Why Your Evaluation Strategy Cannot Be an Afterthought

In March of 2017, one of my proteges, Stephanie, interviewed me for a school assignment regarding the 4 Levels of Evaluation Model. The following Question and Response exchange has been slightly edited for clarity.

Vivian Bringslimark, President and Owner, HPIS Consulting, Inc., started with a very influential quote from a client of hers that spoke of the importance of evaluation: “When you look at the ADDIE model, without the A, you die. And then without the E, you’ve only just added another course to the curriculum.”

She feels it’s very insightful because the evaluation strategy is not an afterthought. It has to be part of the design so that learners can engage with the material and practice in order to be successful with what they’ll be assessed on. Evaluation, while having the potential to be controversial, problematic, and open to a variety of opinions, has to be added to the agenda of the course so that time can be allotted to implement it. 

QUESTION 1: How do you evaluate a learners’ Reaction (or Level 1) to the training, such as [a] Participants’ satisfaction with the training [b] Participants’ level of involvement (or engagement) in the learning experience [c] Level to which participants’ found value (or relevance) to what they learned?

RESPONSE: While there are problems and flaws with smile sheets, Vivian states that they are most popular, easiest to administer, and results can be quantifiable to generate reports. However, while participants can love the course, smile sheets cannot guarantee effectiveness. And participants may not be able to apply what they learned back on the job, even though they liked a course. Another problem with smile sheets is that the emphasis tends to be on the presenter, whether or not they were on target and delivered the right course for the audience.

QUESTION 2: How do you evaluate Learning (or Level 2), such as [a] Degree to which participants acquired knowledge and skill [b] Participants’ beliefs (attitude) that it will be worthwhile to implement what is learned on the job [c] Participants’ level of confidence in doing what they learned on the job

RESPONSE: Tests or quizzes can measure comprehension, knowledge, and recall, but they cannot measure attitude. Ideally, knowledge assessments should be more situation- or scenario-based and problem-solving-oriented, and they should be written to assess higher levels of understanding on Bloom’s Taxonomy. In reality, tests are not ideal because it takes too much time to develop a proper test. SMEs have a lack of experience in writing what they think are knowledge checks but are actually only measuring recall and comprehension. 

QUESTION 3: How do you evaluate Behavior ( or Level 3), such as [a] Degree to which participants apply what they learned during training when they are back on the job [b] Organizational support which reinforces, monitors, encourages, and rewards the performance of critical behaviors on the job [c] Organizational support for on-the-job learning opportunities

RESPONSE: Vivian said, “I’m in love with Level 3!” She calls this level “transfer” and notes a difference between changing behavior and changing performance, which Thomas Gilbert said are not the same. When you focus only on behavior, she explained, it’s implied that it will show itself observably on the job. The focus has to be on performance (with behavior being an aspect of it). One shouldn’t focus only on behavior and think that you’ll get change in performance on the job.

In the pharmaceutical industry, the ultimate measurement of effectiveness is whether you can do what you learned on the job without error.  The concern is not necessarily what you learned or how you learned it – what matters most is if you can perform your job without making a mistake. 

Vivian mentioned a book written by Donald and Jim Kirkpatrick called “Transfer of Learning” which expanded on organizational support and accountability.  She emphasized something very prevalent when you get to level 3 that is, “most failures of transfer have nothing to do with training.” There are “barriers to the transfer of training” that have very little to do with the quality of training or the quality of the trainer, and prevent the ability for employees to transfer learning on the job. Management doesn’t often acknowledge these barriers but blames them on bad training. [See Article – Training Does Not Stand Alone | Transfer Paragraphs]

QUESTION 4: How do you evaluate Results (or Level 4) or the impact of training on organizational performance?

RESPONSE: Vivian mentioned Jack and Patti Phillips, who happened to write Chapter 30 in our ASTD Handbook.  When I told her that I read about the ROI model, she responded that most business managers (not necessarily executives) are not really looking for the bottom dollar figure because it’s a tremendous amount of work. 

