Bringing Better Science to Workforce Learning with Dr. Tessa Forshaw
How can we improve workforce learning?
This is the main theme for our episode today with Dr. Tessa Forshaw. She's a cognitive scientist, faculty chair of the workforce and workplace learning concentration, and director of the workforce learning and innovation initiative at the Harvard Graduate School of Education.
In this episode, we talk about the basics of learning, common mistakes designers make, keeping up with AI, shifting from content production to performance, and why learning myths still persist.
👉 Mentioned Resources:
👉 Recommendations:
📙 My Book
👉 Connect with me:
👉 Support the Channel!
A tremendous thank you to our sponsors! By supporting them, you support this independent podcast.
The Instructional Design, Engagement, and Support (IDEAS) website provides information and resources to help all members of the UMass Amherst community with online teaching and learning technologies.
Brightspace by D2L
Try D2L Brightspace for 30 Days. Get instant access and see for yourself why millions love Brightspace
Transcript
Luke Hobson Hey, Tessa, welcome to the podcast.
Dr. Tessa Forshaw Thanks so much for having me, Luke. How's it going?
Luke Hobson It's going, it's going. I am ready to dive on into the world of workforce learning, by the way. So I can't wait to talk more about you, what you're up to, and everything else of the sort. But before I get ahead of myself, would you mind introducing yourself? Tell us a little bit more about who you are and what it is that you do.
Dr. Tessa Forshaw Yeah, absolutely. So my name is Tessa Forshaw, and I'm based at the Harvard Graduate School of Education. As of July 1, I will be our inaugural chair in Workforce and Workplace Learning, as well as our faculty director of the Workforce Learning and Innovation Initiative, which is a new thing that we're embarking on here at HGSE. Although I think it's actually just officially naming something that was already there in disparate parts and bringing it together and giving it a bigger voice and platform.
Luke Hobson Absolutely. Well, fantastic. Fantastic. Well, speaking of that, that's definitely where I want to go first — to learn more about this new initiative and everything that you're putting together over at HGSE. And then in the email, when you first reached out saying what you wanted to come on the show and talk about, you mentioned one line and I made sure to save it because I was like, that is a great quote. You started off the email by saying: most workforce learning is not built on what we actually know about how people learn. And as someone who lives and breathes and dies by everything focused on adult learners — yeah, I couldn't agree more. We don't take into consideration all the things that we actually should at this point in time in 2026.
And as this was all happening, it just so happened that at work, I had to go through a training that was some type of cheesy voiceover narration, and then at the end there were multiple choice questions — "tell us what you learned." And I'm like, nothing. I don't know what I just did, but I'm trying. How did we get to this point in time of this "check the box" mentality, especially when it comes to employees and adult learners?
Dr. Tessa Forshaw Yeah, I think there are a few reasons. One is that, unfortunately, when we say "learning," I think a lot of people in learning and development and corporate learning think "content." They assume that learning is not a neurobiological process — which it is; it's something that has to happen inside the learner — but instead that it is content. And so if you give people learning content, they will therefore have learned, because they've been through the content. That's my working hypothesis. What do you think about that, Luke? You're just as deep in this as I am.
Luke Hobson It drives me up the wall, because I understand that even the concepts of learning, training, performing — all of these words get kind of mashed together, but they're not. They're different. And then of course, from the HR perspective, it's always wild to me when we have instructional designers sitting in HR, because I'm like, you should know better. You're in the thick of things. You should know that just showing someone a cartoony thing at the end is not going to produce any tangible results. But of course there's no action, there's no follow-up, and a lot of it is just covering themselves from a legal perspective — like, you did this training, so don't sue us. And it's like, yeah, but did you actually want me to get better at something?
Which is always striking when there is a real type of training and it is different. You're like, wait a second — I'm not just clicking the Next button every three seconds. This is actually valuable. Why didn't you make it a big deal? Most trainings I've gone through never actually make a thing of it when it matters. It's always just sitting in your inbox like, "Hey, do this when you can." But when you take a good one, you're like, why didn't you build this up? Market it, hype it up. It was amazing.
