The study employs an annual target sample of 36,000, and meets specific sample requirements in 47 markets. In Québec, the sample consists of approximately 6,400 respondents, including 4,575 in Montreal and 1,000 in Québec City. For a balance of Canada, outside the 47 markets, we sample approximately 3,475 respondents. The survey is conducted among individuals age 14 or older and uses a three-stage process.
In the first stage, potential respondents complete a recruitment questionnaire via telephone (CATI) using probability-based sampling. In the second stage, recruited respondents are asked to complete an online newspaper and magazine readership survey. In the final stage, respondents are asked to complete a detailed online product survey that includes brand, retail, and lifestyle questions.
The sample frame is constructed at the local market level and rolled up to a national level using Statistics Canada census DAs. For each CMA or CA, we cross-reference the DA codes to the Postal Code Conversion File (PCCF) to identify all postal codes that cover the relevant geography. This list of postal codes will be used as input to define the universe for sampling of telephone numbers.
In the major markets, a “Plus Digit” random telephone sampling method is used to generate the sampling frame. The “Plus Digit” method modifies a known live residential telephone number (seed number) by adding a random number to the last digit and is used to improve the probability of including households in the sample that might otherwise be excluded from published directories. The sample for the remaining markets is drawn from the universe of published household telephone numbers without modification to the last digit. Three specific groups receive special recruitment efforts to ensure representative samples: Upper Income/Professional/Managers, Cell-Only Households and 18 – 24 Year Olds.
Recruitment is conducted continually throughout the year, seven days a week. Each interview includes a day code to indicate the day of week the first call must be made, and up to four call-backs may be conducted. Respondent selection is based on the most recent birthday in the home. The recruitment interview gathers basic information including email address and postal codes, and secures permission to email a link to complete the Readership survey online. Respondents are asked to complete the survey within the next two days, and two reminders are sent.
Undeliverable emails trigger a follow-up telephone call, and email reminders are used systematically to maximize the response rate and ensure full completion of the survey. A small percentage of respondents need assistance to complete the readership questionnaire, either by providing internet access to the survey or by assisting the respondent to complete the interview.
All respondents complete the magazine and newspaper readership questionnaire with randomization of magazines and newspaper titles. The survey vehicle incorporates design elements intended to encourage respondent engagement and optimize the quality of responses, and will evolve over time. The exact date of completion is passively recorded and monitored. Incentives range from $15-$30.
The readership metrics are Recent Reading for newspapers and magazines. A title specific ‘brand-centric’ approach is used for measuring publications’ print and digital formats.
The online questionnaire starts with a set of definitions, which remain available via a link throughout the readership section.
Newspapers and Magazines Readership
Respondents are screened for reading a newspaper in the past 3 months, and a magazine in the past year. After screening, they are asked when they last read that publication for each platform (digital and print). Average issue readership is then calculated for each title based on all those responding that they read the publication within the applicable publication interval.
All respondents complete the same magazine and newspaper readership questionnaire with appropriate customizations for randomization of magazine titles and local market newspaper data capture. National newspapers are asked of all respondents. Rotations are used to minimize order bias, including rotation of magazine and newspaper questions. Half the respondents are asked about magazines first, and half about newspapers first.
Reach and Frequency
Publications can generate reach beyond just their regular readers, as respondents answering “longer ago” contribute to the reach build for the publication. Frequency is assessed from a follow up question asking how often they read the publication.
All respondents who complete the readership questionnaire are asked to participate in the follow-up product survey. This product survey is administered online and allows for some local market customization. Respondents who complete the follow-up product questionnaire receive additional incentives, draws for cash prizes and friendly reminders.
All respondents are served a core questionnaire containing questions for ascription hooks, categories requiring a robust base size, and low incidence categories. Each respondent is then invited to complete 4 of 10 remaining partitions.
Imputation is applied to those who did not complete a product questionnaire. There will be no data gaps in the final database.
A single database includes all questions from the readership survey for all media (online and offline) as well as the product data. Quarterly updates of the entire database include the most recent past 12 months of data, across all titles and platforms (accomplished by adding the most recent quarter of data and removing the earliest quarter in the 12-month time series).
The currencies for magazines, national newspapers, and local newspapers in the 22 major markets will be based on a rolling 12-month database. Local newspapers in the remaining 27 markets will have currency established using a rolling 24-month database. The total sample size for the database is more than 42,000.