Google Ad #1 - desktop - 728x90

Google Ad #3 - desktop - 728x90

Google Ad #1 - mobile - 300x200

Google Ad #3 - mobile - 300x200

ATD_Half_Page

Googel Ad #1 - Side Bar - Content pages

ATD Footer Logo

Jay Hall

THE USE OF INSTRUMENTS IN LABORATORY TRAINING

At the beginning of the past two years, I’ve written a post on important points I think you should know. In 2021 I mentioned the importance of context, emotion, and learning to learn. In 2022 I discussed some of the newer cognitive science models of cognition—including distributed, social, and embodied cognition—as well as the role of mental models. This year, I’d like to focus on a trend that we should extinguish: the belief that we are formal, logical, reasoning beings.The cognitive revolution transcended the view of behavioral psychology; we couldn’t talk about what went on in the brain. While many valuable insights were generated by that scientific paradigm, including the role of performance objectives, the move to thinking we could isolate necessary components provided further advances. Results that have stood the test of time—the recognition of sensory, working, and long-term memory as well as core cognitive processes like attention, elaboration, and retrieval—were identified.We tend to adopt existing models as frameworks for thinking about the brain, and the cognitive revolution was no exception. For example, consider the computer as a general-purpose information processing system. It made sense to adopt that model to reflect our abilities to process information and complete a variety of tasks. Thus, the dominant framework for thinking about thinking was to view our brains as formal, logical, reasoning machines.Such models, with some assumptions like mapping the world into formal representations, accomplished impressive tasks. Systems built upon that model could break down problems into components and solve them. They could even learn. Artificial intelligence (AI) advanced as a result. So-called “expert systems” were built upon interviews with the best practitioners. Sadly, these models failed in practical situations, leading to an AI winter. This was a disconcerting time, when behavior naturally observed in humans did not reflect these symbolic systems.The Postcognitive RevolutionOne of the most impressive capabilities of our thinking is the ability to organize the world into flexible frameworks. What David Rumelhart called schemas, Marvin Minsky called frames, and Roger Schank called scripts all reflect our ability to infer knowledge about the world. Thus, we had representations such as wheels that scaled up to vehicles but differently for automobiles versus bicycles. Symbolic logic struggles with this, however. Attempts were made to adapt, such as Lofti Zadeh’s Fuzzy Logic and Douglas Hofstedter’s Slipnets, but suffered in practice.An alternative framework emerged that could be characterized as “subsymbolic,” operating below the symbolic level. John Holland, Keith Holyoak, Richard Nisbett, and Paul Thagard created a genetic algorithm approach in Induction, which borrowed mutation from biology, but wasn’t aligned with neuroscience. Rumelhart, along with Jay McClelland, proposed connectionist models instead in their two-volume set Parallel Distributed Processing with a neuroscientific inspiration. These models were found to exhibit rational behavior similar to humans. They represented the same types of behaviors, including errors, that were empirically obtained from psychology research.These neural net models created not only a psychological revolution, capturing elements of cognition, but also a renewed basis for artificial intelligence, underpinning much of today’s work in machine learning.A Cognitive LegacyOne small legacy of the formal, logical, reasoning model of cognition, however, seems to linger. It appears that many of our practices in learning and development (and business in general), still seem to be based on this idea. This is a potential problem.If we’re formal logical beings, there are several things we should be able to assume. For one, if we give people new information, they’ll appropriately adjust their behavior. Moreover, the person with the greatest access to information should be the one to make a decision, and that decision will be the best one possible. We see this manifest in several ways.For one, our training frequently resembles the presentation of new information with an expectation of behavior change. Too often, we see bullet points of information extracted from (and even presented by) experts. After this, there’s an expectation that we’ll do something differently. Similarly with annual or quarterly reviews. People need feedback at the appropriate time and support for change. We don’t just change when we wish to; it takes time and practice.We also see decisions made by individuals without consultation. Folks will decree a new approach because it’s been determined by the person at the appropriate decision point. Yet the evidence suggests that we get better outcomes when we bring more people into the decision-making process, such as in Keith Sawyer’s Group Genius. When we follow practices that help us suitably explore a problem individually, and then collectively brainstorm solutions, test, and adapt, the outcomes are better.A Problematic LegacySadly, there is also evidence that this approach to work isn’t working. We see it in the outcomes of learning investments, and when we evaluate organizational success as an outcome of our approaches, we see that more informed ones work better.In the 2012 Harvard Business School working paper The Great Training Robbery, Michael Beer, Magnus Finnstrom, and Derek Schrader cite metanalyses that suggest only 10 percent of training investment has an impact. There may be multiple causes for this, but it does suggest that we’re wasting money with our approaches. The Serious eLearning Manifesto suggests that much of traditional learning is based on information presentation, so it’s safe to say that we’re not doing as well as we could.Similarly, efforts like Laura Overton’s Toward Maturity and Lauri Bassi’s Human Capital Capability Scorecard suggest that organizations do better when they’re better aligned with how people think, work, and learn. Organizational performance reliably improves with better employee approaches.Overall, the continued use of ineffective practices undermines our organizational success. We need to effectively achieve our organizational goals, which includes ensuring we can optimally execute what we know we need to do and continually innovate when we have to invent or adapt.Moving ForwardWhat should you do differently?Truly understand how your brain actually operates. Know the basics of the human information processing loop and extensions for learning and performance.Avoid the myths that continue to hinder our industry.Conduct deep needs analyses to find important performance gaps and their root causes. There are many causes of problems besides a lack of knowledge or skills. Stop thinking that if you give people knowledge, behavior will change.Leverage external resources. Decision making is best done collectively and iteratively. To do otherwise is to minimize your potential.Move from legacy practices to evidence-based practices. There are vested interests that continue to exist based on perpetuating approaches that aren’t effective. Many products that are quite lucrative—profiling instruments and development programs —aren’t effective.Start measuring impact. This doesn’t mean what people prefer (there’s no correlation with what people think of a program and its impact). Instead, assess the actual change in the organization. You can’t collect this yourself; you must partner with the business units that require change.Align programs to demonstrable gaps, and then measure until they’ve improved.Ultimately, L&D should be the folks that are responsible for not only optimal execution but also continual innovation. When you’re innovating (problem solving, designing, or researching), you don’t know the answer when you start, so inherently it is learning. It’s informal learning, and no one knows the answer (yet)! Still, it’s an opportunity for L&D to attain a more valuable role in the organization.To make this strategic move and start bringing evidence-based practices from formal to informal learning, take ownership of these practices within your teams. Make it safe to fail; share stories of what you tried; and learn together. Create and understand a learning microculture before taking it to the rest of the organization.Last, identify the organizational legacies that are holding you, and your organizations, back. Then, find alternatives, and implement them. When you’re free of your unexamined past, you can move forward to an evidence-based future that means better outcomes. That’s a trend I’d like to see and encourage you to undertake.

