Compensation Sector Survey
Detailed information for 2001
The purpose of the survey is to obtain a profile of members of the compensation community in the Human Resources community of the federal public service.
Data release - March 11, 2002
- Questionnaire(s) and reporting guide(s)
- Data sources and methodology
- Data accuracy
Statistics Canada conducted the survey on behalf of the Human Resources Community Secretariat (HRCS) of the Treasury Board of Canada Secretariat.
The 1999 Public Service Employee Survey (Survey #4438), carried out with federal employees, gave an overview of employees in the public service, but it was impossible to extract data specific to the compensation community based on the professional group of respondents.
To obtain reliable data that would make it possible to understand the demographic characteristics of this community throughout the federal public service, those responsible for updating the compensation function of the Human Resources Community Secretariat (HRCS) mandated Statistics Canada to conduct a survey.
The purpose of the survey is to obtain a profile of members of the compensation community in the Human Resources community of the federal public service. The results would allow the HRCS to renew recruiting, training and development programs for this community in such a manner that these programs would take into account current data.
- Unionization and industrial relations
Data sources and methodology
This was a census type survey. This means that all public servants in the departments and organizations covered by Schedule 1, Parts I and II of the Public Service Staff Relations Act and members of the compensation group in the Human Resources community who were part of the target population were asked to complete the questionnaire.
The population base was broken down by the HRCS. In August 2001, the HRCS sent a note to all regional and local managers of departments and organizations to determine the number of questionnaires required by them to cover all sectors. The questionnaire was to be completed by all employees who provide payroll and fringe benefit services within the federal public service.
The content of the questionnaire was defined with the help of members of the HRCS. They used the questionnaire for the survey conducted by Statistics Canada with federal employees as a guide (1999 Public Service Employee Survey), as well as the questionnaire used the previous year for a demographic study with the personnel administration group of the human resources community.
The provisional version of the questionnaire was submitted for approval to the members of the Association of Compensation Managers. Their comments were included in the questionnaire and the final layout was designed.
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 collection for this reference period: 2001-09-26 to 2001-10-26
Responding to this survey is voluntary.
Data are collected directly from survey respondents.
The questionnaires were sent by mail to each of the regional and local compensation managers. Each of the mailings sent to the managers included an introductory letter, the number of questionnaires requested, and pre-stamped return envelopes.
Each manager was responsible for distributing the questionnaires to all of their employees working in the field of compensation and to encourage them to complete and return it in the following weeks.
After an initial mailing of 2,130 questionnaires, additional requests for 224 questionnaires were received (for a total of 2,354). Additional questionnaires were sent and the initial collection period was extended. Collection took place from September 26 to October 26, 2001.
Three e-mail reminders were sent during the collection period. In these reminders, we asked regional and local compensation managers to complete the survey if they had not already done so. We also asked them to remind their compensation staff about the survey and to ask them to return the questionnaires as soon as possible.
View the Questionnaire(s) and reporting guide(s) .
A single manual edit was conducted at the reception stage. As clearly identified response categories were essential for scanning purposes, each question in the questionnaire was examined. Then, we intensively scanned the data from the 1,509 questionnaires we received. This method eliminates input errors.
We edited to isolate missing, invalid or incoherent data. We added a sequential number to the questionnaires to identify them and to eliminate any potential duplication. We also reviewed each question to check for the presence of a valid code. If this code was missing, code 9 (not stated) was attributed. We also edited to analyse the link between responses to questions 22, 24, 28 and 30. Redundant data was suppressed in light of what the responses to these questions suggested.
In addition, four types of incoherent data were corrected. In approximately 2% of questionnaires, age group (Q5) in relation to the number of years of experience either in the current position (Q17), in Compensation (Q18) or in the Public Service (Q19), as well as age group (Q5) in relation to the anticipated number of years before retirement (Q21) did not correspond; to resolve this situation, we assigned the code "not stated" as a response to one of the two questions, either the question on age group or the other question.
We also edited so that the value for the department/agency was unique. This question received multiple responses in 25 questionnaires. As there is no relation between the responses, a code of "not stated" was recorded for this question.
This methodology does not apply.
This survey was a census, which means that each record is counted once and has equal weight in the survey results.
As mentioned previously, the number of completed questionnaires for this survey does not necessarily reflect the number of people in the population. Therefore, as the exact size of the population is unknown, it was impossible to adjust for non-response.
The Compensation Sector Survey is a census type survey. Therefore, there can be no error due to sampling variability. However, non-response errors or errors that may occur at almost every stage of the survey process are possible. Respondents may provide erroneous information or errors may occur during the processing and tabulation.
We applied quality control and assurance methods in accordance with current Statistics Canada practices, at each stage of the collecting and processing cycle, in order to verify the quality of the data. We edited particularly to detect missing, invalid or incoherent data.
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.
It should be noted that the "Public Use" microdata files described above differ in a number of important respects from the survey "master" files held by Statistics Canada. These differences are the result of actions taken to protect the anonymity of individual survey respondents. Users requiring access to information excluded from the microdata files may purchase custom tabulations. Estimates generated will be released to the user, subject to meeting the guidelines for analysis and release.
The master file includes explicit geographical identifiers for the given geographical area (17 regions). Where the sample size allows, estimates for the main geographical regions (6 regions) may be obtained. The public use microdata file does not include any indicators below "National Capital Region" and "Outside the National Capital Region".
In order to protect the confidentiality of respondents, the following actions were taken:
1) Suppression of some demographic variables.
Q25 Current area of study
Q26 Working arrangement to pursue studies
Q27 Time expected to complete program of study
2) Three variables were consolidated (Dvq789) to reduce the possibility of respondents being identified.
Q7 Are you an Aboriginal person?
Q8 Are you a person with a disability?
Q9 Are you a person in a visible minority group?
3) Collapsing the answer categories of some variables (i.e., Q1, Q2, Q5, Q10 to Q12, Q14, Q15, Q17 to Q20, Q22, Q23 and Q29).
4) Assessment of risk of disclosure.
Three-dimensional cross tabulations, within the two geographical regions, were produced for every possible combination of indirect identifiers (23 geographical variables). The multiplicity counts, namely the number of times a record is the only one in these three-dimensional tabulations, were taken.
Throughout this analysis, 5,456 tabulations were examined within the two geographical regions, for a total of 10,912 tabulations.
5) Local suppressions.
Procedures were applied to the 5% of the records most affected by multiplicity (75 records). The count of the number of times a variable was part of a single three-dimensional combination was calculated for each record. When the variable with the highest count was not one of the two variables considered "not suppressible" (Q5 and Q12), it was suppressed (given a value of "not stated"). If the variable with the highest count was considered not suppressible, then the variable with the next highest count was chosen for suppression.
Revisions and seasonal adjustment
This methodology does not apply to this survey.
While considerable effort is made to ensure high standards throughout all stages of collection and processing, the resulting estimates are inevitably subject to a certain degree of error. These errors can be broken down into two major types: non-sampling and sampling.
Non-response is an important source of non-sampling error.
The target population consisted of 2,354 individuals. The overall response rate for the 2001 Compensation Sector Survey was 64.1%.
The basis for measuring the potential size of sampling errors is the standard error of the estimates derived from survey results. Because of the large variety of estimates that can be produced from a survey, the standard error of an estimate is usually expressed relative to the estimate to which it pertains. This resulting measure, known as the coefficient of variation (CV) of an estimate, is obtained by dividing the standard error of the estimate by the estimate itself and is expressed as a percentage of the estimate.
Please refer to the User Guide for detailed information.
- Compensation Sector Survey User Guide