Joint impact of multiple observations on a subset of variables in multiple linear regression

Park, Sung H.; Lee, Bum S.; Jung, Hyang S.

doi:10.1080/02664760500054418

Cited by 3 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multiple-case deletion was described by Belsley, Kuh, and Welsch (1980) and in recent years, due to rapid increases in computing speed, has become an active area of research-see, for example, Pena and Yohai (1999), Becker and Gather (2001), Park, Lee, and Jung (2005), Pena (2005), Martin and Roberts (2009), and Martin, Roberts, and Zheng (2010). However, even with today's abundant computing resources, multiple-deletion statistics are computationally expensive for even moderately sized datasets.…”

Section: Introductionmentioning

confidence: 97%

An Adaptive, Automatic Multiple-Case Deletion Technique for Detecting Influence in Regression

2014

View full text Add to dashboard Cite

Critical to any regression analysis is the identification of observations that exert a strong influence on the fitted regression model. Traditional regression influence statistics such as Cook's distance and DFFITS, each based on deleting single observations, can fail in the presence of multiple influential observations if these influential observations "mask" one another, or if other effects such as "swamping" occur. Masking refers to the situation where an observation reveals itself as influential only after one or more other observations are deleted. Swamping occurs when points that are not actually outliers/influential are declared to be so because of the effects on the model of other unusual observations. One computationally expensive solution to these problems is the use of influence statistics that delete multiple rather than single observations. In this article, we build on previous work to produce a computationally feasible algorithm for detecting an unknown number of influential observations in the presence of masking. An important difference between our proposed algorithm and existing methods is that we focus on the data that remain after observations are deleted, rather than on the deleted observations themselves. Further, our approach uses a novel confirmatory step designed to provide a secondary assessment of identified observations. Supplementary materials for this article are available online.

show abstract

Section: Introductionmentioning

confidence: 97%

An Adaptive, Automatic Multiple-Case Deletion Technique for Detecting Influence in Regression

2014

View full text Add to dashboard Cite

show abstract

“…Haslett & Haslett (2007) described some efficient methods for computing deletion diagnostics in the context of conditional residuals from even non‐full‐rank general linear models, without the need for model re‐fitting. The problem of detecting multiple, possibly masked, unusual data points in regression is a topic of considerable recent research interest – see, for example, Peña & Yohai (1999), Becker & Gather (2001), Peña (2005) and Park et al (2005).…”

Section: Introductionmentioning

confidence: 99%

Delete‐2 and Delete‐3 Jackknife Procedures for Unmasking in Regression

Martin¹,

Roberts²,

Zheng³

2010

Aus NZ J of Statistics

View full text Add to dashboard Cite

Single-case deletion regression diagnostics have been used widely to discover unusual data points, but such approaches can fail in the presence of multiple unusual data points and as a result of masking. We propose a new approach to the use of single-case deletion diagnostics that involves applying these diagnostics to delete-2 and delete-3 jackknife replicates of the data, and considering the percentage of times among these replicates that points are flagged as unusual as an indicator of their influence. By considering replicates that exclude certain collections of points, subtle masking effects can be uncovered.

show abstract

Trust, self-efficacy, and leadership perception

Coindreau

Reynoso

2019

Dempresarial

View full text Add to dashboard Cite

This research study examines the extent to which hierarchical superiors are admired as leaders, and their viewpoints and decisions accepted, based on their direct support and influence on their subordinates’ self-concept and sense of personal efficacy. To this purpose, a 25-item inventory was designed and filled in paper by 157 participants from 26 companies. Once the statistical behavior of the items was tested, it was found, through lineal regression and correlation analyses, that superiors are better accepted and admired as leaders, when they display trust in the capabilities of their subordinates, contribute to improve their self-efficacy, and support them in their duties. On the other hand, it was also found that considering the subordinates’ skills and interests, when assigning tasks and challenges, was highly correlated with self-efficacy improvement These findings are consistent with previous research. However, it will be important to validate them in future studies.RESUMENEsta investigación examina la admiración hacia los superiores jerárquicos como líderes, así como la aceptación de sus puntos de vista y decisiones, a partir de su apoyo directo y de su influencia en el autoconcepto y sentido de eficacia personal de los colaboradores. Para este propósito, se elaboró un inventario de 25 reactivos que fue contestado en papel por 157 participantes de 26 empresas. Tras verificar el comportamiento estadístico de los reactivos, se identificó mediante análisis correlacionales y de regresión, que los superiores jerárquicos son más aceptados y admirados como líderes cuando demuestran confianza en las capacidades de sus colaboradores y contribuyen a mejorar su autoeficacia, además de apoyarlos en el trabajo. Por otra parte, el aprovechamiento de las habilidades e intereses de los colaboradores está altamente correlacionado con la mejora de su autoeficacia. Estos resultados son consistentes con la literatura, pero será importante verificarlos en estudios posteriores.RESUMOEsta pesquisa examina os superiores admiração como líderes e aceitação de seus pontos de vista e decisões do seu suporte direto e sua influência sobre o auto-conceito e senso de eficácia pessoal dos funcionários. Para tanto, um inventário de 25 itens foi elaborado e respondido em papel por 157 participantes de 26 empresas. Depois de verificar o comportamento estatístico dos reagentes, identificadas pela análise de correlação e regressão, os superiores são mais aceito e admirado como líderes quando eles mostram confiança nas habilidades de seus funcionários e ajudar a melhorar a sua auto-eficácia, bem como apoiá-los na trabalhar Por outro lado, o uso das habilidades e interesses dos colaboradores está altamente correlacionado com a melhora de sua autoeficácia. Estes resultados são consistentes com a literatura, mas será importante verificá-los em estudos posteriores.

show abstract

Joint impact of multiple observations on a subset of variables in multiple linear regression

Cited by 3 publications

References 4 publications

An Adaptive, Automatic Multiple-Case Deletion Technique for Detecting Influence in Regression

An Adaptive, Automatic Multiple-Case Deletion Technique for Detecting Influence in Regression

Delete‐2 and Delete‐3 Jackknife Procedures for Unmasking in Regression

Trust, self-efficacy, and leadership perception

Contact Info

Product

Resources

About