Canadian Tobacco, Alcohol and Drugs Survey (CTADS)
Detailed information for 2017
Every 2 years
The main objective of the survey is to provide continual and reliable data on tobacco, alcohol and drug use and related issues, with the primary focus on 15 to 24 year olds.
Data release - October 30, 2018
The major objectives of the survey are to: measure the frequency of cigarette smoking, as well as the amount smoked, gain insight into behaviors related to smoking, measure the prevalence and frequency of alcohol use, and measure the prevalence of drug use and the extent of harm related to usage.
Reference period: varies
Collection period: February to December in the reference year
- Environmental factors
- Lifestyle and social conditions
Data sources and methodology
The target population for the CTADS is all persons 15 years of age and over living in Canada with the following two exceptions:
1) residents of the Yukon, Northwest Territories and Nunavut; and
2) full-time residents of institutions.
Because the survey is conducted using a sample of telephone numbers, we are unable to reach households (and thus persons living in households) if all phone numbers for that household are missing from the frame, or if the household does not have any phone numbers associated with it. This means that persons from households with neither landline telephones nor cellular phones are not accounted for and are excluded from the survey population. However, the survey estimates will be weighted to include persons residing in households with neither cellular phones nor landlines.
The questionnaire for CTADS 2017 borrows heavily from the previous CTADS (2013 and 2015) questionnaires and previous iterations of the survey (CTUMS).
The CTADS 2017 questionnaire was developed in coordination with the survey sponsor, Health Canada.
Qualitative testing was used to test new and revised content. A total of fifteen cognitive interviews were conducted, within which the test questionnaire was administered face-to-face with participants, to determine if the content was relevant and well understood.
Collection instrument - The following modules were added to the 2017 questionnaire: Household Vaping (HV), Cessation Methods (CM), Health Professionals (HP), Perception of Harm (POH), Other drugs and substances (ODS), Aboriginal and Education Highest Degree.
Some questions were added or modified in the following existing modules: Electronic Cigarette (ELC), Tobacco Products (TP), Other Cessation Methods (OM) and Cannabis (CAN) modules. Questions in the Purchasing Patterns (PP) module were reordered.
This is a sample survey with a cross-sectional design.
The sample design is a special two-phase stratified random sample of telephone numbers. The two-phase design is used to increase the representation of individuals belonging to the 15 to 19 and 20 to 24 age groups.
In the first phase, households are selected randomly and in the second phase, one or two individuals (or none) are selected, based upon household composition.
In order to ensure that people from all parts of Canada are represented in the sample, each of the 10 provinces were divided into strata or geographic areas. Generally, within each province, a census metropolitan area (CMA) stratum and a non-CMA stratum were defined. In Prince Edward Island, there was only one stratum for the province. In Ontario, there was a third stratum for Toronto. In Quebec, there was a third stratum for Montreal, and in British Columbia there was a third stratum for Vancouver. CMAs are areas defined by the census and correspond roughly to cities with populations of 100,000 or more.
The geographical strata were further divided into 4 household composition strata:
1- Indication of a 15-19 year old.
2- Indication of a 20-24 year old.
3- No indication of either.
4- No information about household composition.
These strata were assigned in a cascading order (first applicable stratum), based on information derived from the Socio-Economic File (SEF) for the 3 first waves, and Census 2016 for the last 2.
The sample for the CTADS was generated using the Household Survey Frame Service. This frame includes up to three telephone numbers for each household, including cellular phone numbers. A random sample of households across the province-strata was produced.
A screening activity aimed at removing not in service and unknown business numbers was performed prior to sending the sample to the computer-assisted telephone interviewing (CATI) unit.
Each telephone number in the CATI sample was dialed to determine whether or not it reached a household. If the telephone number was found to reach a household, the person answering the telephone was asked to provide information on the individual household members. The ages of the household members were used to determine who, in the household, would be selected for the person-level interview. Proxy interviews were not accepted.
To help ensure that sufficient respondents in the younger age groups were reached, a random selection process was set up so that at least one person aged 15 to 19 or 20 to 24 would be selected within a household, if they existed. The reason for this is that about 78% of all households in Canada are comprised only of people aged 25 and over. If all ages were selected with equal probability and retained, the 25 and over age group would be over-represented (refer to the survey objectives). Thus, some of the selected people in the 25 and over age group were screened out and did not receive person-level interview. Two people were selected if more than one of the age groups 15 to 19, 20 to 24, and 25 and over were represented in the household. When two people in the same household were selected, they were always from different age groups. This procedure was meant to minimize any negative impact on the precision of the estimates by age group due to correlation within households.
