Nunavut Government Employee Survey (NGES)
Detailed information for 2016
The Nunavut Government Employee Survey (NGES) collects information from current government employees about their experiences in the workplace, as well as their interest in learning, training and career advancement.
Data release - March 27, 2017
The Nunavut Government Employee Survey (NGES) is an online survey of all current Government of Nunavut (GN) and Government of Canada (GoC) employees working in Nunavut. The survey is conducted by Statistics Canada, and is sponsored by Indigenous and Northern Affairs Canada (INAC) and Employment and Social Development Canada (ESDC). It was developed in consultation with the GN and Nunavut Tunngavik Incorporated (NTI).
The NGES collects information about job experiences, language used at work, access and barriers to job training, interest in career advancement, supervisor support, and issues related to recruitment and retention of Inuit employees. The results of the survey will help the Government of Nunavut (GN) and the Government of Canada (GoC) identify what works well in each department and where improvements should be made. Survey results will also be used to help find ways to increase Inuit employment in the public service, as required by Article 23 of the Nunavut Land Claim Agreement (now referred to as the Nunavut Agreement).
Data sources and methodology
The target population for the survey was all employees of the Government of Nunavut or the Government of Canada working in Nunavut as of April 25, 2016. The survey population for the Government of Nunavut was identified from lists of employees provided by the Nunavut Bureau of Statistics. For the Government of Canada, administrative files were used, updated from lists received directly from various Government of Canada departments.
The 2016 NGES questionnaire was developed in consultation with Nunavut Tunngavik Incorporated, the Government of Nunavut, Employment and Social Development Canada and Statistics Canada. The survey questions were developed to reflect the experience of government employees and particularly Inuit employees working in Nunavut. Questionnaires were tested with government employees in Iqaluit, Nunavut. The questionnaires were also tested on Government of Nunavut and Government of Canada networks before collection began.
The questionnaire is designed as an electronic survey to be completed online at work in English or French. Employees could also choose to download a fillable PDF questionnaire to complete in Inuktitut or Inuinnaqtun, which they could then upload to a secure electronic portal to return to Statistics Canada.
This survey is a census with a cross-sectional design.
NGES was a census of Government of Nunavut employees and Government of Canada employees working in Nunavut, therefore the sampling unit is a person.
Data collection for this reference period: 2016-04-25 to 2016-06-24
Responding to this survey is voluntary.
Data are collected directly from survey respondents and extracted from administrative files.
The data collection period for Nunavut Government Employee Survey 2016 was between April 25, 2016 and June 24, 2016. The initial survey closing date was extended by three weeks, from June 3 to June 24. The collection method was by electronic questionnaire, provided in English and French. A PDF version of the questionnaire was also available in two Inuit languagesInuktitut and Inuinnaqtun. During the collection period, regular e-mail reminders were sent to those who had not yet responded to the survey.
The Government of Canada and the Government of Nunavut provided Statistics Canada with employee email addresses that were used to send email invitations. Only employees with a valid email address received an email invitation. The invitation contained a link to the questionnaire along with a unique code to access the survey. As soon as the respondent submitted their completed questionnaire, the data were transferred through Statistics Canada's internal network and then decrypted for processing. Respondents also had the option of saving their partially completed questionnaires to complete at a later time. The PDF questionnaires that were received by Statistics Canada were captured and edited manually after the end of collection.
The average time required to complete the survey (excluding outliers) was 46 minutes.
The NGES obtained administrative data under the authority of the Statistics Act from the Government of Nunavut in the form of Human Resources files. Employee names, email addresses, and information about location of work were required in order to build the survey frame and support data collected in the questionnaire.
The Government of Canada files were obtained under the authority of the Statistics Act from the Treasury Board Secretariat in the form of the Public Works and Government Services Canada Central Payroll Systems File. Employee names, email addresses, and information about location of work were required in order to build the survey frame and support data collected in the questionnaire.
