Chapter 1 Flashcards
What is empirical research about?
- Identify problems - not taking everything as given
- Think in systems and disentangle potential mechanisms
- Operationalize test settings to generate reliable evidence
- Evaluate results in order to generate predictions and recommendations
What is the research process?
- Formulate a research questions
- Literature review
- Develop a theory
- Formulate hypotheses
- Design empirical tests for the hypothesis testing
- Interpret results
What is the formal definition of theory?
A logically coherent set of ideas that can explain and predict empirical observations.
What does a theory need in order for it not to be only descriptive?
justification
What is the definition of econometrics?
The use of statistical methods to analyze economic data.
What data is used in Economics?
- non-experimental or observation
- it’s rare to have experimental data
What is important to apply to real world data?
economic theory
What are the goals of econometrics?
- Estimting relationships
- Testing economic theories
- Forecasting economic variables
- Evaluating and implementing
What are the steps in econometric analysis?
- Economic model
- Econometric model
What is the role of the Economic model?
formulate qusetion of interest
What is the role of the Econometric model?
operationalize the economic model
Economic vs Econometric model: Table
With what specification do most econometrics deal?
with the specification of the error u
What is the formula for wage?
What are types of Data?
- Cross-section data
- Time series data
- Pooled cross sections
- Panel/longitudinal data
On what do econometrics methods depend?
on the nature of the data used
- may lead to misleading results
Decribe Cross-section data.
- samples of different units at a given point of time/given period
- cross-sectional observations more or less independent
example: pure random sampling - if not random sample, sample-selection problem
Decribe Time Series data.
- observations of a variable or several over time
- typically serially correlated
- ordering of observations conveys important information (lead/lag)
- since not a random set, different problems to consider
Where is Cross-section data typically encountered in?
- accounting
- finance
Where is Time Series data typically encountered in?
- trends
- seasonality
Into what are two or more cross sections combined in?
one data set
How are cross sections drawn in rapport to each other?
independently
When are Pool crossed sections often used?
to evaluate policy changes
Describe Panel or longitudinal data.
- same cross-sectional units are followed over time
- panel daa have a cross-sectional and a time series dimension
- panel data can be used to account for time-invariant unobservables
- panel data can be used to model lagged responses
In what situations is panel or longitudinal data used?
city crime statistics, observed in two years
What is a Construct?
abstract idea which is not directly observable and should be operationalized for empirical testing
What is a Variable?
observable item which can assume different values and is used to measure a theoretical construct
What is the contruct validity?
degree to which a measurement/variable captures the underlying theoretical construct it is supposed to measure
Define internal validity.
degree of condfidence that the causal relationship being tested is trustworthy and not influenced by other factors
Define external validity.
extent to which results from a study can be applied to other situations
What is the construct reliability?
degree to which a measurement provides estimates which are consistent
What is the role of the random variable?
assings unique to the outcome of a random experiment
What are the types of Quantitative variables?
- Discrete random variable
- Continuous random variable
What is the discrete random variable?
random variable with specific set of possible values (ie dice)
Is the continuous random variable discreet?
no (range of possible values)
What is the role of the qualitative (categorical) variable?
take category or label values and place a unit of observations into one of several categories (mutually exclusive) groups (nationality ie)
What describes probability of observing specific values?
Distribution
What can characterize normal distribution?
- mean
- variance/standard deviation
Formula Variance
Formula Standard deviation
square root of Var(X)
List the stages of the basic process of empirical research
- population (parameters)
- producing data/sampling
- sample/data
- data analysis (statistics)
- probability?
+ interference
— all over again
What is empirical research?
usage of statistics generated from a sample to make inferences about the parameters of the population
Define Population
all individuals/research subjects that are of interest to a researcherq
Define Sample
a subset of the population
On what do theoretical properties of statistical estimators usually rest?
on the assumption of random samples
What is the sample of selection bias?
bias resulting from the non-random sampling of a set of observations from the population, where the probability of the oservation being included in the sample potentially affects the observed results in the sample
Sample mean formula
What is the law of large number?
actual probability of an event is estimated by relative frequency with which the event occurs in a long series of trials
What do sample averages converge to as the sample size increases, and what’s the result?
to the population mean nu (with probability 1)
result: estimates get more accurate
What is the mean of all samples?
What is the standard deviantion of all sample means Xi?
What does the standard deviation of sample means do as sample size increases?
it decreases
(law of large numbers)
How will sample of means approximately be?
normal
- even if the distribution is not normal
(central limit theorem)
How can we infer mean of population by calculating mean from a random sample?
either:
- increasing sample size
- repeatedly taking random samples and averaging sample statistics
What do we use to determine accuracy?
confidence interval
Formula confidence interval
On what does the Z multiplier depend?
on the desired level of confidence (comes from normal distribution)
Formula confidence interval if sigma and population mean are unknown
Formula standard error
When do multipliers become nearly the same?
under normal distribution for large samples
What are t-statistics?
for large samples over 500, multipliers for the confidence interval become nearly the same as under normal distribution
When is t-distribution used?
when construction confidence intervals for the population mean, especially when sample is small or standard deviation is unknown
When do we usually reject H0?
if p-value is higher than a significant level a
Summary of testing strategy on the mean of a variable x
- Draw sample of n observations from population
- Calculate sample statistic X
- Calculate standard deviation error of X: s/root(n)
- Calculate test statistic that measure deviation of C from u0 in terms of standard deviation error
- If statistic is larger/smaller than critical values of corresponding t-distribution with -1 degrees of freedom, accept the hypothesis
What are the two types of errors?
Type 1: rejecting a null hypothesis that is actually true
Type 2: accepting a null hypothesis that is actually false
H0 rejected/not rejected
H0 false/true
TABLE
Size of a test:
probability of falsely rejecting the null hypothesis when it’s actually correct
Power of a test:
probability of correctly rejecting the null hypothesis when it is false
What is alpha?
likelihood of a type 1 error <8reejcting a true H0)
What is beta?
likelihood of a type 2 error (accepting a wrong H0)
