Comments on: Sliders with a Purpose

By: Michael Fenton

Michael Fenton — Sat, 05 Mar 2016 03:16:11 +0000

@mathgrip: Great points! I’m now thinking of a blog post that explores what is gained and lost with paper pencil, guess and check with tech, sliders, and regression. I think there is value in a blended approach here, but it would be fun to reflect on the specific advantages of each, as well as what is gained _and_ lost as we move from one approach to another.

By: mathgrip

mathgrip — Sat, 05 Mar 2016 01:22:40 +0000

I love sliders for the unlimited variations you can model. Might also add real-world contextual “targets” along with target points such as, “how do I need to adjust the price or cost-per-item to achieve a target profit or break-even point (systems)?” However, I do like the concrete visual targets in your examples.

That said, reflecting on Andrew’s comment, sliders have the potential to remove the estimation or “guess” part of guess-and-check. There is some nice sense-making that comes from predicting what parameters would achieve the goal and then adjusting based on the feedback and trying again. So, when using sliders, I think I would ask students to make a prediction or series of guess-and-checks before slider play. Or try both approaches and observe.

[Before you slide, estimate/guess values for “a” and “b” that would produce a line passing through the points.]

Now retired from classroom teaching, I miss having a class to try out these ideas.

By: Michael Fenton

Michael Fenton — Fri, 04 Mar 2016 21:31:49 +0000

Andrew, these are great questions! I’m in a similar boat of relying on my gut in some cases, and continuing to develop my explicit thinking in others.

As for when movable points are better than sliders… Hm. I don’t have a comprehensive answer, but I do know that sliders are one layer more abstract (algebraically) than the “change the number” approach you mentioned. And movable points add one more layer of abstraction (again, algebraically) than sliders. Or maybe it’s not a layer of abstraction, but just a “step removed” since it draws the focus away from the expression list and toward the coordinate plane.

One clear use case (given the current set of Activity Builder tools) where I use movable points: When building a screen with graph + note + input (which hides the expression list, and therefore any related sliders).

I’d love to hear more of your thoughts on this topic, either in a comment, or maybe a blog post of your own sometime. 🙂

Thanks for stopping by to chime in!

By: Andrew Knauft

Andrew Knauft — Fri, 04 Mar 2016 21:14:37 +0000

Related question: when is a slider better than “change the number”? Marble slides made the choice to use numbers, which I think was the appropriate move for that activity. But other situations beg for the freedom of “drag and see!” And for those situations, when is a slider better than a draggable point?

I would love to see (or do!) more research on these questions. For now, I rely on my gut and whatever justifications I can hack together on a case-by-case basis!