Bar review

Study groups that work, and the ones that quietly cost you the year

The BATAS editorial team··5 min read

A review year is socially strange: months of solitary work with a distant, high-stakes examination at the end. A study group promises company, accountability and someone to explain what you did not understand. It can deliver all three, or it can become the most efficient anxiety-transfer mechanism ever invented.

This article is about the structure that makes the first outcome likely.

What a group can do that you cannot

Four things, and they are worth organising for:

Mark your answers. This is the big one. You cannot see what you failed to write, because if you knew it was missing you would have written it. Someone reading your script against a rubric can see it in ninety seconds. If your group does one thing, do this.

Test you out loud. Reciting a rule aloud, to a person, without notes, is a different retrieval task from writing it — and it exposes vagueness immediately. "The requisites are… something about notice" does not survive being asked "which requisites?"

Cover a gap you cannot reach. Someone in the room will have understood the tax provision you have read four times. That is not a reason to outsource the subject; it is a reason to spend forty minutes on that provision instead of another hour alone.

Make attendance a fact. For some candidates, a scheduled Tuesday session is the only structure that survives a bad week. That is a legitimate reason to have a group, and it does not need to be disguised as an intellectual one.

What a group cannot do

  • Replace your own retrieval. Listening to a colleague recite cannot substitute for reciting.
  • Cover the syllabus between you. Partitioning eight subjects among members and pooling notes produces eight shallow documents and no one who can answer a question across them.
  • Provide a standard. A group's consensus is not the law; the statute and the decision are. Cite the text, not the group.
  • Guarantee comfort. The best sessions are the ones where you discover three gaps.

The structure that works

A session of ninety minutes, four to six people, with roles:

MinutesSegmentRule
0–10Closed-book recallOne member asks; everyone answers on paper; no notes
10–25Compare and correctRead what you wrote; fix from the text, not from the loudest member
25–55Answer markingTwo scripts swapped, marked against the rubric's three questions, comments written
55–75The hard thingOne member's declared gap, taught by whoever has it, with the provision open
75–90Next weekWho writes which answers; who asks; what each person is fixing

Four features do the work:

  1. Everyone writes, every session. A session where only one person writes is a tutorial, not a study group.
  2. Marking is written. Verbal comments evaporate. Two lines per script: what was missing, and the one thing to fix first.
  3. The gap slot is declared in advance, so nobody has to admit confusion in the moment.
  4. The session ends with commitments, not with a discussion of how much remains.

Size, frequency, and the arithmetic

  • Four to six people. Below four, one person's absence kills the session; above six, two conversations happen and one of them is about the examination's difficulty.
  • Once a week, ninety minutes. More frequent sessions begin to substitute for solitary work, which is where the learning happens.
  • Same day, same hour. The value of a group is partly that it is predictable.
  • Eight months, not twelve. A group formed in October and still meeting in June is a group that has become a social club with reading. Stop meeting for the last six weeks, when the work turns to consolidation and your own flags list — the last-ninety-days rhythm in the last ninety days is solitary by design.

The failure modes, named

The anxiety exchange. Ten minutes of "I have only done Civil twice" and "the examiner will ask about the new jurisprudence" at the start of every session. It feels like mutual support and functions as mutual contagion.

Rule: no status reports. The session begins with recall, not with progress anxiety.

The lecture. One member — usually the most confident — teaches for the whole session while everyone else takes notes. It is comfortable for both sides: the lecturer gets validation, the listeners get the feeling of coverage.

Rule: a maximum of twenty minutes of teaching, always with the text open, always on a topic someone declared.

The pooling. A shared drive of eight outlines, each written by a different member, nobody's gaps recorded.

Rule: share questions, never outlines. If you want someone's understanding of a topic, ask them to mark your answer to a question on it.

The consensus answer. The group agrees on a rule that none of you has checked. This is worse than being wrong alone, because now there are five of you.

Rule: every rule asserted in a session gets a source — provision, rule number, or decision. If nobody can produce it, the item goes on a "verify" list for next week.

The disappearing fifth member. Groups decay by attrition. Decide in advance what happens at three members: usually, merge or fold, rather than drift.

Accountability without surveillance

The useful form is specific and small:

  • Each person names one thing they will be able to do by next week, closed book.
  • Each person brings one written answer to be marked.
  • Missing a session means sending the answer anyway.

That is it. A shared spreadsheet of daily hours is a mechanism for producing guilt at scale, not learning — and hours are precisely the metric how to measure readiness argues against.

A word on explaining

Teaching a doctrine is the strongest test of knowing it. If you can explain, without notes, why an exception exists, you own it. So the "hard thing" slot should rotate, including to the quieter members: the person who has never explained anything is the person whose understanding has never been checked.

What to do if you cannot find a group

You can reproduce most of the value alone, and you should try to:

  • Marking: swap scripts with one other person by message, weekly. A single reliable marking partner is worth more than a group.
  • Testing out loud: record yourself reciting a rule, then listen. Vagueness is audible.
  • Accountability: one weekly commitment, messaged to one person.

What you cannot reproduce is the surprise of another person's question. If you can arrange that once a week in the last six months, you have captured most of the group's value without the group's costs.

For the substance the sessions should be built around, see the answer structure and past bar questions.

study grouppeer markingreviewbar exam

Written by the BATAS editorial team. This article is a study aid, not legal advice, and the law it describes is stated as of 20 October 2026 — verify against the official text before relying on it. How we write and check these articles: editorial standards.