The detailed logic for the selection of individuals was as follows:
1) If everyone in the household is 15 to 19, then one person is selected at random.
2) If everyone in the household is 20 to 24, then one person is selected at random.
3) If everyone in the household is 25 and over, then one person is selected at random; however, this selected person is retained for only a proportion of the cases.
4) If some household members are 15 to 19 and the rest are 20 to 24, then two people are selected at random, one from each age group.
5) If some household members are 15 to 19 and the rest are 25 and over, then two people are selected at random, one from each age group; however, the person selected from the 25 and over age group is retained for only a proportion of the cases.
6) If some household members are 20 to 24 and the rest are 25 and over, then two people are selected at random, one from each age group; however, the person selected from the 25 and over age group is retained for only a proportion of the cases.
7) If all three age groups are represented in the household, then a check is made to see if the 25 and over age group will be retained. If it is, then two of the age groups are selected at random. If not, the 15 to 19 and the 20 to 24 age groups are selected.
Data collection for this reference period: 2017-02-01 to 2017-12-31
Responding to this survey is voluntary.
Data are collected directly from survey respondents.
Data were collected using computer-assisted telephone interviewing (CATI). The CATI system has a number of generic modules which can be quickly adapted to most types of surveys. A front-end module contains a set of standard response codes for dealing with all possible call outcomes, as well as the associated scripts to be read by the interviewers.
View the Questionnaire(s) and reporting guide(s) .
The purpose of processing survey data is to adapt the collected data into a form that is appropriate for analysis and tabulation.
For CTADS, collection was performed using a Computer Assisted Telephone Interview (CATI), which allows for certain edits to be built into the application. For example, Validity Edits, which ensure that the response falls within the allowed range. It also ensured that only character values were entered into character fields or numeric values were entered into numeric fields.
After collection the individual raw data files were appended together and put through a series of standard processing steps designed to clean the data and help ensure its consistency thereby increasing its usefulness. The edits were done on the data both at the micro and macro level.
The flow edits replicated the flow patterns used in the application and set the non-applicable questions to a value of 'Valid Skip'. Non-responses were set to a value of 'Not Stated'. These are questions that were applicable to the respondent but were not answered. In a CATI application these value usually follow a response of 'Refusal' or 'Don't Know'.
In addition, various types of editing were done to detect missing or inconsistent information. For example, edits were performed to check the logical relationship between responses.
New variables were derived using collected variables. A derived variable may be based on one variable by re-grouping or collapsing the categories or based on several variables, by combining them together to define a new concept.
No imputation methods were employed for the CTADS 2017.
The main output of CTADS is two microdata files: one for the household level information, and one for the person level information.
For the microdata files, statistical weights were placed on each record to represent the number of sampled persons that the record represents. One weight was calculated for each household and a separate weight was calculated and provided on a different file, for each person.
The weighting for the CTADS consists of several steps: calculation of a basic weight, followed by several adjustments such as non-response, non resolved telephone numbers and an adjustment for households with multiple telephone lines. Additional adjustments were made to account for the number of persons (0, 1 or 2) selected in the household and the over-sampling of the 15 to 24 age group. The last step of weighting consists of an adjustment to the person weights in order to make population estimates consistent with external population counts for persons 15 years and older (calibration). The following external control totals were used:
1) Monthly population totals for each province, and
2) Population totals by province, sex and the following age groups: 15 to 19, 20 to 24, 25 to 29, 30 to 34, 35 to 39, 40 to 44, 45 to 49, 50 to 54, 55 to 59, 60 to 64, 65 to 69, and 70 and over. These totals were averaged over the survey period.
The method called generalized regression (GREG) estimation was used to modify the weights to ensure that the survey estimates agreed with the external totals simultaneously along the two dimensions.
In order to supply coefficients of variation (CV) which would be applicable to a wide variety of categorical estimates produced from these microdata files and which could be readily accessed by the user, a set of Approximate Sampling Variability Tables has been produced. These CV tables allow the user to obtain an approximate coefficient of variation based on the size of the estimate calculated from the survey data.
The CV in these tables are derived using the variance formula for simple random sampling and incorporating a factor which reflects the multi stage, clustered nature of the sample design. This factor, known as the design effect, was determined by first calculating design effects for a wide range of characteristics and then choosing from among these a conservative value to be used in the CV tables which would then apply to the entire set of characteristics.
