Data Analytics

SRDC analyzes survey and administrative datasets to identify market segments, target population characteristics and needs, behavioural patterns, and statistical relationships. The end goal is to inform clients on issues like program design, targeting, and expected outcomes. Using predictive models, SRDC has used relationships between known random variables in order to predict how an event will likely occur in the future.

Current and Completed Projects:

Student Financial Assistance for Indigenous Learners

SRDC is compiling and reviewing available data on access and usage of CSLP and provincial and territorial student aid programs by Indigenous learners. The broad aim is to support assessment of the effectiveness of existing student financial assistance under CSLP as it relates to Indigenous learners. Federal funding for First Nation and Inuit learners under the Post-Secondary Student Support Program (PSSSP) has been capped since 1996, and program eligibility requirements can be limiting. Indigenous learners who do not receive PSSSP – or whose needs are not fully met – can apply for student financial assistance (SFA) from the federal and provincial/territorial governments, as a mix of student loans and grants. The work considers the demographic characteristics of clients identified as Indigenous learners, their aid received, loan repayment, and default rates. The deliverables will aid these jurisdictions' ongoing review of how effectively the system works to improve PSE access and affordability.

Start-end date: August 2017 - March 2018
Sponsor: Employment and Social Development Canada

Analysis of PIAAC Data and Development of Evidence Base on Adult Learning for Ontario

This study examines the literacy skills proficiency distribution among key subgroups in the province of Ontario including recent immigrants, Aboriginals, Francophone, youth, and older workers. The objective is to inform policy direction with respect to resource targeting for adult learning initiatives in the province. Among other sources, the project utilizes the latest data from the Programme for the International Assessment of Adult Competencies (PIAAC).

Start-end date: March 2015 - September 2015
Sponsor: Ontario Ministry of Training, Colleges and Universities

Essential Skills Social and Economic Impacts

Essential Skills Impacts is the first phase in a larger project to assess the extent to which gains in Essential Skills (ES) are associated with long-term gains on a wide range of social and economic outcomes, including health outcomes, involvement with the criminal justice system, and early childhood education. This project uses the Manitoba Centre for Health Policy’s longitudinal databases to link gains in Essential Skills to a series of socioeconomic indicators of interest. As a preliminary step, the study reviews the reliability and validity of the ES assessments currently being used in Manitoba.

Start-end date: June 2014 - October 2015
Sponsor: Workplace Education Manitoba

Manitoba Works! evaluation

The purpose of the research is to test the Government of Manitoba’s Manitoba Works! employment service model in the real-world setting of Manitoba’s new and evolving employment continuum and to assess its effectiveness for improving the labour market success of individuals who are receiving EIA and/or have complex needs. The research will assess gains on outcomes of interest among participants compared to non-participants, model cost-effectiveness, model implementation, and identify key success factors.

Start date: November 2013
Sponsor: Manitoba Department of Jobs and the Economy

Validation of Manitoba’s Employment and Training Assessment (ETA)

The project aims to assess the predictive validity of Manitoba’s new Employment and Training Assessment tool. As well, it assesses the extent to which adding additional information improves the overall accuracy of the model in predicting long-term unemployment and the extent to which a “scoring” method could be designed that identifies “tiers” or segmentation points, that categorize jobseekers into “categories of risk” or “levels of need” using what has been referred to as a “distance to the labour market approach.” Given the results of the analyses, the project will recommend an optimal approach to needs assessment and jobseeker classification in the Manitoba context.

Start date: November 2013
Sponsor: Manitoba Department of Jobs and the Economy

Further Analysis on the Impacts of Needs Assessment Simplification

This project extends work seeking to identify options to simplify student financial aid (SFA) applications in Canada by assessing the scope for limiting or modifying the data elements currently collected to determine SFA offers. The analysis involves data from actual applications and awards. SRDC simulates aid offers to assess the role played by each data element in determining the level and composition of actual aid offers made and runs sensitivity tests for different student sub-groups to assess their vulnerability to the removal or modification of each element. The simulations and sensitivity tests are undertaken at both a national and a provincial level. The main deliverable is a range of simplification options along with the pros and cons of each, including a range of possible effects on program costs.

Start-end date: July 2013 - December 2013
Sponsor: Human Resources and Skills Development Canada

Manageable Student Debt Threshold Research

This project assesses current understanding on levels of manageable student debt. It begins with a review of existing theoretical and empirical evidence, both national and international, and an environmental scan of the existing practices in defining and measuring manageable student debt. Current private and public sector practices on definition, calculations, and measurements of manageable debt are gathered through key informant interviews with private sector loan providers, credit agencies, and debt counselling representatives. Data analysis of individual level microdata from Statistics Canada is used to estimate various manageable student debt thresholds, corresponding to those identified in the initial review.

Start-end date: January 2013 - March 2013
Sponsor: Alberta Enterprise and Advanced Education

Predicting Student Loan Delinquency and Default

This project develops a model to predict student loan delinquency and default based on borrower characteristics using administrative data. The goal is to improve the targeting of at-risk borrowers and improve the efficiency of program resources devoted to reducing student loan default. The work involves documenting the theoretical and empirical evidence to date, a thorough data assessment to identify key variables, statistical modelling of loan delinquency and default, calculating probabilities for different borrower characteristics, and comparing again to the literature to identify discrepancies and new findings.

Start-end date: October 2012 - March 2013
Sponsor: Human Resources and Skills Development Canada