Emerging Learning and Development Models: Part Four
- 2. In this post we are going to discuss the idea of behavioral currency in training. Behavioral currency refers to designating specific action verbs in learning, such as demonstrate, list, identify or state, as a form of currency and then using this currency to effect behavioral change. Now, without careful differentiation from classic behavioral learning research, the nuances of this system are not readily visible. So we will do our best to explain this system and why we think it’s so important in emergent training systems.
- 3. Classic behavioral learning theory is based upon the following flow: Competencies ——–> Learning Objectives ——–> Instructional Design ——–> Delivery Then we analyze effectiveness using classic Kirkpatrick levels indicated in the chart below: Delivery ——–> Self Evaluation, Knowledge Assessment, Application Assessment, LTOs Kirkpatrick level IV outcomes are learning term outcomes (LTOs), i.e., did training actually produce a change in professional delivery of services? Self-evaluation is used to self-monitor training, knowledge assessment is used to see if the learner mastered content, application assessment determines how the learner puts the knowledge into action, and LTOs, which are very difficult to measure, assess what impact your training dollars had on company sales, solving problems, or other outcomes.
- 4. The term “behavioral objectives” refers to measurable behaviors that result from training. “Understand” is not measurable, we cannot measure understanding. But “demonstrate” is measurable because we can observe it in learners to see if they have learned. So behavioral objectives refer to action verbs rather than process variables such as “know”, “learn about”, “master” or “understand”. So far so good? Okay, now onto behavioral currency. If we designate each behavior in learning as a specific form of currency, let us say, a certain fact has a value of 1, and other facts, which are more important, have a value of 10, then we can weight behaviors according to their intended outcome. Let us use an example. I am teaching you how the heart works and you are a paramedic. So the location of the heart and its anatomy may be given a value of 1, whereas the signs of heart failure might be given a value of 10. Knowing how to detect heart failure and ordering tests to measure it might have a value of 50.
- 5. Knowing how to use diagnostic equipment might have a value of 70 and knowing how to interpret changes in patient status might have a value of 80. So we are placing a value on each behavioral objective to weight it. So far so good again? Okay, next… Now we design a learning system where the total score a learner gets is derived from summation of all weighted learning. Then we can derive a weighted score. They might not know much about the core science, but alot about what to actually do in an ambulance. This in turn is derived from a shift in thinking from knowledge hierarchies to that of situated cognition. That is, that learning is not a recall of first principles, but more a recognition of patterns. So the world of cardiac physiology is so far from the act of stabilizing a heart attack patient that we rarely, if ever, refer to it in practice. It does not mean that knowing how potassium channels work in cardiac tissue is irrelevant, we have just weighted it to be lower than that of knowing how to check blood pressure and other vitals in the ambulance. So we do not use a hierarchy of training:
- 6. Basic Science ——–> Clinical Application ——–> Real World But rather: Real World ——–> Kirkpatrick IV In order to do this and yet weight the background content such as basic science, policies and procedures or response plans, we assign each behavioral objective with a weight. Now that we have weighted the behavioral objective, it can be represented as a form of currency. Currencies can be manipulated to influence behavior in a reciprocal loop. Let us look at some concrete examples. The first is where we assign a “decay component” to a behavior. Suppose the person completes part of the training and gains some weighted points to a score of 10. They then put down the game and fail to return to it within a week. Their behavioral currency decay is set by the designers to be 2 points per week. So if they fail
- 7. to return to the training app, their currency declines to 8, thus incentivizing players to regularly engage in learning. The next component is the “popularity point” system, by which the weighted behavior can gain value by being upvoted by other players. Suppose we put up a challenge problem such as selecting the best option for responding to an emergency at work. We can have users submit their answers and then have other players vote on them. Those answers which gain more votes, similar to “Likes” on a Facebook page, can gain up to 5 additional points, or maybe the instructor can vote their response up. Now the players weighted completed objectives are worth a score of 15 in the game. That indicates that they have been playing regularly, their choices are popular with other players or an authority and that they have completed an objective. This means that behavioral currency, unlike grades or test scores, are dynamic and reflect inputs into the training, such as sharing content or persistent effort. We can also weight behavioral objective packages, such as completing sections A, B and C of a training flow. Those who complete A and B have a lower score than those who complete all three. This means that
- 8. grouped objectives can have a greater value then the sum of individual objectives. So someone who took on learning about the lungs, heart and liver fares better than one who only attempted to learn about the lungs and heart. A pure summative calculation like we see in college courses does not permit this kind of weighting since it can only assign percentage scores. The currency weighted system permits us to reward learners who master more challenging content without penalizing them for making more errors in learning the material that would be incurred if they took on tougher material in a linear grading system. Linear grading systems penalize exploration, because the harder the content, the greater the chance to fail, thus punishing exploration and initiative. By rewarding attempts at higher level content, the learner is incentivized to pursue the material and will not be punished for early failures. Behavioral currency can also be used to obtain items or assets in the game itself. If I have completed 20 questions on how to use Instagram for social media selling, that might provide me with one gem. When you have collected 3 gems, this unlocks new content. Since the currency consists of weighted objectives, our gems are rewarded for good judgment, and not only for factual accuracy. For example, in the question below, we see a typical multiple choice question, one stem and 3 responses.
- 9. The best source of iron in the diet is: a.parsley b.beef c. yogurt There is only one correct answer. But this is not how life works. In reality, parsley has the highest content of iron per gram of dried plant material, but beef is readily available and has other nutrient rich properties. Yogurt is not a great source of iron. So we can weight beef as 10 and parsley as 8 and yogurt as 2. So there is no single correct answer. There are relative answers. Some are better than others. By weighting the behavior (the selection of a best, second best and bad answer), we can evaluate knowledge more holistically. This is not suitable for all situations…surely there is only one answer to the question “what is the best time of year to plant beans?”. But there are relative answers in a question about therapy, such as, “what is the best thing to say to a grieving parent first in an interview?”. By creating a currency around the behavior, we can then weight that behavior according to gradation of responses. This is very difficult to implement without designing the kind of software we have in our toolkit. How does one keep track of weights and more so, changes in weights due to factors like participation variability over time without it? It can be set up so that a correct answer at one point in time becomes the wrong answer later. These take into account relative values of decisions in work or life.
- 10. I trust this discussion has been not too abstract and that it conveys our work in establishing systems of behavioral currency to replace classical evaluation. Once we have “monetized” knowledge in the game in the form of abstracting a currency value for it, we are free to use knowledge in whole new ways. It is our hope that you find this concept helpful in your decision about how to select a sound gamified training system. We cannot live without this feature any more and would dread going back to the days of report cards and static grades. Static grades kind of condemn learners to failure when they do not succeed. Dynamic evaluation rewards context-specific responses and behaviors such as social connection, relative judgement calls and other hard-to-measure metrics in training.
- 11. We continue on our journey of emerging learning technologies next week with a discussion of User Controlled Learning Interfaces. How do we let users control the flow of content, the order of content and their use of content through a smart mobile interface? Let’s find out! Until then, keep on looking up!