A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function

The maximum likelihood parameters estimated in logistic analysis are in terms of a transformation of the original response variable and, although inferences are easily made about the sources of variation in the linearized model using standard procedures applied in regression analysis, the variance-c...

Full description

Bibliographic Details
Main Authors: Rege, J.E.O., Sherington, J.
Format: Journal Article
Language:Inglés
Published: 1996
Subjects:
Online Access:https://hdl.handle.net/10568/28452
_version_ 1855518929436278784
author Rege, J.E.O.
Sherington, J.
author_browse Rege, J.E.O.
Sherington, J.
author_facet Rege, J.E.O.
Sherington, J.
author_sort Rege, J.E.O.
collection Repository of Agricultural Research Outputs (CGSpace)
description The maximum likelihood parameters estimated in logistic analysis are in terms of a transformation of the original response variable and, although inferences are easily made about the sources of variation in the linearized model using standard procedures applied in regression analysis, the variance-covariance structure of the predicted probabilities obtained following back-transformation of logits is complex and estiamtion of sampling variances normally requires inversion of matrices and taking derivatives of the inverse of the link function evaluated at each prediction point. This paper presents a method for estimating sampling variances of such predicted probabilities without the need to invert any matrix or take derivatives of the link function. The method is based on the assumption that the exponent of a linear function of the logits is lognormal. It is demonstrated by way of a numerical example that this approximation is not different from the more complex methods applied by such software as SAS and GENSTAT.
format Journal Article
id CGSpace28452
institution CGIAR Consortium
language Inglés
publishDate 1996
publishDateRange 1996
publishDateSort 1996
record_format dspace
spelling CGSpace284522022-01-29T16:15:21Z A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function Rege, J.E.O. Sherington, J. sampling statistical analysis variants methods The maximum likelihood parameters estimated in logistic analysis are in terms of a transformation of the original response variable and, although inferences are easily made about the sources of variation in the linearized model using standard procedures applied in regression analysis, the variance-covariance structure of the predicted probabilities obtained following back-transformation of logits is complex and estiamtion of sampling variances normally requires inversion of matrices and taking derivatives of the inverse of the link function evaluated at each prediction point. This paper presents a method for estimating sampling variances of such predicted probabilities without the need to invert any matrix or take derivatives of the link function. The method is based on the assumption that the exponent of a linear function of the logits is lognormal. It is demonstrated by way of a numerical example that this approximation is not different from the more complex methods applied by such software as SAS and GENSTAT. 1996 2013-05-06T07:00:38Z 2013-05-06T07:00:38Z Journal Article https://hdl.handle.net/10568/28452 en Limited Access Discovery and Innovation;8(2): 171-180
spellingShingle sampling
statistical analysis
variants
methods
Rege, J.E.O.
Sherington, J.
A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function
title A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function
title_full A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function
title_fullStr A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function
title_full_unstemmed A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function
title_short A method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function
title_sort method for estimating sampling variances of predicted probabilities from maximum likelihood estimates in logit response function
topic sampling
statistical analysis
variants
methods
url https://hdl.handle.net/10568/28452
work_keys_str_mv AT regejeo amethodforestimatingsamplingvariancesofpredictedprobabilitiesfrommaximumlikelihoodestimatesinlogitresponsefunction
AT sheringtonj amethodforestimatingsamplingvariancesofpredictedprobabilitiesfrommaximumlikelihoodestimatesinlogitresponsefunction
AT regejeo methodforestimatingsamplingvariancesofpredictedprobabilitiesfrommaximumlikelihoodestimatesinlogitresponsefunction
AT sheringtonj methodforestimatingsamplingvariancesofpredictedprobabilitiesfrommaximumlikelihoodestimatesinlogitresponsefunction