Dr. Tessa Forshaw Yeah, absolutely. And the other thing I see a lot, related to that, is folks getting what I'd call orders from the business — from their business counterparts, or the CEO, or an engineering leader, whatever it is — and they'll give an order that says something along the lines of, "OK, so in three weeks we have a company offsite, and there are four hours on Tuesday where I want you to teach the team innovation. At the end of it, they all need to be innovative." And most of the learning people I know are like, that's crazy. But I think one of the challenges is that everybody went to school. These CEOs who are amazing at leading a business — and "CEOing" is not a job I could do — they actually think they know so much about learning just because they went to school and learned something once. But there is, as we've discussed, a real science behind how we learn. I'm seeing this moment, especially right now with the broader context of AI and reskilling and job disruption and needing to be able to learn quickly and be adaptable and resilient — I'm starting to hear leaders in L&D finally saying, "I actually can't take your order because it's the wrong thing." I think this phenomenon of pushing back is new, and I hope it gets stronger.
Luke Hobson Yeah, absolutely. I had a conversation with my team the other day and it finally dawned on me — because we're differently structured — that I told them I was actually empowered to say no when it comes to designs, especially with very tenured faculty here at MIT. I look at something and I'm like, this is not right, and of course I push back politely and appropriately. But they did not have that same luxury. So when we're talking about the effectiveness of their designs, it's like, oh, so it's all faculty-driven and you have no say? That's not good for learning design. No matter what, the poor people on that team — what are they going to do? They're fighting with one hand tied behind their back right from the start. It's just not the optimal way of doing things.
Dr. Tessa Forshaw Yeah, that's a really tough dynamic.
Luke Hobson You described learning a moment ago as neurobiological. And from doing research on you and seeing more of your work, I've seen you describe learning in the past as cognitive, relational, situated, effortful, cultural — quite a few different dimensions. That's a lot for us as learning designers trying to put everything together and think about all of them. But one of the things I was thinking about: are there any areas in particular where perhaps we're not focusing enough time? To give you an example — one of my graduate students recently from the University of Miami came back to a class I'm teaching right now and gave a talk on Indigenous learning design. And it was fascinating because he's currently in Alaska, and he said, "Let me share more about what it's like being here." I had no idea. Zero clue. And he had me going into this deep philosophical line of thinking around how I actually assess knowledge from different cultural perspectives — where I have one lens and they have another. So from your perspective, are there certain dimensions that we as learning professionals should be thinking about more, or that we're not currently investing in enough?
Dr. Tessa Forshaw Yeah, that's a great question. I think when it comes to what I always start with — it's neurobiological and it's cognitive. I have to mention it, even though I do think a lot of learning designers know this, but at the end of the day, learning only happens when we have processes of neuroplasticity: the strengthening of connections or the creation of connections between two neurons. There are things we can do that enable that to happen, but there are also things we can do that can shut it down. An example that I think we don't talk about enough: if your learner is hungry and they release hunger hormones from their stomach and their brain, the brain deprioritizes neuroplasticity as a result of those hunger hormones. Same with cortisol — if they're really stressed. Same with sleep deprivation, and several others. There are things in an environment and context that can stop that neurobiological process from happening.
So I think what we don't talk enough about in learning design — especially in web-based or asynchronous learning design — is that you don't necessarily have the ability to control the environment and context the learner is sitting in. But what we do have the ability to do is arm the learner with self-knowledge: if those things are true for them, they can adjust their situation or context to better learn. I think that's something we don't discuss enough. We'll put content out there, but we won't support the learner to be metacognitive — not just about the things they learned, but about their process of engaging in learning: how they're showing up, what they need to adjust or tweak. Are they sitting somewhere really noisy where they keep getting pulled in and out of their attention gates? For me, that's one really big thing I'd love learning designers to think more about: how do I support the learner to adjust their context and environment when I can't control it because it's virtual?
Luke Hobson Absolutely. I love it when you're at a conference sitting in a chair for two hours and a speaker comes up and acknowledges it — like, "No one has moved in hours. Everyone stand up, talk to your partner next to you for five minutes, then come back." And I'm like, thank you. Thank you for allowing me to reset, to do something different and get my mind ready again — especially from those physical perspectives you're talking about. I currently have an 11-month-old baby, by the way, who doesn't sleep. So if you have any tips on sleep, I'm in.
Dr. Tessa Forshaw Amazing, lucky you. I have a nearly three-year-old and a six-year-old, so I can only say I've been there — and yeah, learning would be pretty hard for you right now.
Luke Hobson It's a bit tricky. There's a lot of caffeine in my system just trying to stay awake and somewhat focused. It's a little tough. But it's absolutely worth it — it's been such a blast. And it sounds like it just moves on to the next thing, so it's like, no sleep, but now let's talk about this other crazy milestone. Right. OK, here we go.
Dr. Tessa Forshaw It does end, I promise.
