session 5 Flashcards
How do experiments establish causality and remove confounds
need randomization
internal validity
when you establish causality, when you are on a field experiment sometimes you cant establish validity and cant have internal validity, you establish validity through randomization and controlling your variables
external validity
whether your finding can be generalized to another setting
Ex: in a lab experiment, certain kinds of packaging increase purchase intentions, in na actual retail store will these lead to purchase? Looking at different settings (not only lab setting) different cultural contexts, participants from china and from all over, high external validity means generalizability to other contexts which is good
ecological
whether your findings look at real behaviour. Lab experiments are low in ecological bc you are controlling things whereas field experiments where you don’t know you are part of the experiment are high for ecological. Want high ecological
Trade Off between ecological and internal validity
secondary data
information that has already been gathered
internal: company annual reports, sales/ scanner data, customer database, customer feedback data
external: gvt reports, industry profiles, social media , market overview, case analysis
primary data
data collected for the first time for the study at hand
qualitative and quantitive
Advantages of secondary data
High volume
Low cost
Less time
Helps define your business problem
Already available
May clarify problem
May imprpive primary data collection
Not resource intensive : saves time, money, minimal training
May provide background information
Disadvantages of secondary data
Outdated data, accuracy issues
Not tailored to your business problem exactly
May have ethical issues
Primary data
Data that you collect
Can do this through quantitative or qualitative
Advantages of primary data
More accurate and relevant
Avoid ethical issues
Disadvantages of primary data
Requires training and expertise
Methods vary on expense and speed
expensive
a lot of time