Survey of Advanced Technology (SAT)

Detailed information for 2012 to 2014

The objective of the Survey of Advanced Technology is to collect important information about the extent to which Canadian enterprises use advanced technologies.

Data release - December 11, 2015

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

Description

Statistics Canada has undertaken this survey to provide statistical information on the use of advanced technologies by Canadian firms in the following industry sectors: forestry and logging, oil and gas extraction, mining and quarrying, utilities, manufacturing, wholesale trade, retail trade, transportation and warehousing, and professional, scientific and technical services.

The survey's questions address the following themes: type of technology used including, advanced material handling, supply chain and logistics, advanced business intelligence, advanced design and information control, advanced green technologies, geomatics and geospatial, nanotechnologies and biotechnologies; objectives and obstacles to adoption, capital expenditures and sources of funding; skill requirements and training; expenses tied to advanced technologies; development and implementation of advanced technologies; reasons for not investing in advanced technologies; business practices; product/process/marketing/organizational innovation.

To increase the analytical potential of this survey, Statistics Canada plans to combine the data obtained from this survey with data from other Statistics Canada surveys or administrative data.

The information compiled from this survey will be used by the Canadian and provincial governments to better understand innovation activities linked to the modification and creation of technology as well as to develop policies to help businesses improve their productivity and competitiveness via the technology.

The information compiled from this survey may also be used for market analysis, by industry associations to study the characteristics of advanced technology use within their industry, and by academic researchers to perform research on the characteristics of advanced technology use and its business impacts.

Subjects

• Innovation
• Science and technology

Data sources and methodology

Target population

The target population consists of all enterprises in the following sectors: Forestry and Logging (NAICS 113), Mining, Quarrying, and Oil and Gas Extraction (NAICS 21), Utilities (NAICS 22), Manufacturing (NAICS 31-33), Wholesale Trade (NAICS 41), Retail Trade (NAICS 44-45), Transportation and Warehousing (NAICS 48-49), and Professional, Scientific and Technical Services (NAICS 54) (North American Industry Classification System, Statistics Canada, 2012), with at least 10 employees and at least $250,000 in revenues.

• Additional documentation

Instrument design

The Survey of Advanced Technology was designed by Statistics Canada in consultation with various federal government departments, including Industry Canada, Natural Resources Canada, Agriculture and Agri-Food Canada, Transport Canada, and feedback from industry experts offering a range of skills and interests. Cognitive testing of the electronic questionnaire was carried out with prospective respondents, whose comments contributed to the final version.

Sampling

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

A stratified simple random sample of 11,887 enterprises was selected from the survey population of 84,322 enterprises on the September 2014 version of Statistics Canada's Business Register. The survey population was stratified by industrial grouping, region and three size classes based on number of employees per enterprise:

- Small enterprises (10 to 99 employees)
- Medium-sized enterprises (100 to 249 employees)
- Large enterprises (more than 249 employees)

The sample was selected to meet two sets of objectives:

1) Produce estimates of proportions for defined characteristics for Canada and for selected provinces or regions with a target standard error (quality measure) to satisfy the requirements for estimates by industry and geography.

For the national domains, 87 industrial groupings were identified. For all industrial groupings, a standard error of 5.7% for the proportions to be produced was targeted for all enterprise size categories combined.

The precision requirements for regional domains were based on partnership agreements with the Institut de la statistique du Québec; Ontario Ministry of Economic Development and Trade and Employment and Ontario Ministry of Research and Innovation; and Alberta Innovation and Advanced Education. For Ontario and Alberta, a standard error of 12% was targeted for the proportions to be produced within 48 industrial groupings at the sector (NAICS 2-digit) and subsector (NAICS 3-digit) levels, for all enterprise size categories combined. For Quebec, a standard error of 12% was targeted for the proportions to be produced within all industrial groupings with the exception of the subsector of manufacturing (all NAICS 3-digit within NAICS 2-digit 31, 32 and 33) where a standard error of 6% was targeted, for all enterprise size categories combined. These groupings were constructed to maintain consistency with the industry stratification at the national level. It should be noted that, although not specifically targeted, the remaining provinces and territories were also sampled to ensure precision requirements at the national level were met.

2) Permit microdata analysis within a linked file environment.

Data sources

Data collection for this reference period: 2015-02-25 to 2015-06-25

Responding to this survey is mandatory.

Data are collected directly from survey respondents.

