Summary Quantitative & Design
Methodology in Business Research
Table of Contents
Introduction.............................................................................................................................................................3
Design & Behavioral Research...............................................................................................................................7
Factor Analysis:.......................................................................................................................................................9
Step 1: Set objectives to define what Factor Analysis should be used.................................................................9
Step 2: Check the Assumptions.............................................................................................................................9
Step 3: Select an extraction method...................................................................................................................10
Step 4: Determining the Number of Factors and Rotation.................................................................................12
Step 5: Filtering the data set..............................................................................................................................14
Step 6: Rotating the factors................................................................................................................................14
Step 7: Interpreting factors................................................................................................................................15
Step 8: Using factors in other analyses..............................................................................................................15
Step 9: Determining the Model Fit.....................................................................................................................17
Step 10: Access Validity & Reliability................................................................................................................18
(Co)-Variance Analysis.........................................................................................................................................19
1. One-way ANOVA............................................................................................................................................20
2. N-way ANOVA................................................................................................................................................22
4. ANCOVA.........................................................................................................................................................25
5. MANOVA........................................................................................................................................................26
6. Interpretation..................................................................................................................................................27
Regression Analysis...............................................................................................................................................28
Step 1: Determine the objectives of the Multivariate Regression Analysis........................................................29
Step 2: Think about the Research Design of the Multivariate Regression Analysis..........................................29
Step 3: Check the Assumptions...........................................................................................................................30
Step 4: Estimate the Regression Model..............................................................................................................32
Step 5: Asses the overall fit................................................................................................................................34
Step 6: Interpretation of the model.....................................................................................................................36
Step 7: Validation of the results.........................................................................................................................37
Structural Equation Modeling.............................................................................................................................38
Step 1: Identify Individual Components.............................................................................................................40
Step 2.1: Specify the Structural Model (=Regression).......................................................................................40
Step 2.2 Specify the Measurement Model (Factor Analysis)..............................................................................41
1
, Step 3: Ensuring the requirements.....................................................................................................................42
Step 4: Assessing the measurement model.........................................................................................................43
Step 5: Assessing the Structural Model..............................................................................................................45
Guest lecture: Hierarchic Clustering..................................................................................................................46
2
,Introduction
Types of Analysis:
- Univariate analysis: an analysis based on only one variable (descriptive).
- Multivariate analysis: an analysis based on two or more variables, for more complex
phenomena.
o Dependence technique: when there is at least one independent and a
dependent variable (explanatory).
o Interdependence technique: when all variables have equal status, so no
dependent (Y) and independent (X) variables (exploratory).
Measurement scales:
- Nonmetric or qualitative
o Nominal scale: e.g., industry: no amount nor related
o Ordinal scale: e.g., education: related but no fixed distance
- Metric or quantitative:
o Interval scale: e.g., Likert-scale: equal distance but you cannot multiple (2 is
not double of 1)
o Ratio scale: e.g., turnover: equal distance, you can multiple and there is a
natural 0 point.
Hypotheses:
- Conditional statements
o In situation A B happens
o The higher A the higher B
- Unconditional statements
o A is greater than B
o A is not equal to 0
Measurement Error
- All variables have measurement errors
- Measurement error makes the result less powerful cause true information may still
be hidden
- 2 types of measurement error:
o Random error e.g., Likert scale 1-5: your true value is 2.4 but you can only
choose 2 or 3 this is less reliable.
o Systematic error e.g., double barreled questions: 2 questions in one, which
questions is prioritized and answered? This is less valid.
3
, - Summated scores are used to reduce the measurement error: then several variables
are summed or averaged together to form a composite representation of a concept.
X m=X t + E
X m=X t + Er + E s
Xm Value as measured
Xt True value
E Error
Er Random error
Es Systematic error
Statistical Significance and Power
Hypothesis testing= concluding something about the population based on sample outcomes:
- H 0=there is no effect∨difference
- H 1=there isa effect∨difference
- Type I and type II errors
o Type I error or α = rejecting H 0, even if you shouldn’t: you conclude that there
is a difference/ effect but there is no difference or effect
o Type II error or β = not rejecting H 0, even if you should have: you conclude
that there is no difference/ effect, but there actually is a difference or effect.
- Significance and Power
o Significance = 1−α : By taking a higher significance level (e.g., 95% or 99%),
the chance of Type I error will reduce. A significance level of 95%, leaves a
remaining 5% chance on type I error.
o Power = 1−β : A Power of 80% is still acceptable (so 20% of type II error).
Power: what influences type II error
- Effect size: a larger effect is easier to find*.
- Sample size: the more observations, the higher the power, the lower the chance on
type II error. *Small effects even require larger sample sizes.
- Alpha (α ): the lower the alpha (high significance), the lower the power (so higher
chance on type II error).
