Check out my other extensive summaries for other MADS courses:
Available on:
Contact info:
If you need help or have an inquiry, contact me: https://www.georgedreemer.com
Connect with me on LinkedIn: https://www.linkedin.com/in/georgedreemer/
Donations:
By no means am I looking for fellow students to send me money! But if you feel like sending
me some ETH or BTC, you can do so here:
--> ETH: 0x123e086c6808459e7fC6Ac7F64a77dBA1dDe0149
--> BTC: bc1qgwzc82vph5v8rmzef4ywechjf85772n7m2e22g
, MADS MADLAD |3
wishes you good luck & perseverance.
Grades Testimony:
, MADS MADLAD |4
Table of Contents
Week 1............................................................................................................. 10
Lecture 1: Introduction to Marketing Models ............................................... 10
Model building........................................................................................... 10
3 Type of Models: Iconic, analog and symbolic .......................................... 10
Reasons to use Models .............................................................................. 11
Reading 1: Market Models Chapter 1 ........................................................... 12
1.2 Verhouten Case ................................................................................... 12
1.3 Typologies of Marketing Models .......................................................... 12
1.3.2 Decision Models vs. Models That Advance Marketing Knowledge
............................................................................................................. 12
1.3.3 Degree of Explicitness: Implicit vs. Verbal vs. Formalized vs.
Numerically-Specified Models .............................................................. 13
1.3.4 Intended Use: Descriptive, Predictive and Normative Models..... 15
1.3.5 Level of Demand ............................................................................... 16
1.4 Benefits from Using Marketing Decision Models ................................. 16
1.4.1 Direct Benefits ............................................................................. 16
1.4.2 Indirect Benefits .......................................................................... 17
1.5 The Model Building Process ................................................................. 17
Quiz 1: Questions + Answers......................................................................... 20
Week 2............................................................................................................. 23
Lecture 2: Model Specification (1) – Model types, elements and critera ...... 23
Model Types .............................................................................................. 23
Intended use: Descriptive, Predictive and Normative .......................... 23
Level of demand: Product class/form/industry sales, Brand sales, Market
share .................................................................................................... 23
Amount of behavioral detail: No detail, Some detail, Substantial detail
............................................................................................................. 24
Model elements......................................................................................... 24
Single equation model.......................................................................... 24
, MADS MADLAD |5
System of equations ............................................................................. 25
Disturbance term ................................................................................. 25
Mathematical form .............................................................................. 25
Model criteria ............................................................................................ 26
(1) Simple ............................................................................................. 27
(2) Evolutionary .................................................................................... 27
(3) Complete ........................................................................................ 27
(4) Adaptive.......................................................................................... 28
(5) Robust............................................................................................. 28
Reading 2: Market Models Chapter 2 ........................................................... 29
2.2 Model Criteria ...................................................................................... 29
2.2.1 Implementation Criteria Related to Model Structure .................. 29
2.2.2 Models Should Be Simple ............................................................ 29
2.2.3 Models Should Be Built in an Evolutionary Way .......................... 30
2.2.4 Models Should Be Complete on Important Issues ....................... 30
2.2.5 Models Should Be Adaptive ......................................................... 31
2.2.6 Models Should Be Robust ............................................................ 31
2.3 Model Elements ................................................................................... 31
2.4 Specification of the Functional Form.................................................... 33
2.4.1 Models Linear in Parameters and Variables (linear additive model)
............................................................................................................. 33
2.4.2 Models Linear in Parameters But Not in Variables (nonlinear
additive model) .................................................................................... 34
2.4.3 Models That Are Nonlinear in Parameters, But Linearizable
(multiplicative model) .......................................................................... 35
2.4.4 Models That Are Nonlinear in Parameters And Not Linearizable
(intrinsically nonlinear or intractable) .................................................. 35
2.5 Moderation and Mediation Effects ...................................................... 36
2.6 Formalized Models for the Verhouten Case ......................................... 37
2.7 Including Heterogeneity: aggregated models, unit-by-unit models,