Luke Hobson It is also fascinating — speaking of learning and children — when they hit a new milestone that changes almost overnight. And then their sleep changes depending on that big moment where you're like, you had a different night's sleep because you learned how to stand. It's crazy.
Dr. Tessa Forshaw Yeah. We talk a lot in the context of the neurobiology of learning about metabolic energy. Your brain uses 20% of your metabolic energy, even though it weighs only about 2–5% of your body weight. It's really outpunching itself. And that's even higher for little kids doing all of that learning. So when you really think about it — yeah, no wonder they had a bad night's sleep. But sometimes we distance ourselves from even the science of what's happening to a baby. Of course they're hungry — they've been working really hard, using more metabolic energy, so they need more food. Sometimes we forget to connect those dots.
Luke Hobson Yeah, just a tiny insignificant thing. That's fine. Everything's fine. So, absolutely. Sticking with the theme of learning sciences — transfer is one of the hardest problems to tackle in this entire field. What is the most common design mistake that you see that kills transfer right from the start, before it even has a chance to begin?
Dr. Tessa Forshaw Yeah, so I have so many answers to that question. I'm going to go with three things. The first thing I want to share is that learning experiences like the one you described — clicking through something with a robotic voice — are actually what transfer was studied in for years. They would give a learning experience akin to that, or maybe a teacher would come in and talk for a bit. Then 10 minutes later, they'd give you a task and measure whether you could transfer something learned in abstraction to a different problem — not too dissimilar from what you described, just with multiple choice at the end. And lo and behold, training transfer was considered elusive for about 100 years. There are whole books written about it — Transfer on Trial, The Myth of Transfer. There are scholars convinced transfer isn't even a thing, which is really a profound thing to think about.
But what really changed my mind about transfer was a study done by Dan Schwartz, who is the current Dean of the Stanford Graduate School of Education, around 2006. Students were studying a bald eagle enclosure ecosystem. They weren't mapping what they'd learned in that context — they were drawing on 13 years of lived experience up to that point. And the researchers realized: we can't actually control where transfer is coming from when we give someone a novel problem. What gets cued forward for a learner depends on the emotions they're feeling, the colors in the room, the environment, who they're with, the subject matter. We all have different archival memories with different associations, so different things will come forward for different people.
With that in mind, one of the best things you can do to increase the likelihood of transfer is not to focus so much on "A must be mapped to B," but instead to help learners build associations between their existing knowledge and all the ways A could be used. A great example: if you're doing a coding bootcamp, instead of just saying "go use this language," you could say, "What you've understood here is a symbolic system and a logical way of thinking. This way of thinking can show up here, and here. There are other symbolic systems in the world, and this way of thinking is useful for approaching them. You could use this to do this kind of job, or this kind of job, or on this kind of task." That notion is called expansive framing. It brings all of those associations into working memory and creates links between new learning and other contexts. That way, when those other contexts are cued forward in real life, this new learning is more likely to tag along.
So the biggest mistake we make with transfer is assuming it has to be: we learned A here, therefore we apply A to B here. Instead, anything you've ever learned can be applied to B, and A should be set up to be applicable to B, C, D, E, F, and G.
Luke Hobson It's always interesting with adult learners — we just assume they have no prior knowledge. Like, you know nothing, we're wiping the slate clean. No. What are you talking about? I say that to my EdD students all the time. I'm like, we're going to build on what you know, what you've done, and then think about designing learning experiences from there. It would be so foolish of me to say, OK, I need to start from the very beginning of time. Let's go.
Dr. Tessa Forshaw And I know nearly no one who does build on prior knowledge, which is just wild. That phrase "you can't teach an old dog new tricks" really bothers me. Sure, there are some things in terms of learning that we get a little slower at as we age, but well into our 60s and beyond, the wealth of knowledge we have means we are so able to connect new knowledge to existing knowledge — to start from a place of being asset-rich. And I think that's a really important message right now with all the reskilling conversation in the news every day. Job disruption — just because your job goes away and you're made redundant doesn't mean you are redundant. You have an entire lifetime of skills to transfer forward. We should be supporting learners to facilitate that transfer, rather than saying, "You were an auto worker and your job was made redundant, therefore you have no skills — start at zero to learn this completely unrelated new thing."
