Summary of Quan,ta,ve Research
Week Theme Bo*om line
1 Research design & • Clear, theory-driven RQ is the compass.
ques1ons • Quant methods = toolbox for causal claims & bias-proofing.
• Workflow: RQ → theory → hypotheses → data → test →
conclusion.
2 Variables, validity & • Identify variable type → plot & summarise.
reliability • Validity (internal, external, construct) = accuracy; reliability =
precision (Cronbach’s α).
• SD & 68-95-99 rule quantify spread; watch out for skew.
3 Describing rela1ons • Correlation shows pattern (−1 ↔ +1) but never causation.
• OLS line summarises conditional means & uses all data.
• Add controls in regression to isolate the pure X-Y link.
4 Hypothesis tes1ng & • Sampling variation ⇒ SE; t = βb / SE, p = tail area.
simple regression • Exogeneity needed for unbiased βb; residual ≠ error.
• Model fit: F-test gate, R2/Adj R2 effect size, assumption
checks.
5 Mul1ple regression, • Extra predictors remove confounding; β now means “holding
modera1on & others constant.”
media1on • Moderation via interactions (when/for whom); mediation
shows why.
• Binary DV → linear probability model or logit; χ2 for two
categorical variables.
, Week 1: Research design & questions
Why do we need a quan4ta4ve research method?
• Quan1ta1ve analysis helps us to understand the world.
• It helps us minimizing cogni1ve assump1ons.
• Depending on the state of prior theory and research on the topic, you must use quan1ta1ve
methods to make a useful contribu1on to our understanding of the world
• Only way to establish causal rela1onships
Types of quan4ta4ve research
>>> Depended variables are always on the y-axes!
A theory is: an explana1on of rela1onships among concepts or events within a set of boundary
condi1ons.
A good research ques4on:
• Can be answered and need answering (“so what?”)
• Improve our understanding of how the world works
• Inform theory
How do we change the understanding of the world that we have?
>>> A good research design helps us achieve this by:
1. Using theory, paint the most accurate picture possible of what the data genera1ng process
looks like
2. Use that data genera1ng process to figure out the reasons our data might look the way it
does that don’t answer our research ques1on
3. Find ways to block out those alternate reasons and so dig out the varia1on we need
To do so, we want to op1mize the validity and reliability of our study
Week Theme Bo*om line
1 Research design & • Clear, theory-driven RQ is the compass.
ques1ons • Quant methods = toolbox for causal claims & bias-proofing.
• Workflow: RQ → theory → hypotheses → data → test →
conclusion.
2 Variables, validity & • Identify variable type → plot & summarise.
reliability • Validity (internal, external, construct) = accuracy; reliability =
precision (Cronbach’s α).
• SD & 68-95-99 rule quantify spread; watch out for skew.
3 Describing rela1ons • Correlation shows pattern (−1 ↔ +1) but never causation.
• OLS line summarises conditional means & uses all data.
• Add controls in regression to isolate the pure X-Y link.
4 Hypothesis tes1ng & • Sampling variation ⇒ SE; t = βb / SE, p = tail area.
simple regression • Exogeneity needed for unbiased βb; residual ≠ error.
• Model fit: F-test gate, R2/Adj R2 effect size, assumption
checks.
5 Mul1ple regression, • Extra predictors remove confounding; β now means “holding
modera1on & others constant.”
media1on • Moderation via interactions (when/for whom); mediation
shows why.
• Binary DV → linear probability model or logit; χ2 for two
categorical variables.
, Week 1: Research design & questions
Why do we need a quan4ta4ve research method?
• Quan1ta1ve analysis helps us to understand the world.
• It helps us minimizing cogni1ve assump1ons.
• Depending on the state of prior theory and research on the topic, you must use quan1ta1ve
methods to make a useful contribu1on to our understanding of the world
• Only way to establish causal rela1onships
Types of quan4ta4ve research
>>> Depended variables are always on the y-axes!
A theory is: an explana1on of rela1onships among concepts or events within a set of boundary
condi1ons.
A good research ques4on:
• Can be answered and need answering (“so what?”)
• Improve our understanding of how the world works
• Inform theory
How do we change the understanding of the world that we have?
>>> A good research design helps us achieve this by:
1. Using theory, paint the most accurate picture possible of what the data genera1ng process
looks like
2. Use that data genera1ng process to figure out the reasons our data might look the way it
does that don’t answer our research ques1on
3. Find ways to block out those alternate reasons and so dig out the varia1on we need
To do so, we want to op1mize the validity and reliability of our study
