Nunavut Housing Needs Survey
Detailed information for November 2009 to June 2010
The objectives of the Nunavut Housing Needs Survey are to: collect detailed data on the housing needs of Nunavummiut, quantify and identify the factors related to the housing shortage in Nunavut, provide communities with information on housing conditions within their communities, and assist the Government of Nunavut in planning and providing for the housing needs of Nunavummiut.
Data release - October 29, 2010
In 2009, the Nunavut Housing Corporation (NHC) commissioned the Nunavut Bureau of Statistics (NBS) and Statistics Canada (STC) to undertake a Nunavut-wide Housing Needs Survey. In the spirit of the Nunavut Land Claims Agreement, the NHC requested that the survey collection activity should provide employment opportunities for Nunavummiut and respect local cultural and language preferences.
A partnership was established between Statistics Canada and the Nunavut Bureau of Statistics whereby the NBS would collect the data under the Nunavut Statistics Act and manage data collection activities, including the hiring and training of survey interviewers and collection monitoring, from a northern base. Statistics Canada's role was to provide technical assistance to the NBS throughout the survey development and collection process and to process and analyze the survey results.
The Nunavut Housing Needs Survey data may be used by researchers, Inuit organizations, community planners, governments and the public. Results will provide:
- Accurate and up to date information regarding the housing status of Nunavummiut;
- An understanding of the number of individuals and groups in need of housing;
- The ability to compare the living situations of Nunavummiut to Canadian housing standards;
- Communities with important data on the housing conditions of their residences.
- Households, housing and environment
- Housing and dwelling characteristics
- Indigenous peoples (formerly Aboriginal peoples)
- Population characteristics
Data sources and methodology
The target population consisted of all residential dwellings in the 25 communities in Nunavut (excluding small settlements and isolated outpost camps). Dwellings comprised only of temporary residents with a usual home elsewhere at the time of the interview were not surveyed. Collective dwellings (e.g., jails, group homes, shelters) were also excluded from the survey.
The unit of measurement was both the household and individuals living in the household. One adult person in each dwelling responded to the survey on behalf of all household members.
The survey was modeled on the 2004 and 2009 NWT Community Survey and on the 2006 Census of Population. It was modified to meet the Nunavut Housing Corporation's (NHC) data needs and also for applicability to the housing and local circumstances in Nunavut communities. When applicable, questions used in other Statistics Canada surveys were implemented in this survey.
The questionnaire content was developed in collaboration with the Nunavut Housing Corporation (NHC) and the Nunavut Bureau of Statistics (NBS). Cognitive testing of the initial version of the questionnaire was done by Statistics Canada's Questionnaire Design Resource Centre (QDRC). A focus group session and several one-on-one interviews were conducted in Iqaluit in November 2008. The testing was done in English, with simultaneous translation in Inuktitut when needed. Individuals from a mixture of age groups and from different types of dwellings (public housing, government staff housing and privately-owned housing) were recruited. A second test of the questionnaire was done by the Nunavut Bureau of Statistics involving one on one interviews with residents in Pangnirtung, Nunavut in June and July, 2009.
The sample design consists of a census survey of residential dwellings in 24 communities in Nunavut and a sample survey of dwellings in Iqaluit (the largest community). The frame was created using a list of dwellings provided by local and territorial government administrative sources. In Iqaluit, a systematic sample of 911 dwellings was selected.
Data collection for this reference period: 2009-11-10 to 2010-06-11
Responding to this survey is voluntary.
Data are collected directly from survey respondents.
The survey was conducted in person, using a paper questionnaire. This method is referred to as PAPI (Paper and Pencil Interview). Telephone interviews were permitted when follow-up was required. One adult person in each dwelling responded to the survey on behalf of all household members.
View the Questionnaire(s) and reporting guide(s) .
Once the data were received at Statistics Canada's head office, an extensive series of processing steps was undertaken to examine each record received. A top-down flow edit was used to clean up any question paths that may have been mistakenly followed during the interview.
Responses to questions that applied to the respondent but had missing or invalid data were either imputed or assigned the "not stated" code.
Imputation of the variable HOUSEHOLD INCOME was done using BANFF software, a generalized imputation system developed by Statistics Canada. Donor imputation and other deterministic imputation methods were used.
Missing values for the variables AGE, MARITAL STATUS, SEX and STATUS (usual resident vs temporary resident) were also imputed. Variable SEX was imputed by matching the FIRST NAME to the bank of first names from the complete respondents. Variables MARITAL STATUS, AGE and STATUS were imputed using various deterministic and donor imputation methods based on the information available (including variable RELATIONSHIP to Person 1).
a) Initial Sampling Weight
The principle behind estimation in a probability sample is that each unit in the sample "represents", besides itself, several other units not in the sample. For example, in a simple random 2% sample of the population, each unit in the sample represents 50 units in the population. The initial sampling weight is the inverse of the probability of selection. In this example, the probability of selection is 2%, therefore the initial sampling weight is equal to 50.
