Estimating Effects of Limiting Factors with Regression Quantiles

Cade, Brian S.; Terrell, James W.; Schroeder, Richard L.

doi:10.2307/176999

Cited by 211 publications

(350 citation statements)

References 0 publications

Supporting

Mentioning

341

Contrasting

Unclassified

Order By: Relevance

“…This concept extends to the conditional QR. The conditional quantiles denoted by Q y (q|X) are the inverse of the conditional cumulative distribution function of the response variable, Fy −1 (q|X), where q ɛ [0, 1] denotes the quantiles (Cade, Terrell, and Schroeder 1999;Koenker and Machado 1999). Here, we consider functions of X that are linear in the parameters; for example, Q y (q|X) = β 0 (q)X 0 + β 1 (q)X 1 + β 2 (q)X 2 + ,.., + βp(q)Xp, where the (q) notation indicates that the parameters are for a specified q quantile.…”

Section: Analytical Frameworkmentioning

confidence: 99%

Poverty in maize growing rural communities of southern Africa

Kassie¹,

Abate

Langyintuo

et al. 2014

Development Studies Research

View full text Add to dashboard Cite

This study provides a detailed analysis of the extent and determinants of poverty in the maize-based rural economies of selected districts in southern African countries of Angola, Malawi, Mozambique, and Zambia. Poverty profiles showed that households that were absolutely poor had significantly smaller family sizes, fewer livestock, smaller farm sizes, smaller number of illiterate household members, and less number of important assets such as phones and radios. The study also details the extent and determinants of poverty in the study countries using quantile regression. The importance of maize technology use and resource allocation to the crop in determining the magnitude of poverty is an important finding as the sampled households were essentially semi-subsistent with limited market orientation. This finding also justifies the effort being exerted on the development and deployment of maize and maize-related technologies in rural communities of southern Africa.

show abstract

Section: Analytical Frameworkmentioning

confidence: 99%

Poverty in maize growing rural communities of southern Africa

Kassie¹,

Abate

Langyintuo

et al. 2014

Development Studies Research

View full text Add to dashboard Cite

show abstract

“…Implementation of the regression methodology generally involves some difficulties related to noise in the data and to the choice of the environmental parameters. Noise in the data was addressed by modelling the statistical distribution of the response variable by using generalised linear and/or additive models (e.g., Maravelias and Reid, 1997), regression quantiles (Cade et al, 1999) or cumulative distribution tests (Perry and Smith, 1994). The choice of environmental parameters means building the static statistical model with indirect parameters describing biogeographical pattern or with direct parameters that closely relate the physiological processes at work, making the model more mechanistic and therefore perhaps better fitted but also less transportable (e.g.…”

Section: Introductionmentioning

confidence: 99%

Hydro-plankton characteristics and their relationship with sardine and anchovy distributions on the French shelf of the Bay of Biscay

Petitgas¹,

Massé²,

Bourriau³

et al. 2006

Sci. Mar.

View full text Add to dashboard Cite

SUMMARY:The spatial pattern in hydro-plankton and fish distributions and their relationship were analysed based on the spring 2000 fisheries acoustic survey. The importance of this survey was that it was a multi-disciplinary platform which collected an extensive set of parameters in the hydro-plankton leading to a potentially finer description of hydro-plankton conditions and fish habitats. More than 50 variables were measured on a grid of stations, in four compartments of the ecosystem: hydrology, nutrients, primary producers and meso-zooplankton. First, a joint analysis of all hydro-plankton compartments was performed using multiple factor analysis (MFA). The method was used to estimate a compromise factorial space common to all compartments in which the stations were grouped by hierarchical clustering. The groups were represented spatially and a strong spatial pattern was evidenced. The fish and their spawned eggs were sampled along transect lines using acoustics and CUFES (continuous underway fish egg samplers). The distribution of the fish and their eggs was analysed in relation to the hydro-plankton groups of stations and difference in fish density across hydro-plankton conditions was tested by a pair-wise multiple comparison procedure. Anchovy was associated with a lesser number of hydro-plankton conditions than sardine. Eggs of both species were also associated with a lesser number of conditions than the fish. Finally, the gain provided by using the extensive set of hydro-plankton parameters for mapping large-scale hydro-plankton conditions was analysed in comparison with the situation in which a small set of parameters was available. The extensive set of parameters allowed more hydro-plankton conditions to be identified but only in the coastal area and not on the shelf. Size fractionated chlorophyll was determinant for tracking river plume hydro-plankton condition. However, the fish did not respond to the variety of the coastal hydro-plankton conditions. Keywords: spatial pattern, multiple factor analysis, habitats, hydro-plankton, fish-environment relationship. RESUMEN: CARACTERÍSTICAS HIDRO-PLANCTÓNICAS Y SU RELACIÓN CON LAS DISTRIBUCIONES DE SARDINA Y ANCHOA EN LAPLATAFORMA FRANCESA DEL GOLFO DE VIZCAYA. -Este artículo analiza la relación espacial entre los peces pelágicos y su medio ambiente utilizando datos de la campaña acústica francesa del año 2000. Esta campaña fue multidisciplinaria y permitió la colección de una lista extensa de parámetro en el hidro-plancton. Mas de 50 variables fueron muestreadas sobre la misma red de estaciones. Estas variables se agrupan en cuatro compartimientos del ecosistema: hidrología, nutrientes y materia en suspensión, bacterias y fitoplancton, mesozooplancton. En primer lugar se hizo una análisis conjunto a todas las variables reagrupadas en los cuatros compartimientos aplicando el método MFA (Análisis Factorial Múltiple). El método permitió estimar el espacio factorial común a todos los compartimientos en el que se hizo un agrupamiento de las estaciones. Los grup...

show abstract

“…Usually, hypotheses about the central response of organisms to environmental gradients are tested, although the effects of other variables may influence such response and decrease the fit of the model, which may even become uninformative (Lancaster and Belyea, 2006). In this perspective, quantile regression, calculated at the extreme quantiles, allows the chosen independent variables to be considered as "constraints" to the distribution of biological communities, without compromising the model causal relationship (Cade et al, 1999). Regressions of the extreme quantiles characterize the slope and shape of the boundaries of the data, when biological metrics are tested against environmental gradients (e.g., habitat availability).…”

Section: Development Of Quantitative Habitat Suitability Modelsmentioning

confidence: 99%

“…This statistical tool was introduced in ecology by Cade et al (1999) and can be used to test the role of environmental factors as constraints. Moreover, its application allows the predictions not only of the more probable values of the studied biological metric but also of the maximum or minimum values that could be expected in environmental conditions comparable to the ones used for the model fitting (Cade and Noon, 2003;Doll, 2011;Fornaroli et al, 2015).…”

Section: Introductionmentioning

confidence: 99%