Guidelines for Multivariate Analysis:
4
Methodology in Business Research
Table of Contents
Introduction.............................................................................................................................................................3
Design & Behavioral Research...............................................................................................................................7
Factor Analysis:.......................................................................................................................................................9
Step 1: Set objectives to define what Factor Analysis should be used.................................................................9
Step 2: Check the Assumptions.............................................................................................................................9
Step 3: Select an extraction method...................................................................................................................10
Step 4: Determining the Number of Factors and Rotation.................................................................................12
Step 5: Filtering the data set..............................................................................................................................14
Step 6: Rotating the factors................................................................................................................................14
Step 7: Interpreting factors................................................................................................................................15
Step 8: Using factors in other analyses..............................................................................................................15
Step 9: Determining the Model Fit.....................................................................................................................17
Step 10: Access Validity & Reliability................................................................................................................18
(Co)-Variance Analysis.........................................................................................................................................19
1. One-way ANOVA............................................................................................................................................20
2. N-way ANOVA................................................................................................................................................22
4. ANCOVA.........................................................................................................................................................25
5. MANOVA........................................................................................................................................................26
6. Interpretation..................................................................................................................................................27
Regression Analysis...............................................................................................................................................28
Step 1: Determine the objectives of the Multivariate Regression Analysis........................................................29
Step 2: Think about the Research Design of the Multivariate Regression Analysis..........................................29
Step 3: Check the Assumptions...........................................................................................................................30
Step 4: Estimate the Regression Model..............................................................................................................32
Step 5: Asses the overall fit................................................................................................................................34
Step 6: Interpretation of the model.....................................................................................................................36
Step 7: Validation of the results.........................................................................................................................37
Structural Equation Modeling.............................................................................................................................38
Step 1: Identify Individual Components.............................................................................................................40
Step 2.1: Specify the Structural Model (=Regression).......................................................................................40
Step 2.2 Specify the Measurement Model (Factor Analysis)..............................................................................41
1
, Step 3: Ensuring the requirements.....................................................................................................................42
Step 4: Assessing the measurement model.........................................................................................................43
Step 5: Assessing the Structural Model..............................................................................................................45
Guest lecture: Hierarchic Clustering..................................................................................................................46
2
,Introduction
Types of Analysis:
- Univariate analysis: an analysis based on only one variable (descriptive).
- Multivariate analysis: an analysis based on two or more variables, for more complex
phenomena.
o Dependence technique: when there is at least one independent and a
dependent variable (explanatory).
o Interdependence technique: when all variables have equal status, so no
dependent (Y) and independent (X) variables (exploratory).
Measurement scales:
- Nonmetric or qualitative
o Nominal scale: e.g., industry: no amount nor related
o Ordinal scale: e.g., education: related but no fixed distance
- Metric or quantitative:
o Interval scale: e.g., Likert-scale: equal distance but you cannot multiple (2 is
not double of 1)
o Ratio scale: e.g., turnover: equal distance, you can multiple and there is a
natural 0 point.
Hypotheses:
- Conditional statements
o In situation A B happens
o The higher A the higher B
- Unconditional statements
o A is greater than B
o A is not equal to 0
Measurement Error
- All variables have measurement errors
- Measurement error makes the result less powerful cause true information may still
be hidden
- 2 types of measurement error:
o Random error e.g., Likert scale 1-5: your true value is 2.4 but you can only
choose 2 or 3 this is less reliable.
o Systematic error e.g., double barreled questions: 2 questions in one, which
questions is prioritized and answered? This is less valid.
3
, - Summated scores are used to reduce the measurement error: then several variables
are summed or averaged together to form a composite representation of a concept.
X m=X t + E
X m=X t + Er + E s
Xm Value as measured
Xt True value
E Error
Er Random error
Es Systematic error
Statistical Significance and Power
Hypothesis testing= concluding something about the population based on sample outcomes:
- H 0=there is no effect∨difference
- H 1=there isa effect∨difference
- Type I and type II errors
o Type I error or α = rejecting H 0, even if you shouldn’t: you conclude that there
is a difference/ effect but there is no difference or effect
o Type II error or β = not rejecting H 0, even if you should have: you conclude
that there is no difference/ effect, but there actually is a difference or effect.
- Significance and Power
o Significance = 1−α : By taking a higher significance level (e.g., 95% or 99%),
the chance of Type I error will reduce. A significance level of 95%, leaves a
remaining 5% chance on type I error.
o Power = 1−β : A Power of 80% is still acceptable (so 20% of type II error).
Power: what influences type II error
- Effect size: a larger effect is easier to find*.
- Sample size: the more observations, the higher the power, the lower the chance on
type II error. *Small effects even require larger sample sizes.
- Alpha (α ): the lower the alpha (high significance), the lower the power (so higher
chance on type II error).
Guidelines for Multivariate Analysis:
4