In the case of a census, all the units in the population are selected in the sample. Therefore, the probability of selection is 100% and the initial sampling weight is equal to 1 (this is the case for all communities except Iqaluit).
b) Weight Adjustment for Nonresponse
Regardless of the survey design, all surveys suffer from a certain amount of nonresponse. Nonresponse occurs when all or some of the information requested from sampled units is unavailable for some reason. For example, this may occur when a unit refuses to participate, no contact is made, they cannot be located or the information obtained from them is not usable.
The most common way of dealing with nonresponse is to adjust for it by applying a weight adjustment to responding units in order to compensate for units that did not respond. The nonresponse adjustment process consists of redistributing the initial sampling weights of the nonresponding units to the responding units within response homogeneity groups. This is based on the assumption that the characteristics of the respondents are similar to the characteristics of the nonrespondents, in a given response homogeneity group. In each response homogeneity group, a nonresponse adjustment factor is calculated by taking the sum of the initial sampling weights of both the responding and the nonresponding units, divided by the sum of the responding units.
c) Final Weight
In the NHNS, the final weight is obtained by multiplying the initial sampling weight by the nonresponse adjustment factor.
The final weights must be used to derive meaningful estimates from the survey. For example if the number of dwellings with a given characteristic is to be estimated, it is done by selecting the records referring to those dwellings in the sample with that characteristic and summing the weights entered on those records.
It is important to note that estimates produced from a survey are subject to both sampling and nonsampling errors. Sampling error of a survey occurs when only a portion of the population is enumerated and the sampled units do not have exactly the same characteristics as all of the population units that they represent.
Since it is an unavoidable fact that estimates are subject to sampling error, sound statistical practice is required to provide users with some indication of the magnitude of this sampling error. The most common measure used is the coefficient of variation (CV). It is obtained by dividing the standard error of the estimate by the estimate itself and is expressed as a percentage of the estimate. For example, suppose that, in the NHNS, one estimates that 34.9% of dwellings had a given characteristic and this estimate is found to have a standard error of 0.0051. Then, the coefficient of variation of the estimate is [ (0.0051) / 0.349 ] x 100% = 1.46%.
The coefficient of variation is used to reflect the quality level of an estimate. The quality level of an estimate is indicated with a letter or a note in the published tables.
To ensure that the survey met its objectives, a Steering Committee comprised of authorities from the Nunavut Housing Corporation, the Nunavut Bureau of Statistics and Statistics Canada determined the concepts and focus.
The survey questionnaire was reviewed by a Statistics Canada Questionnaire Review Committee as well as the Questionnaire Design Resource Centre at Statistics Canada. Several rounds of testing were carried out in Iqaluit and in Pangnirtung, Nunavut to evaluate questionnaire content before the questionnaire was finalized.
Local interviewers with local knowledge of their community were hired as both Senior and Community Interviewers. Senior Interviewers were trained by Statistics Canada staff and provided with interviewer procedure checklists. Community Interviewers were later trained in the communities and supervised by Senior Interviewers
The survey was broadly announced before and during collection to raise awareness of the project and the importance of participation. The media provided the opportunity to do several interviews which were broadcast across the territory. Local interviewers did radio announcements on community radio stations.
A comprehensive list of dwellings was compiled from local and territorial sources of administrative data. Senior Interviewers reviewed and updated the dwelling lists during collection.
Throughout the collection process, control and monitoring measures were in place and corrective action was taken to minimize non-sampling errors. These measures included response rate evaluation and regular contact with interviewers in the field. Where interviewers were having problems with concepts or understanding sections of the questionnaire, they were given feedback over the telephone. When necessary, instructions were circulated more widely to make sure that everyone was clear on the correct procedures.
Interviewers made a number of attempts to contact persons not at home and refusals were followed up to encourage respondents to participate in the survey.
Prior to sending questionnaires to Statistics Canada Head Office, Senior Interviewers reviewed each questionnaire completed by themselves and their Community Interviewers. If necessary, respondents were re-contacted for clarification.
Once received at Statistics Canada and prior to being sent for data capture, paper questionnaires underwent another grooming process to verify the accuracy of responses and to determine whether interviewers were having any difficulty understanding the questionnaire concepts and procedures.
During survey processing, survey records were edited according to prespecified edit rules to check for errors, gaps and inconsistencies in the survey data. For example, checks were made to ensure that numerical answers to certain questions fell within acceptable logical ranges. Checks were also made to ensure that portions of the questionnaire that were to be skipped in the interview because of a previous answer were in fact, skipped. Where errors or inconsistencies were found, the erroneous information was blanked out and replaced by a "not stated" response.
Editing was mostly "top-down" meaning that when a flow question was encountered, the flow pattern indicated by the response of that question was accepted as true.
Finally, results from the NHNS were compared with data from other sources (eg., population estimates and Census of population) in an attempt to identify any large inconsistencies.
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.
Data based on a count of fewer than 10 respondents were suppressed to ensure confidentiality of respondents. To further reduce risks of disclosures, all estimates were rounded to the nearest 10 units.
Revisions and seasonal adjustment
This methodology does not apply to this survey.
The final response rate (the number of responding dwellings as a fraction of all dwellings occupied by usual residents in Nunavut) was 79%. Response rates were lower in larger communities such as Iqaluit, Rankin Inlet and Cambridge Bay. If we exclude those communities from the overall response rates, the final response rate for the remaining communities' increases to 87%.