JSM Preliminary Online Program
This is the preliminary program for the 2006 Joint Statistical Meetings in Seattle, Washington.

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2006 Program page

Activity Number: 109
Type: Contributed
Date/Time: Monday, August 7, 2006 : 8:30 AM to 10:20 AM
Sponsor: Section on Statistics and the Environment
Abstract - #307414
Title: Over-Estimation of Trend Caused by Negative Binomial Regression Fit to Zero-Inflated Count Data
Author(s): Mihoko Minami*+
Companies: The Institute of Statistical Mathematics
Address: 4 6 7 Minami Azabu, Minatoku Tokyo, 106 8569, Japan
Keywords: zero-inflated negative binomial regression model ; bycatch data ; size parameter ; partial dependence ; temporal trend
Abstract:

We show that applying negative binomial regression model to data with many zero-valued observation could cause severe overestimation for effects of covariates. In some situations, count data may contain many zero-valued observations, but also include large values. For example, catch (bycatch) of non-target species by a set in fisheries could be mostly zeros, but might take a large value when aggregations of animals are caught. Negative binomial regression model is a widely used regression model to overdispersed count data. For data with many zero-valued observations, negative binomial regression model might look fit adequately well. However, it might overestimate effects of covariates and result in false warning of trend. We investigate this phenomenon theoretically and show some examples in a real situation.

  • The address information is for the authors that have a + after their name.
  • Authors who are presenting talks have a * after their name.

Back to the full JSM 2006 program

JSM 2006 For information, contact jsm@amstat.org or phone (888) 231-3473. If you have questions about the Continuing Education program, please contact the Education Department.
Revised April, 2006