Hedonic Wine Price Predictions and NonNormal Errors

Research output: Contribution to journalArticle

8 Citations (Scopus)
5 Downloads (Pure)

Abstract

This paper considers using nonnormal error distributions for the hedonic wine price function, to predict 'average' wine prices and hence identify over and under-priced wines. In addition to the log-normal distribution, the Weibull, gamma, exponential and log-logistic distributions are employed for a comprehensive data set of over 6,000 Australian wines. Estimates of marginal attribute impacts do differ with distribution employed with the Weibull being the most different. On average for all wines, the log-logistic distribution produces the most accurate predictions, however, the performance of distributions differs according to price range. Combining the predictions can produce substantially lower absolute errors on average, because of the varying accuracy performance of the distributions.
Original languageEnglish
Pages (from-to)519-535
Number of pages17
JournalAgribusiness
Volume26
Issue number4
DOIs
Publication statusPublished - Aug 2010

Fingerprint

Pleasure
Wine
wine
wines
prediction
logistics
Normal Distribution
distribution
price
Prediction
performance

Cite this

@article{6c881f2061fc44c5b18f7a99f0bfcd2b,
title = "Hedonic Wine Price Predictions and NonNormal Errors",
abstract = "This paper considers using nonnormal error distributions for the hedonic wine price function, to predict 'average' wine prices and hence identify over and under-priced wines. In addition to the log-normal distribution, the Weibull, gamma, exponential and log-logistic distributions are employed for a comprehensive data set of over 6,000 Australian wines. Estimates of marginal attribute impacts do differ with distribution employed with the Weibull being the most different. On average for all wines, the log-logistic distribution produces the most accurate predictions, however, the performance of distributions differs according to price range. Combining the predictions can produce substantially lower absolute errors on average, because of the varying accuracy performance of the distributions.",
keywords = "Open access version available, Non-normal errors, Wine pricing",
author = "Edward Oczkowski",
note = "Imported on 12 Apr 2017 - DigiTool details were: month (773h) = Aug 2010; Journal title (773t) = Agribusiness (New York): an international journal. ISSNs: 1520-6297;",
year = "2010",
month = "8",
doi = "10.1002/agr.20237",
language = "English",
volume = "26",
pages = "519--535",
journal = "Agribusiness",
issn = "0742-4477",
publisher = "Wiley-VCH Verlag",
number = "4",

}

Hedonic Wine Price Predictions and NonNormal Errors. / Oczkowski, Edward.

In: Agribusiness, Vol. 26, No. 4, 08.2010, p. 519-535.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Hedonic Wine Price Predictions and NonNormal Errors

AU - Oczkowski, Edward

N1 - Imported on 12 Apr 2017 - DigiTool details were: month (773h) = Aug 2010; Journal title (773t) = Agribusiness (New York): an international journal. ISSNs: 1520-6297;

PY - 2010/8

Y1 - 2010/8

N2 - This paper considers using nonnormal error distributions for the hedonic wine price function, to predict 'average' wine prices and hence identify over and under-priced wines. In addition to the log-normal distribution, the Weibull, gamma, exponential and log-logistic distributions are employed for a comprehensive data set of over 6,000 Australian wines. Estimates of marginal attribute impacts do differ with distribution employed with the Weibull being the most different. On average for all wines, the log-logistic distribution produces the most accurate predictions, however, the performance of distributions differs according to price range. Combining the predictions can produce substantially lower absolute errors on average, because of the varying accuracy performance of the distributions.

AB - This paper considers using nonnormal error distributions for the hedonic wine price function, to predict 'average' wine prices and hence identify over and under-priced wines. In addition to the log-normal distribution, the Weibull, gamma, exponential and log-logistic distributions are employed for a comprehensive data set of over 6,000 Australian wines. Estimates of marginal attribute impacts do differ with distribution employed with the Weibull being the most different. On average for all wines, the log-logistic distribution produces the most accurate predictions, however, the performance of distributions differs according to price range. Combining the predictions can produce substantially lower absolute errors on average, because of the varying accuracy performance of the distributions.

KW - Open access version available

KW - Non-normal errors

KW - Wine pricing

U2 - 10.1002/agr.20237

DO - 10.1002/agr.20237

M3 - Article

VL - 26

SP - 519

EP - 535

JO - Agribusiness

JF - Agribusiness

SN - 0742-4477

IS - 4

ER -