How the Southam Literacy Survey was done

The literacy survey by Southam News is based on face-to-face interviews with 2,398 adults in 148 Canadian communities and rural areas. Respondents were selected at 418 sampling points.

Three samples were drawn for the survey: a random cross-section of the entire population, extra samples for some cities and regions, and a special oversample of 21- to 25-year-olds.

Excluded from the survey are those under 18, residents north of the 60th parallel, transients, members of the armed forces, natives living on reserves and anyone in an institution, such as a prison, hospital or nursing home.

Interviews were conducted in English or French in the respondents' homes between May 6 and June 10, 1987. The results have been weighted to ensure that each segment is represented in proportion to its share of the population.

In particular, the results were weighted to reflect the most recent Statistics Canada data on how far Canadians went in school.

Following general industry practice, all respondents were given two $1 lottery tickets to be interviewed.

Literacy was defined as reading and writing skills in either of Canada's two official languages. Extra steps were taken to learn as much as possible about potential respondents who spoke neither English nor French.

An average interview took 80 minutes and involved two parts: a background and activity questionnaire followed by a test of reading, writing and numbers skills using five dozen items based on everyday life.

These items were adapted from a 1985 U.S. survey for the National Assessment of Educational Progress by substituting Canadian spellings, materials and topical references.

Respondents could take as long as they liked to tackle the literacy items. The tasks were divided into two smaller and overlapping sets, each of 40 items answered by half the respondents.

People were handed copies of the items to be read, which they could hold on to while answering. They didn't have to rely on memory.

Two specially printed newspapers were used to measure the literacy in reading newspaper articles -The Independant and L'independent

Literacy in reading newspaper articles was measured using two specially printed newspapers, in English and French. The papers were printed by the Ottawa Citizen, with cooperation from Le Droit, Ottawa's French language daily.

An outside literacy consultant, Jean-Paul Hautecoeur, advised on the French-language adaptation. Ten other experts in literacy, adult education and journalism provided extensive advice on the design of the background questionnaire.

The standards for literacy and illiteracy were determined by a 25-member panel.

Although the total sample is 2,398, the national picture is based on a weighted sample of 1,503. In 19 cases out of 20, results based on a theoretical random sample of this size will differ by no more than 2.5 percentage points in either direction from the results obtained by interviewing all adult Canadians.