random effects model

(redirected from Random-effect model)

random effects model

A statistical model that may be used in meta-analysis, in which both within-study sampling error (variance) and between-studies variation are included in assessing the uncertainty or confidence interval of the results of the meta-analysis.
References in periodicals archive ?
Two approaches are developed to capture the unobserved heterogeneity: fixed-effect model and random-effect model.
A two-way random-effect model, however, overcomes this shortcoming.
78 and p-value = 0 supports the preference for the random-effect model rather than for the pooled model.
The overall result shows the random-effect model and the fixed-effect model are preferred to the pooled model for Europe country group in supporting H1 through H5.
However, the estimator from the random-effect model is more efficient than that of a fixed-effect model due to higher degrees of freedom saved by random-effect specifications.
Therefore, the random-effect model needs consideration.
In such cases, the random-effect model may be appropriate.
Even more complicated hierarchical random-effect models for both vessel and catch length-composition effects could be considered.
The fixed- and random-effect models were used to investigate these relationships.
In particular, the fixed--and random-effect models are used to investigate the dynamics between government spending and private consumption for the sample countries.
We used Poisson regression in a GAM model, and combined individual city regression coefficients using fixed and random-effect models.
The basic idea behind the use of the fixed- and random-effect models is that the effects of omitted variables that either may stay constant over time for a given cross-sectional unit or are the same for all cross-sectional units at a given point in time (or a combination of both) can be absorbed into the intercept term of a regression model (see Hsiao, 1993).