Survey of Earned Doctorates (SED)
Detailed information for 2007-2008
This survey is designed to determine such factors as: labour market and mobility plans after graduation, how graduates funded their doctoral studies and how much, if any, debt they accumulated during their studies and the time required to complete a doctoral degree. In addition, information on educational history and socio-economic background is collected.
Data release - May 29, 2009
The Survey of Earned Doctorates (SED) is an annual census of doctorate recipients in Canada that was conducted for the first time on a national basis during the 2003-2004 academic year. The basic purpose of this survey is to gather data about all doctoral graduates in Canada to inform government, associations, universities and other stakeholders on the characteristics and plans of these very highly qualified graduates as they leave their doctoral programs.
These data are important in improving graduate education by providing governmental and private agencies with the information necessary to make program and policy decisions. Data about an institution's own doctorate recipients are also provided to, and used by, research offices of institutions who participate in the survey.
The survey's key data objectives are to evaluate the impact of the various sources of institutional funding; to gather information on the retention of doctoral students in Canada; to gain a better understanding of postgraduate education financing and debt level; to allow labour market planners to assess the additions to the domestic stock of highly qualified human resources in various fields; and, to allow an examination of the path to receipt of doctoral degrees and the impact of foreign students.
The Survey of Earned Doctorates consists of a questionnaire given to all graduates earning a doctorate degree (research-oriented or other) in one of the academic institutions granting doctorates throughout Canada.
The survey collects data about the graduate's postsecondary academic path, funding sources, field of study and his/her immediate postgraduate plans.
The data from the SED can be used by universities and governments to make policy decisions that affect graduate education throughout Canada, by federal agencies to inform parliament and to make decisions about financial commitments that affect graduate education throughout Canada; and, in the evaluation of graduate education programs, strategic planning at the provincial level, labour force projections, and affirmative action plans at all levels.
- Education, training and learning
- Fields of study
- Outcomes of education
Data sources and methodology
The target population is all doctoral graduates from Canadian postsecondary education institutions who have obtained their degree during the reference period. The survey population excludes institutions that did not participate in the survey during the reference period.
The target population is identified from the list of Canadian postsecondary institutions granting doctoral degrees. This list is compiled and kept up-to-date by the Centre for Education Statistics of Statistics Canada. Every listed institution was invited to participate in this survey. Institutions with no doctoral graduates for the survey reference year were excluded from the target population.
A paper questionnaire was tested by a pilot test conducted from November 2002 to June 2003. Changes to the questionnaire were made following the pilot. These changes, related to overall layout and question-specific instructions to respondents, focused on reducing respondent burden and improving data quality.
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: 2007-07-04 to 2008-11-03
Responding to this survey is voluntary.
Data are collected directly from survey respondents.
All doctoral graduates from participating institutions are invited to fill in a paper SED questionnaire, which is distributed by their institutions. The graduates can return the completed questionnaire directly to Statistics Canada or to their institutions. Institutions mail back the completed questionnaires to Statistics Canada. Follow-up calls with non-respondents are made by Statistics Canada. Data is not gathered from proxies.
View the Questionnaire(s) and reporting guide(s) .
Data capture was performed by traditional keying on microcomputers. Verification of the accuracy of the keyed data was done by having a different operator capture non-text fields of all the questionnaires a second time. Quality control was achieved by comparing the two sets of captured information against the paper questionnaire.
The data editing phase of processing involves the identification of erroneous or inconsistent values in the survey data, and the modification of such conditions.
The first type of error treated were errors in questionnaire flow. Conflicting questionnaire information would sometimes indicate that a respondent had answered questions in a section that in fact did not apply to him. In these cases, the superfluous data was eliminated.
The second type of errors treated involved editing the survey records according to pre-specified edit rules to check for logical inconsistencies in the survey data. In these cases a strategy was established to eliminate the minimum amount of data, establishing priorities so as to eliminate the values that were most likely to be in error. An example of this would be when a respondent indicated that he was Canadian by birth at one question, while also indicating that he was a landed immigrant in another question and also giving the year he became a landed immigrant. In this case, the responses to the questions on immigrant status were considered more likely to be correct and therefore the response to the question on citizenship (Canadian by birth) was recoded.
The third type of error dealt with assigning a code of "not stated" whenever the respondents did not provide any information to questions that should have been answered.
No adjustment is made for item non-response.
To compensate for the non-response of individuals, weighting is done within weighting classes to ensure that the survey population totals agree with known totals. Larger institutions are adjusted within institution, and institutions with a small number of graduates are grouped geographically to form weighting classes.
Weighting classes are also split by gender, citizenship or immigration status where appropriate. In most cases response rates differ by gender. Other variables are not reported consistently by institutions and therefore, cannot be used as often as the gender to form weighting classes.
Data was compared with other sources such as the Postsecondary Student Information System (PSIS) and the previous Survey of Earned Doctorates (SED) release.
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
This methodology does not apply to this survey.
This census survey has been designed to control errors and reduce their potential effects but the results remain subject to errors. Non-response error is the most significant source of error. Response error may be due to misinterpretation of the questions or definitions, or inability of the respondent to recall requested information. Coverage error results from incomplete listing of the population of interest. Processing errors include mistakes in entering and coding data. Captured data are verified to minimize these kinds of errors.
While SED is intended to cover all institutions offering doctoral degrees, four institutions was contacted but did not participate. It is estimated that less than 0.5% of all graduates, graduated from these institutions. No adjustment was made for these graduates. Since the undercoverage is negligible for the 2007/2008 survey, cross-sectional estimates should not be affected. Moreover, a direct comparison of this year published levels with previous years levels is possible but only if the undercoverage of those years was inexistent or negligible.
2,517 usable questionnaires were received. There were a total of 4,925 graduates in the participating institutions, which corresponds to a 51% response rate.
The response rate has two components: factors common to an institution or institutions; and factors dependent on the individual. For example, if one institution distributes questionnaires in a graduation application package, while another distributes them at convocation, the first institution will have a higher response rate from graduates.
Although weighted to represent all graduates at participating institutions, results can be applied with more confidence to graduates living in Canada. Tabulations of graduates living outside of Canada, foreign students, and similar variables cannot be released due to risk of disclosure.
There is no adjustment for item non-response as almost all questionnaires were received fully completed.
The unit non-response rate of 49% for SED decreases the reliability of estimates based on the survey data. The existence of non-response in a census survey creates variance and potential bias in the estimated characteristics. The degree to which an estimate of a characteristic is affected depends on how similar respondents and non-respondents are with respect to this characteristic, and the extent to which dissimilarities are accounted for by the survey weights. The estimation methodology used in SED assumes that all persons within a weighting class (respondents and non-respondents) have the same propensity to respond and that this propensity is independent of the characteristics measured by the survey. The validity of these assumptions determines the quality of the survey estimates and may vary from one characteristic to another.
For the 2007/2008 SED, auxiliary information on the frame that could be used to create weighting classes was very limited. Thus, it was not possible to construct weighting classes to adjust for all of the expected sources of non-response bias. In particular, estimates of error do not account for potential bias introduced by the lower proportion of responding graduates who had moved outside of Canada. Data users are advised to apply caution in extrapolating results of the 2007/2008 SED to the population of graduates who moved out of Canada immediately after graduation.