Vivian said that there is a concept called Return On Expectation (ROE), which she found very appealing. Demonstration that objectives were achieved could be a measure of effectiveness. She has become a huge fan of Robert Brinkerhoff’s “Success Case Method” that suggests Level 4 can be measured with qualitative data. 

QUESTION 5: How do you evaluate Return on Investment (or Level 5)?

RESPONSE: Jack and Patti Phillips did a lot of research pointing to a lack of support for evaluation studies and that not everything you deliver is worthy of a level 4 and 5 evaluation. The reason is that it’s very expensive and companies do not have the finances, people, etc. to commit to it. Jack Phillips said only 5% of courses and programs are really worthy of an ROI.

QUESTION 6: Please provide an example of the kind of data you would collect for each level?

RESPONSE: For clients desperate for metrics, the first and easy metric is completion against training requirements. But, she notes it’s a false metric because simply being at 100% in satisfying requirements doesn’t mean you won’t have deviations and errors on the job.

Another metric can be knowledge checks (targeting level 2) such as verbal or paper-and-pencil tracking scores. Metrics can also be obtained through a performance review or performance demonstration (targeting level 3).  Vivian has not seen a solid level 4 package in her specific industry but hopes to see one in the future.

QUESTION 7: Which level(s) is the most important to you? 

LEVEL 3 (see Question 3 Response)

QUESTION 8: Which level(s) is the least important to you? 

RESPONSE: Vivian has seen the use of the Level 2 written assessment abused by serving as a metric for terminating employees. This is not what tests were created for. Level 2 written assessments can lose their educational intention and value when turned into a performance issue under the guise that employees are not getting the knowledge. 

Remember, the goal of the training is for employees to perform their procedures without human errors after they return to work. It’s less about receiving 100% on a knowledge recall immediately after the session is over.

Vivian Bringslimark, HPIS Consulting, Inc.

QUESTION 9: Is there ever a time when training does not require evaluation?

RESPONSE: Yes, evaluation is not supported by employers when they only need documentation that employees attended the training. She has heard an HR director say, “I just need a checkmark. I need to produce training rosters to show the [agency] that we delivered what we promised”.

Vivian implements activities more indicative of transfer (of learning) than a written evaluation such as practice with simulated scenarios. One unique activity has learners develop marketing posters for future attendees and present them in class as their final activity, which is extremely engaging for learners and serves as a measure of effectiveness for the trainer. 

QUESTION 10: What do you think about The Kirkpatrick Model or the New World Kirkpatrick Model? What do you like or dislike and why?

RESPONSE: Vivian admits to mixed feelings about the Kirkpatrick Model. She noted that Donald Kirkpatrick, whom she has met, was intending to write a research paper (for his dissertation), but the model went rogue because people were grasping for anything to make evaluation tangible.  She feels the Kirkpatrick model can be a very plausible solution when management is demanding proof of learning. The Kirkpatrick Model works well when you’re in a classroom setting and you have a course.

Vivian explains the model’s limitations: it cannot be solely relied upon when expanding beyond the course into larger-scale program evaluations for performance improvement. The Kirkpatrick Model becomes challenging to scale up when enlarging the scope of the training programs to include site-wide initiatives.

However, Vivian likes the New World Model’s critical behaviors and leading indicators that were added because it makes the model more practical and even useful in program initiatives. By identifying the critical behaviors, the organization can likely sustain the new behaviors and therefore improve the performance outcome(s).  She said that leading indicators can serve as post points or check-in intervals that are more advantageous than waiting for the end when the ROI study is done.

RE: Assignment 6: Training Manager Interview Comments

1.) Wow, Stephanie, you did a great job with this interview on a very interesting interviewee. So interesting that Jack and Patti Phillips feel that not every course is worthy of a Level 4 or 5 evaluation, and that Jack said only 5% are really worth a level 5. I have never done or designed even a level 4 so I don’t have first-hand experience with the effort and expense involved but this is eye-opening – so it sounds part of the analysis might be to assess whether a course is worth going beyond level 3, making it a conscious, careful decision. This would be somewhat determined by the organization too – it would interesting to know what typical decision points are to conduct (finance) the more advanced assessments.

