Trucking Commodity Origin and Destination Survey (TCOD)

Detailed information for 2018

The purpose of the Trucking Commodity Origin and Destination Survey is to measure the commodity movements and the outputs of the Canadian trucking industry.

Data release - February 21, 2020

• Questionnaire(s) and reporting guide(s)
• Description
• Data sources and methodology
• Data accuracy

Description

The objective of the Trucking Commodity Origin and Destination Survey is to measure the commodity movements and the outputs of the Canadian trucking industry. The survey data are used by federal and provincial governments, trucking associations, members of the industry, universities and research institutions to assess the industry's growth rate and contribution to the Canadian economy and to measure the volume of provincial and inter-provincial trade transported by trucking companies. In addition, the statistics are used by planning boards to help determine the volume of traffic on highways and by trucking companies that are investigating expansion opportunities. The TCOD survey estimates are used as an input to Statistics Canada's System of National Accounts.

Collection period: From September of the reference period to the following March

Subjects

• Transportation
• Transportation by road

Data sources and methodology

Target population

The TCOD survey extracts its frame from Statistics Canada's Business Register (BR). The BR statistical structure of a business contains from top to bottom four levels of statistical entities: enterprise, company, establishment and location. The survey population consists of all companies on the BR with at least one trucking establishment (NAICS: 484XXX) that has at least $1.3 million in annual revenue. The local trucking sector NAICS (48411X, 48422X) was added to the previous survey coverage. The frame is created on January 1st of the reference year to allow the interviewers to start collection early in the year. However, a sample of births is selected at the end of the reference year from the list of companies that were not in the TCOD survey population on January 1st of the reference year but that appeared in the survey population for at least one day during the reference year.

In the survey, all shipments made by the companies on the survey frame are in-scope. Shipments of less than 25 kilometres are no longer excluded.

• Additional documentation

Instrument design

The TCOD survey uses the following three collection methods: electronic data reporting; on-site visits, via computer assisted personal interviews (CAPI) and profiles, via computer-assisted telephone interviews (CATI).

Computer-assisted Personal Interviewing (CAPI) involves the use of a laptop computer by a Statistics Canada Interviewer to capture responses to a survey questionnaire during a face-to-face interview. The interview is normally carried-out in the respondent's business or other agreed upon location. The survey responses are encrypted on the laptop's hard disk. Once the interview is completed, the survey responses are transmitted to Statistics Canada headquarters for processing. This method is stable, and all application changes undergo rigorous review and testing prior to implementation, with particular focus on ensuring the confidentiality of survey data.

For centralized Computer-assisted telephone interviewing (CATI), a Statistics Canada Interviewer, working from a centralized Statistics Canada premise, telephones the respondent and captures responses to the interview on a laptop. The survey responses are entered directly on a computer within Statistics Canada's secure, private network. New surveys using CATI applications are developed and tested prior to implementation with particular focus on ensuring the confidentiality of survey data.

Sampling

This is a sample survey with a cross-sectional design.

The survey is an annual survey that employs a four-stage sample design where a stratified simple random sample of trucking companies is selected without replacement at the first stage. All companies in the must-take strata are in the sample. The companies that are not part of any of the must-take strata are then stratified by three NAICS sectors (long-distance, local and moving companies), 13 activity types based on historical data (for example, transportation of forest products, dry bulk, etc.), 13 provinces or territories of domicile and by annual revenue. The larger companies are part of take-all strata and the smaller companies are part of take-some strata. The Lavallée-Hidiroglou method (1988) is used to optimally determine, in each cell, the threshold between take-all and take-some revenue strata.

At the second stage, for each company selected in the first-stage sample, a period of time (e.g., January to June, July to December or January to December) is randomly selected within the reference year. At the third stage, all shipments of companies in EDR (Electronic Data Reporting) must-take strata are selected. For the rest of the sampled companies a systematic sample of shipping documents is selected via personal, on-site visits, for the given second-stage period of time. When there is more than one shipment on the shipping document selected at the third stage, a fourth-stage sampling process is involved, in which a systematic sample of shipments is selected from the shipping document.

Data sources

Responding to this survey is mandatory.

Data are collected directly from survey respondents.

The TCOD survey uses the following three collection methods: (1) electronic data reporting; (2) on-site visits and (3) profiles, via computer-assisted telephone interviews (CATI).

(1) Electronic Data Reporting (EDR). A small number of companies send their data electronically. All of their data are processed using fully automated coding and imputation systems.

(2) On-site visits. The on-site visits are the most frequent mode of collection in the survey. Statistics Canada interviewers visit each company, select a systematic sample of shipping documents, select a sample of shipments on each document and finally transcribe the data from the documents onto laptop computers.

(3) Profiles via CATI. This collection method is used when neither of the other methods (on-site visits or EDR) can be used for a given company. Instead of visiting the company to collect data, the interviewer will collect, through a CATI interview, information about each "typical shipment" and note the number of each "typical shipment" that was made by the company during the reference period.

View the Questionnaire(s) and reporting guide(s).

Error detection

Once all necessary information for the survey is collected, a series of verifications takes place to ensure that the records are consistent and that collection and capture of the data do not introduce errors. Reported data are examined for completeness and consistency using automated edits coupled with manual review. Outliers, i.e., respondents reporting extremely large values, are processed manually.

Imputation

Missing values and data found in error are imputed by an automated system. The system imputes the data using different imputation rules depending on the shipment, available information and the type of data to be imputed. The imputed data are again examined for completeness and consistency.

Estimation

Since the sample is selected based on a four-stage sampling design, the sampling weight is calculated in four steps. The first-stage sampling weight is calculated for each company belonging to the sample. The second-stage weight is then calculated for the company's selected reference period. The third-stage weight is calculated for each shipping document selected during the company's reference period. Finally, the fourth-stage weight is calculated for each shipment selected from the shipping document during the company's reference period. The final weight is obtained by multiplying these four weights together. In addition, an adjustment factor based on the total revenue for the reference period for which the company reported is integrated into the calculation of the estimates. The weighted values, obtained by multiplying the final weights to the survey data, are then aggregated to produce the required estimates.

In order to produce estimates of variance by domain, a Horwitz-Thompson variance estimator for a four-stage design is used.

Quality evaluation

The survey results are analyzed before dissemination. In general this includes a detailed review of the individual responses (especially for the largest enterprises), a review of general economic conditions as well as historic trends and comparisons with industry averages and ratios.

Disclosure control

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 for a specific industry or variable may be suppressed (along with that of a second industry or variable) if the number of enterprises in the population is too low.

Revisions and seasonal adjustment

Annual estimates are provided for the reference year. The data for the previous reference year are revised if necessary. As this is an annual program, seasonal adjustments are not applicable.

Data accuracy

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-sampling error is not related to sampling and may occur for many reasons. For example, non-response is an important source of non-sampling error. Population coverage, differences in the interpretation of questions, incorrect information from respondents, and mistakes in recording, coding and processing data are other examples of non-sampling errors.

Non-sampling errors are controlled through a careful design of the questionnaire, the use of a minimal number of simple concepts and consistency checks. Coverage error was minimized by using multiple sources to update the frame. Measures such as response rates are used as indicators of the possible extent of non-sampling errors.

Sampling error occurs because population estimates are derived from a sample of the population rather than the entire population. Sampling error depends on factors such as sample size, sampling design, and the method of estimation. An important property of probability sampling is that sampling error can be computed from the sample itself by using a statistical measure called the coefficient of variation (CV). The assumption is that over repeated surveys, the relative difference between a sample estimate and the estimate that would have been obtained from an enumeration of all units in the universe would be less than twice the CV, 95 times out of 100. The range of acceptable data values yielded by a sample is called a confidence interval. Confidence intervals can be constructed around the estimate using the CV. First, we calculate the standard error by multiplying the sample estimate by the CV. The sample estimate plus or minus twice the standard error is then referred to as a 95% confidence interval.