Flashcards in Chapter 1 Deck (13)
Describe the key difference between experimental and non-experimental data
Experimental data collected under controlled conditions in a laboratory environment through an experiment, whereas non-experimental data are not accumulated in such a manner.
What is empirical analysis in economics?
A study that uses data in a formal econometric analysis to test a theory, estimate a relationship, or determine the effectiveness of a policy.
List the steps in empirical economic analysis
1. The formulation of question of interest
2. The construction of a formal economic model
3. The construction of an econometric model
4. The postulation of hypothetical values for the unknown parameters (or betas)
5. The gathering of data
6. The estimation of the parameters in the econometric model (the betas)
Describe the features of a cross-sectional data set
- A sample of individuals or a variety of other units taken at a given point in time (minor timing differences are ignored)
- Random sampling (each observation is drawn at random from the population, and that each draw is independent of all other draws + draws are equiprobable)
Describe the potential issues that may arise with cross sectional data
- Sampling may not be random after all due to unavoidable biases i.e wealthier families less open about their income (a sample selection problem)
- Observations may be related to each other
Describe the features of a time series data set
- Consists of observations on a variable or several variables over time i.e stock prices
- Data collected at a set time frequency
Describe potential issues that may arise with time series data
- The need for de-seasonalization
- Economic observations are not independent across time
Describe the features of a pooled cross section data set
- Contains both cross-sectional and time series features
Example: a number of schools are sampled for certain characteristics in 1960 and, after an education policy change, are again sampled for the same characteristics in 1980.
- The aggregation of two cross-sections makes for a much larger sample size.
- Useful for analysing how a key relationship has changed over time
Describe the potential issues that may arise with pooled cross section data
- Variables can still change over time irrespective of the event being analysed
Describe the features of a panel data set
- Consists of a time series for each cross-sectional member
Example: Financial information for certain individuals over time
- The same cross-sectional members are observed over time
Advantages of using a panel data over pooled cross-sectional data?
- Panel data allows for control over the unobserved characteristics of units.
- Panel data is usually of a longer timespan, thereby allowing for the study of lags in behavior
How is a causal effect established between two variables?
If there is sufficient correlation, not only association