Electronic or E-questionnaires are used to collect data from respondents. Before questionnaires are sent out, all sampled enterprises are contacted to collect the name and email for the respondent with enough knowledge of the enterprise and advanced technologies to complete the survey (e.g. entrepreneur, CEO or senior manager). Invitations to complete E-questionnaires are sent to respondents with email addresses. Access codes are mailed to respondents for sampled units with no email address. Intensive non-response follow-up is conducted by email and telephone as appropriate.

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

Error detection

Error detection is an integral part of both collection and data processing activities.

During collection, edit rules were applied to the survey data as they were entered into the electronic questionnaires by respondents in order to identify capture and reporting errors.

Prior to imputation, subject matter specialists used a variety of tools and approaches to detect and resolve inconsistencies, invalid responses and outliers in the collected data. They reviewed questionnaires missing key variables for imputation and populated these variables where possible. Responses written in by respondents for "other please specify" questions were coded to the list of options supplied to respondents for the question whenever possible. Other edit rules were applied the collected data to identify and correct records with inconsistent, incomplete or invalid responses in an automated way. Finally, rules were applied to the data to detect inconsistencies in the patterns of variable responses and outlier values requiring review and resolution.

Imputation

Imputation was used to determine plausible values for missing or inconsistent variables in the collected data which could not be resolved through editing.

Donor imputation was performed using the generalised system BANFF. This involved identifying respondent records (donor) that were as similar as possible to the records requiring imputation (recipient) based on information that was available for both enterprises.

Instead of imputing each variable independently, variables were grouped into blocks. These blocks were defined based on relationships amongst variables. For a given recipient, all missing variables within the block were imputed from the same donor, thereby maintaining these relationships.

Donor records were required to pass imputation edits and were selected from the same imputation class as the recipient. Imputation was only performed if there were at least 3 donors in the imputation class and at least 5% of all records in the imputation class were donors.

Imputation classes were defined based on industrial groupings derived from sampling, business size, if the enterprise is a user of advanced technology or non-user, and region.

Recipient and donor matches were attempted at the most detailed level of imputation class first to ensure the highest possible degree of similarity.

Recipients that failed to find a donor were matched to records in successively less detailed imputation classes.

Imputation was not performed on out-of-scope units, out-of-business units, partial units who failed to answer the mandatory questions, or non-responding units.

Estimation

A complete file of weighted micro data was created for all sampled enterprises in the survey population for which data were reported or imputed. Weights were adjusted by a factor to account for total non-response so that the final estimates would be representative of the entire survey population. Weighted estimates were produced using the Generalised System of Estimation.

Quality evaluation

The survey estimates for 2014 were compared with those from the 2007 iteration by industry, business size and region (where possible).

The largest differences were investigated and explained or resolved.

Disclosure control

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. 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.

Direct disclosure or primary confidentiality occurs when the value in a tabulation cell is composed or dominated by few enterprises.

Residual disclosure occurs when confidential information can be derived indirectly by piecing together information from different sources or data series.

Revisions and seasonal adjustment

This methodology does not apply to this survey.

Data accuracy

Data quality is assessed based on measures of non-sampling errors and sampling errors. Non-sampling error is not related to sampling and may occur for various reasons during the collection and processing of data. For example, non-response is an important source of non-sampling error. Under or over-coverage of the population, differences in the interpretations of questions and mistakes in recording, coding and processing data are other examples of non-sampling errors. To the maximum extent possible, these errors are minimized through careful design of the survey questionnaire, verification of the survey data, and follow-up with delinquent respondents to maximize response rates.

The response rate for the survey was calculated as the total number of completed questionnaires as a percentage of the total active, in-scope survey sample. The overall response rate for the survey was 68.8% for a total of 7,912 completed questionnaires.

Sampling errors occur as a result of taking a sample of the population. The sample drawn for this survey was only one of many possible samples that could have been drawn. The standard error is a commonly used statistical measure indicating the error of an estimate associated with sampling and was calculated for use in assessing the reliability of estimates are expressed as a percentage. Coefficient of variations are the standard errors expressed as a percentage of the estimate which have been calculated for use in interpreting the reliability of estimates based on an average of responses.

Data quality indicators for the survey are based on the standard error (SE) and the imputation rates. Quality indicators indicate the following: A is very reliable (SE between 0% and 2.49%); B is reliable (SE between 2.50% and 4.99%); C is good (SE between 5.00% and 7.49%); D is acceptable (SE between 7.50% and 9.99%); E is use with caution (SE between 10.00% and 14.99%); and F is too unreliable to be published (SE greater than or equal to 15.00%).