Non Doctrinal Research: Merits and Demerits
Chapter Ninety
Syllabus topic 5, "Research Methods"
Pages 329 to 331 of 543
In one line
Non doctrinal research is the only way to know what is actually happening, and it is slow, expensive, hard to check, and easy to do so badly that its findings are worse than none.
In the wording a student can write in an exam: the merits of non doctrinal research are that it reveals the gap between law as enacted and law as it operates, identifies causes rather than symptoms, supplies the evidence on which reform proposals must rest, discovers what the researcher did not anticipate, and makes the law answerable to those it affects; its demerits are that it is expensive in time and access, requires skills legal training does not supply, produces findings limited to their sample, is much harder to verify than doctrinal work, and can mislead badly when the design is poor.
Merits
It reveals the gap. Chapter 850's criticism of doctrinal work is answered only here. The right in Khatri and Suk Das either reaches people or does not, and nothing but empirical work can say which.
It identifies causes. Chapter 740: a reform aimed at the wrong cause achieves nothing, and only investigation separates four candidate causes.
It supplies evidence that persuades. Chapter 730: a proposal must survive a ministry that has no obligation to act, and a finding is harder to refuse than an argument.
It discovers what was not anticipated. This is the merit most worth naming, and chapter 700 has the example: the Sheela Barse investigation was directed at an allegation of assault and found that there was no adequate arrangement for legal assistance to women prisoners.
It makes the law answerable to those it affects. A rule studied only through its own sources is judged by the people who made it; a rule studied empirically is judged by what it does to people.
And in India it is nearly always new, chapter 860.
Demerits
It is expensive in time. Six weeks of observation is six weeks, and it cannot be compressed.
It depends on access. A refusal by a court, a jail, a college or an authority can end a study, and there is often no remedy.
It requires skills law teaching does not supply. Sampling, instrument design, coding and analysis, chapter 320.
Its findings are limited to their sample. A study of one taluk is evidence about that taluk, chapter 1020, and the temptation to write as though it were about the country is very strong.
It is hard to verify. A reader can check a doctrinal proposition in minutes by opening the Act. Checking an empirical finding means repeating the fieldwork, which almost nobody does, so empirical claims circulate unchallenged in a way doctrinal ones do not.
Non Doctrinal Research: Merits and Demerits
And a poor design misleads. This is the serious one. A doctrinal error can be spotted by a reader who knows the law; an error of sampling or instrument design produces a number that looks exactly like a good number, and the reader has no way of telling.
The demerit that deserves its own section
Bad empirical work is worse than none, and a student should understand why.
A wrong doctrinal proposition is refutable. Anyone can open the section and see that it does not say that.
A wrong empirical finding is not. If a dissertation reports that seventy per cent of respondents were unaware of legal aid, and the respondents were forty people who happened to be at a bus stop, the number is meaningless and looks identical to a number that is not.
And numbers travel. They are quoted, they enter other people's literature reviews, and they acquire authority by repetition.
Which is why the safeguards are not optional: state the method, state how the sample was chosen, state the size, state the refusals, state the limits. Chapter 790's eight characteristics exist for exactly this.
What makes non doctrinal work reliable
A defined universe and a stated sampling method, chapters 1000 to 1020.
A pre-tested instrument, chapter 690, step six.
Recorded departures from the plan. Substitutions, refusals, days missed.
Separation of what was observed from what was concluded.
Triangulation where possible, that is, checking a finding by a second method: records against interviews, or observation against what officials say happens.
And admitted limits, chapter 1280.
Distinctions
| Doctrinal | Non doctrinal | |
|---|---|---|
| Answers | What the law is | What is happening under it |
| Sources | Statutes, judgments, rules | Records, people, observation, non-legal documents |
| Cost | Low | High in time and access |
| Checkable by a reader | Yes, in minutes | Only by repeating the work |
| Error is | Refutable from the source | Invisible in the result |
| Generalisable | Within the jurisdiction | Only within the sample |
A worked example
Two studies reporting the same finding, and why one is worth something.
Study A. The researcher spoke to people in a settlement and found that most were unaware of free legal services. Conclusion: awareness is low in Maharashtra.
Study B. A schedule of six questions administered by the researcher, because most respondents could not read, to every fifth household in three named settlements of one taluk, over two weeks in a stated month, with 12 refusals out of 96 approaches recorded, producing a stated proportion. Conclusion: in these three settlements, at that time, the stated proportion of respondents had not heard of free legal services; the finding does not extend beyond them; and two respondents said they had heard of it from a neighbour who had been turned away, which suggests a mechanism worth investigating.
Non Doctrinal Research: Merits and Demerits
A is unusable and unfalsifiable. B is usable within its limits and tells a District Authority where to hold its camps.
And B costs two weeks and a printed sheet, which is the point this whole module is making: reliable empirical work is within a student's reach, and the difference between A and B is discipline, not resources.
Quick revision
Merits: reveals the gap between law and operation; identifies causes; supplies evidence that persuades; discovers what was not anticipated; makes the law answerable to those it affects; and in India is nearly always new.
Demerits: expensive in time; depends on access; needs skills law teaching does not supply; findings are limited to the sample; is hard for a reader to verify; and a poor design misleads invisibly.
The special demerit: a wrong doctrinal proposition is refutable from the source, and a wrong number is not, and numbers travel by repetition.
Safeguards: a defined universe and stated sampling; a pre-tested instrument; recorded departures, substitutions and refusals; observation kept separate from conclusion; triangulation; and admitted limits.
Test yourself
1. Give four merits of non doctrinal research. It reveals the gap between the law as enacted and the law as it operates, which doctrinal method cannot detect; it identifies causes rather than symptoms; it supplies evidence that is harder for a ministry to refuse than an argument; and it discovers what the researcher did not anticipate, as the Sheela Barse investigation did.
2. Why is a bad empirical finding worse than a bad doctrinal proposition? Because a doctrinal error is refutable: any reader can open the provision and see that it does not say what was claimed. An error of sampling or instrument design produces a number indistinguishable from a sound one, so it cannot be spotted by a reader, and numbers acquire authority by being quoted.
3. Name five safeguards that make empirical work reliable. A defined universe with a stated sampling method; a pre-tested instrument; recorded departures from the plan, including substitutions and refusals; separation of what was observed from what was concluded; and triangulation, checking a finding by a second method such as records against interviews.
4. What is the practical difference between an unusable and a usable empirical study? Discipline rather than resources. A usable study states its instrument, how respondents were selected, how many were approached and how many refused, over what period and in what places, and confines its conclusion to those bounds; an unusable one reports an impression and generalises it.
The rest of this subject
These notes are cut from the University's printed syllabus. Open the syllabus itself, or the past papers, for the same subject.