Survey of Federal Government Expenditures in Support of Education (FEDEX)

Detailed information for 2014/2015

The purpose of this survey is to collect data on federal government expenditures in support of education.

Data release - July 4, 2017

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

Description

The Federal Government Expenditures in Support of Education survey enables Statistics Canada to determine, in detail, direct federal government financial support for education at all levels by department and by province/territory. A standard questionnaire is sent to all federal government departments and agencies with programs related to education and training.

Users of these data include stakeholders from both inside and outside of Statistics Canada. The information collected from this survey, as well as the other finance surveys, gives governments and various associations involved in the field of education, information which assists in the formulation of education policy. External users of these data include federal government departments which of course are also data providers. Data collected are also used to reconcile financial data from other sources. Moreover, these data are included as part of the inputs under ''Total education revenues and expenditures in Canada'' which is required by many organizations.

Other divisions within Statistics Canada also use these data "indirectly" as inputs into their statistical programs, for example, annual benchmarks are produced for sub-annual surveys of labour income and to feed the System of National Accounts and the Public Institutions Division. Data from this survey are also used to feed information to international organizations such as the Organization of Economic Co-operation and Development (OECD). Moreover, these data are also used as inputs to the Pan-Canadian Education Indicators Program as well as the Canadian Social Trends publication.

Reference period: April 1 - March 31

Collection period: March to December

Subjects

• Education finance
• Education, training and learning
• Government

Data sources and methodology

Target population

The target population is federal government departments, agencies, commissions, boards and crown corporations.

Instrument design

The questionnaire was originally designed many years ago by Statistics Canada and other federal government departments which were directly involved in the education sector. Since the original design dates back to the early 1980s, information on details of the questionnaire design, testing or focus groups is not available.

A few minor changes were made in the questionnaire design in the early 2000s, including an "Authorization to release Statement" as well as statement indicating to respondents that if they choose to send data by fax or other electronic means, Statistics Canada cannot guarantee their confidentiality. Only when the data are actually received can the Statistics Act guarantee confidentiality.

The questionnaire with reference year 2003-04 was re-designed in 2004-05 as cell numbers had to be added in order to be compatible with the new "Blaise" data capture program.

Sampling

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

Data are collected for all units of the target population, therefore, no sampling is done.

Data sources

Data collection for this reference period: 2015-03-24 to 2015-12-31

Responding to this survey is voluntary.

Data are collected directly from survey respondents and derived from other Statistics Canada surveys.

Questionnaires and Guideline Booklets are sent to all Federal Departments, Agencies, Commissions, Boards and Crown Corporations. Information on Language Training is obtained from Treasury Board of Canada and the FSEP (Federal Science Expenditures and Personnel) file is received from Science and Technology Division of Statistics Canada, the aforementioned is entered in the appropriate questionnaires when the forms are returned to the division, the data are manually verified with the previous year's questionnaires, validated and entered into the system.

Programs in support of university research, research grants or fellowships, are collected by the Federal Scientific Expenditures and Personnel Survey (Record Number 4212) and integrated into the FEDEX expenditure data.

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

Error detection

The Blaise system applies pre-programmed edits and flags potential errors in the early stages of the data capture. At this stage, Blaise's capture edits ensure that certain cells are not left blank, that others are numeric or textual only. Some edits ensure that answers are restricted to a certain size to meet the requirements of the answers being sought. For example, the year variable can only have four characters. Furthermore, some cells are hard-coded with an answer so that the respondent can see what answer was given in the past and can make the necessary changes to them, i.e. education program names if required.

After the capture edits, the Blaise system does the edit checks to ensure that "Recipient Type", "Expenditure Item", "Expenditure Type" and "Level of Education" are completed for all respective cells and all years. If not, a warning message pops up. In addition, the Blaise system does numerous checks and control totals for the actual dollar amounts in the cells of the questionnaire. For example, "flags" are raised if the total of certain cells exceeds, or shows a blank, in some other corresponding cell within the same category.

The edit failures are corrected accordingly and re-entered in the questionnaire. These corrections are done by either simple means or having to contact the respondents.

Imputation

Imputation for missing, invalid or incomplete (i.e. after thorough follow-up attempts) is generally done by the survey manager and is based on historical estimation using the same department's responses from the previous year(s) and in(de)flation. Other published sources are rarely but sometimes used.

Estimation

This methodology type does not apply to this statistical program.

Quality evaluation

Data collected with this survey are compared to several other published sources during the processing phase. Initially, the "actual" data being disseminated in the reference year are compared to the "preliminary" data reported by the respondent in the previous year. Values reported as transfers by the respondent are also compared to the amounts reported by the recipient within their published reports. For example, transfers from the federal government to school boards are validated against the amounts reported as revenue in the school boards' financial statements. Any inconsistencies highlighted in the process, along with any large changes from the values reported in the previous year, are reviewed by the assigned analysts.

Disclosure control

Statistics Canada is prohibited by law from releasing any information it collects that 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 any data disclosure, confidentiality analysis is done using the Statistics Canada Generalized Disclosure Control System (G-Confid). G-Confid 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

No seasonal adjustment or calendarization is necessary for this survey. The survey manager makes revisions to the Final year figures if necessary and material.

Data accuracy

Non-sampling errors may have the largest impact on the accuracy of the data collected. The data capture process involves a large amount of manual data processing which may lead to material errors. This risk is managed by the "Blaise" system, developed by Collection Systems Division, which incorporates various levels of tabulation and outlier checks. Any value flagged by this system is then examined by the survey team. In the case that an inconsistency is found, the team contacts the respondent to request supplemental information.

Coverage error also has the potential to affect the accuracy of the survey data. Due to potential program and name changes in the various federal departments, the assigned analysts begin each cycle with independent research to determine if the complete population is covered by the survey.

RESPONSE RATE
The response rate is 85%.

NON-SAMPLING ERROR
The methodology of this program has been designed to control non-sampling errors and to reduce the potential effects of these. Measurement errors may exist due to the difficulty of isolating the specific costs of education within mixed function programs. To control non-sampling errors, as well as the effect of non-responses, the survey team deals directly with the survey respondents to clarify the information provided.

NON-RESPONSE BIAS
The main contributing factor to non-response can be attributed to the definition of education and its variations within the federal programs or departments. To mitigate this, the survey team makes presentations to departments so as to clarify concepts. The survey team also cross references the aggregate provincial totals with other data sources to determine if a material level of under-coverage exists.

COVERAGE ERROR
During each survey cycle, the collected aggregated values of federally funded programs in support of education in Canada are compared with similar data from other published sources. This analysis provides a benchmark for comparison with the survey data. Any cases of under-coverage are investigated by the assigned analysts.

OTHER NON-SAMPLING ERRORS
This methodology does not apply.