Change in Injury Rates over time: Our Customers vs. Industry Benchmarks - ON-254

Desired discipline(s): Engineering - biomedical, Engineering, Engineering - mechanical, Computer science, Mathematical Sciences, Economics, Social Sciences & Humanities
Company: eCompliance
Project Length: 4 to 6 months
Preferred start date: As soon as possible.
Language requirement: English
Location(s): Toronto, ON, Canada
No. of positions: 1

About the company: 

In October 2012, supported by a group of successful entrepreneurs from across the U.S. and Canada, Adrian Bartha and his business partner, Josh LeBrun, assumed the reins of a 12 employee online training company from previous owners and began building eCompliance into the leading cloud product protecting the frontline workforce in hazardous industries.

eCompliance is a SaaS business dedicated to eliminating 1 million workplace incidents. We do this by boosting safety particpation on the frontlines through the industry’s best mobile app. By exposing risk, we enable better executive decisions based on solid data.

Please describe the project.: 

Our mission is to eliminate 1 million workplace incidents by 2020. Previous empirical research from our proprietary dataset has revealed that customers with top quartile safety participation reduce 3.5x more incidents then their bottom quartile peers. See white paper here: https://bit.ly/2lWc22a

The results of this study have profoundly impacted our business. We are incorporating the learnings into our new product development and accelerating the number of workplace incidents eliminated. This study was conducted by a third party researcher from Stanford University. The researcher randomly sampled our customers, and interviewed them to collect a history of incidents.  

Study 2.0 can be broken into three parts:

  1. We want to update this study using a larger dataset. The original study had a smaple size of 40. We would like a sample size of at least 100. We have 400 customers.
  2. We want to expand the study to prove/disprove the hypothesis that safety is good business, and the safest companies enjoy better margins and higher profitability thanks to their strong safety record. This would include a study of insurance premiums, revenues and other costs over time and a correlation with incident rates. This data would then be overlayed against our measurement of safety participation from in-app use to prove or disprove the hypothesis that higher levels of participation are leading indicators of safety and financial performance.
  3. We want to continue our machine learning experiments on this dataset by tagging and categorizing the particular activities that comprise Safety Participation, overlaying this data against Participation quartile, and identifying similar or different patterns of behaviour by quartile rank. 

Research Objectives:

  • Test that safety participation is a leading indicator of safety performance
  • Collect good data on safety performance (lost time injuries, incidents, fatalities, medical aids, first aids, etc.)
  • Test that strong safety performance is a leading indicator of business performance (revenue, margins, costs)

Methodologies to be used:

  • Determine statistically significant sample size
  • Randomly identify target clients for study
  • With our support, approach clients and secure their participation in study
  • Develop information collection template and process
  • Suggest other data points that would illustrate the ROI from safety investments and improved safety performance: non-financial (attracting talent, reducing turnover, improved productivity), financial (insurance premiums, litigation)
  • Interview clients and collect information by phone
  • Review customer activities and tag forms for Machine Learning experiments
  • Analyze data and summarize findings

Required expertise/skills: 

The intern will have expertise in the following areas:

  • Research methods
  • Quantitative skills
  • Interviewing skills