pooled models, partially pooled models .................................................... 39
, MADS MADLAD |6
2.8 Marketing Dynamics ............................................................................ 40
2.8.1 Introduction ................................................................................ 40
2.8.2 Modelling Lagged Effects: One Explanatory Variable................... 40
2.8.3 Modeling Lagged Effects: Several Explanatory Variables ............. 44
2.8.4 Lead Effects ................................................................................. 45
Reading 3: Market Models Chapter 3 ........................................................... 47
3.1 Introduction ......................................................................................... 47
3.2 Data Structures .................................................................................... 47
3.3 “Good Data” ........................................................................................ 48
3.3.1 Availability ................................................................................... 48
3.3.2 Quality ......................................................................................... 48
3.3.3 Variability .................................................................................... 48
3.3.4 Quantity ...................................................................................... 48
3.4 Data Characteristics and Model Choice................................................ 49
3.5 Data Sources ........................................................................................ 50
3.5.1 Introduction ................................................................................ 50
3.5.2 Classification ............................................................................... 50
3.5.3 Internal Data ............................................................................... 51
3.5.4 External Data ............................................................................... 51
3.5.5 Household Data vs. Store Level Data ........................................... 52
3.5.6 Big Data ....................................................................................... 53
3.5.7 Subjective Data............................................................................ 53
Week 3............................................................................................................. 55
Lecture 3: Specification (2) ........................................................................... 55
Function Form: Recap – SCAN*PRO Model ................................................ 55
Multiplicative model .................................................................................. 55
Advantages vs. Disadvantages.............................................................. 55
Multiplicative model: Log transformation = linearization (before
estimation) ........................................................................................... 55
Multiplicative model: Antilog transformation (after estimation).......... 56
, MADS MADLAD |7
Example: model for Karvan Cevitam sales at Albert Heijn .................... 57
Model Specification: Model Choices .......................................................... 60
Aggregate Model (A) ............................................................................ 60
Unit-by-unit model (B) ......................................................................... 61
Pooled model (C) .................................................................................. 61
Partially pooled model (D).................................................................... 62
Dynamic Effects: How to deal with dynamics ............................................ 63
Dynamic Effects/Distributed-lag models .............................................. 64
Role of a third variable (z) .......................................................................... 66
Model Assumptions: Preliminary ............................................................... 68
Quiz 3: Questions + Answers......................................................................... 69
Week 4............................................................................................................. 72
Lecture 4: Estimation (1) – Some unresolved questions ............................... 72
Types of Data ............................................................................................. 72
Rule of thumb for Data .............................................................................. 72
Scale of data .............................................................................................. 72
Ordinary Least Squares – Bivariate Regression Analysis............................. 73
Multivariate Regression Analysis ............................................................... 75
OLS in R: the ‘lm’ command ....................................................................... 77
How to run a regression for a different chain? ..................................... 79
How to include multiplicative model? .................................................. 79
Interaction effects in a multiplicative model.............................................. 80
Parameters are elasticities......................................................................... 80
Multiplicative Model: Recap ...................................................................... 81
How to in R – Lagged variables and per chain ............................................ 83
How to in R – Model specified per chain .................................................... 84
How to in R – Pooled model....................................................................... 85
How to in R – Partially Pooled model ......................................................... 86
Quiz 4: Questions + Answers......................................................................... 89
Week 5............................................................................................................. 91
, MADS MADLAD |8
Lecture 5: Specification issues – Omitted variables bias, wrong functional form,
