Jump to ContentJump to Main Navigation
International Smoking StatisticsA collection of historical data from 30 economically developed countries$

Barbara Forey, Jan Hamling, Peter Lee, and Nicholas Wald

Print publication date: 2002

Print ISBN-13: 9780198508564

Published to British Academy Scholarship Online: September 2009

DOI: 10.1093/acprof:oso/9780198508564.001.0001

Show Summary Details

(p.789) Appendix III

(p.789) Appendix III

Source:
International Smoking Statistics
Publisher:
Oxford University Press

Consumption category estimation

The method used in this report to estimate the means of consumption categories of numbers of cigarettes smoked per day was introduced by Todd in 19861 This was a modification of an earlier method given by Todd in 19782.

The 1978 method used a standard distribution based on detailed frequency distributions of male and female cigarette smokers that had applied in the UK. This distribution separated those who smoked a number of cigarettes ending with five or zero from those who smoked a number in the intervening categories. The standard distribution was modified when detailed distributions from the New Zealand Censuses of 1976 and 1981 became available. By combining the two censuses, data on over 700 000 male and 600 000 female smokers were available in 11 consumption categories (1-4, 5-9, …, 45-49, and 50+).

This enabled better distributions to be constructed for those smoking larger numbers of cigarettes.

The 1986 standard distribution is given below:

1986 standard distribution

Consumption category

Male smokers

Female smokers

Assumed mean of consumption category

%

%

1-4

7

10

2.5

5

4

5

5

6-9

6

10

7.5

10

13

17

10

11-14

7

8

12.5

15

8

11

15

16-19

4

4

17.5

20

25

20

20

21-24

6

5

22.5

25

4

3

25

26-29

2

1

27.5

30

5

2

30

31-34

1

1

32.5

35

1

0

35

36-39

0

0

37.5

40

4

2

40

41-44

0

0

45

0

0

46-49

0

0

50

2

1

50

51-59

0

0

60

1

0

60

61+

0

0

100

100

Use of the standard distribution assumes that, for the country, age, and sex group in (p.790) question, the shape of the distribution within each consumption category (e.g. 16-24) was the same as in the standard distribution. This procedure seems reasonable where consumption categories are small, e.g. 1-5, 6-10, 11-20. 21-30, and 31+. Its validity where few consumption categories were used, e.g. 10-14 and 15+, is less certain. On rare occasions when consumption categories used non-standard dividing points (e.g. 8-12) we split the standard category by assuming a uniform distribution over the standard category.

References

Bibliography references:

1. Todd GF. Cigarette Data Collection: A record of six cigarette smoking parameters in 20 countries. Internal report. 1986.

2. Todd GF. Cigarette consumption per adult of each sex in various countries. J Epidemiol Community Health 1978;32:289-93.