Testing for Bias in Forecasts for Independent Multinomial Outcomes

This paper deals with a test on forecast bias in predicting independent multinomial outcomes where the predictions are probabilities. The new Likelihood Ratio (and Wald) test extends the familiar Mincer Zarnowitz regression to a multinomial logit model instead of a linear regression. The test is eva...

Full description

Saved in:
Bibliographic Details
Main Authors: Philip Hans Franses, Richard Paap
Format: Article
Language:English
Published: MDPI AG 2025-01-01
Series:Forecasting
Subjects:
Online Access:https://www.mdpi.com/2571-9394/7/1/4
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850204467856670720
author Philip Hans Franses
Richard Paap
author_facet Philip Hans Franses
Richard Paap
author_sort Philip Hans Franses
collection DOAJ
description This paper deals with a test on forecast bias in predicting independent multinomial outcomes where the predictions are probabilities. The new Likelihood Ratio (and Wald) test extends the familiar Mincer Zarnowitz regression to a multinomial logit model instead of a linear regression. The test is evaluated using various simulation experiments, which indicate that the size and power properties are good, even for small sample sizes, in the sense that the size is close to the used 5% level, and the power quickly reaches 1. We implement the test in an empirical setting on brand choice by individual households.
format Article
id doaj-art-d7c41d00ca8c4b4dba00cbbb960f0436
institution OA Journals
issn 2571-9394
language English
publishDate 2025-01-01
publisher MDPI AG
record_format Article
series Forecasting
spelling doaj-art-d7c41d00ca8c4b4dba00cbbb960f04362025-08-20T02:11:17ZengMDPI AGForecasting2571-93942025-01-0171410.3390/forecast7010004Testing for Bias in Forecasts for Independent Multinomial OutcomesPhilip Hans Franses0Richard Paap1Econometric Institute, Erasmus School of Economics, Burgemeester Oudlaan 50, 3062PA Rotterdam, The NetherlandsEconometric Institute, Erasmus School of Economics, Burgemeester Oudlaan 50, 3062PA Rotterdam, The NetherlandsThis paper deals with a test on forecast bias in predicting independent multinomial outcomes where the predictions are probabilities. The new Likelihood Ratio (and Wald) test extends the familiar Mincer Zarnowitz regression to a multinomial logit model instead of a linear regression. The test is evaluated using various simulation experiments, which indicate that the size and power properties are good, even for small sample sizes, in the sense that the size is close to the used 5% level, and the power quickly reaches 1. We implement the test in an empirical setting on brand choice by individual households.https://www.mdpi.com/2571-9394/7/1/4multinomial outcomesprobability forecastsforecast biasmultinomial logit model
spellingShingle Philip Hans Franses
Richard Paap
Testing for Bias in Forecasts for Independent Multinomial Outcomes
Forecasting
multinomial outcomes
probability forecasts
forecast bias
multinomial logit model
title Testing for Bias in Forecasts for Independent Multinomial Outcomes
title_full Testing for Bias in Forecasts for Independent Multinomial Outcomes
title_fullStr Testing for Bias in Forecasts for Independent Multinomial Outcomes
title_full_unstemmed Testing for Bias in Forecasts for Independent Multinomial Outcomes
title_short Testing for Bias in Forecasts for Independent Multinomial Outcomes
title_sort testing for bias in forecasts for independent multinomial outcomes
topic multinomial outcomes
probability forecasts
forecast bias
multinomial logit model
url https://www.mdpi.com/2571-9394/7/1/4
work_keys_str_mv AT philiphansfranses testingforbiasinforecastsforindependentmultinomialoutcomes
AT richardpaap testingforbiasinforecastsforindependentmultinomialoutcomes