Language Industry Survey

Detailed information for 2004

The objective of this survey was to collect new statistical information on the economic production of the Canadian language industry and its sectors for the purpose of creating an industry profile and collecting benchmark data for future surveys to track industry performance.

Data release - March 3, 2006

• Questionnaire(s) and reporting guide(s)
• Description
• Data sources and methodology
• Data accuracy

Description

The objective of this survey was to collect new statistical information on the economic production of the Canadian language industry and its sectors, including translation, interpretation and language training, for the purpose of creating an industry profile and collecting benchmark data for future surveys to track industry performance.

Data is intended to help businesses and associations involved in the language industry gain a better understanding of their industry and subsequently allow them to make informed business decisions.

The results of this survey will be used by businesses and trade associations as well as by government departments and agencies to understand the economic output of the language industry in Canada, to study industry performance and structure, and to assist in policy development.

Collection period: The most recent 12 month fiscal period (business year) for which the respondent had complete financial records or 2004.

Subjects

• Business, consumer and property services
• Languages
• Professional, scientific and technical services

Data sources and methodology

Target population

The target population for the Language Industry Survey was all establishments in the North American Industrial Classification System (NAICS) codes 541930 'Translation and Interpretation Services' and 611630 'Language Schools'. In order to increase the likelihood of surveying medium and large businesses (the population of interest) rather than self-employed a minimum employee cut-off of one employee was imposed.

• Additional documentation

Instrument design

The Language Industry Survey questionnaire was initially developed by Industry Canada. The Small Business and Special Surveys Division reviewed and revised the questionnaire with input from Industry Canada based on feedback from questionnaire testing with Statistics Canada's Questionnaire Design Resource Centre.

Sampling

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

A census was conducted of establishments with a minimum of 1 employee for NAICS 541930 Translation and Interpretation and NAICS 611630 Languages Schools.

Data sources

Data collection for this reference period: May 2005 to July 2005

Responding to this survey is voluntary.

Data are collected directly from survey respondents.

Data was collected using a mail-out, mail-back questionnaire. Prior to the start of the collection period, we attempted to contact all companies in the survey in order to verify and update the information on the frame as well as to identify a 'knowledgeable' contact person to complete the survey. This process is known as 'pre-contact'. The questionnaire was subsequently mailed to the attention of the person identified during pre-contact. Telephone follow-up began on June 13th for approximately 6 weeks.

View the Questionnaire(s) and reporting guide(s) .

Error detection

Questionnaires were initially manually edited and outliers detected as they were received during collection. A series of edit rules were developed and invalid or inconsistent entries were corrected using these rules. Telephone follow-up was conducted for missing entries that could not be manually edited.

Imputation

Data were imputed through the use of the hot-deck imputation method. For the most part donors were selected based upon a match on the response to question 1 on the questionnaire (type of language industry service provided by the business). For the revenue and expenses quantitative variables an additional nearest neighbour criteria was used in selecting a donor so that the donor record selected was the one having the closest revenue value as reported on the Business Register (and also matching on the response to question 1). Likewise for the number of employees quantitative variable a similar nearest neighbour strategy was used but with the number of employees reported on the Business Register.

All quantitative variables were calculated through the use of ratios using the values of the donor and recipient records, and auxiliary variables from the Business Register for revenue and number of employees when needed. For the quantitative variables all totals if they were missing were resolved first followed by the resolution of the parts of the totals. Qualitative variables would just take the value of the donor.

Estimation

Estimates for this survey were calculated using the Generalized Estimation System (GES) software. Prior to entering the data into GES, the data were post-stratified by NAICS and by a derived revenue size variable based upon the revenue reported from the BR.
Estimates were produced at the overall level as well as by sector, region, revenue size category and sector and region together. The quality of the estimates was much better for the percentages than for the totals.

Furthermore, the quality of the estimates were much lower when produced in the domains (groupings) mentioned above compared to at the overall level. This is because the number of establishments contributing to each estimate within each grouping was not large enough to produce stronger estimates.

Quality evaluation

The Language Industry Survey is the first survey of its kind. There was little in the way of existing comparable data. The data was reviewed by industry specialists for validation prior to release.

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.

Aggregated tabular data was released for this survey. Cells for which a domain had less than 6 respondents were suppressed.

Revisions and seasonal adjustment

This methodology does not apply to this survey.

Data accuracy

The data indicators used for the Language Industry Survey were 'Standard Error' and 'Coefficient of Variation'. The standard error is a commonly used statistical measure indicating the sampling error of an estimate. The standard error and the coefficient of variation (standard error expressed as a percentage of the estimate) were used in the data tables to provide an indication of the level of data quality of the estimate.

Alphabetic symbols representing the size and range of the standard error and coefficient of variation were used in the statistical tables to provide an indication of the extent of the sampling error of the estimates.

The response rate for this survey was 36%.

• Additional documentation