endogeneity, non-constant parameters, multicollinearity & Pooling ............ 91
Table 5.1: Violations of the assumptions about the disturbance term ....... 91
Violations of assumptions .......................................................................... 92
Causes of violations of assumptions (overview)......................................... 93
Omitted variable bias................................................................................. 93
Wrong functional form .............................................................................. 94
Endogeneity ............................................................................................... 94
Non-constant parameters .......................................................................... 94
Pooling....................................................................................................... 95
Chow Test: To pool or not to pool? ...................................................... 95
In Practice – Example: Pooled vs Unpooled & Chow Test ........................... 96
In Practice – Example: OLSDV Pooled Model vs U-by-U Unpooled & Chow Test
.................................................................................................................. 98
Multicollinearity ........................................................................................ 99
Causes of multicollinearity ................................................................... 99
Typical symptoms of multicollinearity ................................................ 100
Detection of multicollinearity............................................................. 100
Solutions to multicollinearity ............................................................. 101
Possible solution: Recode variables .................................................... 101
Multicollinearity Conclusions ............................................................. 103
Week 6........................................................................................................... 104
Lecture 6: Disturbance term assumptions - Autocorrelation,
Heteroscedasticity, Non-normality ............................................................. 104
Distribution of Residuals .......................................................................... 104
OLS – When all lm assumptions are satisfied ........................................... 104
Autocorrelation: dependence of disturbance term.................................. 105
Remedies of Autocorrelation: GLS...................................................... 109
Heteroscedasticity ................................................................................... 109
Non-normality ......................................................................................... 111
, MADS MADLAD |9
In Practice: Chocolate case (Verhouten) .................................................. 114
1: Autocorrelation .............................................................................. 114
2: Heteroscedasticity.......................................................................... 116
Quiz 6 – Questions + Answers ..................................................................... 120
Week 7........................................................................................................... 124
Lecture 7: Face validity, Statistical validity, Predictive validity, Exam hints . 124
Face validity ............................................................................................. 124
Statistical validity ..................................................................................... 124
Model fit: R-squared & Adjusted R-squared ....................................... 125
Overall test of model significance ...................................................... 126
Information-criteria: AIC, AIC3, BIC, CAIC ........................................... 126
Testing significance of individual parameters: t-test .......................... 127
Diagnosis of violation of linear model assumptions: residual plots .... 127
Checking unusual data points: visual examples & how to detect ....... 128
Checking unusual data points: Overview of process & Influence plot 131
Methods for outlier-robust estimation: Examples and How to in R .... 132
Predictive validity .................................................................................... 132
Predictive validity measures: APE, ASPE, RMSE, MAPE, RAE/Naïve
benchmark model .............................................................................. 133
How to in R: Prediction and Predictive validity criteria....................... 134
Exam Hints ............................................................................................... 135
Practice Exams /w Answers: 2016-17, 2017-18, 2018-19 ............................... 137
, M A D S M A D L A D | 10
Week 1
Lecture 1: Introduction to Marketing Models
Model building
➢ What is a model?
o A stylized representation of reality
➢ What is the goal of models?
o Understand this reality
➢ What are the basic elements of every model?
o A model should contain the most important elements, but is never
complete.
o Make as simple as possible
3 Type of Models: Iconic, analog and symbolic
➢ Iconic Models: resemble reality but use other materials or another scale:
for example to capture design ideas.
o Examples: sketches, prototypes, virtual, reality and scale models.
➢ Analog Model: specific characteristics of an idea or system
o Focus on key elements
o Do not contain details
o Do not resemble reality but are helpful in analyzing its functions
o Examples: flow charts, circuit diagrams
➢ Symbolic Models: represent ideas using code, an abstract
representation of reality
o Examples: numbers, mathematical formulas, words, music notes
ALL LECTURES + SELECTED READINGS
+ ALL AVAILABLE WEEKLY QUIZZES
+ EXAMS (2016-17, 2017-18, 2018-19)
E n h a n c e d w i t h a d y n a m ic t a b le o f c o n t e n t s .
, MADS MADLAD |2
Note from MADS MADLAD:
Thank you for buying my summary. I sincerely hope it helps you excel and learn
from this course. When I was writing these I sometimes struggled with this
program, but there were no summaries available.