Luke Hobson And now of course we have to talk about AI, because we've made it 20-something minutes before getting there — a record for any podcast I've done recently. Even speaking to my brother-in-law, who is an electrician — when I asked him, "Have you used AI for anything?" He said, "Nah." But now it's like, "Can I show you what I can do with Claude? I've been scanning things and asking how to fix this problem. It's crazy." Of course, I still need his expertise, because I'm not going to touch something and just assume what the AI gave me will solve the problem without blowing a fuse or blowing up my house. But to be able to say — this might be your sidekick going forward, for jobs where it's just you and you can't figure out a problem. Now you've got another buddy. And of course, as we've now seen with many folks, even within academia — people who perhaps weren't so keen on learning something new — they absolutely can upskill and reskill in today's world. That whole phrase about not being able to teach old dogs new tricks is so silly, by the way.
Dr. Tessa Forshaw Yeah, it's crazy.
Luke Hobson I have an older dog, and I have taught him new things. So there's your evidence. Now — how are you positioning AI learning experiences today? Because many stakeholders and employers I've been talking to don't know what they're looking for, but they're saying, "I need my people to know about AI." I feel like that's evolved too — we've been making AI courses here, and originally it was about machine learning to big data to analytics to generative AI to agentic AI to RAG and everything else. How are you keeping up, and how are you trying to position AI for today's workforce?
Dr. Tessa Forshaw Yeah, great question. I've been doing quite a few studies recently on AI coaching agents. I think one of the reasons I'm interested in that space is because that seems to be the use case I hear about most. People are using LLMs a lot independently. I'm hearing phrases like "ambient coaching" — coaches embedded in everything, popping up. This idea that they'll replace the need for a development manager, or L&D, because you just have coaches who can give you everything in real time all the time. My colleague and I decided to study this directly.
We had a really effective coaching intervention for leadership skill development in managers — happening in the flow of their work, with a human coach, based on the International Coaching Federation's methodology plus a few specific practices we had identified. It's been quite successful. So we created an AI coaching agent that was meant to completely instantiate this human coach. It had to ask the same questions in the same way in the same order — really trained to be that specific human, trained on the human's approach as well as what the human was trained on. We weren't trying to build the best AI coach we could; we were trying to replicate this particular human as closely as possible.
We ran the study with consultants at the manager level — that point in consulting where you start to really develop managerial skills for junior people below you, but you're not yet at the senior manager level managing large, complex things. And what we found is that on the performance tasks — a simulated onboarding session — there was really no difference in performance between those coached by the AI and those coached by the human. But they were required to use the AI coach. When we spoke to them about it, those who had the human coach said things like, "It was the best coaching experience I've had. They asked such good questions. They really pushed me. They wouldn't let me go until we got to the juicy bits." The same people about the AI coach said things like, "It was a sadistic kind of therapy." They said, "I wanted it to stop asking me questions. It just needed to give me the answer. Why did it ask me so many questions? I just wanted a script."
So what we took away from this is: even if you make someone go through that AI intervention and they come out in a comparable place, no one would choose to do it voluntarily. They think it's crazy and they never want to use it again. You're in this catch-22 where you've built something that works because it's Socratic and based on good coaching principles, but nobody wants to use it. And I think we've seen that with other Socratic-style AI products that have since been sunset. In fact, a colleague of mine at the MIT Media Lab, upon hearing this story, said to me: "So Tess, what I'm hearing is that Socrates wasn't killed because he was corrupting the youth — he was killed because he was an LLM that was driving everyone mad." And I was like, yes. I think that's it.
Luke Hobson Yes, that's it. Amazing. It's fascinating. That reminds me of an experiment at Georgia Tech — this was pre-AI in terms of how people were thinking about it — where they disguised a chatbot inside an LMS as a peer tutor. There were two real peer tutors, and then they put in a third one powered by IBM Watson — that was the whole Jill Watson project. They didn't tell the students. It was set up around a computer science course, so it fit the context. And the students thought, "Wow, Jill is an incredible TA." But eventually they started to notice something was up, because they'd message her at 2 a.m. or 3 a.m. and she'd respond immediately, every time. They thought, "That's a bit weird. But maybe she's just phenomenal." Sure enough, at the end some students were trying to nominate her for course TA of the year — and then they revealed it was AI. And they were like, "Wow. It can actually be wonderful when done correctly." And it's only evolved from there. But I would bet that if they'd said upfront, "Here's the AI tutor and here are the two real people," the students would have said, "Give me the people."
