Nunavut Air Exit Survey & Nunavut Cruise Exit Survey

Detailed information for third quarter 2025

The Government of Nunavut conducts the Nunavut Air Exit Survey (NAES) and the Nunavut Cruise Exit Survey (NCES) in partnership with Statistics Canada to better understand and serve travellers visiting Nunavut.

Data release - To be determined

• Description
• Data sources and methodology
• Data accuracy

Description

The NAES and NCES are conducted by the Government of Nunavut in partnership with Statistics Canada to measure the characteristics and economic impact of the tourism activities of visitors in Nunavut.

The objective of the survey is to provide information about the number of trips, nights, activities, and expenditures in Nunavut by visitors to the territory, by region of residence, main trip purpose, trip duration, mode of transportation, types of accommodations used, and communities visited, among other characteristics.

Results from the NAES and NCES are used by the Government of Nunavut and other tourism stakeholders to better understand and serve travellers visiting Nunavut.

Reference period: Quarter

Collection period: July to September (Q3) each year.

Subjects

• Domestic travel
• International travel
• Travel and tourism

Data sources and methodology

Target population

The NAES targets all non-Nunavut residents visiting the territory and departing by commercial air, except for air crew, military members on active duty, and diplomats. The observed population is comprised of travellers from the target population who are passing through the airports in Iqaluit, Rankin Inlet, and Cambridge Bay, and who are 18 years of age or older.

The NCES targets all non-Nunavut residents visiting the territory and departing by cruise ship, except for boat crew, military members on active duty, and diplomats. The observed population is comprised of travellers from the target population who disembark in the communities of Iqaluit, Pond Inlet, and Cambridge Bay, and who are 18 years of age or older.

Respondents from the observed population report for all members of their travelling party.

All trips to Nunavut made by respondents in the target population are in-scope for the NAES and NCES, although both surveys only ask questions about the trip that the respondent and their travelling party are currently undertaking. Therefore, if a travelling party is intercepted a second time during the same trip to Nunavut, they are not asked the survey again; however, if the same individual or travelling party visits Nunavut multiple times during the reference period, they can respond to the survey multiple times, once for each trip.

Instrument design

The content of the NAES and NCES is based on the National Travel Survey and Visitor Travel Survey questionnaires, with a few adjustments to accurately measure travel and tourism characteristics in the Nunavut context.

A previous iteration of the Nunavut exit survey project prior to the COVID-19 pandemic fully tested the NAES and NCES questionnaires in 2018 and 2019, although no data was produced from this iteration of the project. For the current iteration of the NAES and NCES, which began collection in 2024, the same questionnaires are used.

Data collection for both surveys is done by Government of Nunavut staff on paper questionnaires during face-to-face interviews.

Sampling

The NAES and NCES are cross-sectional convenience sample surveys that target all non-resident visitors in Nunavut (with the exceptions mentioned previously).

The sampling unit for the NAES and NCES is the travelling party, which is a group of individuals travelling together who can jointly report spending for their trip. Although individual travellers respond to the survey, they do so on behalf of their travelling party and care is taken not to interview multiple individuals from the same travelling party.

Data sources

Data collection for this reference period: 2025-07-01 to 2025-09-30

Responding to this survey is voluntary.

Data are collected directly from survey respondents.

During data collection, Government of Nunavut interviewers approach travellers to determine if they are in-scope for the surveys, and complete face-to-face interviews during which the selected traveller responds on behalf of their entire travelling party. Interviews are completed in English, and interviewers are provided with copies of the questionnaire in French, Inuktitut and Inuinnaqtun, as well as German for the NCES, which can be used as a reference if the respondent does not speak English. The responses are recorded on paper questionnaires.

For the NAES, potential respondents are approached in the departure lounges of the Iqaluit, Rankin Inlet, and Cambridge Bay airports, and the interviews last approximately 5-10 minutes.

For the NCES, respondents are intercepted as they disembark from cruise ships in Iqaluit, Pond Inlet or Cambridge Bay, and the interviews last approximately 3-6 minutes.

Error detection

Once the paper questionnaires for NAES and NCES have been received at Statistics Canada, they are scanned and the responses are captured electronically through an automated process. The accuracy of the automated capture is verified by a Statistics Canada data capture expert and corrected as needed, which is the first stage of error detection.

The data then passes through several data processing steps, including cleaning and editing, during which several verifications are completed to detect errors. The errors are typically the result of misunderstanding or mistakes during the face-to-face interview, and some examples of errors that are verified and corrected are:

• Trip start and end dates that do not align with the collection period
• Sum of nights in visit locations total more than the trip duration
• Accommodation expenses are declared for a same-day visit
• No accommodation type is declared for an overnight visit

In addition to the automated edits and cleaning during data processing, validations are executed throughout the process to identify unexpected results, which are verified by subject matter experts in tourism statistics using the original questionnaire images to create coherent data.