I really enjoyed this – in fact, I am about to read it the 3rd time!

2.) There is so much great content in your interview summary. I particularly liked the section on how Ms. Bringslimark is “in love with Level 3.” The distinction between changing behavior and changing performance is an important one. It really is all about how you can perform your job, not how you learned it. Over the last couple of weeks, I have also gained a measurable appreciation for Level 3 evaluations. I have been developing training for a workshop and feeling stressed over how to perform an effective level 2 evaluation. The idea of focusing more on the transfer of information and performance in the workplace has allowed me to concentrate my efforts on Level 3 evaluations.

Looking at performance evaluations for performance demonstration also resonates. In a new competency-based training (being developed) both performance evaluations and training need analyses based on the individual competencies will be used to determine transfer of learning and behavioral changes in the workplace. By using both the performance evaluations and the training needs analysis, both non-training and training issues can be determined.

Amazing job, Stephanie. Truly educational.

3.) Excellent thorough interview and post.  It looks like Vivian has built very strong evaluation processes in her design and execution.  

A few comments:

Level One – I agree “smile sheets” have many downsides, however, if they only identify a logistic item (the room was too small for effective learning) or as you indicated about the facilitator (the facilitator wasn’t boring) both of these bring value in the set up for the next training event.  

Level Three – Absolutely agree with her assessment of performance versus behavior.  I thought her take on the barriers to training was incredibly insightful.

Q # 8 – Her answer surprised me about level 2 being the least important. It is very hard to see which ones are least important because they all have value if used correctly.  I have found level 2 is critical to assess whether learning is actually taking place and hard to move on to 3 without a strong level 2 execution. 

Q # 10 – Great way to end the interview.  It was great to hear her perspective on the new world model. 

Again an excellent post.  Thank you.

Do you have an evaluation procedure or a training effectiveness process? Yes, you need to make this formal. Allow me to explain why.

(c) HPIS Consulting, Inc.

Do I really need a Go-Live Strategy for Redesigned SOPs?

Develop Your Go-Live Strategy Before Opening up a Change Control Request

Don’t underestimate what Go-Live will look like the day the new SOPs go into effect.  The overall project plan may reflect one lined task – [Go-Live] and the date, but who is responsible for making it happen?  Doc Control administration hits a button and the eDoc system sends out a notification or perhaps folks find out about the revisions and effectiveness date from their LMS notification. What else needs to be done, you ask? 

Actually, a lot of things and without a detailed plan of assigned tasks or a set of instructions, the first week of Go-Live could be disastrous if not managed well.  Whether it is broken down further in the project plan or contained in the communication strategy, the plan needs to drill down into specifying how tasks and activities will be managed, by whom, and by when. 

5 Basic Steps for a GO-LIve Plan
Steps to take BEFORE draft SOPs are sent to Doc / Change Control

Go-Live: Step 1 – Transition Options

Most quality system redesigns include more than one SOP.  Sometimes the project includes multiple work instructions that are designed to work in tandem.  Releasing one without the other automatically creates a departure from the new SOP.  One approach design teams take is to release the package all at once.  Other teams decide to release serially, one after the other with timed intervals to manage the learning curve. Other approaches include pairs or high priority clusters similar to a wave or a phased tactic.  All have their advantages and disadvantages. 

And all need an impact assessment that examines the details about what will change, how that change will be handled and how many users will be involved.  In other words, how many tasks could be out of compliance when the procedure(s) goes into effect?  Will the release shut down the organization on Day 1?

Now vs. Later – Timing is Everything!

A few years ago, a client received a pretty involved FD-483.  One of the corrective actions was to enhance the Training Quality System.  So, a project team was deployed, process maps verbally approved, and the revised SOPs were in various stages of being ready for a field test.  The design team was concerned that SOP decisions for in-process drafts could impact the final drafted SOPs significantly. The team needed to be able to update a final draft SOP immediately without getting bogged down in their change control process.  Not to mention managing a new round of awareness training that could change a week later!

