Multilevel Regression

Vocabulary flashcards from the lecture notes on multilevel regression, covering key terms and their definitions.

Multilevel Regression

A statistical technique used to analyze data that is organized at more than one level.

Level 1 and Level 2

In a multilevel model, Level 1 indicates individual units (i) and Level 2 indicates group units (j).

Homoscedasticity

An assumption that the variance of errors is constant across levels of an independent variable.

Fixed Coefficients

Coefficients that are treated as fixed effects, resulting in one-way ANOVA with j parameters.

Random Coefficients

Coefficients that are treated as random, assuming they follow a normal distribution.

Intraclass Correlation Coefficient (ICC)

A statistic that measures the proportion of variance that is attributable to group differences.

Between Group Variability

The variance that exists between different groups in a multilevel model.

Within Group Variability

The variance that exists within each group in a multilevel model.

Design Effect

The factor by which the sample size must be inflated to obtain the same level of precision as simple random sampling due to clustering.

Random Intercept Model

A model where the intercept is allowed to vary randomly across groups.

Parameter Variance

The variability of the estimated parameters across different groups.

Reliability

The degree to which an assessment tool produces stable and consistent results.