This is why I decided to write something that is truly complete with a lot of
effort put into it. It helped me and my friends get good grades, but I also
always had you in mind, the future reader. When necessary, I always went the
extra mile to make this summary, more readable, organized and complete.
If you feel like it, leave me a review of how the course is going using this
summary, it will make my day to hear your opinion good or bad!
Check out my other extensive summaries for other MADS courses:
Available on:
Contact info:
If you need help or have an inquiry, contact me: https://www.georgedreemer.com
Connect with me on LinkedIn: https://www.linkedin.com/in/georgedreemer/
Donations:
By no means am I looking for fellow students to send me money! But if you feel like sending
me some ETH or BTC, you can do so here:
--> ETH: 0x123e086c6808459e7fC6Ac7F64a77dBA1dDe0149
--> BTC: bc1qgwzc82vph5v8rmzef4ywechjf85772n7m2e22g
, MADS MADLAD |3
wishes you good luck & perseverance.
Grades Testimony:
, MADS MADLAD |4
Table of Contents
Week 1............................................................................................................. 10
Lecture 1: Introduction to Marketing Models ............................................... 10
Model building........................................................................................... 10
3 Type of Models: Iconic, analog and symbolic .......................................... 10
Reasons to use Models .............................................................................. 11
Reading 1: Market Models Chapter 1 ........................................................... 12
1.2 Verhouten Case ................................................................................... 12
1.3 Typologies of Marketing Models .......................................................... 12
1.3.2 Decision Models vs. Models That Advance Marketing Knowledge
............................................................................................................. 12
1.3.3 Degree of Explicitness: Implicit vs. Verbal vs. Formalized vs.
Numerically-Specified Models .............................................................. 13
1.3.4 Intended Use: Descriptive, Predictive and Normative Models..... 15
1.3.5 Level of Demand ............................................................................... 16
1.4 Benefits from Using Marketing Decision Models ................................. 16
1.4.1 Direct Benefits ............................................................................. 16
1.4.2 Indirect Benefits .......................................................................... 17
1.5 The Model Building Process ................................................................. 17
Quiz 1: Questions + Answers......................................................................... 20
Week 2............................................................................................................. 23
Lecture 2: Model Specification (1) – Model types, elements and critera ...... 23
Model Types .............................................................................................. 23
Intended use: Descriptive, Predictive and Normative .......................... 23
Level of demand: Product class/form/industry sales, Brand sales, Market
share .................................................................................................... 23
Amount of behavioral detail: No detail, Some detail, Substantial detail
............................................................................................................. 24
Model elements......................................................................................... 24
Single equation model.......................................................................... 24
, MADS MADLAD |5
System of equations ............................................................................. 25
Disturbance term ................................................................................. 25
Mathematical form .............................................................................. 25
Model criteria ............................................................................................ 26
(1) Simple ............................................................................................. 27
(2) Evolutionary .................................................................................... 27
(3) Complete ........................................................................................ 27
(4) Adaptive.......................................................................................... 28
(5) Robust............................................................................................. 28
Reading 2: Market Models Chapter 2 ........................................................... 29
2.2 Model Criteria ...................................................................................... 29
2.2.1 Implementation Criteria Related to Model Structure .................. 29
2.2.2 Models Should Be Simple ............................................................ 29
2.2.3 Models Should Be Built in an Evolutionary Way .......................... 30
2.2.4 Models Should Be Complete on Important Issues ....................... 30
2.2.5 Models Should Be Adaptive ......................................................... 31
2.2.6 Models Should Be Robust ............................................................ 31
2.3 Model Elements ................................................................................... 31
2.4 Specification of the Functional Form.................................................... 33
2.4.1 Models Linear in Parameters and Variables (linear additive model)
............................................................................................................. 33
2.4.2 Models Linear in Parameters But Not in Variables (nonlinear
additive model) .................................................................................... 34
2.4.3 Models That Are Nonlinear in Parameters, But Linearizable
(multiplicative model) .......................................................................... 35
