Cook’s distance for generalized linear mixed models

Pinho, Luís Gustavo Bastos; Nobre, Juvêncio S.; Singer, Júlio M.

doi:10.1016/j.csda.2014.08.008

Cited by 22 publications

(19 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

Section: Discussionmentioning

confidence: 99%

“…Following a conservative criterion, observations with Cook's distance over percentile 99 for both fixed and random effects (Pinho et al, 2015) were excluded (n = 7) and models were then re-fitted.…”

Section: Discussionmentioning

confidence: 99%

“…Evaluation of model residuals was used to identify outliers and/or influential observations (i.e., data points with high leverage). We calculated Cook's distance for GLMM (Pinho, Nobre, & Singer, ) to detect influential observations. Following a conservative criterion, observations with Cook's distance over percentile 99 for both fixed and random effects (Pinho et al, ) were excluded ( n = 7) and models were then re‐fitted.…”

Section: Methodsmentioning

confidence: 99%

See 2 more Smart Citations

Facilitation beyond species richness

2018

View full text Add to dashboard Cite

Facilitation studies have previously focused on the effects of plant–plant interactions on species richness and, more recently, on functional traits or phylogenetic aspects. Little is known, however, about the simultaneous effects that facilitation have on overall biodiversity, jointly considering taxonomic, functional and phylogenetic diversity. In this study, we investigated shrub facilitation on herb communities in a Mediterranean grassland over a 9‐year period (2007–2015), each year representing different water availability conditions. Taxonomic, functional and phylogenetic diversity indices were estimated using generalised Hill numbers within the integrated framework of attribute diversity. We used generalised linear mixed models to analyse the effect of shrub cover on the different diversity indices along a water availability gradient. Shrub cover had a positive effect on all attribute diversity indices, yet greater for abundant species. Shrubs also favoured survival of certain species (26.5% of total species), but above all they affected relative abundances of common species, filtering growth of dominant species (e.g., Poa bulbosa) and promoting evenness. Diversity indices accounting only for abundant (i.e., high coverage) species showed the strongest facilitation effects that peaked at mid‐range water availability conditions. Synthesis. Our results suggest that shrubs are key in the maintenance of all three dimensions of biodiversity in the Mediterranean grasslands, promoting more even communities, and to a lower extent, also species richness. We show that shrub facilitation on taxonomic, functional and phylogenetic diversity of herbaceous communities is maximum for abundant species at mid‐range values of water availability, in accordance with the humped‐back model. This study illustrates the need to evaluate different biodiversity dimensions from an integrated perspective, such as attribute diversity, to disentangle the role of facilitation in community assembly.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Facilitation beyond species richness

2018

View full text Add to dashboard Cite

show abstract

“…First, we used Cleveland plots between each covariate and the response variable (Zuur, Ieno & Elphic ). Secondly, we calculated Cook's distances with formulas proposed for GLMM (Xiang, Tsea & Lee ; Pinho, Nobrea & Singerb ). Thirdly, we conducted an influential‐cases analysis to quantify the extent to which model parameters are affected by removal of observations (Van der Meer, Te Grotenhuis & Pelzer ).…”

Section: Methodsmentioning

confidence: 99%

The modulating role of group stability on fitness effects of group size is different in females and males of a communally rearing rodent

Ebensperger

Correa

León

et al. 2016

Journal of Animal Ecology

View full text Add to dashboard Cite

Group size may influence fitness benefits and costs that emerge from cooperative and competitive interactions in social species. However, evidence from plural breeding mammals indicates that group size is insufficient to explain variation in direct fitness, implying other attributes of social groups were overlooked. We studied the natural population of a social rodent during 5 years to test the hypothesis that social stability - in terms of group composition - modulates the effects of increasing number of breeding females (a proxy of communal rearing) and males on the number of offspring weaned (sired) and on the number of offspring weaned (sired) surviving to breeding age (two proxies of direct fitness). We quantified the effects of social stability (measured as changes in female or male group members between mating and the onset of lactation) on these fitness measures. We used live trapping, telemetry and DNA markers to determine social and fitness measures. Social stability in degus was variable in terms of the number of changes in group composition across groups. Low stability was mostly due to mortality and emigration of group members. Results supported a modulating role of social stability on the relationship between group size and the number of offspring weaned (sired). Stability in female and male group composition were both modulators of fitness to females and males. The modulatory role of stability was sex specific, where high social stability was often fitness beneficial to the females. Instead, low social stability was fitness enhancing to the males.

show abstract

“…Thus, it is necessary to use a measure that is useful for evaluation in such cases. In this sense, an extension of the measure of influence proposed by Xiang, Tse, and Lee (2002) was proposed by Pinho, Nobre, and Singer (2015). This extension takes into account the aspects described above.…”

Section: Diagnostic Analysismentioning

confidence: 99%

Longitudinal modeling using log-gamma mixed model: case of memory deterioration after chronic cerebral hypoperfusion associated with diabetes in rats

et al. 2019

View full text Add to dashboard Cite

In recent years several longitudinal studies have been conducted in the field of pharmacology. In general, continuous response variables occur frequently in these situations and tend to present asymmetric characteristics, as well as being restricted to the set of positive real numbers. Therefore, using the normal model would be incorrect. In this conjecture, generalized linear mixed models (GLMM) are used to analyze data characterized in this way, aiming to accommodate inter-and intra-individual variations. Thus, we propose a mixed gamma model (LGMM) with a log link function and random effects normally distributed to evaluate data from a longitudinal experiment, where the effects of cerebral ischemia associated with diabetes on the performance of long-term retrograde memory were evaluated in rats. Based on the results obtained, the random intercept model presented a good fit and accommodated the correlation inherent to the data. It was possible to observe that normoglycemic animals, when compared to hyperglycemic animals, whether submitted to ischemia or not, had their cognitive capacity partially preserved, indicating that hyperglycemia ('diabetes') aggravates the cognitive effects of brain ischemia.

show abstract

Cook’s distance for generalized linear mixed models

Cited by 22 publications

References 24 publications

Facilitation beyond species richness

Facilitation beyond species richness

The modulating role of group stability on fitness effects of group size is different in females and males of a communally rearing rodent

Longitudinal modeling using log-gamma mixed model: case of memory deterioration after chronic cerebral hypoperfusion associated with diabetes in rats

Contact Info

Product

Resources

About