Dr. Tessa Forshaw Yeah, totally. And I think there's a human premium emerging with learning — there are some things that still require humans to hold the space for. Like in the coaching I just described, that's a lot of internal work; leadership coaching involves what we'd call productive struggle, and that still requires a human to hold it. And when you step back from that finding and place it in the context of 150 years of educational literature, it's not that surprising. We're now talking about, in different language, Vygotsky's Zone of Proximal Development as it was originally discussed. We're talking about Bronfenbrenner's Dynamic Dyads. Stuff we've known about for a really long time — we're just now seeing, again, that the human part was quite important. That isn't to say AI isn't great. I think AI has tremendous potential. I just don't think it lies in replacing the role of a human.
Luke Hobson Yeah. The role of AI in learning design has fascinated me and will continue to fascinate me as I experiment with things. For some things I'm like, wow, this is incredible. And other cases honestly terrify me — because now we've given some folks who don't fully understand the power of this technology a button that says "generate 20 questions for my class" and they just upload it like that. And you're like, can we talk about how we construct questions? But you hear from different organizations saying, "That's what we want," and companies saying, "Great, we'll build that and give it to you." Of course there's a lot more to it than that, which is something I'm very closely monitoring and trying to share with people.
But to put a positive spin on things — I've been able to create simulations that would have cost tens of thousands of dollars, but now I can make on my own and give directly to learners. Making websites and apps like that. Just insane. What is your favorite instance right now of using AI for good, for learning professionals?
Dr. Tessa Forshaw I think one of my favorites in the L&D context is the ability for facilitators to have rich data about the people in their classes that they can access in real time. In the context of a leadership development paper I've read recently — not my own work — they had user dashboards built on HRIS data, LMS data, performance review data, 360s, and a range of other assessments. And they used that data to help the facilitator find the moment of transfer for each individual learner. So they could walk up and say, "Luke, sounds like this is a tough moment. Let's chat. How are you approaching it? Didn't you do something that involved project planning for a private equity situation? Why don't you think about how the skills you used in designing and executing that could be really helpful in what you're supporting your team with right now?" Those kinds of bridges do what we talked about earlier — they bring forward things the learner already knows into working memory and give them access to it. That's kind of incredible, because to do that before you'd need to sit with a learner, understand their entire work history, their personal history, every extracurricular — to find the relevant connections. Now you have dashboards that give you a sense of all of that. It's amplifying the human premium, if you will.
Luke Hobson Absolutely. The more data we have, the more learning analytics we have, the better we can design learning experiences — especially personalized ones with different learning pathways. Many years ago I taught as part of a MOOC for an instructional design principles course, and the team did a phenomenal job building out different learning paths by sector. When you said, "I'm going to be a corporate L&D person," it went into a corporate pathway focused on readings, discussions, and a mini cohort of corporate L&D folks. And if someone said, "Actually, I want to do higher ed," they could flip the switch and go into a higher ed mode. Back then, that was all the team's incredible work building things behind the scenes. Now I'm thinking — we could do that with AI. And at such a faster rate. Sectors of sectors within higher education. You could just keep going deeper and have it speak to each individual as personally as possible. Crazy to think about.
Dr. Tessa Forshaw Yeah, completely. It is crazy.
Luke Hobson Now let's talk about the dark side of AI — because we do both work in higher education. One of the most requested webinars I've been getting for the past year has been: "How do you rethink assessments in the age of AI?" Specifically, making something "AI-resistant" is the key phrase I keep hearing, which is interesting. With the rise of agentic browsers — where people can share screen access with something like Perplexity's Comet and it's actually completing their coursework inside Canvas — we're like, yeah, that's a problem. How do we rethink assessments when AI can literally do everything? It's not impossible — there are many ways: journals, reflections, teach-back sessions, walkthroughs. But what's your take on rethinking assessments in the age of AI?
Dr. Tessa Forshaw Yeah. I start in a slightly more esoteric place with it. Looking at what just happened at Princeton this past week — removing the honor code for folks who haven't followed it and now making every exam proctored, which hasn't been the case in hundreds of years. And I think there are a few things going on. One is that humans are always going to optimize for the thing that requires less metabolic energy — that's a natural bias the human brain has. So you really have to think about positioning the point of the learning and the point of the assessment, and not just have people do things because you said so.
I say that as a parent who has a six-year-old and sometimes just wants her to do things because I said so — and as an educator who sometimes wants students to just do what I ask. But if they don't see a clear value proposition, if they don't see that they're going to get something of value from the interaction, then no matter what we do, we're going to keep finding ways to overcome restrictions — with the next technology and the next one after that, all the way to quantum. We're just going to keep finding ways around the guardrails.
