About prose literacy l The prose literacy map of Canada l Using the map l Some definitions l About the data sources
About the data sources
Prose literacy data at the community levels
The approach uses the 2006 Canadian census data to create a file for each province that includes a record for every individual in the province that completed the Census 2B long form, roughly 20% of the population aged 16 and over. An estimate of a person’s outcome (in this case their prose literacy score) for all people on the file is estimated using a regression-based multiple imputation technique. Multiple scores (25) are generated for each individual and subsequently grouped into the IALSS proficiency levels. A separate regression is performed to account for local variation in the relationships. The imputations control for age group, gender, educational attainment, immigrant status, aboriginal status, and mother tongue. The imputation process faithfully reproduces the distributions of scores and proportions at each proficiency level at the Census Sub-Division and higher levels within this matrix of variables. The principle source of error in the estimates is associated with the fact that the variables used in the imputation do not account for all of the variation in proficiency observed at the individual level. These errors do not have a material impact on most uses of the data as the aggregation of scores into proficiency levels and of individuals into groups circumvents the problem.
Please contact CCL for more detailed information on the methodology.