The Government of Nunavut and Government of Canada Human Resources files were used for some imputation (as outlined below in the section on imputation) and to create some derived variables.
Respondents were advised by Statistics Canada that their survey data may be linked to information from the Government of Nunavut and Government of Canada Human Resource administrative files for the purposes of enhancing survey data and reducing respondent burden.
Data integration was conducted on a case-by case basis. In some instances, frame variables from administrative sources were inconsistent with survey responses. In these situations, survey responses were considered to be the most accurate information and estimates were produced based on the survey responses.
Some imputation was also conducted, where frame variables were used to supplement survey responses. This occurred if there was a frame variable with the same type of information and the survey response was missing.
View the Questionnaire(s) and reporting guide(s) .
Some information on the frame file was also obtained on the survey questionnaire. In these cases, the survey responses were considered to be the most accurate. If the information was missing from the survey response, some imputation was conducted to replace the non-response on the survey file using information from the survey frame. A total of 38 missing values were imputed from the matching frame variables.
The NGES is a census rather than a sample. If all units had an e-mail address available and if the response rate to the survey had been 100%, the weight would have been one for each respondent. However, not all units had an e-mail address and not all units with an e-mail address responded to the survey. For this reason, the survey weights were adjusted to account for both types of sample loss (non-contact due to not having an e-mail address and non-response).
The weight adjustment for non-contact consisted of inflating the weights of the units having an e-mail address to represent those for whom an e-mail address was not available. Similarly, the weights of respondents were inflated to represent the individuals who didn't respond to the survey. The adjustment factors for non-contact and non-response were calculated within non-contact and non-response weighting classes respectively, and formed using auxiliary variables available on the frame such as department, employee group type, NOC code, hire date, region, etc. The weights were then calibrated so that the sum of the weights equaled the department totals.
An accurate estimation of the variance must take into account the sample loss due to non-contact and to non-response. For NGES, a generalized bootstrap method was applied to generate bootstrap weights which can be used to calculate the variance of the estimates produced.
Data validation was conducted by the subject matter team in the Social and Aboriginal Statistics Division at Statistics Canada through the verification of survey responses against expected trends and outputs. As this questionnaire has not been conducted in the past, estimates cannot be assessed against historical data.
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.
NGES implemented a series of measures to prevent disclosure of confidential data. A minimum unweighted count of ten respondents is required in every cell of any given cross-tabulation; any estimate based on fewer than ten respondents must be suppressed or combined with other categories. For a bivariate statistic, both contributing variables must have at least this minimum number of observations to contribute, for example, both the numerator and denominator of a ratio must be based on at least 10 observations.
Another measure to protect confidential data is the prohibition of releasing unweighted data. All data must be release only in aggregate form.
For the NGES all cells in a statistical table are rounded to the nearest ten. The same procedure is used for the marginal sub-totals and totals as well as for the sums and differences of aggregates. It is recommended that researchers proceed in the same manner.
This survey is a census and therefore, there is no error due to sampling. However, it is subject to non-sampling errors, such as non-response and coverage errors, which can induce bias and increase sampling variance.
The frame was created using administrative files and lists of employees. Staff changes between fiscal years could impact the quality of the frame. A total of 67 duplicate records and a total of 67 out-of-scope units were identified before collection and removed from the frame. A total of 29 units were identified as out-of-scope for the survey, based on their survey responses. They were treated as respondents and removed after the weighting phase.
The proportion of individuals on the NGES frame with an e-mail address was 70.8% for the Government of Nunavut (GN) and 87.7% for the Government of Canada (GOC). Among the units with an e-mail address, the NGES reached a final response rate of 35.5%.
A measure for the reliability of the estimations can be the value of the associated CV (coefficient of variation). The extent of the survey error is approximated by the CV with the following guidelines:
-a CV value of 16.5% and below is considered acceptable
-a value of CV between 16.6% and 33.3% is considered marginal and requires cautionary use
-a CV value of 33.3% and above is considered too unreliable to be published.