Multivariable Modeling - Poisson & Negative Binomial

These flashcards cover key concepts related to the Poisson and Negative Binomial models, focusing on their definitions and applications in a multivariable context.

Poisson Distribution

A statistical distribution that describes counts of 'rare' binary events when the number of trials is large and the probability of success is small.

Negative Binomial Regression

A statistical method used to model count data that exhibits overdispersion, which occurs when the variance is greater than the mean.

Likelihood Ratio Test (LRT)

A statistical test used to compare the goodness of fit between two models; one being nested within the other.

Deviance

A measure used in model fitting to evaluate how well a model fits a given set of observations, analogous to residuals in linear regression.

Expected Value (E(Y))

The average or mean value we would expect from a probability distribution, often calculated as lambda (λ) multiplied by the amount of person-time.

Overdispersion

A condition in statistical models where the variance is greater than the mean, indicating that the data has a greater level of variability than expected.

Underdispersion

A situation in statistical data where the variance is less than the mean, indicating that data points are less variable than anticipated.

Covariate Patterns

The different combinations of factors or variables (covariates) that can affect the outcome being studied, important for multivariable modeling.

Rate Ratio (RR)

A measure used in epidemiology to compare the frequency of an event occurring in two different groups.

Person-Time

The amount of time that individuals are at risk for an event or outcome in a study, typically used in the calculation of incidence rates.