Lecture 1: Introduction
Introduction to quantitative research
Quantitative research: resolves around answering a particular research question by
collecting numerical data that are analyzed using statistics
- In essence, quantitative research methods provide us with a toolbox to
study the (social) world around us by the use of the scientific method
- Helps in minimizing cognitive assumptions that may distort our
interpretation
- Depending on the state of prior theory and research on the topic, you
have to use quantitative methods to make
a useful contribution to our understanding
of the world
- Only way to establish causal relationships
o Causality: refers to a relationship
where a change in one variable (X)
directly causes a change in another
variable (Y), holding all other
factors constant
Types of research: descriptive vs. inferential
Types of quantitative research:
- Descriptive (what?)
o Interested in a quantitative answer: ‘How many X are Y?’
o Interested in a numerical change: ‘Are the X of Y rising compared to Z?’
- Inferential (why?)
o Test relationships: ‘what is the relation between X and Y?
o Explain something: ‘what factors cause changes in X over Y?’
Validity in research
Validity: do we measure what we want to measure?
- Internal validity: the extent to which you are able to draw the correct
conclusions about the causal relationships between variables
- External validity: the extent to which your findings are generalizable to the
broader population (of individuals, firms, etc.) and diRerent settings
Threats to validity:
- Omitted variables bias: occurs when you leave out (omit) an independent
variable (IV) that is a determinant of the dependent variable (DV) and correlated
with one or more of the includes IV’s
o Leaving out this IV will lead to an over- or underestimation of the relation
between your variables of interest
o In our analysis we need to control or adjust for these variables
, - Reverse causality: occurs when the direction of the arrow in our theoretical
model goes the other way
o DiRicult to empirically rule this out à sometimes logical reasoning can
help us (e.g., ice cream sales and temperature)
- Sample selection bias: selection of data for analysis in such a way that proper
randomization is not achieved, leading to an unrepresentative sample of the
population intended to be analyzed
o Random sample: to avoid these issues, we need to draw an independent
and identically distributed sample from the population
§ Identically distributed: there are no overall trends the distribution
doesn’t fluctuate and all items in the sample are taken from the
same probability distribution
§ Independent: sample items are all independent events, they aren’t
connected to each in any way
- Measurement error: data analyst makes random mistakes when imputing data
in a database (no problem since the errors are distributed randomly) or
diRiculties in measuring skill level of employees (less problematic, will only make
you estimates more ‘noisy’)
Lecture 2: Understanding and describing variables / Correlations
Measurement and reliability
Reliability: the degree to which a measure produces stable/repeatable and consistent
results à ‘do we measure consistently?’
- Inter-item reliability: measures if items on a scale are internally consistent, has
to be done with any scale before aggregating the scale into a score (e.g.,
extraversion) à most widely applied measure is called Cronbach’s ⍺
o Cronbach’s ⍺: measure of inter-item reliability, meaning it assesses
whether multiple items on a scale consistently measure the same
underlying construct à in most social sciences, a value of α ≥ 0,7 is
considered to be ‘acceptable’
§ α closes to 1 (higher consistency), α too low (items may not
measure the same thing
Reliability coeRicients: (TYPE à WHAT IS IT? à WHAT DO YOU NEED à COEFFICIENT)
- Test-retest: measure of stability à give same test twice to the same people à
correlation coeRicient
- Parallel form: measure of equivalence à give diRerent forms to the same people
à correlation coeRicient
- Inter-rater: measure of agreement à have two people rate the same behaviors
à # of agreements / # of observations
- Inter-item: measure of item consistency à single test administration à
Cronbach’s alpha / McDonald’s omega
Types of variables in quantitative research
Introduction to quantitative research
Quantitative research: resolves around answering a particular research question by
collecting numerical data that are analyzed using statistics
- In essence, quantitative research methods provide us with a toolbox to
study the (social) world around us by the use of the scientific method
- Helps in minimizing cognitive assumptions that may distort our
interpretation
- Depending on the state of prior theory and research on the topic, you
have to use quantitative methods to make
a useful contribution to our understanding
of the world
- Only way to establish causal relationships
o Causality: refers to a relationship
where a change in one variable (X)
directly causes a change in another
variable (Y), holding all other
factors constant
Types of research: descriptive vs. inferential
Types of quantitative research:
- Descriptive (what?)
o Interested in a quantitative answer: ‘How many X are Y?’
o Interested in a numerical change: ‘Are the X of Y rising compared to Z?’
- Inferential (why?)
o Test relationships: ‘what is the relation between X and Y?
o Explain something: ‘what factors cause changes in X over Y?’
Validity in research
Validity: do we measure what we want to measure?
- Internal validity: the extent to which you are able to draw the correct
conclusions about the causal relationships between variables
- External validity: the extent to which your findings are generalizable to the
broader population (of individuals, firms, etc.) and diRerent settings
Threats to validity:
- Omitted variables bias: occurs when you leave out (omit) an independent
variable (IV) that is a determinant of the dependent variable (DV) and correlated
with one or more of the includes IV’s
o Leaving out this IV will lead to an over- or underestimation of the relation
between your variables of interest
o In our analysis we need to control or adjust for these variables
, - Reverse causality: occurs when the direction of the arrow in our theoretical
model goes the other way
o DiRicult to empirically rule this out à sometimes logical reasoning can
help us (e.g., ice cream sales and temperature)
- Sample selection bias: selection of data for analysis in such a way that proper
randomization is not achieved, leading to an unrepresentative sample of the
population intended to be analyzed
o Random sample: to avoid these issues, we need to draw an independent
and identically distributed sample from the population
§ Identically distributed: there are no overall trends the distribution
doesn’t fluctuate and all items in the sample are taken from the
same probability distribution
§ Independent: sample items are all independent events, they aren’t
connected to each in any way
- Measurement error: data analyst makes random mistakes when imputing data
in a database (no problem since the errors are distributed randomly) or
diRiculties in measuring skill level of employees (less problematic, will only make
you estimates more ‘noisy’)
Lecture 2: Understanding and describing variables / Correlations
Measurement and reliability
Reliability: the degree to which a measure produces stable/repeatable and consistent
results à ‘do we measure consistently?’
- Inter-item reliability: measures if items on a scale are internally consistent, has
to be done with any scale before aggregating the scale into a score (e.g.,
extraversion) à most widely applied measure is called Cronbach’s ⍺
o Cronbach’s ⍺: measure of inter-item reliability, meaning it assesses
whether multiple items on a scale consistently measure the same
underlying construct à in most social sciences, a value of α ≥ 0,7 is
considered to be ‘acceptable’
§ α closes to 1 (higher consistency), α too low (items may not
measure the same thing
Reliability coeRicients: (TYPE à WHAT IS IT? à WHAT DO YOU NEED à COEFFICIENT)
- Test-retest: measure of stability à give same test twice to the same people à
correlation coeRicient
- Parallel form: measure of equivalence à give diRerent forms to the same people
à correlation coeRicient
- Inter-rater: measure of agreement à have two people rate the same behaviors
à # of agreements / # of observations
- Inter-item: measure of item consistency à single test administration à
Cronbach’s alpha / McDonald’s omega
Types of variables in quantitative research