A Trend to Supersede

Jay Cross is CEO of the Internet Time Alliance. He may be reached at jaycross.com

Jay Cross

March 2018 - TD Magazine

Book Capsules, March 2018

This article from SHRM.org explores how to verify a job candidate's resume.

Verify Degrees and Protect the Company from Resume Fraud

Marla designs help documents for the software applications her company produces. She sits across from Sean who develops monthly training webinars. Marla would like to “move up” to building instruction like Sean does, but Sean tells Marla that he thinks that the work that Marla does is actually more useful for the company’s users.We could argue the merits of documentation versus training, but there’s always the question of what each user needs at each moment of time. Context is everything.The science of learning is a science that works to understand how we learn and how we can improve learning and instruction. But to be fair, humans have been learning on their own since, well, forever.For instance, if Corinne wants to know how to do something, she often asks the person next to her and then gets on with her job. Yan mentors Morey and they meet most Tuesday afternoons. Marlena uses self-study materials and is going to take the certification exam in less than three weeks.So why do numerous disciplines (neuroscience, psychology, education, computer science, and others) expend enormous energy and resources to figure out how we learn and how to best build instruction? Can’t we just leave people alone?We Learn All the TimeAn interesting study by D.W. Livingstone shows that the majority of workplace learning happens not in formal settings but in informal ones. Numerous studies, including those by Livingstone, Jay Cross, Jane Hart and others, show informal learning as a proportion of all workplace learning at greater than 70 percent.If you haven’t already looked into informal learning, you should, because training professionals need to understand it more. Saul Carliner’s book, Informal Learning Basics, on the subject is a good primer, as are writings by Jay Cross and Jane Hart.You might think that the definitions of learning and instruction are self-explanatory, but actually they aren’t so simple. And they help us see exactly what we are trying to do.Learning, DefinedThere are many definitions of learning. A favorite comes from De Houwer:  “Learning is changes in behavior that result from experience.”Some definitions don’t require changes in behavior, only changes in knowledge. For me, this doesn’t cut it. If you can’t do what you are learning, the learning isn’t quite cooked yet. For example, if you can’t use what you are learning, you are on your way but not there yet.Applied learning requires knowledge and experience with the range of possibilities, of course. As my colleague Catherine Lombardozzi reminded me, people need deep knowledge, not only to follow a script. So if a customer service agent is following a script for handling returns, but she cannot handle the variety of returns, she only knows how to follow a script. She hasn’t really learned the whole returns process.One definition of learning discusses how learning requires the ability to correct errors. Now this is truly deep learning. One of the biggest problems people have is correcting things that go wrong. Misconceptions and mistakes are a big problem in learning. So when defining learning, we may want to determine how far into expertise we are talking about. Are they proficient (competent or skilled) or are they expert?Why Build Instruction?People can and do learn on their own. But we also know that when there are very specific things we want people to be able to do, there are efficient ways to help them. If learning occurs when we interact with our environment, then instruction is the creation of special environments specifically to promote learning.Dr. Richard Mayer, professor of psychology at the University of California, in his book Applying the Science of Learning reminds us we are trying to change behavior as a result of instruction and there are methods to do so that are more effective than others (that’s what the science of learning and instruction are about).This view has very far reaching implications. It means that if we are building “instruction,” we have to impact what people do. That’s a high bar. Although it is may not be practical to expect that everything we do reach that high bar, we ought to reach that bar often.So that leads to two question that we should ask ourselves:How do we infer (assume changes in) learning after instruction?How do we determine if the change has made it from instruction to the job?We will answer these critical questions soon.This is article 1 of Science of Learning 101, and I very much want your comments. Is this helpful to you? Does it answer your questions? Your comments can change the course of this series.ReferencesCross, J. (2007). Informal learning: Rediscovering the Natural Pathways that inspire innovation andPerformance. San Francisco: Pfeiffer.De Houwer, J., Barnes-Holmes, D. & Moors, A (2013). What is learning? On the nature and merits of a functional definition of learning. Psychonomic bulletin & review.Livingstone, D.W. (2001). Adults’ informal learning: Definitions, findings, gaps and future research. NALL Working Paper 21-20010. Centre for the Study of Education and Work, Ontario Institute for the Study of Education, University of Toronto. Retrieved from  Mayer, R. E. (2011). Applying the science of learning. Upper Saddle River, NJ: Pearson.

Science of Learning 101: What’s the Point of Instruction?

categories.lvl0

Search results for 'Jay Hall'

Categories

Refine By