Why designing experiments is so hard for students & what we can do to help!

We’ve all heard all about how inquiry is important in science classrooms.  So why is it so hard to do?  And why are there still so teachers reluctant to do it?  Well, for lots of reasons actually…but this post will just talk about one of them – because it’s really hard to do it AND have students learn science content.

Now before all the comments go flying around contradicting that, just hang on and hear me out.  What I mean is that it’s hard for students to learn about the process of doing science (which includes a lot of things, only one of which is designing experiments) while at the same time learning complex science content.  Not saying you can’t (and shouldn’t) do it.  I do.  I’m just saying we all need to understand WHY it’s so difficult and scaffold students when necessary to get there!

Cognitive Load Theory

John Sweller developed Cognitive Load Theory in the 1980’s and much research has been done on it since then.  If you’re interested, please take the time to look into it what the research can tell you about efficient learning, but I’ll give you a super quick explanation of it.

Stimulus (sight, sound, smell, etc.) must be paid attention to enter our working memory.  People have between 5-9 working memory “slots.”  Once something is in working memory then it activates things related things in our long-term memory (this is where prior knowledge and developing diverse connections to that knowledge come into huge play!)  Once the related long-term memory is activated, it can be used in working memory to process.  Long-term memory is vast…but processing can only occur in the very limited working memory.


There are three types of “load” placed on working memory:

Cognitiveloadtheory2Extraneous load is what we can do without – it’s what we try to limit by writing clear instructions, limiting distractions, and not putting a graph on one page and then having a discussion about it several pages later making the reader flip back and forth (I hate that!).

Intrinsic load is the load placed on our working memory that’s due to the problem or task itself.  You can reduce this by breaking a problem into steps.  This is also reduced as the person develops related long-term schema and chunks together information.  Chunking is taking various pieces of information or tasks become so well practiced and known that it begins to be “chunked” together.

Germane load is the load of learning.  It takes working memory space to learn – to create new schema in long-term memory.

Balancing types of cognitive load

The problem with cognitive load is that it is additive and it must all be taken care of in our limited working memory.  AND, the load of learning (Germane Load), is the last to get taken care of!  So basically, if the total of Extraneous + Intrinsic is equal to or greater than a person’s working memory capacity, then there is no room for Germane load and they won’t learn anything from the experience.  This is when a someone is “cognitively overloaded.”

We’ve all had those moments with our students where we’re working through a problem with them, they’re actively engaged, they’re “with you” every step along the way and you feel like you’re making progress with them – and then you get to the end and they look at you and in all honesty say “I have no idea what we just did and there’s no way I could do it again.”  That’s when they’ve been overloaded and there was no room for them to learn – to develop schema in their long-term memory about what they just did.CognitiveLoadTheory3

So you can help someone that’s been overloaded by lowering Extraneous and/or Intrinsic load.

Lowering Extraneous might mean taking the “story” out of the “story problem” and going back to just plain problems, taking out extraneous information, cutting out that graph and setting it right next to the discussion about it rather than making them flip (with has a technical name called “Split Attention Effect“).

Lowering Intrinsic load might mean breaking the problem into smaller steps, going back and taking time to increase the students’ prior knowledge related to the current goal or scaffolding the student in some way – providing a template, guiding questions, hints, backwards-faded worksheets, reading guides, etc.  Those scaffolds can be “official” – tech tools, worksheets, etc. – or they can be “spur of the moment” guiding questions a teacher asks a student as they struggle with something.

So you’re saying everything has to be simple and striped down?

Nope, not at all!  The world is full of Extraneous Load and students must be able to operate within it.  And the world is full of Intrinsically difficult problems and we need to solve them!  But people can’t learn if they’re cognitively overloaded.  So lower the Extraneous or Intrinsic enough to allow Germane – which leads to them developing schema (“prior knowledge”) – their enhanced schema/prior knowledge now lowers the Intrinsic load of a problem, which means they can tackle something with either higher Extraneous or Intrinsic next time – and this is the “efficient learning loop” that’s the sweet spot of learning!

Of course there’s a catch…

What is good for one students is not necessarily good for all!  What will lower cognitive load for a novice at a specific skill or content might actually INCREASE cognitive load for someone that already has relevant schema developed.  This is called the Expertise-Reversal Effect.  For example, if you have students that struggle with math skills in a chemistry class, Backwards-Faded Worksheets can really help them out and get them to learn the skills much more efficiently.  However, for a student that already has fairly well-developed schema for using those math skills, the scaffolding provided by those worksheets may actually confuse them.  And this is where the fabulous-ness that is differentiation comes into play! 🙂

So how does this all relate to students designing experiments?

If students do not have well-developed schema for designing an experiment AND they don’t have well-developed schema for the science content involved in the experiment they’re assigned to design, they will most definitely be overloaded!  And this happens much more easily than most of us with these schema developed realize!

Therefore you can either lower the load of the design aspect – by scaffolding or breaking into step-wise pieces – or you can lower the load of the science content.

How much scaffolding a student needs will be a combination of their unique prior knowledge (their schema related to both the design process and the science content involved) and the science content itself.  And it will shift from activity to activity.  A student that feels fairly comfortable with the design process during one activity may be need more scaffolding on that aspect if faced with another activity that really ramps up the science content.

So how can you scaffold students designing their own experiments?

Backwards design! 🙂

Students are trying to design their experiment in the same order they typically read them in – title, purpose/problem/question, background information and then materials.  But how can they know what materials they need if they haven’t thought about what they’re doing yet (the procedure)?  And how can they design the procedure if they haven’t thought about what data they need?  And how can they know what data they need if they haven’t thought about what format the results will be in to answer/address the purpose/problem/question?

This was the basis of my PhD dissertation (along with the creation of a software tool for students that’s described next) – which if you need to get yourself sleepy you can read it here.  But the short version is that it worked – with an effect size (difference in averages divided by standard deviation) of 1.5 which is a HUGE effect in education! 🙂

So how can I take this backwards-design scaffolding to the next level?

There’s still a lot extraneous things going on in an experiment design activity even if students are scaffolded in the backwards manner.  That’s why I created the software application during my PhD dissertation.  It combines all sorts of research on how to minimize Split-Attention Effect (like not having a rubric with all the information for the entire lab report on a separate sheet of paper that students have to consult – it’s built into the software and only shows the rubric for the section they’re working on), and lots of research-based tech design about how to design scaffolding tools so that they fade away and guide students through the process itself…maybe I’ll post about all that kind of stuff another time (or you could read about it in my dissertation)

Where can I get this software?

It’s free, and available for Mac and Windows at www.studentdesignedlabs.com

There’s a school named “Test” with a teacher named “Test” with a student named “Test” that all have the password “Test” that you can use to play around with it.  Or you can email me to create your own teacher account and then add whatever student accounts you’d like.

What’s next for this software?

I have a whole list of user-requested upgrades and want to deploy for tablets.  I’d love for you to help out with this project here at my Kickstarter Project.

Did you enjoy this article?
Get Free Updates

Leave a comment

Your email address will not be published.

10 thoughts on “Why designing experiments is so hard for students & what we can do to help!”