Using heteroskedasticity-consistent standard error estimators in OLS regression: An introduction and software implementation

By Andrew F. Hayes, Li Cai. Behavior Research Methods, 2007

1,298 citations recorded by the source corpus.

Subfields: Economics and Econometrics, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/bf77cf9bd8432afb754f8570a395b510421c4ac3

How researchers see this paper

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

In Economics and Econometrics

In Statistics and Probability

Citations in the corpus

Neglected biological patterns in the residuals (2011, methodology): “In a traditional regression, the homoscedasticity of errors is assumed and the variance–covariance matrix of the errors, φ, is taken to be σ(2)I, where I represents an identity matrix of appropriate dimensions, and σ(2) represents the variance of model residuals (Hayes and Cai 2007).”
Two-condition within-participant statistical mediation analysis: A path-analytic framework. (2017, background): “Anything that affects the quality of these point estimates such as heteroscedasticity (Hayes & Cai, 2007; Long & Ervin, 2000; Yuan & MacKinnon, 2014) could potentially affect the inference.”
An eye‐tracking study on information processing in risky decisions: Evidence for compensatory strategies based on automatic processes (2011, neutral): “9We used STATA regression options ‘‘cluster(id)’’ and ‘‘robust’’ (see Gould, Pitblado, & Sribney, 2006; Hayes & Cai, 2007; Rogers, 1993).”
The Rise of an Athlete Brand: Factors Influencing the Social Media following of Athletes (2020, neutral): “…which violated the assumption in OLS regression that the error term had constant variance across all values of the independent variables (Hayes & Cai, 2007), we calculated heteroscedastici-ty-consistent standard errors before we performed the regression analyses (Hill et al., 2011).”
Coherence Shifts in Probabilistic Inference Tasks (2010, methodology): “To account for the repeated measurement design SEs were adjusted for 74 clusters of observations and robust standard errors were used to account for violations of the homoscedasticity assumption (Gould, Pitblado, & Sribney, 2006; Hayes & Cai, 2007).”
On the Unnecessary Ubiquity of Hierarchical Linear Modeling (2017, methodology): “…are unaffected by the statistical correction to the standard error estimates and the computational formulas are equivalent to a single-level model (Hayes & Cai, 2007).4 This makes CR-SEs an 3 A negative ICC may be present, for instance, if a school district intentionally tries to balance the…”
Persuasion Through Emotion? An Experimental Test of the Emotion-Eliciting Nature of Populist Communication (2018, methodology): “The variables were meancentered for products, and HC3 correction was used to obtain heteroskedasticity-consistent standard errors (Hayes & Cai, 2007).”
The positive and negative semantic dimensions of relationship satisfaction (2013, methodology): “To correct for this, we employed the heteroskedasticityconsistent standard error estimate procedure developed by Cribari-Neto (2004), and computed the analyses using the SPSS macro and syntax provided by Hayes and Cai (2007).”
The Relative Trustworthiness of Inferential Tests of the Indirect Effect in Statistical Mediation Analysis (2013, background): “In principle, anything that could influence the accuracy of those standard errors, such as heteroscedasticity (see e.g., Hayes & Cai, 2007; Long & Ervin, 2000), could influence the performance of these methods.”
Unraveling the “passion orchestra” in academia (2016, methodology): “Therefore, we reran our analyses using the HC3 estimator for robust standard errors, described in Long and Ervin (2000) and Hayes and Cai (2007).”
How Hot Is Your Bottom Line? Linking Carbon and Financial Performance (2011, methodology): “To comply with the homoskedasticity assumption of OLS regressions, we used heteroskedasticity-consistent standard error estimators (Hayes & Cai, 2007).”