What I'm trying to say is: I agree that these technologies have made cheating easier, and that's really bad — not just in universities, but in the workplace too. People having someone else take remote control of their computer to do their job while they sit there with a cup of tea. People cheating in coding assessments for engineering roles. It's widespread and common. But that says to me that the underlying problem is that people don't necessarily understand the value of the exchange.
So in my own classes, the strategy I've been using — and it's early, only two years in, so I can't guarantee it will keep working — is to really talk about what is the value of this assessment and why are we doing it. It's a portfolio assessment. I teach creativity and innovation, specifically how adults develop creative and cognitive skills. We do a lot of creation, divergent thinking, portfolios, solutioning, prototyping. And I talk a lot about how this is a practical thing you can actually show people to get a job — the ethical implications of it not being your own work if you ever want to use it. We have a real discussion about why this is the assessment, what the guardrails are, what it looks like. Sometimes students offer up alternative assessment ideas, and I genuinely consider them — because the point is that the assessment gives me the information I need, but also has to be useful to the learner. If it's not, we'll end up in this same state no matter what technology comes next. I know that's not a very satisfying answer — everyone wants the silver bullet.
Luke Hobson It's the real answer, though. The real answer is to talk to other humans, explain the merit and integrity of it, discuss the underlying cause and effect. And that's truly the real answer — which I would love to be able to give people, because they want so much more. They're like, "Give me some magical thing that blocks all AI from now until infinity." And I'm like, I don't have it. It's not a thing.
Dr. Tessa Forshaw Yeah, and then they're going to put on their Meta glasses and it's going to show them the answer right on the screen in front of them and they're just going to type it in. Or whatever the next thing is.
Luke Hobson I'm so glad you said that, because I wrote about that and got interviewed on Inside Higher Ed about it. The next wave is going to be when it's literally in their glasses. Are you going to ban glasses from the classroom? And of course I had plenty of more traditional academics write about me being like, "this guy's wrong." And I was like — this is reality. I don't know what to tell you. It's in your glasses now. Yeah. Sorry.
The only time I've actually caught something blatant — and I always, like you, have that conversation with students upfront, show them how I use generative AI in my own work designing experiences, give the pros and cons and the full perspective — the only time I actually caught it was when someone's prompt was literally in the answer. It was like, OK. I had a conversation: "What prompted this? What was going on? I'd love to talk about it." And sure enough — "Dr. Hobson, I was rushing, I submitted at 11:59 PM, I just panicked." And I said, "It's OK. Let's understand. Don't do it again. Resubmit. I want to hear from you, not from ChatGPT." And that was it.
Dr. Tessa Forshaw Yeah. And another common case I come across is students whose English is a second or third — or sometimes even tenth — language, because they're amazing. They feel self-conscious about their writing, so they run things through ChatGPT because it looks more like polished prose, and they think that will reflect better on them. I had a student raise this in conversation, and I made a commitment: I will not mark anyone down for grammar, punctuation, accidentally wrong word choice, or sentence flow — because that's not the point of this class. The point is the thinking and the work. And I think that made people feel quite safe, like you could make mistakes here and we'd fix them together.
Luke Hobson Unless you use an em dash — then you fail. How could you?
Dr. Tessa Forshaw Oh yeah, how could you? And the colon! The colon has been taken away from me. I love a colon!
Luke Hobson That's everything. It just doesn't matter anymore. Whenever I see people on LinkedIn — there are a handful of phrases that if you ask ChatGPT to write a "fantastic viral LinkedIn post," it will use every single time. And it's painful because I read something and I'm like, but I actually say that. I actually say "honestly" a lot in my writing, which is now flagged as a clear sign of AI. Like, "And honestly, that's a great point." Or "And if it's not this, it's that." Great. I'll just have to change how I talk. Fantastic.
Dr. Tessa Forshaw Hahaha.
Luke Hobson Going back to cognitive science for a second — if you were to name one design principle from that space that learning professionals are ignoring, either by choice or simply because they're not aware of it, when trying to connect learning to performance — what would it be?
Dr. Tessa Forshaw Something we're ignoring. My go-to is normally building on existing knowledge — but we've discussed that. And the second one I'd normally offer would be something around metacognition — and we've covered that too. So for the third, I think from a cognitive science perspective, people really do forget that we can't multitask. In a lot of web-based learning experiences, I see designers really trying to engage learners through the screen and through video, but also trying to have them engage actively in a chat or a poll that's adjacent — not in the flow of attention of the course. If there's a side conversation happening in the chat, that's great from a social, peer-building perspective, but it's also really tricky. You're taking a person out of their focus on the screen, pulling them into the chat, and then they're coming back and forth. If you think about your attention and perception gates, that's a bit like making someone walk out their front door and back in again every time they need to change tasks. It's quite disruptive. It prohibits flow state. So my third one is: remember that multitasking is a myth. We can't process things serially.
