Premium
Using Multilevel Models to Model Heterogeneity: Potential and Pitfalls
Author(s) -
Duncan Craig,
Jones Kelvyn
Publication year - 2000
Publication title -
geographical analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.773
H-Index - 65
eISSN - 1538-4632
pISSN - 0016-7363
DOI - 10.1111/j.1538-4632.2000.tb00429.x
Subject(s) - categorical variable , multilevel model , scale (ratio) , population , computer science , econometrics , management science , data science , sociology , geography , mathematics , cartography , machine learning , demography , economics
Within the last few years, geographers and researchers in other cognate disciplines with geographic concerns have begun to use multilevel models. While there are several useful existing introductory accounts of these models in the geographical literature, this paper seeks to extend them in three main ways to clarify and emphasize further the substantial opportunities they afford. First, it focuses on how multilevel models are centrally concerned with modeling population heterogeneity as a function of predictor variables. Second, it considers and illustrates a number of specific interpretive issues that can arise when conducting multilevel analyses of place effects. Lastly, it traces some more general, conceptual issues surrounding the use of multilevel models in geographical research. The arguments made are illustrated through an analysis of variations in drinking behavior using data from a typically complex, large‐scale survey; particular attention is given to the inclusion of categorical predictors.