Annual Survey of Research and Development in Canadian Industry (RDCI)

Detailed information for 2016

Status:

Active

Frequency:

Annual

Record number:

4201

The Annual Survey of Research and Development in Canadian Industry collects research and development expenditures and personnel data used to monitor science and technology related activities of companies and industrial non-profit organizations in Canada.

Data release - December 18, 2017 (preliminary)

Description

The Annual Survey of Research and Development in Canadian Industry (RDCI) is a cross-economy survey of businesses and industrial non-profit organizations in Canada that 1) perform or fund research and development (R&D) or 2) have previously reported R&D expenditures and have recent payments or receipts for technology. The survey comprises businesses and industrial non-profit organizations in all NAICS industries other than universities (NAICS 61131) and all levels of government (NAICS 91 public administration).

The concepts and definitions employed in the collection and dissemination of research and development (R&D) data are provided in the Frascati Manual 2015 (Organisation for Economic Cooperation and Development (OECD), 2015). According to this definition:

"Research and experimental development (R&D) comprise creative and systematic work undertaken in order to increase the stock of knowledge - including knowledge of humankind, culture and society - and to devise new applications of available knowledge."

The RDCI collects in-house R&D expenditures and personnel, outsourced R&D expenditures and payments and receipts for technology.

In-house R&D expenditures include current costs (comprised of wages and salaries of permanent, temporary and casual employees; services to support R&D; R&D materials; and all other current costs) and capital costs (comprised of software; land; buildings and structures; and equipment, machinery and all other capital costs). In-house R&D expenditures are characterized by their geographic distribution (provinces and territories), sources of funds (originating sector inside or outside Canada), fields of research and development, and nature of R&D activity (basic research, applied research and experimental development).

In-house R&D personnel include researchers and research managers; R&D technical, administrative and support staff; and other R&D occupations. These data are available by geographic distribution (provinces and territories).

Outsourced R&D expenditures comprise payments made to other organizations to perform R&D, and may be directed to other organizations (including companies; private non-profit organizations; industrial research institutes or organizations; hospitals; universities; federal government departments or agencies; provincial or territorial government departments, ministries and agencies; provincial or territorial research organizations or other organizations or individuals) inside or outside Canada.

Technology payments include payments made or received for patents, copyrights, trademarks, industrial designs, integrated circuit topography designs, original software, packaged off-the-shelf software, databases with a useful life exceeding one year, other technical assistance, industrial processes and know-how. Technology payments can be made to, or received from affiliated or unaffiliated organizations within or outside Canada.

Statistical activity

The survey is administered as part of the Integrated Business Statistics Program (IBSP). The IBSP program has been designed to integrate approximately 200 separate business surveys into a single master survey program. The survey instrument conforms to the common look, structure and content for business surveys in the integrated program.

Reference period: The fiscal year for fiscal year end date between April 1, RY and March 31, RY+1

Collection period: December to April after the reference period

Subjects

  • Research and development
  • Science and technology

Data sources and methodology

Target population

The target population comprises all companies and industrial non-profit organizations that perform and/or fund research and development (R&D), or have had R&D expenditures in the past and continue to make or receive technology payments within the reference period. The survey is a cross economy survey and includes all NAICS codes except NAICS 61131 (universities) and NAICS 91 (public administration).

Instrument design

The RDCI uses two questionnaires: one for companies and another for industrial non-profit organizations. These questionnaires were developed to conform to international standards for research and development concepts (OECD, Frascati Manual 2015). Electronic questionnaires (EQ) are the principal mode of collection and these were tested with company respondents in English and French to confirm respondents' understanding of terminology, concepts and definitions as well as their ability to provide the requested data and to navigate the EQ applications. Questionnaire content testing occurred in March 2014 in English in Ottawa, Toronto and Montreal and in French in Gatineau and Montreal. This first round of testing concentrated on validating respondents' understanding of concepts, questions, terminology, the appropriateness of response categories and the availability of requested information. The subsequent round of testing in June 2015 occurred in English in Toronto and French in Montreal.