To estimate variances directly, bootstrap weights were also created and made available.
While rigorous quality assurance mechanisms are applied across all steps of the statistical process, validation and scrutiny of the data by statisticians are the ultimate quality checks prior to dissemination. Many validation measures were implemented. They include:
a. Analysis of changes over time
b. Verification of estimates through cross-tabulations
c. Confrontation with other similar sources of data
d. Coherence analysis based on Quality Indicators
Statistics Canada is prohibited by law from releasing any information it collects that could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data that are released or published to prevent publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.
A Public Use Microdata Files (PUMF) is available for the CTADS. It should be noted that the PUMF may differ from the survey "master" files held by Statistics Canada. These differences usually are the result of actions taken to protect the anonymity of individual survey respondents. The most common actions are the suppression of file variables, grouping values into wider categories, and coding specific values into the "not stated" category. Users requiring access to information excluded from the microdata files may purchase custom tabulations. Estimates generated will be released to the user, subject to meeting the guidelines for analysis and release.
Household File and Person File
The survey's master data files include explicit geographic identifiers for province and stratum (census metropolitan area (CMA), non-CMA, Toronto, Vancouver or Montreal).The survey's public use microdata files only contain an identifier for province.
Household Age Composition:
Household age composition is available as the number of household members (capped at two) in the following age ranges: 0 to 14, 15 to 24, 25 to 44, and 45 and over.
Other Modifications to the Household File and Person File:
In order to avoid potential identification of respondents resulting from an unusual combination of characteristics, 178 records on the household and person files had a demographic variable recoded.
Additionally, when the sum of household members derived from the information about their age ranges exceeded five - the maximum value of the household size variable (HHSIZE), the age range variables (15 to 24, 25 to 44 and 45 and over) were modified. On those records, all the age ranges present in the household were maintained, but some of them had the value "two or more" replaced with "one".
Person File Only
Starting with Cycle 1 of 2002, the master data file contains the first three digits of the respondent's postal code. Since Cycle 2 of 2003, the master and the public use microdata files contain an urban/rural variable (DVURBAN). This variable is based on the urban/rural status of the enumeration area (defined by Statistics Canada) in which the majority of the postal codes fall. Urban areas have minimum population concentrations of 1,000 people and a population density of at least 400 people per square kilometre based on the 2011 Census population counts. All the territory outside the urban areas is considered rural.
The detailed marital status variable (six categories) is available on the master file only, while on the public use microdata file this variable has been grouped into three categories.
Cases were identified where the derived variable for the respondent's age (DVAGE) in conjunction with the number of years they have been a smoker (DVYRSSMK) and the age they had their first cigarette (PS_Q30) was greater than 85 (the maximum derived age). The number of years smoked was decreased so that the number of years smoked plus the age they had their first cigarette cannot be greater than 85.
Revisions and seasonal adjustment
This methodology type does not apply to this survey.
"Current smokers" are defined as those who currently smoke every day or occasionally. Since this question is asked to all CTADS respondents, it provides an accurate estimate when inferred upon the target population. As detailed further in this document, measures were taken to eliminate or mitigate the impact of any sources of error.
The estimated percentage of current smokers, and its 95% confidence interval, are provided in the additional document link below for Canada and for each of its provinces, for both the most current edition of CTADS (2017) and for the previous edition (2015).
Non-sampling errors can be defined as errors arising during the course of virtually all survey activities, apart from sampling. They are present in both sample surveys and censuses (unlike sampling error, which is only present in sample surveys). Non-sampling errors arise primarily from the following sources: non-response, coverage, measurement and processing.
The non-response rate was higher in 2017 than in 2015, meaning that non-response bias could also be higher. However, adjusting non-response during processing should help minimize the impact of this bias.
The new database for CTADS was created using several linked sources, such as the Census, administrative data and billing files.
Undercoverage may be possible, but there is much less overcoverage if we compare it to the random digit dialing strategies used in the past.
Before collection, Statistics Canada sent telephone numbers not associated with a household to a pre-dialing service for validation.
CTADS data were processed in a new system, the Social Survey Processing Environment (SSPE). This generalized, structured system aims to establish best practices.
Qualitative tests and cognitive interviews were done to determine whether the questionnaire content, particularly the new and revised sections, were relevant and understood.
The household response rate for the CTADS 2017 was 50.7%, the person response rate was 70.4% and the overall response rate was 35,7%