Answer Strategy Flashcards
What is an Answer Strategy?
Statistical or qualitative methods to summarise and interpret data (cleaning, transforming, estimating data.)
What are the five elements for Answer Characterisation?
Domain
Unit
Outcomes
Conditions
Uncertainty
What are the two main approaches to address uncertainty in answers?
Bayesian: Focuses on the probability of an estimate being the true value (estimand).
Frequentist: Focuses on procedures for arriving at an estimate.
What are 5 types of Answer Strategies?
- Point Estimation
- Hypothesis Testing
- Bayesian Formulations
- Interval Estimation
- Extreme Bounds
What is Point Estimation?
Provides a single value as an estimate of a population parameter (e.g., mean, regression coefficient). This estimate can be derived using techniques like maximum likelihood estimation.
Uncertainty: standard error
What is Hypothesis Testing?
Hypothesis testing is a statistical method using a null hypothesis to assess the likelihood of observing an outcome as extreme as the one observed, assuming the null hypothesis is true.
What is a p-value?
P-value: Probability of getting an estimate as extreme as ours if there’s truly no effect
p > 0.1 = statistically insignificant
p < 0.05 = statistically significant
p < 0.001 = statistically highly significant
What are the two types of uncertainty/errors in a hypothesis test?
Type I Error (False Positive): Rejecting a true null hypothesis.
Type II Error (False Negative): Failing to reject a false null hypothesis.
What is the Bayesian Formulation?
The Bayesian formulation is a statistical method that combines prior beliefs with data to generate probabilities for a range of possible answers/estimands.
It aims to generate “rational beliefs” about the values you’re looking for
What is interval estimation
Provides a range of values where the true value is likely to lie with a certain level of confidence
What is extreme bounds?
Estimates the “best” and “worst” case scenarios based on the data.