Television Viewing Databank

Detailed information for Fall 2004

The databank permits users (including government departments and agencies, specifically the Canadian Radio-Television and Telecommunications Commission, associations, academics, and market researchers) to analyse the use of television by Canadians. The information is used for programming, planning, monitoring of policies and for market research.

Data release - March 31, 2006

• Description
• Data sources and methodology
• Data accuracy

Description

The databank permits users (including government departments and agencies, specifically the Canadian Radio-Television and Telecommunications Commission, associations, academics, and market researchers) to analyse the use of television by Canadians. The information is used for programming, planning, monitoring of policies and for market research.

The survey data collected by BBM come from a stratified sample of Canadians aged two years and over. Their viewing during a specified seven-day period is collected by means of a diary-type questionnaire and is recorded in quarter-hour intervals. Households included in the sample are sent one diary for each working television in the home. The diaries are designed to record the tuning of each individual household member two years of age or older. Demographic characteristics (including age group, sex, education, occupation and linguistic characteristics) are also collected on this document. The results of four seven-day periods during the fall are combined to produce the Fall Sweep Survey.

Reference period: Fall

Collection period: Fall

Subjects

• Culture and leisure
• Television viewing and radio listening

Data sources and methodology

Target population

Data collected by the BBM survey represent the television viewing of all Canadians aged 2 years and older with the following exceptions:
- residents of the Yukon
- residents of the Northwest Territories
- residents of Nunavut
- residents of communal institutions (e.g. prisons, hospitals, college dormitories)
- residents of Indian Reserves
- residents of households where no member speaks either English or French
- residents of households which have no telephone service

Instrument design

This methodology does not apply.

Sampling

This is a sample survey with a cross-sectional design.

The ten provinces of Canada are geographically divided into approximately 400 mutually exclusive units called cells. For the most part, these cells are counties, census divisions, census agglomerations or other areas defined by Statistics Canada.

For each cell BBM determines the number of respondents necessary to provide statistically significant data. Both the estimated size of the cell population and the level of detail (e.g. demographic groups) required are considered in calculating the required sample size. Records of response rates and statistics on the composition of families in the cell are used to calculate the number of households to which diary packages must be mailed to achieve the required sample.

Telephone directories and lists are used to draw a systematic sample independently within each cell. Where there is more than one Post Office serving a given cell, samples are selected so that each area served by a separate Post Office is considered independently.

Approximately five to twelve weeks prior to the survey start date, specially trained and supervised telephone interviewers contact the selected households. The interviewer verifies the mailing address and records the home language, and the number of television sets in the house. Households that cannot be contacted the first time are re-called up to as many as five additional times.

The list of successfully "enumerated" households becomes the sample for a specific survey.

Data sources

Responding to this survey is voluntary.

Data are extracted from administrative files.

Basic data are extracted from administrative files provided by an external organization, Bureau of Measurement (BBM), that collected them for statistical purposes.

The final databank, which is retained by Statistics Canada, indicates the network affiliation and geographic location of each television station. It contains also the characteristics of each program: content (news, public affairs, drama, religious programming, etc.), language, country of origin, and percentage of Canadian content.

Although the Fall Sweep Survey reporting period is four weeks, each household is asked to provide just one week's information. To achieve the four-week coverage, the enumerated households are randomly apportioned over all four weeks of the survey.

Each enumerated household is mailed a package with a separate diary booklet for each television set. The diary has separate pages for each day of the week and is designed so that the viewing of each household member (and visitor) can be recorded separately. Respondents are requested to register their television viewing by recording the station call letters, the program name, and the channel number for each quarter-hour period of viewing during the day (6:00 a.m. until 2:00 a.m. the following morning). A separate section of the diary requests the respondent to provide background information (e.g. age, sex, language, education, occupation, ownership of VCR, etc.) for each of the household members.

Error detection

Information from returned diaries is first manually checked for completeness, then transferred to computer tape. These data are then subjected to a careful and extensive series of manual and computer checks. About 50% of diaries require further manual editing. If necessary, respondents are called to confirm information. Estimates of television viewing are then made for each of the defined areas of the country.

In cases of simultaneous substitution on cable television systems, recorded tuning is attributed to the actual station signal transmitted over the cable regardless of which station's call letters were indicated by the respondent in the diary. Simultaneous substitution may occur when two stations are broadcasting the same program at the same time. Under certain conditions, the cable company may substitute one signal for the other. This most commonly occurs when one station is Canadian and the other is American. In such a case, at the request of the Canadian station, the signal of the Canadian station would be transmitted on the channel normally carrying the American station as well as on the channel carrying the Canadian station.

To compensate for possible differences in response rates, the data are weighted by age/sex groups. In the Montréal and Ottawa-Gatineau (formerly Ottawa-Hull) areas the data are further weighted by "home language" group (i.e. English, French, other).

Imputation

This methodology does not apply.

Estimation

This methodology type does not apply to this statistical program.

Quality evaluation

Basic viewing data are acquired from BBM Bureau of Measurement. Survey results such as the average hours of television viewing by demographics of survey respondents are compared with data from BBM Bureau of Measurement.

Survey results related to the characteristics of each program: content (news, public affairs, drama, religious programming, etc.), country of origin and percentage of Canadian content are subject to a detailed review in an annual meeting involving staff from Statistics Canada and CRTC.

Disclosure control

Statistics Canada is prohibited by law from releasing any data which would divulge information obtained under the Statistics Act that relates to any identifiable person, business or organization without the prior knowledge or the consent in writing of that person, business or organization. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.

BBM has established a number of reporting standards or minimum sample requirements. Neither BBM nor Statistics Canada will release data where the number of valid respondents for a specific area or demographic group falls below the reporting standard.

Revisions and seasonal adjustment

This methodology does not apply to this survey.

Data accuracy

While the return rate, at 32.8%, is modest by Statistics Canada standards, it is in line with Canadian and international broadcasting industry practice for audience measurement.

Sampling error tables are not available.