ANALYSIS OF THREE NATURAL EXPERIMENTS RELATING TO FIRE DEATH RATE.

Jerry Banks, Douglas Montgomery

Research output: Contribution to journalArticle

Abstract

This study relates the incidence of fire deaths to a number of regressor variables that may have predictive or explanatory value. The data for these analyses arise in the setting of 'natural experiments'. That is, the authors analyze data as it arises in nature. Regression analysis is a statistical procedure for modeling and investigating the relationship between variables in this setting. Three experiments are presented. The first experiment shows that there is a significant difference in fire death rates for warm and cold climates. The second experiment 'blocks out' the effect of cold weather by analyzing data from regions having similar cold weather climates. The third experiment uses data from the 48 contiguous states to investigate the hypothesis that minority population is related to fire death rates. It is shown that states with large minority populations have significantly higher fire death rates than states that have small minority populations.

Original languageEnglish (US)
Pages (from-to)33-53
Number of pages21
JournalJournal of Fire Sciences
Volume1
Issue number1
StatePublished - Jan 1983
Externally publishedYes

Fingerprint

Fires
Experiments
Regression analysis

ASJC Scopus subject areas

  • Engineering (miscellaneous)
  • Safety, Risk, Reliability and Quality
  • Materials Science(all)

Cite this

ANALYSIS OF THREE NATURAL EXPERIMENTS RELATING TO FIRE DEATH RATE. / Banks, Jerry; Montgomery, Douglas.

In: Journal of Fire Sciences, Vol. 1, No. 1, 01.1983, p. 33-53.

Research output: Contribution to journalArticle

@article{64c200618c9a40f5b57a5de20a094322,
title = "ANALYSIS OF THREE NATURAL EXPERIMENTS RELATING TO FIRE DEATH RATE.",
abstract = "This study relates the incidence of fire deaths to a number of regressor variables that may have predictive or explanatory value. The data for these analyses arise in the setting of 'natural experiments'. That is, the authors analyze data as it arises in nature. Regression analysis is a statistical procedure for modeling and investigating the relationship between variables in this setting. Three experiments are presented. The first experiment shows that there is a significant difference in fire death rates for warm and cold climates. The second experiment 'blocks out' the effect of cold weather by analyzing data from regions having similar cold weather climates. The third experiment uses data from the 48 contiguous states to investigate the hypothesis that minority population is related to fire death rates. It is shown that states with large minority populations have significantly higher fire death rates than states that have small minority populations.",
author = "Jerry Banks and Douglas Montgomery",
year = "1983",
month = "1",
language = "English (US)",
volume = "1",
pages = "33--53",
journal = "Journal of Fire Sciences",
issn = "0734-9041",
publisher = "SAGE Publications Ltd",
number = "1",

}

TY - JOUR

T1 - ANALYSIS OF THREE NATURAL EXPERIMENTS RELATING TO FIRE DEATH RATE.

AU - Banks, Jerry

AU - Montgomery, Douglas

PY - 1983/1

Y1 - 1983/1

N2 - This study relates the incidence of fire deaths to a number of regressor variables that may have predictive or explanatory value. The data for these analyses arise in the setting of 'natural experiments'. That is, the authors analyze data as it arises in nature. Regression analysis is a statistical procedure for modeling and investigating the relationship between variables in this setting. Three experiments are presented. The first experiment shows that there is a significant difference in fire death rates for warm and cold climates. The second experiment 'blocks out' the effect of cold weather by analyzing data from regions having similar cold weather climates. The third experiment uses data from the 48 contiguous states to investigate the hypothesis that minority population is related to fire death rates. It is shown that states with large minority populations have significantly higher fire death rates than states that have small minority populations.

AB - This study relates the incidence of fire deaths to a number of regressor variables that may have predictive or explanatory value. The data for these analyses arise in the setting of 'natural experiments'. That is, the authors analyze data as it arises in nature. Regression analysis is a statistical procedure for modeling and investigating the relationship between variables in this setting. Three experiments are presented. The first experiment shows that there is a significant difference in fire death rates for warm and cold climates. The second experiment 'blocks out' the effect of cold weather by analyzing data from regions having similar cold weather climates. The third experiment uses data from the 48 contiguous states to investigate the hypothesis that minority population is related to fire death rates. It is shown that states with large minority populations have significantly higher fire death rates than states that have small minority populations.

UR - http://www.scopus.com/inward/record.url?scp=0020495620&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0020495620&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0020495620

VL - 1

SP - 33

EP - 53

JO - Journal of Fire Sciences

JF - Journal of Fire Sciences

SN - 0734-9041

IS - 1

ER -