Sampling

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

The RDCI is a stratified sample of companies classified by: 57 unique industry groups, R&D size and geography.

Data sources

Data collection for this reference period: 2017-10-24 to 2018-03-31

Responding to this survey is mandatory.

Data are collected directly from survey respondents and extracted from administrative files.

Electronic questionnaire with non-response follow-up and failed edit follow-up.

Administrative data are those data that have been collected for administrative purposes (ex: tax activities of companies or individuals) as opposed to statistical purposes. The use of administrative data reduces data collection costs and respondent burden. Concepts or definitions of administrative data variables differ from those identified in survey design. The administrative data source does not cover the entire target population, and as such, sampling error will be present. The portion not covered by tax data have been identified as "must-take" units to address possible sampling error issues. Non-sampling errors and bias may be present as a result of data collection methodology.

Administrative data are used for many different statistical purposes: replacing or complementing direct data collection to reduce costs and respondent burden; achieving efficiencies in statistical operations, such as the creation of survey frames, design of survey samples, imputation, and estimation. In collaboration with data providers, Statistics Canada uses its mandate under Section 13 of the Statistics Act to access administrative data for statistical purposes.

The confidentiality of administrative data relating to individual persons, companies or organizations (referred to as identifiable administrative data) must be strictly maintained as required by Subsection 17(1) of the Statistics Act.

Scientific Research and Experimental Development (SR&ED) tax incentive program data are used for data replacement for "take-none" units, however SR&ED does not collect: capital R&D expenditures and lease costs, R&D expenditures in the social sciences and humanities or payments for R&D performed by organizations outside Canada.

Corporation income tax return data (T2) provided by the Canada Revenue Agency (CRA) is used to provide revenue information previously collected on the survey.

Payroll deduction tax data (PD7) provided by the CRA is used to generate employment size categories (based on number of employees variable) for dissemination purposes.

In addition to data collected through the survey, the RDCI uses administrative data from the Canada Revenue Agency (CRA)(approved SR&ED tax credit applications) for the "take-none" component of the sample in order to reduce response burden for smaller companies. These data are also used to assist in imputation for non-response.

Records are matched by Business Number root (BN). The Statistics Canada definition of R&D differs from that of the CRA in the inclusion of: all capital expenditures related to R&D, current costs for rental of capital goods and R&D in the social sciences and humanities.

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

Error detection

Error detection is an integral part of both collection and data processing activities. Automated edits are applied to data records during collection to identify reporting and capture errors. These edits identify potential errors based on year-over-year changes in key variables, totals, and ratios that exceed tolerance thresholds, as well as identify problems in the consistency of collected data (e.g. a total variable does not equal the sum of its parts). During data processing, other edits are used to automatically detect errors or inconsistencies that remain in the data following collection. These edits include value edits (e.g. Value > 0, Value > -500, Value = 0), linear equality edits (e.g. Value1 + Value2 = TotalValue), linear inequality edits (e.g. Value1 = Value2), and equivalency edits (e.g. Value1 = Value2). When errors are found, they can be corrected using the failed edit follow-up process during collection or via imputation. Extreme values are also flagged as outliers, using automated methods based on the distribution of the collected information. Following their detection, these values are reviewed in order to assess their reliability. Manual review of other units may lead to additional outliers being identified. These outliers are excluded from use in the calculation of ratios and trends used for imputation, and during donor imputation. In general, every effort is made to minimize the non-sampling errors of omission, duplication, misclassification, reporting, and processing.