2.4.4 Models That Are Nonlinear in Parameters And Not Linearizable
(intrinsically nonlinear or intractable) .................................................. 35
2.5 Moderation and Mediation Effects ...................................................... 36
2.6 Formalized Models for the Verhouten Case ......................................... 37
2.7 Including Heterogeneity: aggregated models, unit-by-unit models,
pooled models, partially pooled models .................................................... 39
, MADS MADLAD |6
2.8 Marketing Dynamics ............................................................................ 40
2.8.1 Introduction ................................................................................ 40
2.8.2 Modelling Lagged Effects: One Explanatory Variable................... 40
2.8.3 Modeling Lagged Effects: Several Explanatory Variables ............. 44
2.8.4 Lead Effects ................................................................................. 45
Reading 3: Market Models Chapter 3 ........................................................... 47
3.1 Introduction ......................................................................................... 47
3.2 Data Structures .................................................................................... 47
3.3 “Good Data” ........................................................................................ 48
3.3.1 Availability ................................................................................... 48
3.3.2 Quality ......................................................................................... 48
3.3.3 Variability .................................................................................... 48
3.3.4 Quantity ...................................................................................... 48
3.4 Data Characteristics and Model Choice................................................ 49
3.5 Data Sources ........................................................................................ 50
3.5.1 Introduction ................................................................................ 50
3.5.2 Classification ............................................................................... 50
3.5.3 Internal Data ............................................................................... 51
3.5.4 External Data ............................................................................... 51
3.5.5 Household Data vs. Store Level Data ........................................... 52
3.5.6 Big Data ....................................................................................... 53
3.5.7 Subjective Data............................................................................ 53
Week 3............................................................................................................. 55
Lecture 3: Specification (2) ........................................................................... 55
Function Form: Recap – SCAN*PRO Model ................................................ 55
Multiplicative model .................................................................................. 55
Advantages vs. Disadvantages.............................................................. 55
Multiplicative model: Log transformation = linearization (before
estimation) ........................................................................................... 55
Multiplicative model: Antilog transformation (after estimation).......... 56
, MADS MADLAD |7
Example: model for Karvan Cevitam sales at Albert Heijn .................... 57
Model Specification: Model Choices .......................................................... 60
Aggregate Model (A) ............................................................................ 60
Unit-by-unit model (B) ......................................................................... 61
Pooled model (C) .................................................................................. 61
Partially pooled model (D).................................................................... 62
Dynamic Effects: How to deal with dynamics ............................................ 63
Dynamic Effects/Distributed-lag models .............................................. 64
Role of a third variable (z) .......................................................................... 66
Model Assumptions: Preliminary ............................................................... 68
Quiz 3: Questions + Answers......................................................................... 69
Week 4............................................................................................................. 72
Lecture 4: Estimation (1) – Some unresolved questions ............................... 72
Types of Data ............................................................................................. 72
Rule of thumb for Data .............................................................................. 72
Scale of data .............................................................................................. 72
Ordinary Least Squares – Bivariate Regression Analysis............................. 73
Multivariate Regression Analysis ............................................................... 75
OLS in R: the ‘lm’ command ....................................................................... 77
How to run a regression for a different chain? ..................................... 79
How to include multiplicative model? .................................................. 79
Interaction effects in a multiplicative model.............................................. 80
Parameters are elasticities......................................................................... 80
Multiplicative Model: Recap ...................................................................... 81
How to in R – Lagged variables and per chain ............................................ 83
How to in R – Model specified per chain .................................................... 84
How to in R – Pooled model....................................................................... 85
