Introduction

The primary prevention workforce comprises two groups with varying roles²:

Practitioners

Those who specialise in designing, implementing and monitoring actions to prevent family violence. Practitioners must understand the drivers of family violence and are engaged in activity that focuses on actions to prevent violence before it starts. This group was the target audience for the Census.

Contributors

Those located within specific sectors or disciplines where participation in preventing family violence may be a part of their role, but is not their primary focus. This group was not in scope for the Census.

For the purposes of the Census, references to the primary prevention workforce relate only to primary prevention practitioners.

The questionnaire (request a copy by emailing the Centre for Workforce Excellence at cwe@familysafety.vic.gov.au) was developed by ORIMA Research and the FSV project team, with content informed by the previous Census. The questionnaire was designed as a single instrument, with screening questions at the start to categorise respondents into one of the three workforces and route them through to the applicable survey questions.

Learnings from the 2017 Census led to the following changes being implemented in the current Census:

Changes to workforce definition to improve data quality.

In 2017, respondents were categorised into four “tiers”. For the 2019-20 Census, this approach was carefully revised and replaced with the three workforce categories outlined earlier in Table 1.

It should be noted that the current Census results for the primary prevention workforce are unable to be benchmarked against results from the 2017 Census, due to the substantial changes in the way workforces were defined in 2019-20.

Extensive consultation with the sector to accurately inform the design and development of the questionnaire.

ORIMA Research and FSV conducted a series of consultative workshops, meetings and interviews with representatives from the target workforces between August and October 2019.

The survey dissemination method and elements of the survey design, including the initial screening questions, were tested amongst the target workforces and iteratively refined.

A consultation summary report was prepared in November 2019, summarising participants’ feedback and recommendations regarding survey communication, design and dissemination.

Sector consultation was made possible with the assistance of the Victorian Council of Social Services, Domestic Violence Victoria and No to Violence.

Ethics approval was granted for this project by the ORIMA Research Human Research Ethics Committee on Thursday 31 October 2019 (Approval Number: 0112019). Research approvals were also granted by the Victoria Police Research Committee and the Victorian Department of Education and Training Research in Victorian Government Schools and Early Childhood settings committee, to conduct research with their staff.

The survey was administered using an online self-completion methodology. As part of the questionnaire finalisation process, a pilot was conducted between Monday 11 November and Friday 15 November 2019, to assess the suitability of survey design and content, and to test the online system and survey length.

Pilot participants were volunteers recruited by the FSV project team, and included individuals representing each of the three workforces. A total of n=16 individuals completed the pilot survey, from a pilot contact list of N=30.

Participants were asked to provide feedback via email. Comments made within the survey were also analysed and feedback was clarified directly with participants as required. Overall, the pilot was assessed as being successful as there were no substantial difficulties raised or improvement suggestions provided in relation to any aspect or question of the survey, and no critical survey issues were uncovered. A pilot testing outcome summary was provided to the FSV project team which detailed some suggestions for improvement primarily in relation to optimising the clarity of some response options. Some feedback was also provided in relation to accessibility.

Following the pilot survey, the online survey was revised to incorporate pilot feedback, and was finalised in consultation with the FSV project team in preparation for the main fieldwork phase. The online survey underwent comprehensive internal testing by the ORIMA project team, as well as User Acceptance Testing by the FSV project team, prior to launch.

The main survey was conducted between Monday 18 November 2019 and Friday 28 February 2020. A small extension was also granted for certain workforces until Friday 13 March 2020 in order to boost final response numbers. Participation in the survey was voluntary, and responses to the survey were private and confidential.

Survey dissemination (via Survey Coordinators)

The Victorian workforce intersecting with family violence is vast, and there is no central or reliable record of contact details for all individuals employed in this sector. Therefore, in order to conduct the Census, ORIMA Research and the FSV project team relied on sector, departmental, and organisational (or similar) representatives to assist in a controlled dissemination of a generic Census survey link.

