Understanding Regression Assumptions

Berry, William

doi:10.4135/9781412986427

Cited by 398 publications

(269 citation statements)

References 0 publications

Supporting

Mentioning

253

Contrasting

Unclassified

Order By: Relevance

“…For a detailed discussion of regression assumptions, see Berry (1993) and Hayes and Darlington (2017). Here we address the normality and homoscedasticity assumptions.…”

Section: Linearity Homoscedasticity and Other Assumptionsmentioning

confidence: 99%

A Tutorial on Testing, Visualizing, and Probing an Interaction Involving a Multicategorical Variable in Linear Regression Analysis

Hayes

Montoya

2017

Communication Methods and Measures

332

198

View full text Add to dashboard Cite

“…For a detailed discussion of regression assumptions, see Berry (1993) and Hayes and Darlington (2017). Here we address the normality and homoscedasticity assumptions.…”

Section: Linearity Homoscedasticity and Other Assumptionsmentioning

confidence: 99%

A Tutorial on Testing, Visualizing, and Probing an Interaction Involving a Multicategorical Variable in Linear Regression Analysis

Hayes

Montoya

2017

Communication Methods and Measures

332

198

View full text Add to dashboard Cite

“…OLS measures are problematic metrics of sprawl, however. Density gradients attenuate or grow with distance by their very nature; they inherently succumb to spatial autocorrelation, thereby invalidating OLS regression results (Berry 1993). We employ a spatially-adjusted regression model to get around these problems (Fotheringham et al 2000), which includes a vector of adjacent means (a spatial lag) of the dependent variable as an extra independent variable in the analysis.…”

Section: Density Characteristicsmentioning

confidence: 99%

“…Sprawl is a dynamic phenomenon, yet work on sprawl often focuses on a single temporal snapshot or disjointed snapshots, rather than following longitudinally in synchrony with urban evolution. The methodology most commonly employed in analysis relies heavily on descriptive and multivariate statistics that are prone to unreliable results owing to spatial autocorrelation (Berry 1993;Fotheringham et al 2004;Moran 1950). Use of geospatial metrics to avoid spatial autocorrelation problems (usually a fatal roadblock when encountered in analysis) is exceptional when measuring sprawl.…”

Section: Introductionmentioning

confidence: 99%

A Toolkit for Measuring Sprawl

Torrens

2008

Appl. Spatial Analysis

202

163

View full text Add to dashboard Cite

Debate regarding suburban sprawl in urban studies is contentious. It is fair to say that the phenomenon is not fully understood to satisfaction in the academic, policy, or planning communities and there are a host of reasons why this may be the case. Characterization of sprawl in the literature is often narrative and subjective. Measurement is piecemeal and largely data-driven. Existing studies yield contrary results for the same cities in many cases. The partial appreciation for the intricacies of sprawl is problematic. In practice, city planning agencies and citizen advocacy groups are scrambling to suggest and develop "smart growth" strategies to curb sprawl, without a strong empirical basis for measuring the phenomenon. Yet, sprawl is extremely popular with consumers. In this paper, we develop an innovative approach to diagnosing sprawl, looking across the full range of its characteristic attributes in a comprehensive fashion that is robust to some well-known challenges. This proves to be very useful in sweeping the parameter space of the phenomenon, enabling the visualization and valuation of sprawl surfaces across attributes, allowing us to check the pulse of a developing city. We apply the work to Austin, TX, a controversial exemplar of American sprawl, with the surprising result that sprawl and "smart growth" are found to co-exist and co-evolve. This raises questions about relationships between the two, with consequences for planning and public policy.

show abstract

“…The scale data were ordinal, yet we treated them as continuous for the purpose of our analysis because the Petty Tyranny in Organizations Scale and Satisfaction Scale produce quantitative discrete ordinal variables with a sufficiently wide range of values [19]. Most of the demographics were categorical.…”

Section: Resultsmentioning

confidence: 99%

The Impact of Prevalent Destructive Leadership Behaviour on Subordinate Employees in a Firm

Tran¹,

Tian²,

Sankoh³

2013

AJIBM

View full text Add to dashboard Cite

This paper examines the impact of destructive leadership behaviours experienced by subordinate employees. Structured questionnaires based on the petty Tyranny in organizations scale to explore the scope and nature of destructive leadership were used. The study further explores the relationship among leadership experiences, various measures of subordinates' satisfaction and turnover level. The results showed that despite the central role of leadership holds for many firms in Hanoi, Vietnam, subordinate employees reported experiencing toxic destructive leadership. There was a significant negative relationship among destructive leadership, all measures of satisfaction and turnover level. Surprisingly, there was not a significant negative impact on turnover level (inclination to remain in the employment) among the subordinate population.

show abstract

Understanding Regression Assumptions

Cited by 398 publications

References 0 publications

A Tutorial on Testing, Visualizing, and Probing an Interaction Involving a Multicategorical Variable in Linear Regression Analysis

A Tutorial on Testing, Visualizing, and Probing an Interaction Involving a Multicategorical Variable in Linear Regression Analysis

A Toolkit for Measuring Sprawl

The Impact of Prevalent Destructive Leadership Behaviour on Subordinate Employees in a Firm

Contact Info

Product

Resources

About