Research Project
Lecture 8 – Experimentation
Todays Plan
• Causation
• Experimental design
• Developing experimental stimuli
• Experimental validity
Causation
Correlations vs. Causation
• Correlation is similar lines but they do not necessarily influence each other; causation
• E.g. visitors to Disneyland Paris and precipitation in Louisiana have similar lines but they do no cause or
influence each other
Reasons why X and Y can correlate
• X causes Y
• Y causes X = reverse causality
• Z causes both X and Y = third variable
• Spurious correlation
Reverse Causality = X appears to cause Y, but it is actually Y that causes X
• E.g. diversified firms tend to profit more; more profitable firms need to find ways to re-invest of its rather
than because diversification causes profitability.
,Third Variable = X appears to cause Y, but both X and Y are actually caused by Z
• The more toys a child has, the higher his or her IQ; both number of toys and IQ may be caused by family
resources such as income
Three conditions for causality
• Relationship between X and Y; X and Y vary together
• Time order; X cannot happen after Y
• Elimination of other possible causal factors; causes constant and controlled
Experimental Design
Basic features of between-subjects design
• Independent variable = manipulated across groups or between-subject
- Between-subject = one participant is assigned to one experimental conditions of IV
- Within-subject = one participant is assigned to several experimental conditions
• Dependent variable = measured
• Context = laboratory, online, field
• Controlling extraneous factors
- All things but independent variable are the same
- Participants randomly assigned
- Measurement of other variable under control
,Developing experimental stimuli
Stimuli = event or object to which a response is measured
• Can be visual, textual, verbal, space etc.
• E.g. visual stimuli – nutritional labelling
Confounding Variables = variables that have affected the results (DV) apart form the IV
• Can be extraneous that has been controlled
Experimental Validity
Internal validity
• Conclusions about the effects of IVs on DVs are valid
• Correct implementation of principles
• Lab studies are high in internal validity
External validity
• Conclusions generalized outside the experiment
e.g. lab participants to consumers
• Field studies are higher in external validity
Randomized Control Trial
• RCT in parallel groups = participants are randomly allocated into one of several conditions
• RCT in crossover = random allocation into different groups + DV measured several times
, • RCT in cluster = pre-existing groups of participants are randomly selected to receive and intervention
Designs facing threats to internal validity
• Quasi-experiment = assignment to the experimental conditions is not random
e.g. nutria-score in supermarket A vs. No nutri-score in supermarket B
• Pre-post = the DV is measured before and after an IV category
Replication experiments
1. See whether the effect is robust across situations
2. Build on previous experiments see whether your explanation makes sense
Field experimentation
• Testing customers reactions to changes
Online A/B testing
= refers to a randomized experimentation process wherein two versions of a thing are compared against each toher
to determine which performs better
Lecture 9 – introduction to linear regression
Today’s plan
• Intro into Regression Models
• Simple linear regression model
• Multiple linear regression model
• Individual assignment 3
Intro into Regression Models
Regression = an approach for modelling the relationship between a dependent variable (DV) and one or more
independent variables (IVs)
• Linear regression: DV is continuous / quantitative
• Logistic regression: DV is binary
Used to
Lecture 8 – Experimentation
Todays Plan
• Causation
• Experimental design
• Developing experimental stimuli
• Experimental validity
Causation
Correlations vs. Causation
• Correlation is similar lines but they do not necessarily influence each other; causation
• E.g. visitors to Disneyland Paris and precipitation in Louisiana have similar lines but they do no cause or
influence each other
Reasons why X and Y can correlate
• X causes Y
• Y causes X = reverse causality
• Z causes both X and Y = third variable
• Spurious correlation
Reverse Causality = X appears to cause Y, but it is actually Y that causes X
• E.g. diversified firms tend to profit more; more profitable firms need to find ways to re-invest of its rather
than because diversification causes profitability.
,Third Variable = X appears to cause Y, but both X and Y are actually caused by Z
• The more toys a child has, the higher his or her IQ; both number of toys and IQ may be caused by family
resources such as income
Three conditions for causality
• Relationship between X and Y; X and Y vary together
• Time order; X cannot happen after Y
• Elimination of other possible causal factors; causes constant and controlled
Experimental Design
Basic features of between-subjects design
• Independent variable = manipulated across groups or between-subject
- Between-subject = one participant is assigned to one experimental conditions of IV
- Within-subject = one participant is assigned to several experimental conditions
• Dependent variable = measured
• Context = laboratory, online, field
• Controlling extraneous factors
- All things but independent variable are the same
- Participants randomly assigned
- Measurement of other variable under control
,Developing experimental stimuli
Stimuli = event or object to which a response is measured
• Can be visual, textual, verbal, space etc.
• E.g. visual stimuli – nutritional labelling
Confounding Variables = variables that have affected the results (DV) apart form the IV
• Can be extraneous that has been controlled
Experimental Validity
Internal validity
• Conclusions about the effects of IVs on DVs are valid
• Correct implementation of principles
• Lab studies are high in internal validity
External validity
• Conclusions generalized outside the experiment
e.g. lab participants to consumers
• Field studies are higher in external validity
Randomized Control Trial
• RCT in parallel groups = participants are randomly allocated into one of several conditions
• RCT in crossover = random allocation into different groups + DV measured several times
, • RCT in cluster = pre-existing groups of participants are randomly selected to receive and intervention
Designs facing threats to internal validity
• Quasi-experiment = assignment to the experimental conditions is not random
e.g. nutria-score in supermarket A vs. No nutri-score in supermarket B
• Pre-post = the DV is measured before and after an IV category
Replication experiments
1. See whether the effect is robust across situations
2. Build on previous experiments see whether your explanation makes sense
Field experimentation
• Testing customers reactions to changes
Online A/B testing
= refers to a randomized experimentation process wherein two versions of a thing are compared against each toher
to determine which performs better
Lecture 9 – introduction to linear regression
Today’s plan
• Intro into Regression Models
• Simple linear regression model
• Multiple linear regression model
• Individual assignment 3
Intro into Regression Models
Regression = an approach for modelling the relationship between a dependent variable (DV) and one or more
independent variables (IVs)
• Linear regression: DV is continuous / quantitative
• Logistic regression: DV is binary
Used to
