Panel Models in Sociological Research: Theory into Practice

Halaby, Charles N.

doi:10.1146/annurev.soc.30.012703.110629

Cited by 802 publications

(637 citation statements)

References 62 publications

Supporting

Mentioning

629

Contrasting

Unclassified

Order By: Relevance

“…For example, the use of random-effects modeling (via HLM), which is showing a clear upward trend, is worrisome because it appears that most who use the method do not know that they may have reported seriously compromised estimates by ignoring endogeneity and not having examined if their estimator is consistent (Antonakis, et al, 2010;Halaby, 2004). There are very simple remedial procedures to take when the random-effects estimator is inconsistent and we hope that researchers (and journal editors) will pay more attention to these issues.…”

Section: The Future Of Methodsmentioning

confidence: 99%

“…These models are called random-effects/coefficients models (or oftentimes HLM models, named for the popular program used to estimate them). However, these models have a major shortcoming that has been repeatedly documented in the literature (Antonakis, et al, 2010;Bollen & Brand, 2010;Cameron & Trivedi, 2005;Halaby, 2004;Mundlak, 1978) and calls to correctly test these models have mostly fallen on deaf ears. As mentioned by Halaby (2004, p. 508), "Key principles that ought to routinely inform analysis are at times glossed over or ignored completely."…”

Section: Structural Equation Modeling (Sem)mentioning

confidence: 99%

See 1 more Smart Citation

What makes articles highly cited?

Antonakis

Bastardoz

Liu

et al. 2014

The Leadership Quarterly

109

View full text Add to dashboard Cite

We examined drivers of article citations using 776 articles that were published from 1990-2012 in a broad-based and high-impact social sciences journal, The Leadership Quarterly. These articles had 1,191 unique authors having published and received in total (at the time of their most recent article published in our dataset) 16,817 articles and 284,777 citations, respectively. Our models explained 66.6% of the variance in citations and showed that quantitative, review, method, and theory articles were significantly more cited than were qualitative articles or agentbased simulations. As concerns quantitative articles, which constituted the majority of the sample, our model explained 80.3% of the variance in citations; some methods (e.g., use of SEM) and designs (e.g., meta-analysis), as well as theoretical approaches (e.g., use of transformational, charismatic, or visionary type-leadership theories) predicted higher article citations. Regarding the statistical conclusion validity of quantitative articles, articles having endogeneity threats received significantly fewer citations than did those using a more robust design or an estimation procedure that ensured correct causal estimation. We make several general recommendations on how to improve research practice and article citations.

show abstract

Section: The Future Of Methodsmentioning

confidence: 99%

Section: Structural Equation Modeling (Sem)mentioning

confidence: 99%

What makes articles highly cited?

Antonakis

Bastardoz

Liu

et al. 2014

The Leadership Quarterly

109

View full text Add to dashboard Cite

show abstract

“…The alternative random effects model (REM) is capable of producing estimates for world-system position, but the consistency of REM estimates hinges crucially on the validity of the assumption that the country-specific error term is uncorrelated with the right-hand covariates. Diagnostic (Hausman) tests show that this assumption is violated by these data (Halaby, 2004;Wooldridge, 2002). Monte Carlo simulations suggest the FEVDM is preferable to the REM model when the assumption of uncorrelated unit effects is not met, and to FE models when the between case variation is sufficiently large relative to the within variance, as is the case here (Plumper and Troeger, 2007).…”

Section: Notesmentioning

confidence: 99%

Global models of networked organization, the positional power of nations and economic development

Mahutga

2013

Review of International Political Economy

View full text Add to dashboard Cite

Interdisciplinary literature on global commodity chains (GCCs)/global value chains (GVCs) and global production networks (GPNs) contends that interfirm power differentials within globally networked forms of economic organization have implications for the developmental trajectories of nation-states. In this article, I advance these literatures in three ways. First, I bridge the two approaches by elaborating an exchange-theoretic conceptualization of interfirm power that is latent in the two literatures. This conceptualization focuses narrowly on the determinants of inter-firm power asymmetries and is useful for explaining why actual production networks vary in terms of the relative power of buyers and producers. Second, I develop an empirical framework to advance basic research on the link between globally networked forms of economic organization and national economic development. In particular, I derive cross-nationally and temporally comparable country-level measurements of the average bargaining power of a country's resident firms using industry-specific international exchange (trade) networks. I demonstrate the validity of these indices through a historical analysis of trade networks in the transport equipment and garment industries and by analysing crossnational variations in wages in the two industries. Finally, I conclude by charting a parallel path for GCC/GVC and GPN research that implicates global models of network organization in macro-comparative analyses of economic development.

show abstract

“…In the analysis of panel data, fixed effects regression models are usually to be preferred over random effects models (Halaby 2004). Fixed effects models estimate the relation between changes in one variable and changes in an outcome variable within the same respondent.…”

Section: The Effect Of Volunteering On Trustmentioning

confidence: 99%

Trust and Volunteering: Selection or Causation? Evidence From a 4 Year Panel Study

Bekkers¹

2011

Polit Behav

113

View full text Add to dashboard Cite

Participation in voluntary associations is often believed to make citizens more trusting of others. This paper reports longitudinal analyses of a time intensive form of participation-volunteering-and generalized social trust using data from three waves of the Giving in the Netherlands Panel Study spanning 4 years (2002)(2003)(2004)(2005)(2006); n = 692) refuting this belief. Trust is relatively stable over a 4 year period (0.73). Changes in volunteering are not related to changes in trust. Trust is higher among volunteers mainly because of selective attrition: persons with low trust are more likely to quit volunteering.

show abstract

Panel Models in Sociological Research: Theory into Practice

Cited by 802 publications

References 62 publications

What makes articles highly cited?

What makes articles highly cited?

Global models of networked organization, the positional power of nations and economic development

Trust and Volunteering: Selection or Causation? Evidence From a 4 Year Panel Study

Contact Info

Product

Resources

About