Skip to main content

Table 5 Indicators differentiating occasional or daily users of residential wood heating from non-users: multivariate analysis corrected for stratified sampling

From: Use of residential wood heating in a context of climate change: a population survey in Québec (Canada)

Variables OR1 CI95% 1 P value2 c index3 Rank
Model1     0.8176 14
Sociodemographic characteristics      
Income before tax, from all sources (last 12 months):    0.0056   
• < $45,000 reference group    
• ≥ $45,000 1.1 0.9;1.5    
• not disclosed 0.6 0.4 ; 0.9    
First language learned at home:    0.0257   
• French only reference group    
• English only 0.3 0.1 ; 0.7    
• other language in addition to French or English 1.0 0.4 ; 2.8    
• languages other than French and English 0.6 0.3 ; 1.6    
Dwelling      
Type of dwelling:    < 0.0001   
• apartment reference group    
• house 10.8 6.7 ; 17.4    
Perceived efficiency of the dwelling's insulation against heat:    0.0009   
• very good reference group    
• good 0.7 0.5 ; 0.9    
• average 0,6 0,4 ; 0.8    
• poor 0.3 0.1 ; 0.6    
Dwelling built before 19831:    0.0159   
• yes reference group    
• no 3.1 1.4 ; 6.7    
• unknown 1.0 0.8 ; 1.3    
Region of residence      
Region lived in:    < 0.0001   
• Eastern Québec reference group    
• Northern part of southern Québec 1.2 0.7 ; 2.1    
• Central Québec 1.0 0.6 ; 1.7    
• Québec City region 0.7 0.4 ; 1.1    
• North of Montréal 0.7 0.5 ; 1.1    
• South of Montréal 0.7 0.5 ; 1.1    
• Montréal and Laval 0.2 0.1 ; 0.4    
Consultation of smog warning in the media    0.0112   
• always reference group    
• often 0.9 0.6 ; 1.4    
• sometimes 1.2 0.8 ; 1.7    
• rarely 1.4 0.9 ; 2.1    
• never 0.8 0.5 ; 1.1    
Beliefs      
Belief of the contribution of anthropogenic causes to climate change in the last fifty years:    0.0388   
• a lot reference group    
• average 0.8 0.6 ; 1.1    
• not much 1.0 0.7 ; 1.5    
• not at all 0.5 0.3 ; 0.9    
Prohibition of wood heating when there is winter smog:    < 0.0001   
• completely agree reference group    
• somewhat agree 1.8 1.3 ; 2.6    
• somewhat disagree 2.6 1.9 ; 3.6    
• completely disagree 5.2 3.6 ; 7.5    
Model2     0.8029 645
Type of dwelling:    < 0.0001   
• apartment reference group    
• house 10.0 6.2 ; 16.2    
Region lived in:    < 0.0001   
• Eastern Québec reference group    
• Northern part of southern Québec 1.1 0.7 ; 1.9    
• Central Québec 1.0 0.6 ; 1.7    
• Québec City region 0.7 0.4 ; 1.1    
• North of Montréal 0.7 0.4 ; 1.0    
• South of Montréal 0.7 0.4 ; 1.0    
• Montréal and Laval 0.2 0.1 ; 0.3    
Prohibition of wood heating when there is winter smog:    < 0.0001   
• completely agree reference group    
• somewhat agree 1.8 1.3 ; 2.6    
• somewhat disagree 2.5 1.8 ; 3.4    
• completely disagree 4.4 3.1 ; 6.2    
  1. 1OR: odds ratio; IC95%: 95% confidence interval
  2. 2p value associated with the Wald test by means of logistic regression.
  3. 3Area under the ROC curve; between 0.8 and 0.9: good model. No collinearity between the independent variables was observed.
  4. 464 models had a c index over 0.8. This model ranked first.
  5. 564 models had a c index over 0.8. This model ranked last.