Yesterday we had an in service training session on internal assessment. A group of students presented their experiences of good IA processes. None of them do physics so physics wasn't mentioned but from the feedback I get from my students I know that the way I organise the individual investigation is a winner. Here are 8 reasons why:

Scaffolding

I make sure that students have developed the skills necessary to do their own investigation. The activities performed before the investigation include introductions to the use of Excel, Loggerpro, Algodoo, GeoGebra and Capstone. Students are also introduced to the use of dataloggers by Vernier and Pasco. They perform many experiments where they plot curves, linearise, use error bars, place best fit lines and interpret gradients and intercepts.

Practice

My students do a mini IA to introduce the idea of research question and hypothesis. I mark this according to the criteria and give individual feedback.

Early

My students do their investigation in March of the 1st year before they get bogged down with other IA's and university applications.

Class time

I use three weeks of class time for the investigation.

Feedback

I give feedback via Google classroom throughout the process.

Fast turnaround

I return the first draft with feedback within one or two days.

Good first draft

Due to ongoing feedback my students produce good first drafts This means that they need little time to finish it off. I give them a week.

Safety net

Because I do the process so early, students have time to do another one. I only allow this in special circumstances for example if the student has weak English or, for no fault of their own, the experiment just didn't work. They do the second IA during class time which is only possible due to the student centred nature of the course. I expect them to get ahead of the normal classwork before starting.