So, as part of the Go-Live strategy, a decision was made to schedule the release of the procedures all at once when the final set of SOPs were ready to go to Change Control.  It was determined (assessed), that too many little changes for the Training System would result in confusion, retention issues, and a paperwork nightmare controlling which version of the form was correct.  This is a popular option for design teams because the coordination of new changes happens on the same day and is much easier to track. Albeit the learning curve can be high as the site works on closing the size of their change gap.

During the development of the last set of SOPs, FDA conducted a follow-up visit for the FD-483.  One of the training CAPAs was not yet completed. It was waiting for the redesigned set of procedures to go into effect.  The new version would encompass this observation and essentially close it out.   Unfortunately, this incomplete CAPA created a hold for product to be released.  As a result, the priority to complete a retrospective qualification of 700(+) “trainers” derailed the Training Quality System project for close to 4 months.  Additional immediate fix CAPAs were also generated.  Ironically, these CAPAs could have been avoided if the design team chose the serial release option. 

Go-Live: Step 2 – Who Will Do What, When?

Establish roles, responsibilities, and priorities for the transition period.  This means identifying the beginning and end date for the transition from what we used to do to what we will do by what date. When do we anticipate being fully operational with the new procedures?  Is the end date automatically the day the procedures go into effect? Or is there a sequence of tasks that need to be executed before the Go-Live date? Have those tasks been assigned to individuals with clear expectations and due dates?

We’re Still Open During Our Construction Phase!

One client began their Go-Live Strategy with Step 2 to help assess the impact of the changes to her organization which dictated the decision to release each SOP serially; one at a time.  Using a worksheet, each targeted SOP was listed vertically while impact criteria were listed horizontally.  The team need to “see” associated form numbers, cross-referenced SOPs, the actual impact, mitigation steps, how to transition, other impacted documents, disposition status and dates of those other documents and of course next steps. 

With careful examination of what tasks, forms, and steps were actually changing, the Project Lead was able to realistically determine how much time it would take to mitigate the impact.  Subtasks such as final form revisions and revisions to other documents were being monitored and project tracked as well. 

This information shaped a pragmatic and timed sequence of events. Resulting in the conclusion that releasing one SOP at a time was the most effective way to transition into the future state with minimal disruption despite that the overall policy document would have to be revised multiple times during the transition period.  The project lead volunteered to be the taskmaster on those subsequent policy changes and updates for the stakeholder briefings.

What else needs to be done, you ask? Actually, a lot of things and without a detailed plan of assigned tasks or a set of instructions, the first week of Go-Live could be disastrous if not managed well.

Vivian Bringslimark, HPIS Consulting, Inc.

Included in this impact assessment were numerous immediate response CAPA corrective actions.  It was vital to ensure that future state procedures did not eliminate commitments that were made to the agency.  These CAPAs resulted in subprojects that also needed to be managed within the overall quality system redesign project.  Given the scope of these projects, completion dates were projected to be 6 – 9 months out.  Yet, the future state quality system procedures were expected to be operational way before these subprojects came to fruition. 

In essence, the day the SOPs went into effect, the site would automatically be out of compliance and numerous deviations would have to be initiated.  To manage the SOP change gap, an overarching CAPA was approved to allow the organization to meet agency milestones and move forward with the new process while working on the subprojects.   With each planned release of a future state SOP, the impact assessment worksheet was updated with feedback including future release notes.

Catch up on the previous redesigned Quality Systems blogs below:

  • Blog # 1 – Redesigning Quality Systems: Achieving User Adoption
  • Blog # 2 – Manage Your Stakeholders and Users Expectations
  • Blog # 3 – Gap Assessments are Necessary for Redesign Projects but so is the right level of support
  • Blog # 4 – What to Expect When Processing Map with SMEs
  • Blog # 5 – Field Test Your SOPs before they go into effect
  • Blog # 6 – Change Management and It’s Little Cousin Training
  • Blog # 8 – Is an Awareness Training Only Session Enough for Successful User Adoption?
  • Blog # 9 – So, We Went Live, What Happens Next?

Who is Vivian Bringslimark?

© HPIS Consulting, Inc.  All rights reserved.