Some Tips on Starting Student Self-Designed Labs

I often get asked about how I have my students design their own labs.  It's always a hard question for me to answer, mainly because I have been doing this with students for many years; it's one of those things where the process is so ingrained in my brain that it takes a while for me to articulate it.  This means that when people ask me about how, exactly, to have students design their own labs for inquiry, they often get a long and uncomfortable stare from me as my brain tries to slowly drag out that information.  
To avoid giving people that weird stare and putting  any more "crazy" in the title of this blog, I have decided to write it out for everyone to avoid any more awkward encounters.
Planning for a student self-designed lab:
First, make sure you have determined what skills and concepts you want students to show in the lab, and communicate those to your students in whatever way works for you.  Ideally, the lab should allow students to connect and hook together several concepts and skills rather than show one skill or concept in isolation.  This is why, in my classroom, I use these as my students' "Level 5 Opportunity" on my scoring scale so they can show me they know, own, and can use the skills in this new situation.  After that's been clarified, then you can do one of two things:
1) Get out an existing lab or kit you have buried in one of your cabinets.  (This is exactly what I did for our current Newton's Laws lab.)  Throw away the directions and any accompanying worksheets and put out the materials for students to use.  Let students bring in any additional materials they need to design their lab that do not violate any local ordinances, school policies, or the boundaries of good taste.
OR
2) Think about your essential question for the unit and design a lab around it.  Brainstorm materials students could use that you already have or what could be acquired by a quick trip to the dollar store.  Again, let students bring in any additional materials they need to design their lab within reason.
After this,  students need to generate specific testable questions in order to focus their design--mainly so you won't have about thirty 14- and 15-year-olds in a potentially hazardous lab setting not knowing what they are supposed to do, but also because students need to be able to generate good questions to start their own learning.   I always make them generate questions first as a team, and these questions must begin with the word "does."  Why?  Because a special education teacher I once co-taught with found that the quality of the questions students generated increased exponentially when they used this specific sentence starter.  We have no idea why this works--just that it does (pun totally intended).  
But please be aware that if you have students who aren't used to generating questions to get the learning started (being used to teachers supplying questions), your students are not going to be good at it at first.  To be completely honest, most of their questions will be complete stinkeroonies at first.  But that's OK--learning doesn't start with perfection.  They will generate questions they already know the answer to ("Does a car move slower down a ramp with carpet on it?"  Yes.  Find another question, please.) or they will generate questions that are vague and general and all fluffy and fuzzy ("Does Newton's laws have anything to do with two air pucks colliding?").  They do this because those questions are safe--they either confirm something they already know or they are so general that no matter what they do they can view what they did as "correct."  Most of my students are all about getting right answers rather than the process of learning, so generating specific questions where they take a risk and try and find out something they don't know is a scary struggle for them at first.  And it's only something they can get better at through repeated practice...the main reason why we decided to do one self-designed lab per unit in our Physical Science class this year.
After students generate their "Does" questions as a team, we then put them in a shared Google Doc so the entire class can choose from everyone's questions.  This is when I get involved in getting students to refine some questions, asking more questions to get them to see where their questions need revision.   Just like students have a hard time asking good questions, this is where I have a hard time and need some work.  You see, the questions I have to ask here are ones where I lead students to their own ideas and revisions of their questions--not lead them to what my ideas of their questions should be.  The whole point of a self-designed lab is that it be designed by the student "selves," not indirectly designed by the teacher self.  The last thing you want to do here is take on the role of the all-knowing learning puppet-master and pull strings to lock students in to what you want the lab to look like--let students be creative in their thinking by letting their thinking dominate here, not yours.
After they generate their questions and choose which question they would like to test, you can lay out any lab details and specifics that also need to be considered.  I usually like to write them out on my website, but I have also started recording these instructions so students can have the joy of pausing me, rewinding me, or setting the recording up to loop so they can go to sleep to me.  Below is my most recent example using a tool called Present.me:
(There are two things I learned from making the presentation above: 1) I need more sleep.  2) Present.me is not for us kinesthetic types that move a lot while they talk, unless you want to look like an 8-bit video game character.  Next time, I think I'll use Knovio.)
I usually have them fill out a pre-lab form next--this is where they answer their question with a hypothesis, and then start laying out the specifics of their lab experiment.  This pre-lab form is for rough-drafting purposes only--students only turn this in to get feedback from me before they set foot in my lab area, and I don't slap a number on it for a score.  I used to give scores for it, and found my students slipping into the zombie-like worksheet-filling-out mode, not taking it seriously and having very little ownership in it, even though it is supposed to be their lab.  To further this ownership, don't be tempted to add your own questions they have to answer in their final write-up to "help" them with their data analysis.  You're not helping them; you're doing the thinking for them.  Students will struggle with synthesizing all of the concepts and skills in order to explain their data at first, but that struggle is an inherent part of learning and remembering.  As long as you are giving them multiple opportunities to practice this skill throughout the year, they will eventually get better and better at it.
The final step here is to be ready to be impressed when the students come up with awesome labs you never even thought of.   This is one of the main reasons why I let students run with their labs: because I get to see students be awesome (and learn how to be awesome) in their own right. And, as I've said before, my best days at work are spent climbing over students labbing it up all over my lab area floor, not yip-yapping at kids from the front of the room.  
So, you can see why I give long, silent stares when people ask me how I do my student self-designed labs--there's a lot to it. And I've even left out the bits about what to do after they gather their data; that will have to be saved for another post.   While I've done this for many years, I'm no expert at this (I keep catching myself trying to lead students to what I want them to do.  Bad teacher).  It also takes a lot of time.  But, considering that this time is spent watching my students learn how to take their learning to some amazing places, I consider it time well spent.