Cross-classified multilevel models

George Leckie

arxiv: 1907.02569 · v1 · pith:33XHG7QJnew · submitted 2019-07-04 · 📊 stat.ME · stat.AP

Cross-classified multilevel models

George Leckie This is my paper

Pith reviewed 2026-05-25 08:50 UTC · model grok-4.3

classification 📊 stat.ME stat.AP

keywords multilevel modelscross-classifiedhierarchical linear modelsnon-hierarchical datasocial science data

0 comments

The pith

Cross-classified multilevel models extend standard multilevel modelling to non-hierarchical data with cross-classified structures.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper is trying to establish that standard multilevel models are insufficient for many real-world datasets because they assume strict hierarchical nesting. Social and behavioural data frequently have cross-classified structures where a lower-level unit belongs to more than one higher-level unit. Cross-classified hierarchical linear modelling is the approach described to analyze such data. A reader would care because this allows more accurate statistical inferences when groups overlap in complex ways.

Core claim

Cross-classified multilevel modelling is an extension of standard multilevel modelling for non-hierarchical data that have cross-classified structures. Traditional models involve exact nesting but social reality is more complicated, so this method handles cases where units are classified by multiple factors.

What carries the argument

Cross-classified hierarchical linear modelling, which accounts for the effects of multiple crossed classifications on the response variable.

If this is right

Accurate variance partitioning across crossed factors becomes possible
Effects from multiple overlapping groups can be estimated simultaneously
Standard models may produce biased results if applied to cross-classified data

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Applications may arise in epidemiology with patients belonging to multiple hospitals and regions
Extensions to generalized linear models for non-normal outcomes could follow the same structure
The description suggests a need for educational resources on implementing these models in software

Load-bearing premise

Social and behavioural data often do not follow pure or strict hierarchies.

What would settle it

Observing no difference in conclusions between standard and cross-classified models when applied to data with known cross-classifications.

read the original abstract

Cross-classified multilevel modelling is an extension of standard multilevel modelling for non-hierarchical data that have cross-classified structures. Traditional multilevel models involve hierarchical data structures whereby lower level units such as students are nested within higher level units such as schools and where these higher level units may in turn be nested within further groupings or clusters such as school districts, regions, and countries. With hierarchical data structures, there is an exact nesting of each lower level unit in one and only one higher level unit. For example, each student attends one school, each school is located within one school district, and so on. However, social reality is more complicated than this, and so social and behavioural data often do not follow pure or strict hierarchies. Two types of non-hierarchical data structures which often appear in practice are cross-classified and multiple membership structures. In this article, we describe cross-classified data structures and cross-classified hierarchical linear modelling which can be used to analyse them.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

This is a clear but unoriginal exposition of cross-classified multilevel models with no new methods or results.

read the letter

This paper is an expository description of cross-classified multilevel models. It explains that standard multilevel models assume strict nesting but many social datasets have units that belong to multiple overlapping groups at once, such as students who attend one school but live in one neighborhood. The write-up walks through the basic idea and notes that extensions exist to handle these structures. That is the main content. It does a decent job keeping the explanation accessible and avoiding heavy notation, which makes the distinction between hierarchical and cross-classified data easy to follow. The abstract is consistent with long-standing definitions in the field. Nothing in the piece is new. The concepts and the modeling approach it describes have appeared in textbooks and papers for years. There are no fresh derivations, no simulation checks, no applied examples with fitted models or software code, and no comparison against other ways of handling non-nested data. The paper simply restates established material. The citation pattern is conventional and points to the usual references, which is fine for an overview but does not add evidence of independent verification. This kind of write-up can be useful for students or applied researchers who have never seen the distinction before and want a short primer. It is less useful for anyone already working with multilevel models. Because the paper contains no original technical contribution, simulation study, or empirical demonstration, it does not look like a candidate for peer review as a research article. If a journal has a dedicated section for tutorials or teaching pieces, that would be a better fit.

Referee Report

0 major / 2 minor

Summary. The paper claims that cross-classified multilevel modelling is an extension of standard multilevel modelling for non-hierarchical data that have cross-classified structures. It contrasts traditional hierarchical nesting (e.g., students within schools within districts) with more complex social and behavioural data structures, identifies cross-classified and multiple-membership structures as common non-hierarchical cases, and states that the article will describe cross-classified data structures together with the cross-classified hierarchical linear modelling approach used to analyse them.

Significance. If the exposition is accurate and complete, the manuscript supplies a concise, accessible introduction to an established extension of multilevel modelling that is directly relevant to analysts working with social and behavioural data. No novel theorem, derivation, or empirical result is advanced, so the contribution is pedagogical rather than methodological; its value rests on clarity of presentation for readers who encounter non-nested structures.

minor comments (2)

[Abstract] The abstract and opening paragraphs repeatedly use the phrase 'cross-classified hierarchical linear modelling' without immediately distinguishing it from the more common term 'cross-classified multilevel model'; a single clarifying sentence would reduce potential reader confusion.
No equations, worked numerical example, or software syntax is referenced in the provided abstract; if the full manuscript likewise omits a concrete illustration, the description remains purely verbal and may limit utility for practitioners.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive review and recommendation to accept the manuscript. The report accurately summarizes the paper's scope and contribution as a pedagogical introduction to cross-classified multilevel models.

Circularity Check

0 steps flagged

No significant circularity; purely expository

full rationale

The paper is an explanatory tutorial describing cross-classified multilevel models as an established extension of hierarchical linear models. No equations, derivations, predictions, fitted parameters, or novel claims are advanced. The abstract and text consist of definitions and standard distinctions between hierarchical and cross-classified structures, with no load-bearing steps that could reduce to self-citation or construction. This matches the reader's assessment of score 0.0.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

The paper is descriptive and introduces no new free parameters, axioms, or invented entities beyond standard statistical modeling assumptions already present in multilevel literature.

pith-pipeline@v0.9.0 · 5683 in / 1025 out tokens · 49162 ms · 2026-05-25T08:50:48.245356+00:00 · methodology

Cross-classified multilevel models

Core claim

What carries the argument

If this is right

Where Pith is reading between the lines

Load-bearing premise

What would settle it

discussion (0)