Luke Hobson I love it. Even before jumping on with you, I had to shut off Slack and email. There's nothing else open. I'm only focused on this conversation, or it just won't work. Nope. Nothing.
Dr. Tessa Forshaw Me too, me too. Yeah. Notifications, same deal. My husband actually berates me because I have all my phone notifications turned off during work hours. And sometimes he really needs to reach me. So we have a deal now — he has to call three times in a row to get through. Because OK, fine. Yeah.
Luke Hobson Yes, exactly. The one answer I was also going to share — which I'm sure you'd appreciate — is how shocked I am by how little attention is spent on reflection, and just letting people sit with information and think. Especially for adults, it's always on to the next thing. And I'm like, no, no, no. Wait. Stay here. Stay with me. Let's keep thinking about this — connecting it to a past experience, thinking about how we're going to apply it in the future. Don't just jump to the next thing. We're trained to say, "What's next?" But sometimes we just need to stay.
Dr. Tessa Forshaw Yeah. And I think often we expect learners to do that on their own — it's a hidden curriculum — but then we don't give them the time. And coming back to the multitasking point: you can't be metacognitive and cognitive at the same time. You need time and space to do that.
And the other thing about reflection I notice a lot is that — this is happening with Socratic LLMs too — people think that if you just make learners reflect, they'll learn. But not all reflection is created equal. Reflection that's passive and just recounting something doesn't drive any significant benefit that I'm aware of. Maybe a little bit of recall practice, but not much. What does move outcomes is metacognitive reflection — thinking about: What were my strategies? How did I make sense of that? Why do I think that? Where am I going to use this? How is this similar to something I've learned before? Did I jump too quickly to a conclusion? Yesterday when I was in class I had a hard time learning — what should I do differently today? What's my goal for this particular learning experience right now? Is it just to have enough awareness of this topic that I can have a five-minute conversation with someone who works in this area once in a while? Or is this a core part of my job that I really need to understand and apply? Because the strategies you use for those two situations are going to be quite different. So yeah, the misconception that all reflection predicts improved outcomes is something I'd love to see shift. We need to be encouraging learners to engage in specifically metacognitive reflection.
Luke Hobson Speaking of the shift — perfect segue to my next question. I know from your past you've worked with internal L&D teams, external practitioners, and everything in between, building certificates, trainings, upskilling programs, and so on. Does the shift from content production to performance enablement look different depending on which side of that line you're on?
Dr. Tessa Forshaw Sure. In internal L&D, I think the performance enablement piece is arguably easier because you have the ability to influence and develop the managers and leaders where the performance actually happens, and you can influence the conditions of the environment. But from the professional certificates and upskilling programs side, I know that for a long time, people working in this space have talked about how important employer partnerships are. I think they're becoming even more important now — because for learning to be rapid, and for rapid upskilling and rapid performance to happen, we need managers and more experienced people in organizations who are influential in the development of those junior to them to really set the conditions for dynamic upskilling to happen. That was less critical when we didn't need learning to be applied so quickly. But for that to happen — based on the work I've done on how humans rapidly transfer learning from prior knowledge to a novel task — the manager level is the most influential. Their behaviors and practices are essential to get right.
Luke Hobson Have you found that when working with these folks — because this is something I've experienced — the L&D team wants to do everything correctly and is capable of doing so, but they aren't actually given the ability until an outside voice comes in and says, "Yeah, they're right. You should follow that." And then the big boss says, "OK. Do it that way." And I'm like — they were ready. They knew what to do. But instead you were like, "I need this module by the end of the day." Just give them the ability to do things.
Dr. Tessa Forshaw Yeah, there are a lot of organizations where the dynamic around L&D really needs to shift. I think it comes from this idea that every business leader is like, "I went to school, so I know about learning." But there is a real science behind it. There's a lot that works and, unfortunately, a lot that doesn't. And we spend a lot of money on things that science has repeatedly told us don't work. I agree with you — we need to lift up, enable, and support our L&D teams. That's how I see us getting through the next stretch. Certainly not by getting rid of them, as I'm hearing some people suggest.
