Using Generalized Estimating Equations for Longitudinal Data Analysis

By Gary A. Ballinger. 2004

1,560 citations recorded by the source corpus.

Subfields: Computer Networks and Communications, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/70e5c2060902b67591a245f7174ac1b8ecf2a8dd

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Networks and Communications

In Statistics and Probability

Citations in the corpus

Desirability bias in foresight: Consequences for decision quality based on Delphi results (2011, background): “It is crucial to account for the clustered data structure and to correct for potential correlations in order to obtain meaningful coefficients, reliable standard errors, and to avoid a loss of information [18,67].”
CEO Career Variety: Effects on Firm-Level Strategic and Social Novelty (2014, background): “Thus, a unit change in the independent variable is equivalent to the expected change in the average response, or dependent variable (Ballinger, 2004). found evidence that firms led by CEOs with higher levels of career variety display greater strategic dynamism (resource reallocation and corporate…”
Examining the “I” in Team: A Longitudinal Investigation of the Influence of Team Narcissism Composition on Team Outcomes in the NBA (2020, methodology): “We conducted the analysis using generalized estimating equations (GEE) (Ballinger, 2004; Liang & Zeger, 1986) in SPSS (version 22).”
DunedinPACE, a DNA methylation biomarker of the pace of aging (2021, methodology): “…DNA methylation data in the NAS, we used generalized estimating equations to account for nonindependence of repeated observations of individuals (Ballinger, 2004), following the method in previous analysis of those data (Gao et al., 2018), and econometric fixed-effects regression (Wooldridge,…”
Rethinking ‘Top-Down’ and ‘Bottom-Up’ Roles of Top and Middle Managers in Organizational Change: Implications for Employee Support (2017, positive): “…data using Generalized Estimating Equations (GEE) – which is an extension of the Generalized Linear Model that allows for non-independence of observations caused by nesting, clustering, repeated measures, and/or longitudinal observations (Ballinger, 2004; Echambadi, Campbell, and Agarwal, 2006).”
Slack resources, firm performance, and the institutional context: Evidence from privately held European firms (2017, methodology): “With a pooled dataset consisting of up to four years of data for each firm, we tested our hypotheses using Generalized Estimating Equations (GEE) which accommodates the analysis of panel data with repeated, within-subject measures (Ballinger, 2004).”
Team Locus-of-Control Composition, Leadership Structure, Information Acquisition, and Financial Performance: A Business Simulation Study (2005, methodology): “A drawback of GEE, however, is that the residuals from these models are correlated, making the development of summary goodness-of-fit statistics for GEE models problematic (Ballinger, 2004; Zorn, 2001).”
On the Unnecessary Ubiquity of Hierarchical Linear Modeling (2017, methodology): “With GEE, there is some researcher input required when selecting the working correlation structure; however, it does not have to be properly specified and estimates are robust to fairly large misspecifications (Ballinger, 2004; Zeger et al., 1988).”
Outcome of asthma and wheezing in the first 6 years of life: follow-up through adolescence. (2005, methodology): “General estimating equations were used to assess the relation between the preschool wheeze phenotypes and longitudinal wheeze from age 8 to 16 yr (12).”
A Tale of Two Assets: The Effects of Firm Reputation and Celebrity on Earnings Surprises and Investors' Reactions (2010, methodology): “As a robustness check, however, we also utilized a generalized estimating equations (GEE) regression model, a method found suitable for panel data because it measures both within- and between-firm variance and generates robust estimates of standard errors (Ballinger, 2004; Wade et al., 2006).”
Modeling Clustered Data with Very Few Clusters (2016, methodology): “For example, Bauer and Sterba (2011) noted that 94% of psychology studies account for clustering with MLMs whereas Peterson (2009) found that only 3% of studies in economics account for clustering with MLMs.”
Networks, Technology, and Entrepreneurship: A Field Quasi-experiment among Women in Rural India (2017, neutral): “…suited for the structure of our data; it not only allows for the modeling of time (here, year), but also accounts for the systematic biases, nonindependence, person-level, and person-year variance introduced by repeated measures from the same participants over time (see Ballinger, 2004).”
Endogeneity: A Review and Agenda for the Methodology-Practice Divide Affecting Micro and Macro Research (2020, background): “Arellano & Bond, 1991; Ballinger, 2004; Bergh, 1993; Blundell & Bond, 1998 Using Exogenous Events – Quasiexperiments where intervention or exogenous event is used to establish direction of causality.”
Job Characteristics and Job Satisfaction: Understanding the Role of Enterprise Resource (2010, methodology): “GEEs can be used to test main effects, interactions, and categorical and continuous independent variables (Ballinger 2004).”
Should Entrepreneurially Oriented Firms Have Narcissistic CEOs? (2016, methodology): “Following the suggestions from Ballinger (2004) to fit GEE models, we chose an identity link function and the Gaussian link family.”
Group territoriality and the benefits of sociality in the African lion, Panthera leo (2009, methodology): “We used generalized estimating equations (GEE) to analyse the relationships between the dependent and independent variables (PROC GENMOD, SAS 9.1, SAS Institute Inc., Cary, NC, U.S.A.) (Zeger et al. 1988; Ballinger 2004).”