Imputation

When a respondent did not provide enough information for any of the spending variables, imputation is used to replace the missing data, for both package costs and non-package spending. If the respondent indicated that their trip included a package, but they did not provide a package cost, the package cost is imputed. For the categorical non-package spending, imputation of each implicated spending category is required in the following four cases:

1. A respondent provided answers for some of the non-package spending categories but did not provide a cost for a spending category where a cost was expected (given the other information provided about the trip)
2. An NAES respondent did not provide any answers for the non-package spending questions
3. A respondent provided only a total for the non-package spending, with no categorical spending breakdown
4. The cost for a non-package spending category was determined to be an outlier

A form of nearest neighbour imputation is used, wherein the value from a similar respondent is used to replace the missing value. Imputation is done using the per person per day value for the relevant spending category, so that the travel party size and the trip duration are controlled.

Estimation

The weighting for the NAES and the NCES are done independently since each survey covers a different target population.

For the NAES, the respondents are weighted to represent the target population of all non-residents of Nunavut that departed Nunavut by air in the reference quarter. The total number of air travellers in the target population is obtained from an external aviation analytics provider and the estimate of the number of non-Nunavut residents departing Nunavut by air is obtained by adjusting this external total using information collected from an airport tally survey. The tally survey is performed throughout collection and is used to estimate the residency of air travellers that were on flights that departed Nunavut.

For the NCES, the respondents are weighted to represent the target population of all travellers that visited Nunavut by cruise ship during the reference quarter. The total number of cruise travellers is obtained from administrative post-trip forms that each company is required to complete in order to operate voyages in Nunavut.

Estimates are produced using G-EST, a generalized estimation system created by Statistics Canada and using the survey weights.

Bootstrap weights are created to estimate the quality of each weighted estimate. A total of 500 bootstrap replicates are created for each of the NAES and NCES, with each replicate being a random subsample of the respondents from each survey. Statistical software can then use each replicate to produce the estimate of interest. It is the variability amongst these 500 estimates that can be used to estimate the sampling variance, which can then be used to produce an indicator of data quality.

Quality evaluation

Data quality is systematically evaluated with each release using expert subject matter knowledge from representatives in the Tourism Statistics Program and collaborators in the tourism sector of the Government of Nunavut, and investigations undertaken as needed.

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.

For the NAES and NCES aggregate tables, categorical responses are collapsed together to ensure that any published estimate is based on a sufficient number of questionnaire responses to avoid identification of any individual respondent. Any cells in aggregate tables that are based on an insufficient number of questionnaires (less than three) are suppressed to ensure the confidentiality of all respondents is protected.

Microdata files are not produced for the NAES or NCES.

Revisions and seasonal adjustment

Data for the NAES and NCES are not seasonally adjusted.

Revisions to published estimates are not a routine part of the NAES and NCES process, however they will be undertaken if issues are identified with previously released data.

Following the release of 2024 NAES data, an issue was identified with the passenger volumes reported by airlines operating a specific route departing Nunavut that had been used in the calibration process. This caused overestimation of several statistics, so the auxiliary data was corrected in collaboration with the aviation analytics provider, and revised data was released.

Data accuracy

Sampling variability is the error in the estimates caused by the fact that the survey is conducted on a sample of respondents rather than on the entire population. Standard error and the related concepts of coefficient of variation (CV) and confidence interval provide an indication of the magnitude of sampling variability.

For the NAES and NCES, the data quality of each estimate is obtained by estimating the coefficient of variation (CV) using the bootstrap weights. The CV and the domain size are then used jointly to produce a letter grade for the data quality, as follows:

• A domain with less than three questionnaires is assigned the letter "F" and suppressed
• A domain with between three and nine questionnaires is assigned the letter "E" and should be used with caution
• A CV from 0.00% to 4.99% is assigned the letter "A" and is considered excellent
• A CV from 5.00% to 14.99% is assigned the letter "B" and is considered very good
• A CV from 15.00% to 24.99% is assigned the letter "C" and is considered good
• A CV from 25.00 % to 34.99% is assigned the letter "D" and is considered acceptable
• A CV of 35.00% or higher is assigned the letter "E" and should be used with caution

The standard error and coefficient of variation do not measure systematic biases in survey data that might affect estimates. Rather, they assume that sampling errors follow a normal curve.

For the NAES and NCES, there are several possible sources of non-sampling error or bias, including incorrectly recorded responses on paper questionnaires, capture issues and nonresponse bias caused by certain categories of travellers being less inclined to respond to the surveys. Issues with data capture are addressed to the extent possible through error detection and imputation, however none of the sources of non-sampling error are quantified for these surveys.

Response rates
Response rates are not measured for the NAES or NCES, as there is no sampling frame, and refusals and other types of nonresponse are not tracked during data collection.

Non-sampling error
Issues with data capture are addressed to the extent possible through error detection and imputation, as described in the relevant sections.

Non-response bias
Nonresponse bias is not measured for the NAES or NCES, as there is no sampling frame, and therefore no information is collected about nonrespondents.

Coverage error
Efforts are made to limit coverage errors by conducting data collection as widely as permitted within the limits of logistical constraints and by intercepting as many visitors in Nunavut as possible.

For the NAES, interviews are conducted in the three major gateway airports of Iqaluit, Rankin Inlet, and Cambridge Bay. There are very few flights that depart Nunavut from other communities, so most visitors should be covered by these locations.

For the NCES, a master itinerary of all cruise vessels arriving in Nunavut is available, and efforts are made to attend as many as possible for interviews. Data collection is conducted in the communities of Iqaluit, Cambridge Bay, and Pond Inlet, and there are few vessels that visit Nunavut that do not dock in at least one of these three communities.