Presentation
Last updated: February 24, 2015.
Consider data sources and communication.
Data Sources
Fundamental aspect of empirical research in social science. (Also
physical science, but difficulties are different.)
Data may not be trustworthy.
- People forget.
- People don't measure, they estimate or perceive.
- Technical statistical issues:
- Sample selection issues (choice; correlation with unobserved
influences).
- Sample size.
Data must correspond to the theme.
- Macro (social flows) vs. micro (individual choice).
- Role alignment (e.g., demand-side vs. supply-side in industry).
- Behavior (choice) vs. performance (outcome).
Measurement: cause -> effect.
If the cause increases, so does the effect in some proportion, but
how much of each affects cost and performance, and the coefficient
is important for prediction.
- Measurement can result in values, changes, orders, or mere a
judgment of "sameness".
- Values include things like money, price, and order quantity.
- Changes are things like speed or growth rates.
- Ordering: "bigger", "preferred".
- "Sameness": red vs. blue.
Commmunication
Listener comes first!!
- Unless you're a teacher, boss, or very famous. ;-)
Repeating your words doesn't work.
- Listeners hear you, but don't connect your words (which they
understand) to your meaning (which they don't).
- People understand things in different ways ("Satir's sensory
modes"). Prepare multiple explanations, and examples in different
sensory modes:
- visual, audible, tactile, (smell), (taste).
Listeners make assumptions.
- They expect you to say what they expect to hear. "Cognitive
dissonance." Speak to the audience's expertise.
- E.g., profs assume you took their class.
- And remember what they taught!
- Study your AG's research!
So read, Read, READ!
- AG's research!
- citations in papers you've already read.
- Anything you cite, read it twice!
- Look for:
- Famous researchers
- Researchers at famous universities and institutes (but avoid
"house journals")
- High impact factor journals (Google for "impact factor" -- use
comparisons not absolutes! there are different definitions)
- Frequent citations
- Basic research, textbooks