Anyone who knows me knows that I am something of a stickler about objectives. I advocate for a set of tri-level or “cascading” objectives and insist that these objectives be established at the very beginning of a design project. My logic is that in order to create something that is effective, you have to start by knowing what your client is trying to achieve in business terms (business objectives), the context for how the learning you engender is meant to be applied (performance objectives), and what your learning solution will address (learning objectives).
But I’ve been teaching a course on constructivist learning environments, and constructivism is a perspective that matches a lot of what I believe about adult learning. In his book about constructivist instructional design, Jerry Willis has this to say about objectives:
The beginning of a project is probably the worst time to create specific detailed objectives. That is when there is the least agreement about what should be learned. The beginning is also the worst time to complete learner, task, and concept analyses. Such work can begin there, but that understanding will emerge across the design process and will be of much higher quality than information and perceptions gained primarily at the beginning. Knowledge and understanding will emerge across the design and development process.
Mmm… That may sound easy to dismiss when you’ve been trained to conduct thorough front-end assessment and to define your objectives as the starting place for the design process. And I want to dismiss it – truly I do – my friends at work will gasp aloud if I don’t dismiss it and argue, as I always have, for the idea that objectives come first. But…
We recently held a course review session at work that also really caught my attention on the subject of objectives. The course that we discussed was really terrific… it was performance-based; the activities were interesting and engaging (and not the same-old same-old); the discussions were designed to go deep; and the course had a nice e-simulation at the end that brought everything together. There was only one problem. On the question of whether the course met the objectives as defined, the answer was an unequivical “no.” As written, the course’s objectives required that participants would be able to perform complex tasks, and there wasn’t near enough individual practice activities to achieve that end.
In discussions about this particular project, a couple of things came to light… it was possible that the objectives for the course were actually written after the course was designed (kind-of similar to how middle school students go back and write the required outline after they’ve finished the essay). Of course, if you’re writing objectives after the fact, it should be easier to write objectives that match what you did. But regardless, this terrific course was likely the result of having a real strong understanding of the performance we were going after, and designing learning activities that would help employees get there, maybe without ever solidifying the learning objectives in behavioral terms. Mmm…
As I read further into Jerry Willis’ conceptualization of constructivist design, I found that his description of the design process rang true for my own practice of doing design. I’ve always recognized that ADDIE (or any ADDIE-like instructional design process you might pick) is iterative, not linear. As such, while I craft objectives to get started, I constantly compare how the end product is shaping up against those objectives. And when the design and the original objectives don’t match, I’m more likely to change the objectives than I am to change the design. I’ll add in some objectives that I find I need to address but failed to account for up front. Or I’ll decide I don’t have time to go as deeply into an area that I might have liked, and so I have to cut back on more ambitious objectives. By the time I’m done, the course meets the objectives – or is that the objectives effectively preview the course I have created?
Mabye it’s semantics, but maybe we don’t have objectives when we start. (Gasp!) Maybe the objectives “emerge across the design and development process.”
As I’ve taught this constructivist design course, I’ve advised students that we need to understand the business and performance objectives of the learning solution we’re designing, but we don’t have learning objectives in the same sense that we do when we are creating and instructional solution. (The specific constructivist techniques taught in my course include: experiential learning, reflective practice, action learning, job rotation, developmental relationships, communities of practice, online collaboration, webquest, problem-based learning, simulation, case study, and more.) The constructivist perspective is that learners construct their own meaning, we can’t guarantee they will learn specifically what we intend, but we can craft an environment that is likely to help them to develop ideas and practices that are in alignment with what we hope. (I probably just got myself in trouble with a lot of constructivists… I’m sort-of a practical constructivist.)
So here’s where I think I’m landing…
It’s ridiculous to think that we can nail everything we need to know about a project before we even begin to design the solutions. The world is too complicated for that, and the variables that influence decisions too numerous. That doesn’t mean we work without objectives. Developing a strong understanding of business and performance objectives has to be among our first priorities – they help us to decide whether the project is worth the investment of time and energy, and they give us a starting place for analyzing whether learning is an important input to the desired performance. It’s an important practice to draft learning objectives early and to constantly compare them against the emerging program so that by the end of the project, the objectives are a true reflection of what is accomplished in the program.
I’m going to stop this post here (it’s already too long), but there are tons of implications…related to handoffs (to my colleagues at work: like between Consult and Produce?!), decision-making authority, validation of learning, client relationship management, and more. I’d be pleased to hear reactions, questions, and other perspectives…
Quote source: A General Set of Procedures for Constructivist Instructional Design by Jerry Willis (p. 317). In Constructivist Instructional Design: Foundations, Models, and Examples (2009) edited by Jerry W. Willis.
To regular readers… thanks for your patience regarding the long pause since my last post. So much to do, so little time…