When I first tried small group writing conferences, I grouped by need: Writers who were stuck in the early stages of the writing process worked together with me to discuss ways to become “unstuck”, those who needed more support with organization met at a different time, and those who were ready for greater challenges met in yet another group.
There is certainly nothing wrong with grouping students by need. The conferences become more efficient: you can often provide feedback to 4 or 5 students in about 15 or 20 minutes. There are possible downsides, however. Writing is complicated, and some writers may have more than one need. If you aren’t careful, small (lower-case-g) groups can become (all-capitals) GROUPS that meet too regularly, and writers may start to label themselves as struggling or advanced. My most successful needs-based small groups have been very short term and laser-focused.
An alternative to pulling small groups based on like needs is to just ask small groups of students to work on their writing at a table while you monitor their progress regardless of their needs. If the purpose of a conference is feedback, then providing feedback as close as possible to the actual act of writing can increase the power of the conference.
When I pull small groups of students in this way, I say something like this to the whole class before I begin:
“For the next couple of days, I am going to pull some small groups just to see if I can give you some support while you are writing. It’s kind of like the coaching I did when I taught swimming: I would give advice while my swimmers were in the pool, while they were actually swimming. This worked better than waiting til they got out of the pool. Don’t worry: I won’t interrupt your thinking and you can certainly ask me to let you work a bit longer before I talk with you, but I think having you close by in small groups might be a way to give you better feedback.”
When I have spoken with students about this method of conferring in small groups, their feedback to me has been very positive. They feel they can ask quick questions, and they are more likely to listen to my feedback if they haven’t written so much text that any potential problems seem too big to solve. It also provides an opportunity to give frequent positive feedback to your more resistant writers: you can acknowledge even small progress, and because you aren’t sitting alone with one student for the entire time, you are less likely to create a dependent writer. Another benefit is that writers may overhear what you saying to another writer. This is a good thing! They may try a strategy you teach someone else.