Shared Data Environment

The Breaking Down Barriers Shared Data Environment (BDB-SDE) addresses the need for accessing better data that will allow for a deeper understanding of the extent, nature and causes of socio-economic disadvantage. The environment will bring together data from a range of sources to enable deep analysis of core issues and the testing and evaluation of ideas to address community, household, and individual poverty.

Four pillars

Built around four pillars, the Breaking Down Barriers Shared Data Environment (BDB-SDE) aims to make it easier and faster to link analysis with practice and policy.

  • Access to data

    There is currently a large array of data generated by Government, private sector, non-profits and individual citizens. Many datasets, however, remain difficult to access, and are thus underutilised by analysts to guide critical social and economic policy decisions and research.

    Access to these datasets is key to developing a deep and comprehensive understanding of the drivers and issues surrounding disadvantage. The BDB-SDE aims to offer access to this data that will give more researchers, analysts, and policy makers information for better data-driven decision making.

    Access also requires a respect and adherence to the sensitivities and private nature of unit record data.  The BDB-SDE will be housed and accessed through a secure data laboratory (MIDL: Melbourne Institute Data Laboratory) which has been certified at a level equivalent of Commonwealth government protected status.

  • Curation of data

    Often when working with administrative data, although ready for use it may not be easily understood by the analyst. The Melbourne Institute’s Data & Analytics team have the capabilities to curate data and develop programs and additional features to assist in the understanding of datasets within the BDB-SDE. This feature aims to allow for faster and more effective analysis.

    The BDB-SDE will also provide approved users with a range of features including:

    • Memoranda, codebooks, and other materials to assist the users in their understanding of the data.
    • Access to a sub-sample data set (what we refer to as a “research ready” and/or “working data set”) that contains the popular variables amongst similar researchers.
    • Access to programs used to transform and create sub-samples of data sets including the above mentioned “research ready” versions of a dataset.
    • Access to memoranda, sample code and derived variables other approved users (within the same SDE) have shared from prior analysis.

    This model sets the researcher up for faster analysis of the data, saving them significant time on data cleaning and allowing them to jump straight into vital analysis leading to deeper insights and stronger collaborations. Additionally, the development of shared Memorandums, codebooks and programs developed by MI staff increase the usability, understanding and ease of access to datasets within the BDB-SDE.

    Approved users can also contribute to this shared body of knowledge which is housed inside MIDL’s information management system (or MIDL Wiki). This unique feature ensures that the evolving and growing body of knowledge will continue help researchers and analysists into the future.

  • Research and Analysis

    By housing data in a single secure environment and undertaking curation of these data, researchers and analysts will be able to undertake deeper studies and evaluations that focus on assessing the key drivers of disadvantage and how policy and/or practice may affect and improve disadvantage in Australia. The BDB-SDE enables an information sharing ecosystem that enables researchers to advance their analyses based on work undertaken by others in the same field and/or using the same data sources.

    The “shared” component of the BDB-SDE is a critical feature in that it encourages cross-disciplinary and broader collaboration across a range of stakeholders. This collaboration will result in better curation of data, easier linking of measures across data sets, and shorter periods between question development and analysis and reporting.

    The BDB-SDE will also lead to increased studies and evaluations of the issues that underpin disadvantage.  In addition, through the development of the BDB-SDE, it will be easier to develop relevant data visualisations and enhanced capabilities that will allow increased engagement with service providers and policy influencers that will also permit better evidence-based decision making.

  • Connection with stakeholders

    The BDB-SDE is a unique environment that will drive a deeper understanding of the drivers that force many Australians in and out of disadvantage. Its focus on sharing knowledge and working together to solve and identify key themes and problems will create a much-needed connection between data and analysis with decision and policy makers at a state and national level.

    This connection with stakeholders is critical to increasing data-driven policies and programs to better service those in disadvantage and thus and enabling a positive change and step towards eliminating entrenched and persistent disadvantage in Australia.

Governance

The BDB-SDE will be governed by the Melbourne Institute’s robust data governance policies and procedures including processes around data protection, user training and all other safeguards to ensure the effective management of data in the environment.

Thorough processes are in place to develop appropriate agreements with each data custodian. These agreements will ensure that the data custodian will be engaged at all points throughout the process and have full oversight and approval of how their data is being accessed and used. Additionally, the data custodian will be required to approve all testing of the data ingress and the clearance and vetting of data from the environment as per the data sharing agreement.

Over time new data sets will be added to the BDB-SDE. When this occurs, approval will be sought from each data custodian to combine the new data set with their existing data in the environment. Data custodians will have full control over which datasets can be accessed together and at what level. Addendums will be added to their individual data sharing agreements based on these decisions.

All users must undergo training and complete an authorisation process prior to data access. They will also be required to sign a User Agreement that contains a confidentiality agreement that will stipulate their requirements to maintain data privacy and confidentiality whilst using MIDL and the BDB-SDE. MIDL and the BDB-SDE ensures all governance processes meet the Five Safes Framework to effectively manage disclosure risk.

Inside the secure environment

The BDB-SDE will sit within the Melbourne Institute Data Lab (MIDL) a secure, purpose built, data enclave that enables virtual access to micro-level data for curation, analysis, and visualisation. MIDL has the ability to house multiple shared data environments which can be described as large ‘super-projects’ that can involve individual users or teams working on projects with either single or multiple datasets albeit on a specific research theme. These large workspaces can see multiple data sets combined and analysed safely inside the environment and allow users to undertake deep study of critical issues important to Australian society.

The key services of MIDL include:

  • A virtual computing environment with an increased security posture to host a range of sensitive data assets to approved users for the purpose of research that informs Australian economic and social policy. Data sets may include assets from the Australian Federal Government, state and local governments, service providers and industry. Datasets with security classification up to PROTECTED (or similar in cases where data custodians are outside Government) may be stored and hosted for approved users through MIDL.
  • Safe and secure processes to protect data assets of data custodians minimising risk of disclosure.
  • Access to a secure information management system (MIDL Wiki) from inside the environment which provides an increased value proposition for both researchers and data custodians. This Wiki page provides users with additional data documentation, “research ready” and a “working data set” versions of data assets as curated by research staff at the Melbourne Institute enabling faster research.
  • Ability for researchers to use/combine a range of data sets for analyses fitting a specific research theme through a Shared Data Environment (SDE).
  • Access to additional support, data services and engaged research opportunities using Melbourne Institute research and professional staff.

MIDL is serviced by a team at the Melbourne Institute with technical services provided by Cyconsol and the University of Melbourne Business Services. All staff involved in the project undergo security checks and are bound by non-disclosure agreements. Personnel in the Melbourne Institute Data & Analytics team providing data support will hold appropriate security checks and/or security clearances as arranged between the Melbourne Institute and the data custodian.

MIDL and the BDB-SDE has additional security controls in place for reviewing and extracting files from the environment. All data leaving the BDB-SDE will be vetted and approved by the data custodian to ensure no sensitive data are released. These controls may vary between datasets as agreed within data sharing agreements.