These representatives, known as Survey Coordinators, were carefully recruited to ensure good coverage of all areas of the workforces that intersect with family violence in Victoria. Coordinators were asked to either email the survey link directly to their contacts, or act as an intermediary, by asking their contacts to share the link to relevant cohorts within their extended network.

A total of 22 Survey Coordinators assisted in promoting and disseminating the Census across the three workforces. Coordinators were provided with support materials to assist them in both identifying in-scope workforces and participants, and also to share to such individuals. This ensured that a consistent and clear invitation and message was communicated across the sector.

Pre-registration

Prior to the main survey period, a Census pre-registration page was set up by ORIMA Research support survey dissemination. This page allowed individuals to voluntarily register their email address to receive an invitation to the survey upon launch.

Response rate

In advance of fieldwork, the FSV project team undertook an extensive data collection exercise to estimate the population size for each of the relevant workforces for this project (see details in the following section). Figures were collated via consultation with various organisational representatives across the sector. Estimated headcounts and/or full-time equivalent (FTE) figures were provided by key occupation groups. These figures were used to monitor response rates and are the basis for response rate figures below.

Overall, a total of 5,021 responses were received for the Census, including 517 from the primary prevention workforce (see Table 2).

As illustrated in Table 2, the number of survey responses received for the primary prevention workforce is greater than the estimated population (147% response rate). This inconsistency may be due to a number of reasons, including:

Population figures that were used to calculate response rates are estimates of the true population size. These figures were collated by the FSV project team via various workforce contacts and do not represent the estimated overall headcount at a single point in time (as different workforce contacts provided figures at different points over 2019). Although the available population figures are assumed to provide a good estimate of the size of the workforce, in the absence of any single and reliable source of data it cannot be known how closely these figures mirror that of the true population.

A key consideration regarding the estimated figures is that there is likely to have been some employee turnover or restructuring of roles / organisations since this estimate was collated. There is also a small risk that the original estimates may not have covered all in-scope areas of the workforce (though it should be noted that the FSV project team undertook a substantial amount of work to ensure all areas of the workforce were covered).

Respondents were responsible for classifying themselves into the correct survey (by answering the screening questions), hence there is a risk that incorrect self-classification has occurred for some. The project teams very carefully designed the screening questions in collaboration with the sector and provided a range of key roles as examples to assist respondents in classifying themselves correctly. However, there is still a small risk that this was done incorrectly according to FSV’s understanding of in-scope roles. This was further mitigated where possible at the data processing stage, with any respondents who were identified as clearly answering the wrong survey (through their role and organisation type) were allocated back to the correct cohort.

It should be noted that respondents’ interpretation of their own role may also be misaligned to FSV’s understanding / classification. Role examples were provided to mitigate this however some respondents may have disagreed with their classification.

Some primary prevention contributors may have also participated in the Census. As outlined on page 3, it should be noted that practitioners were the key target audience within the primary prevention workforce, and contributors were out of scope.

As this survey was an attempted census of workforces that intersect with family violence (i.e. all those in scope for the survey were assumed to have been invited to participate, via either a personalised or generic survey link), the survey results are not subject to sampling error.

However, the survey is subject to potential non-sampling error, including coverage error and non-response error. Unlike sampling error, non-sampling error is generally not mathematically measurable. ORIMA Research uses several strategies to address sources of non-sampling error to the extent possible, including careful questionnaire construction and data processing quality control.

Percentages in this report are based on the total number of valid responses made to the particular question being reported on. In most cases, results reflect those participants who expressed a view and for whom the questions were applicable. ‘Don’t know / can’t say’ and ‘prefer not to answer’ responses are included only where they aid in the interpretation of results. Results presented as percentages throughout the report may not add up to 100% (particularly where displayed in chart form) due to rounding, or where participants were able to select more than one response.

Results for demographic cohorts (e.g. age, organisation type, etc.) are only presented in this report where notable differences are observed. Suppression rules have been implemented throughout this report whereby groups of individuals with fewer than 10 respondents have not been reported on to protect respondent confidentiality.

Please note that all results are self-reported by respondents and have not been verified against any external secondary data.