Luke Hobson Hey, speaking of things that don't work — learning styles. I love to end interviews with this question, because as a cognitive scientist, I'm sure this drives you absolutely up the wall. What I've found from teaching adults for many years is that people hold this belief very strongly. And I wasn't fully aware of how entrenched it was until I started researching it more — like, why does this keep coming up? Going back to the original work from Fleming, it seemed like the perfect storm: teachers were being told to focus more on the individual and weren't given many different ways to do so, and then along comes this person saying, "Hey, I have this idea — VARK." And they're like, we can use that. And then it gets written into educational textbooks everywhere. That explains a little bit of why it's so deeply ingrained in our culture, despite lacking any real research credibility. When I tell people this, it literally ruins their day. Especially if you're watching on YouTube — feel free to tell me I'm a terrible person. I made a learning styles video and it was very fun to read those comments.
Dr. Tessa Forshaw Hahaha!
Luke Hobson How do you, as a cognitive scientist from Harvard — to give the listeners a strong credibility anchor here — approach debunking learning myths with adults?
Dr. Tessa Forshaw So firstly, I am in the same boat. I've completely experienced the shock of people's reactions when I say learning styles are a myth. What I've learned over the years is that these myths — particularly this one, along with the right brain/left brain, creative-or-analytical myths — are so deeply integrated into our social fabric, and people really build an identity around them. That's one of the reasons I think they're so problematic. The number of times I've shared that there are widespread misconceptions about how our minds work, I've had people come to me and say, "But I am a visual learner. That's what I am." It's a deep identity. And that's really hard, because what's happening is that person is likely limiting the learning experiences they engage with or open themselves up to. They probably have less natural motivation to engage with learning experiences delivered in a different format — despite the fact that there are dozens of studies at dozens of universities by hundreds of academics showing learning styles are not a thing. Not only are they not cognitively a thing in the sense that it doesn't make any actual neurological sense that you'd have a dominant way of learning through your eyes, body, or hearing — that doesn't fit how your brain actually works. But even if it did — even setting all of that aside — the research also shows that when a learner self-identifies as one particular style and is then taught across a range of styles, there's no correlation between what style they think they are and what style they actually learn from. Same with teachers — when you ask a teacher what style a learner is, there's no correlation between what the learner thinks they are and what the teacher thinks they are. There's just nothing there.
What I try to do when I talk about it is to help people think about not just learning styles, but a broader series of neuromyths — and then ask: what would be possible if we actually did know how to use our own minds? Imagine if you could make design choices in your own life and work and learning experiences — like you and I do, shutting off our notifications. Imagine if everyone were equipped with these simple pieces of knowledge, in the same way that we all know you eat carbs before a long run. Why don't we know the equivalent things about our minds? I think that's the question.
Luke Hobson Interesting. And I know the learning styles myth isn't going away either. I love your framing of, "What if?" — it's something we'll just keep fighting the good fight on and hopefully it will shift at some point. Which is funny too: whenever folks talk about their learning style, they're always a visual learner. Everyone. Always visual. And then I watched the rise of podcasts happen, and all of a sudden people are like, "Auditory learner?" And I was like, yeah, OK. We're just shifting from one medium to another, but it all goes to the same place. Great times. Just so much fun.
But Tessa, this has been a fantastic conversation. Absolute blast. Love getting to meet you and learn more about everything from your perspective. Where can folks go to learn more about you, your work, your new initiative, and everything of the sort?
Dr. Tessa Forshaw Absolutely. You can always find me on LinkedIn — feel free to reach out. And in terms of Harvard Graduate School of Education, you can find the Workforce Learning and Innovation Initiative there. We also have a new certificate program focused on experienced learning and development professionals, or adjacent roles. It's a mid-to-late career certificate program at Harvard, and I'd love to see any of your listeners there. If you want to come and have more of a discussion about learning styles, I'm happy to do that too. I can give you all those links.
Luke Hobson Awesome. And you mentioned there's a code for people to sign up with?
Dr. Tessa Forshaw Yes, we've got a coupon code to make sure your subscribers get a discount. And we also have further discounts for folks who want to come as a team. So feel free to get in touch via the email on the website.
Luke Hobson Awesome. I'll be sure to include all those links in the show notes below for folks to go check things out. And please, feel free to show up to Dr. Tessa and talk about learning styles. I'm sure she would love that. Well, Tessa, once again, thank you so much for being here. This was an absolute blast.
Dr. Tessa Forshaw Thanks so much for having me. I really appreciate it.
