1
00:00:01,160 --> 00:00:04,550
'Artificial intelligence and
law' may be an interesting

2
00:00:04,560 --> 00:00:07,570
topic. It is not yet a manageable dissertation

3
00:00:07,580 --> 00:00:11,030
question. A research question tells you what you

4
00:00:11,040 --> 00:00:13,530
are trying to find out, which material you

5
00:00:13,540 --> 00:00:15,870
need, and what a reasoned answer could look

6
00:00:15,880 --> 00:00:18,860
like. Begin with your programme's requirements.

7
00:00:19,320 --> 00:00:21,280
Is the project required or optional?

8
00:00:21,700 --> 00:00:22,740
What is the word limit?

9
00:00:23,240 --> 00:00:26,540
What topic approval, supervision
and ethics rules apply?

10
00:00:27,100 --> 00:00:30,350
There is no universal UK dissertation template.

11
00:00:30,360 --> 00:00:33,210
Your handbook and supervisor
come before an online

12
00:00:33,220 --> 00:00:37,160
checklist. Now narrow the
topic along three dimensions:

13
00:00:37,720 --> 00:00:40,560
the legal problem, the
jurisdiction and the scope.

14
00:00:41,320 --> 00:00:43,170
Instead of trying to cover every effect of

15
00:00:43,180 --> 00:00:46,650
AI, you might investigate
how one specified legal

16
00:00:46,660 --> 00:00:49,620
framework addresses one clearly defined use.

17
00:00:49,900 --> 00:00:52,510
That is a direction to discuss, not an

18
00:00:52,520 --> 00:00:55,000
assertion that the framework succeeds or fails.

19
00:00:55,820 --> 00:00:58,760
For our original planning
example, imagine the question:

20
00:00:59,140 --> 00:01:02,670
how should a specified university
policy distinguish permitted

21
00:01:02,680 --> 00:01:05,460
study support from replacement authorship?

22
00:01:05,680 --> 00:01:08,450
This is a policy research exercise, not a

23
00:01:08,460 --> 00:01:11,060
description of every university's current rules.

24
00:01:11,460 --> 00:01:13,800
You would still need to
identify the institution,

25
00:01:13,980 --> 00:01:17,500
relevant policy version and scope
before treating it as a real project.

26
00:01:18,440 --> 00:01:20,430
Check feasibility before
falling in love with the

27
00:01:20,440 --> 00:01:23,070
title. Can you access the primary documents and

28
00:01:23,080 --> 00:01:25,570
scholarship? Can the question
be answered within the

29
00:01:25,580 --> 00:01:28,790
permitted length? Does it
invite an argument, rather

30
00:01:28,800 --> 00:01:32,030
than a catalogue? If answering it would require

31
00:01:32,040 --> 00:01:35,810
interviews or private data,
discuss ethics and access

32
00:01:35,820 --> 00:01:39,600
before collecting anything.
Make a small research log.

33
00:01:40,160 --> 00:01:42,700
Record the source, the question it helps answer,

34
00:01:42,780 --> 00:01:45,200
its limits, and what to investigate next.

35
00:01:45,920 --> 00:01:48,840
Keep actual passages and your
interpretation distinct.

36
00:01:49,320 --> 00:01:52,980
An impressive bibliography is not evidence
that you have answered the question.

37
00:01:53,760 --> 00:01:55,750
Then design a provisional argument map.

38
00:01:55,760 --> 00:01:57,840
One section defines the problem.

39
00:01:58,040 --> 00:01:59,800
Another tests a possible response.

40
00:02:00,020 --> 00:02:02,700
A third examines objections and alternatives.

41
00:02:03,220 --> 00:02:06,010
Each section should earn its place by answering

42
00:02:06,020 --> 00:02:07,760
part of the main question.

43
00:02:08,240 --> 00:02:09,900
The map can change as you learn;

44
00:02:10,100 --> 00:02:12,190
it is a working tool, not a promise

45
00:02:12,200 --> 00:02:13,940
to defend your first idea.

46
00:02:15,040 --> 00:02:17,790
Before your next supervisor
meeting, bring the proposed

47
00:02:17,800 --> 00:02:21,710
question, a short rationale,
a realistic source plan

48
00:02:21,720 --> 00:02:24,400
and two precise uncertainties.

49
00:02:24,800 --> 00:02:26,910
Ask for feedback you can act on within

50
00:02:26,920 --> 00:02:29,700
the supervision rules. Your self-check:

51
00:02:30,160 --> 00:02:34,320
could two sensible researchers give
different reasoned answers to your question?

52
00:02:38,880 --> 00:02:41,200
If yes, there may be room for evaluation.

53
00:02:41,940 --> 00:02:44,110
If you only need to list facts, consider

54
00:02:44,120 --> 00:02:46,280
whether the question is still too descriptive.

55
00:02:47,060 --> 00:02:48,860
The free dissertation guide is linked below.