Harvey Goldstein's working treatment of random-coefficient / multilevel modelling for clustered data. Covers two- and three-level models, cross-classified and multiple-membership structures, the discrete-response extensions, model diagnostics, and Bayesian approaches. First published by Arnold; carried through multiple editions. Arnold maintained a companion support site for the book.
The text moves from the basic two-level random-intercept model through random-slope and three-level extensions, into the cross-classified and multiple-membership structures that arise when units belong to more than one higher-level grouping (pupils across schools and neighbourhoods, patients across hospitals and general practices). Discrete-response extensions cover binary, ordered and count outcomes; later chapters address model checking and Bayesian estimation routes.
Goldstein worked at the UCL Institute of Education and the University of Bristol's Centre for Multilevel Modelling, and co-authored the MLwiN software that turned multilevel modelling from a curiosity into routine practice for education research, epidemiology and survey statistics.
About the author
Harvey Goldstein— Professor of Social Statistics
Biostatistics & Multilevel Modelling · 1939–2020
Affiliations: UCL Institute of Education — Professor of Statistical Methods (historical) · University of Bristol — Professor of Social Statistics; Centre for Multilevel Modelling (with Jon Rasbash and colleagues) · Royal Statistical Society — Guy Medal in Silver (1998) · British Academy — Fellow from 1996
Intended audience
- applied statisticians
- social scientists working with clustered data
- biostatisticians
Subject classification
- Multilevel modelling
- Applied statistics
- Hierarchical models
Scope
- two- and three-level random-coefficient models
- cross-classified and multiple-membership structures
- discrete-response multilevel models
- Bayesian multilevel estimation
Bibliographic details
| Format | ISBN | Year | Notes |
|---|---|---|---|
| Print edition (Arnold) | 0340806559 | 2003 | Open Library edition record (publisher line matched Arnold/Hodder). |
| ISBN-13 (EAN-13) | 9780340806555 | 2003 | Standard 13-digit form of the print ISBN above (same catalogue line). |
| Print edition (Arnold) | 0340595299 | 1995 | Open Library edition record (publisher line matched Arnold/Hodder). |
| ISBN-13 (EAN-13) | 9780340595299 | 1995 | Standard 13-digit form of the print ISBN above (same catalogue line). |
Source notes
- Open Library work editions — Edition list filtered to Arnold/Hodder imprints for ISBNs shown in the table.
Within Engineering, Technology & Statistics on the Arnold list
An Introduction to Randomized Controlled Clinical Trials
First published 2000 · clinical researchers new to trial design
Carpentry & Joinery — Work Activities
First published 2000 · construction-trades apprentices
A Practical Approach to Motor Vehicle Engineering
First published 2000 · apprentice technician
Metal Machining — Theory and Applications
First published 2000 · mechanical engineering undergraduates
Compact Numerical Methods for Computers
First published 1979 · scientific programmers
The Hydraulics of Open Channel Flow — An Introduction
First published 1999 · civil engineering graduates
Frequently asked
- Who is Multilevel Statistical Models written for?
- The book is written for applied statisticians, social scientists working with clustered data (pupils within schools, patients within hospitals), biostatisticians, and quantitative researchers in education, epidemiology and public health.
- What does the book cover?
- It covers two- and three-level random-coefficient models, cross-classified and multiple-membership structures, discrete-response extensions for binary and count outcomes, model checking, and Bayesian estimation routes — moving from the basic random-intercept model through the modelling structures used in current educational and epidemiological practice.
- Is the book still in print?
- The Arnold-imprint editions of the book are out of print as standalone Arnold editions; later editions of Multilevel Statistical Models are published by John Wiley & Sons in the Wiley Series in Probability and Statistics.
Last reviewed: