The New Teaching Elementary Science

A great resource for unstructured science activities. Loads of ideas, as well as help and inspiration for teaching science by the play-debrief-play method.

A summary of the method, with my own small additions:

Stage 1: Play (Gathering Knowledge)
Set up enough stations so that students have enough space and materials to work without competing for space and materials. Tell the students how much time you will give them for a particular centre. Wassermann and Ivany suggest 10 to 15 minutes. Ideally, allow students to work as long as most of them are engaged - then they have time to test many ideas and explore more deeply.
While the students are manipulating, circulate with generally encouraging comments, but none that are directive or evaluative. You want the ideas of what further to research to come from your students. For many free experimentation activities, I like to add note-taking, so that students can refer to them later when they may have forgotten exactly what they tried.
Stage 2:  Debrief (Promoting Understanding)
Call your students to a different space to debrief, bringing their notes with them.
This time should be a safe time for all students.  All students should be able to share what they discovered without judgement. If some students disagree on a result, you help them engage in respectful debate, and discuss how to proceed (e.g. repeat the experiment to confirm results, try the experiment in a different way).  Ask the students how they might test their ideas as to why something is working a certain way, and write the ideas on the board. At this point, I pair and group students with similar ideas/interests to work together for the Replay, and give them a question(s) to focus on.
Stage 3: Replay (Applying Knowledge)
Students go back to the materials, focusing on addressing the ideas that came up during Debriefing. Students may choose one of the ideas to test, or test several if they have time - this may need to be organized by the teacher. This is a time for students to do duplicate experiments, use controls, make sure of fair testing etc, to be as rigorous as they can - these methods should be familiar to the students or explained to them before and during their experimentation.
Stage 4 (optional): Final Debrief
At a final debrief, students report on what they found and the class concludes what they have learned so far. (There will be many questions still unanswered, which is real science!)