• The study population comprised people of at least 18 years of age, living in the 10 Canadian provinces in private homes, able to express themselves in either French or English.
  • Interviews were conducted between May 19 th and June 14 th, 2004
  • 3,014 telephone interviews were completed: 2,011 in English conducted from Research House’s Toronto call centre and 1,003 in French conducted from CROP’s Montreal call centre.
  • The margin of error for a sample of 3,014 respondents is 1,8%, 19 times out of 20
    • 2,1% for Internet users (n=2,173)
    • 3,4% for Internet non-users (n=839)
    • 3,1% for French speakers (n=1,003)
    • 2,2% for English speakers (n=2,014)
  • A probabilistic sampling model was used, which ensures a random and representative selection from all eligible individuals belonging to the study population
    • For the random selection of respondents, the method of the “next birthday” was applied using the following question:”Of those in your household who are at least 18 years of age (and speak English/French), may I speak with the person who will next celebrate his/her birthday?”
    • No substitution was authorized and 5 call-backs were made so as to be able to question the individual selected from the household
  • The sample was built using a mix of directory listed and unlisted numbers. This methodology, offered by ASDE Canada Survey Sampler, gives equal chance to unlisted households to participate in the survey. 60% of the sample was randomly drawn from the electronic telephone directory, and 40% was constituted from a random digit dialing process.
  • As the sample was stratified in order to obtain a large Francophone sample and in the absence of socio-demographic quotas, the final sample featured greater proportions of Francophones and women and lower proportions of individuals 65+ years of age than in the Canadian population. Results were therefore weighted to ensure they are representative of the Canadian population, i.e. by ensuring that the weight of each sub-group of respondents in the sample reflects this same sub-group within the Canadian population (weighting data provided by Statistics Canada, 2001 census). Example: As men were proportionally fewer in the study than they are in the general population, each response given in the study by a male individual carries a little more weight in the final results than one given by a female respondent
  • The data were therefore weighted by:
    • Region (Atlantic Anglophone, New Brunswick Francophone, Quebec Anglophone, Quebec Francophone, Ontario Anglophone, Ontario Francophone and West Anglophone)
    • Age (15-34 years old, 35-64 years old and 65 years old and over)
    • Sex

Before weighting %

After weighting %

GENDER

Male 40 49
Female 60 51

AGE

18-34

25

30

35-64

60

51

65 +

12

17

REGION

Atlantic anglophone

12

7

New Brunswick francophone

3

1

Quebec anglophone

4

4

Quebec francophone

28

20

Ontario anglophone

25

37

Ontario francophone

3

1

Western Canada

25

30

  • The average lengths of the interviews were:

Research House

CROP

Overall

22.2 minutes

22.2 minutes

Internet users

25.3 minutes

26.1 minutes

Internet non-users

13.2 minutes

14 minutes

  • Out of 3,014 respondents, 2,095 agreed to participate in a future study.
  • Community size variable:
    • Urban: 1,000,000 inhabitants and over
    • Semi-urban: 100,000 to 999,999
    • Semi-rural: 5,000 to 9,999
    • Rural: Less than 5,000