How to in R – Partially Pooled model ......................................................... 86
Quiz 4: Questions + Answers......................................................................... 89
Week 5............................................................................................................. 91
, MADS MADLAD |8
Lecture 5: Specification issues – Omitted variables bias, wrong functional form,
endogeneity, non-constant parameters, multicollinearity & Pooling ............ 91
Table 5.1: Violations of the assumptions about the disturbance term ....... 91
Violations of assumptions .......................................................................... 92
Causes of violations of assumptions (overview)......................................... 93
Omitted variable bias................................................................................. 93
Wrong functional form .............................................................................. 94
Endogeneity ............................................................................................... 94
Non-constant parameters .......................................................................... 94
Pooling....................................................................................................... 95
Chow Test: To pool or not to pool? ...................................................... 95
In Practice – Example: Pooled vs Unpooled & Chow Test ........................... 96
In Practice – Example: OLSDV Pooled Model vs U-by-U Unpooled & Chow Test
.................................................................................................................. 98
Multicollinearity ........................................................................................ 99
Causes of multicollinearity ................................................................... 99
Typical symptoms of multicollinearity ................................................ 100
Detection of multicollinearity............................................................. 100
Solutions to multicollinearity ............................................................. 101
Possible solution: Recode variables .................................................... 101
Multicollinearity Conclusions ............................................................. 103
Week 6........................................................................................................... 104
Lecture 6: Disturbance term assumptions - Autocorrelation,
Heteroscedasticity, Non-normality ............................................................. 104
Distribution of Residuals .......................................................................... 104
OLS – When all lm assumptions are satisfied ........................................... 104
Autocorrelation: dependence of disturbance term.................................. 105
Remedies of Autocorrelation: GLS...................................................... 109
Heteroscedasticity ................................................................................... 109
Non-normality ......................................................................................... 111
, MADS MADLAD |9
In Practice: Chocolate case (Verhouten) .................................................. 114
1: Autocorrelation .............................................................................. 114
2: Heteroscedasticity.......................................................................... 116
Quiz 6 – Questions + Answers ..................................................................... 120
Week 7........................................................................................................... 124
Lecture 7: Face validity, Statistical validity, Predictive validity, Exam hints . 124
Face validity ............................................................................................. 124
Statistical validity ..................................................................................... 124
Model fit: R-squared & Adjusted R-squared ....................................... 125
Overall test of model significance ...................................................... 126
Information-criteria: AIC, AIC3, BIC, CAIC ........................................... 126
Testing significance of individual parameters: t-test .......................... 127
Diagnosis of violation of linear model assumptions: residual plots .... 127
Checking unusual data points: visual examples & how to detect ....... 128
Checking unusual data points: Overview of process & Influence plot 131
Methods for outlier-robust estimation: Examples and How to in R .... 132
Predictive validity .................................................................................... 132
Predictive validity measures: APE, ASPE, RMSE, MAPE, RAE/Naïve
benchmark model .............................................................................. 133
How to in R: Prediction and Predictive validity criteria....................... 134
Exam Hints ............................................................................................... 135
Practice Exams /w Answers: 2016-17, 2017-18, 2018-19 ............................... 137
, M A D S M A D L A D | 10
Week 1
Lecture 1: Introduction to Marketing Models
Model building
➢ What is a model?
o A stylized representation of reality
➢ What is the goal of models?
o Understand this reality
➢ What are the basic elements of every model?
o A model should contain the most important elements, but is never
complete.
o Make as simple as possible
3 Type of Models: Iconic, analog and symbolic
➢ Iconic Models: resemble reality but use other materials or another scale:
for example to capture design ideas.
o Examples: sketches, prototypes, virtual, reality and scale models.
➢ Analog Model: specific characteristics of an idea or system
o Focus on key elements
o Do not contain details
o Do not resemble reality but are helpful in analyzing its functions
o Examples: flow charts, circuit diagrams
➢ Symbolic Models: represent ideas using code, an abstract
representation of reality
o Examples: numbers, mathematical formulas, words, music notes
