Log in
Log in
  • Home
  • October 2019 Event Summary: From Training to Application: Bridge the Gap with Predictive Learning Analytics (All-Day Workshop)

October 2019 Event Summary: From Training to Application: Bridge the Gap with Predictive Learning Analytics (All-Day Workshop)

November 03, 2019 11:59 AM | Anonymous

By Susan Camberis

Editor, Training Today

Most Talent Development (TD) professionals are familiar with the concept of scrap learning – the idea that too frequently a high percentage of learning is lost between delivery and application. 

ATDChi’s October full-day workshop featured Ken Phillips, CPLP, an expert in the area of learning measurement and evaluation.  Familiar to many TD professionals, Phillips is the Founder and CEO of Phillips Associates, a consulting and publishing company that specializes in Predictive Learning Analytics (PLA). 

According to Phillips, “PLA is a methodology for pinpointing the underlying causes of scrap learning associated with a learning program, using predictive analytics and data, so that targeted corrective actions can be taken to maximize learning transfer.”  It involves three phases: data collection and analysis, solution implementation, and reporting results.  Each phase involves multiple steps all directed at the end goal of reducing scrap learning. 

Scrap learning, the opposite of training transfer, is a significant business issue because of the amount of wasted participant and Learning & Development (L&D) department dollars it represents.  Research by Brinkerhoff and CEB (now Gartner) suggests that scrap-learning rates typically range from 45% to 85%.  Brinkerhoff also found that typically 15% of any training population would try to apply new learning, regardless of what we do (or don’t do) as TD professionals. 

PLA “bridges the gap” between training delivery an application by connecting Kirkpatrick’s five levels of learning evaluation.  As shared by Phillips, more than 80% of organizations conduct Level 1 (Reaction, Satisfaction, and Intention) and Level 2 (Knowledge and Retention) evaluations.  About 60% measure Level 3 (Application & Implementation) and about 35% measure Level 4 (Business Impact).  Only 15% attempt to measure Level 5 (Return on Investment).  

“The beautiful part of PLA is having the opportunity to determine what is driving scrap learning,” according to Phillips.  The methodology can be applied to training, conferences, eLearning, and other types of formal training. 

If you are interested in reducing scrap learning in your organization, here are 5 smart ideas to consider:

  1. Focus on high-visibility programs. PLA is not needed or appropriate for every program or learning event. Consider events that are planned (not informal), high profile (either costly or strategically important), and will be delivered to a large number of participants (ideally 40 – 60; minimum 20-25).  You will need enough data for initial data calibration.
  2. Take a baseline.  The PLA methodology starts with calculating the organizational cost of scrap learning, including wasted participant and L&D department dollars.  The baseline includes factors such as number of programs delivered per year, average number of participants attending, and estimated percentage of scrap learning.  Phillips has found that most organizations experience scrap-learning rates close to 60%.  If you are applying PLA to an existing program, you can ask learners and managers what percentage of training they believe is actually being used on the job, as a way to more closely estimate up front.
  3. Understand what contributes to learning application.  The PLA methodology creates a Learner Application Index (LAI) for each learner based on 12 factors that center around learning program design, learning attributes, and the learner work environment (i.e., manager support).  By understanding which factors impede and which ones support training transfer for a given program, you can begin to prioritize which factors to focus on to make improvements. 
  4. Remember Ebbinghaus.  The Ebbinghaus Forgetting Curve suggests that up to 90% of what people have learned is forgotten after 31 days.  When you have LAI data available, you can develop targeted learning reinforcement based on each learner’s needs.    
  5. Determine how you will measure scrap learning.  According to Phillips, there are generally two methods to estimate scrap learning.  The first method and the most popular is simply called “Estimation.”  The ROI Institute purports that estimation can be done using random sample focus groups, interviews, or surveys.  After reminding people what was covered in the training, you ask people to estimate the percentage of program material applied back on the job (0 to 100%), the confidence level of their estimate (0 to 100%), and what obstacles got in the way of applying what they learned.  The second method is to use Level 3 evaluation data (Application & Implementation), however, these results are not always available, and, if the data is not numeric, will not be usable. 

If your organization is not actively addressing scrap learning, consider learning more about PLA so you can help lead the way.

To learn more about Ken Phillips and the PLA method visit:  


CARA Group Logo

Phone: 872-228-7476  |   Email:
Privacy Statement
© 2014-2024, ATD Chicagoland Chapter
Powered by Wild Apricot Membership Software