Imputation

When non-response occurs, when respondents do not completely answer the questionnaire, or when reported data are considered incorrect during the error detection steps, imputation is used to fill in the missing information and modify the incorrect information. Many methods of imputation may be used to complete a questionnaire, including manual changes made by an analyst. The automated statistical techniques used to impute the missing data include:

- deterministic imputation (e.g., adding components to create a total),
- replacement using previously reported anticipated values for the current period values (the survey asks for reference year (RY), RY+1 and RY+2 values for key variables),
- replacement using historical data (with a trend calculated, when appropriate),
- replacement using auxiliary information available from other sources,
- replacement based on known data relationships for the sample unit, and
- replacement using data from a similar unit in the sample (known as donor imputation).

For the research and development surveys, the key question on expenditures for in-house R&D is verified or imputed first; these values are then used as anchors in subsequent steps to impute other related variables.

Imputation generates a complete and coherent microdata file that covers all survey variables.

Estimation

The sample used for estimation comes from a one-phase sampling process. An initial sampling weight (the design weight) is calculated for each unit of the survey and is the inverse of the probability of selection. It is then adjusted to take into account outliers that might have been misclassified.

The weight calculated for each sampling unit indicates how many other units it represents. The final weights are usually either one or greater than one. Sampling units which are selected for certainty (must-take units) have sampling weights of one and only represent themselves; outlier units with larger than expected size are seen as misclassified; their weight is usually adjusted so as to only represent themselves.

The sampling unit, being the company, is considered an estimation unit. The characteristics of the estimation units are used to calculate aggregate estimates, including industrial classification. Estimation for the survey portion is done by simple aggregation of the weighted values of all sampled companies found in the domain of estimation. Estimates are computed for several domains of estimation, such as industry groups, country of control, company size, and are based on the most recent classification information available for the company and the survey reference period.

In the case of an ineligible sampling portion (take-none portion) of the target population, a model estimate is produced using two adjustments: the first is derived from the relationship between two closely related variables - current in-house expenditures from the questionnaire and current in-house expenditures from tax data; the second adjustment is used to model all other variables based on either the in-house R&D expenditures or the outsourced R&D expenditures in Canada. The overall estimate is composed of estimates from both the surveyed and modeled portions.

Quality evaluation

Prior to the data release, combined survey results are analyzed for comparability. This analysis includes a detailed review of:

- individual responses (especially for the largest organizations),
- general public funding announcements and initiatives,
- coherence with results from other R&D surveys,
- prior period intentions, and
- information from other external sources (e.g. annual reports, news articles).

The survey estimates are also analyzed with trends observed in data from previous collection cycles and media reports.

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.

In order to prevent data disclosure, confidentiality analysis is done using the Economic Disclosure Control and Dissemination System (EDCDS). EDCDS is used for primary suppression (direct disclosure) as well as for secondary suppression (residual disclosure). Direct disclosure occurs when the value in a tabulation cell is composed of or dominated by few enterprises, while residual disclosure occurs when confidential information can be derived indirectly by piecing together information from different sources or data series.

Revisions and seasonal adjustment

Data for RY - 1 reference period are revised in the following way:
- Inclusion of new units identified using tax data for companies not in the frame at the time the sample was drawn.
- Use of new tax data received for imputation of the non-response units within the sample portion.
Revised RY-1 data are released with the RY preliminary data.

Data accuracy

There are two types of errors to which survey data can be subject: sampling errors and non-sampling errors. As a census, this survey is not subject to sampling error. Non-sampling error is not related to sampling, and may occur for various reasons during the collection and processing of data.

Non-sampling errors include:

- non-response (both total and partial),
- under- or over-coverage of the population,
- differences in the interpretations of questions and mistakes in reporting, and
- coding and processing errors.

To the greatest extent possible, these errors are minimized through careful design of the survey questionnaire, verification of the survey data, and follow-up with respondents when needed to maximize response rates.

Quality rating codes are estimated using imputation rates. The imputation rate is calculated based on the contribution of imputed values to the total estimate. The quality indicator code ranges from A to F, where an 'A' rating indicates excellent data quality, and estimates with an 'F' rating are too unreliable to be published.

Documentation

Date modified: