Canadian Armed Forces Members and Veterans Mental Health Follow-up Survey (CAFVMHS)
Detailed information for 2018
The Canadian Armed Forces Members and Veterans Mental Health Follow-up Survey is conducted by Statistics Canada in collaboration with the Department of National Defence and the Canadian Armed Forces (DND-CAF), Veterans Affairs Canada (VAC) and the University of Manitoba. The purpose of this survey is to re-assess the mental health of respondents who participated in a similar survey conducted in 2002.
Data release - April 23, 2019
The Canadian Armed Forces Members and Veterans Mental Health Follow-up Survey (CAFVMHS) collects information about the mental health and well-being of Canadian Armed Forces (CAF) members who had previously responded to the 2002 Canadian Community Health Survey - Mental Health and Well-being - Canadian Forces (CCHS-CF). It will also gather information on the factors that affect their health and their use of health care services.
This survey will measure changes to the mental and physical health of current and former CAF members since the survey was originally conducted in 2002. As a result, for the first time there will be data that describes the impact of mental health disorders, on a range of outcomes among military personnel, over a period of time. This follow-up survey targets this gap by re-interviewing the original respondents from the 2002 survey.
Information on a variety of topics related to mental health and well-being will be collected. The survey asks questions about:
- mental health disorders such as Post Traumatic Stress Disorder (PTSD), generalized anxiety disorder, and depression;
- chronic conditions, physical activity, stress and work related stress;
- social support, childhood experiences, deployment and military experiences;
- access to, and use of, formal and informal mental health services and supports; and
- perceived and unmet needs for services and supports.
Reference period: Calendar year
Collection period: January to May 2018
- Diseases and health conditions
- Health care services
- Mental health and well-being
Data sources and methodology
The target population for this survey is a subset of the target population for the 2002 Canadian Community Health Survey (CCHS) - Mental Health (MH) and Well-being - Canadian Forces (CAF). It includes full time regular members of the Canadian Forces during the 2002 survey reference period.
The observed population will consist of respondents to the 2002 CCHS-CF, who were full time regular members of the Canadian Forces during the 2002 survey reference period, that are living within the 10 provinces.
This follow up survey excludes the reservists who have paraded at least once in the six months preceding the 2002 survey reference period. The population being studied will include members that have been released since the 2002 study.
This study is a follow up survey, therefore, the content is primarily based on the original 2002 survey which was partly based on a selection of mental disorders from the World Mental Health Survey (WMH2000) and other content from the Canadian Community Health Survey Cycle 1.1. The CAFVMHS is also based on content from existing sources such as the 2012 CCHS - Mental Health Survey, the 2013 Canadian Forces Mental Health Survey, and the 2016 Life After Service Survey.
An expert group of mental health professionals guided the content development and strategic direction of the new follow up survey. The entire survey is 75 minutes in length; 60 minutes for survey content and 15 minutes for the entry and exit portions of the survey.
Qualitative testing involves administering cognitive interviews to selected participants that are representative of the targeted population for the survey. This testing was conducted by StatCan's Questionnaire Design Resource Centre (QDRC) to ensure that respondents understood the questions in the way that they are intended to be understood.
In addition, Qualitative testing was conducted in English and French with cognitive interviews administered to 40 participants composed of regular current and released members of the CAF. The personal interviews, half English and half French took place in Kingston, Ottawa, Quebec City and Halifax. Recommendations were given by the Questionnaire Design Resource Centre (QDRC) and revisions based on these recommendations were made to the content of the questionnaire.
Some of the existing modules have been modified to include an additional reference period by repeating lifetime reference period questions with the addition of "since 2002". These modifications were added in order to analyze mental health changes in respondents since the original survey in 2002. In addition, some modules have been dropped, some modules from other surveys have been added and new content has been developed and added.
This is a sample survey with a longitudinal design.
The 2002 survey was a sample survey from a DND database with a cross-sectional design. This survey is a longitudinal follow-up to the 2002 survey. Respondents from 2002 will be linked to data sources such as records from the DND records, the VAC records, tax files, the Census and a mortality database to determine if they are still alive and living in the 10 provinces. A maximum of 4,300 respondents from 2002 that are living in the 10 provinces, based on the updated contact information was selected for the 2018 survey.
The sampling unit is former or current Canadian Armed Forces regular members. The units sampled are respondents from the 2002 survey that were regular force members. This follow up survey excludes the reservists who have paraded at least once in the six months preceding the 2002 survey reference period.
Data collection for this reference period: January 08, 2018 to May 31, 2018
Responding to this survey is voluntary.
Data are collected directly from survey respondents.
The survey collection period was from January 2018 to May 2018 and was administered by Computer Assisted Personal Interview (CAPI) by interviewers hired from four reginal offices of Statistics Canada. The regional offices are located in Halifax, Montréal, Toronto and Edmonton.
Respondents were mailed a bilingual introductory letter approximately 1 to 2 weeks prior to collection. In most cases, interviews took place at the respondent's place of residence. There were no proxy interviews.
In the case of non-response, either by no contact, missed appointment, or refusal, the method of follow up was three fold. Interviewers made attempts to reach the respondent at home, by telephone, or by follow up letter. A refusal or no contact letter was sent to respondents who have refused to answer the survey or who StatCan interviewers were unable to contact. If respondents were not at home when the interviewer stopped by, a missed appointment letter with StatCan contact information was left.
A response booklet of answer categories was provided to respondents as a visual aid to help facilitate responding to questions where responses are repeated in a series of questions, where there is a long choice of response categories, and for questions and answer categories that are sensitive. Respondents were also given a handout at the start of the interview that provided DND or VAC mental health resources and contacts for support if needed.
The computer-assisted interviewing method (CAI) was used and the questionnaire was programmed in BLAISE. The CAI method offers a number of data quality advantages that are built directly in the application. Question text, including reference periods and pronouns, are customized automatically based on factors such as the age and sex of the respondent, the date of the interview and answers to previous questions.
Second, edits to check for inconsistent answers or out-of-range responses are applied automatically and on-screen prompts are shown when an invalid entry is recorded. Immediate feedback is given to the respondent and the interviewer is able to correct any inconsistencies.
Third, questions that are not applicable to the respondent are skipped automatically.
View the Questionnaire(s) and reporting guide(s) .
Some editing is done directly at the time of the interview. The computerized questionnaire reduces transcription errors and data transmission. Where the information entered is out of range (too large or small) of expected values, or inconsistent with the previous entries, the interviewer is prompted, through message screens on the computer, to modify the information. However, for some question interviewers have the option of bypassing the edits, and of skipping questions if the respondent does not know the answer or refuses to answer. Therefore, the response data are subjected to further edit processes once they arrive in head office.
The editing during processing consist, first of all, to identify and eliminate potential duplicate records and to identify non-response and out-of-scope records. Then, editing consists in modifying the data at the individual variable level. The items from the survey output which need to be kept on the survey master file are determined. Subsequently, invalid characters are deleted and the data items are formatted appropriately. Text fields are stripped off the main files and written to a separate file for coding.
Due to some technical problems in certain skip patterns of the suicide module, some respondents were not asked the question required for the calculation of the derived variables '12-month suicidal thought' and '12-month suicidal attempt'. These two variables have been imputed.
Longitudinal weighting is designed to generate estimates that are representative of the defined target population at the time the longitudinal sample is selected. Then, the estimates based on longitudinal sample are representative of people who were released prior to 2002.
For CCHS-CF 2002, the sample was selected from 5 strata. Therefore, in each stratum, the initial sampling weight was a result of the ratio between the number of persons in the population and the number of persons in the sample. Then, as for CAFVMHS 2018, a non-response adjustment and an adjustment for the out of scope cases were performed.
A longitudinal master weight is given to each person included in the final longitudinal sample, which includes, the sample of persons having responded to the survey and people out-of-scope (OOS) for collection in 2018 but that were in-scope for collection in 2002.
The initial weight for each longitudinal unit is assigned using the final weight from the 2002 CCHS-CF in order to represent the same initial population as in 2002. Weights of the units with an unknown location, which no tracing source or the province of residence are available, are assigned to the units with a known location. Weights of the non-responding units are redistributed to responding units and to those identifies during collection as OOS for collection. Units that are non-respondents have their weights assigned to the responding and OOS for collection units with similar characteristics within response homogeneity groups (RHGs).
While rigorous quality assurance mechanisms are applied across all steps of the statistical process, validation and detailed review of the data are the ultimate quality checks prior to dissemination. Many validation measures were implemented.
Non-sampling errors can be defined as errors arising during the course of virtually all survey activities, apart from sampling. They are present in both sample surveys and censuses (unlike sampling error, which is only present in sample surveys). Non-sampling errors arise primarily from the following sources: non-response, coverage, measurement and processing.
Non-response errors result from a failure to collect complete information on all units on the selected sample. Non-response produces errors in the survey estimates in two ways. First is that non-respondents often have different characteristics from respondents, which can result in biased survey estimates if non-response is not corrected for effectively. Secondly, it reduces the effective size of the sample, since fewer units than expected answered the survey. As a result, the sampling variance increases and the precision of the estimated decreases.
The main method used to reduce the non-response bias was a series of weight adjustments to the survey weights to take in consideration the non-response.
Coverage errors consist of omissions, erroneous inclusions, duplications and misclassification of units in the survey frame. Since they affect every estimate produced by the survey, they are one of the most important type of error; in the case of a census they may be the main source of error. Coverage errors may cause a bias in the estimates and the effect can vary for different sub-groups of the population. This is a very difficult error to measure or quantify accurately.
For CAFVMHS, the frame was updated with information from the Department of National Defence of Canada (DND), census and the Amalgamated Mortality Database (AMDB). Despite these updates, it has not been possible to obtain the tracing information or the province of residence for 30 cases.
Several measures are taken to reduce the level of response error. These measures include questionnaire review and testing using cognitive methods, the use of highly skilled interviewers and extensive training of interviewers with respect to the survey procedures and content.
Data processing for CAFVMHS has been done by steps, including data entry, validation, coding and control. This process permitted to improve processing activities.
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.
Revisions and seasonal adjustment
As the data are based on a sample of persons, they are subject to sampling error. That is, estimates based on a sample will vary from sample to sample, and typically they will be different from the results that would have been obtained from a complete census. More precise estimates of the sampling variability of estimates can be produced with the bootstrap method using bootstrap weights that have been created for this survey. The bootstrap method was used to estimate the sampling variability for all of the estimates produced based on the data from CAFVMHS 2018.