Quantitative and Qualitative Research
Chapter Eighty-Five
Syllabus topic 5, "Research Methods"
Pages 314 to 316 of 543
In one line
Quantitative research counts and qualitative research listens, and the honest position for a law student is that most legal problems need both, because a number tells you how much and only an account tells you why.
In the wording a student can write in an exam: quantitative research measures phenomena in numerical terms and analyses the resulting data by statistical or arithmetical means, seeking to establish magnitude, frequency, distribution and relationship; qualitative research investigates the meaning, quality and context of a phenomenon through non-numerical material such as interviews, observation and documents, seeking to understand rather than to measure; the two answer different questions and are commonly combined.
Quantitative research
What it produces. Numbers: how many, what proportion, how often, how much, and whether two things vary together.
Its methods. Structured questionnaires and schedules with fixed responses, counts of records, and statistical analysis.
Its strengths. It permits comparison, it can be summarised, it supports a claim about magnitude, and it is checkable by anybody with the same data.
Its weaknesses. It measures only what was decided in advance to measure, so it cannot discover anything the researcher did not anticipate; it strips context; and it can be very precise about something unimportant.
Its use in law. Counting undertrials, chapter 700; counting how many colleges examine the clinical papers practically; counting how many remand productions included the required information.
Qualitative research
What it produces. Accounts: what happened, how it was experienced, what people meant, why they acted as they did.
Its methods. Unstructured and semi-structured interviews, participant and non-participant observation, case studies, and the close reading of documents.
Its strengths. It finds what the researcher did not anticipate, which is its single greatest advantage; it supplies the mechanism behind a number; and it works where the population is small, hard to reach or unwilling to be counted.
Its weaknesses. It does not support claims about magnitude; it is harder to check, because the material is not reducible to a table; the researcher's own judgment enters at every stage; and it is very time-expensive per respondent.
Its use in law. The interviews in Sheela Barse, chapter 700; the questions people actually asked at a camp, chapter 670; and any study of why a rule is not being followed.
The pairing, and why it is usually the right answer
A number without a mechanism is not actionable. Knowing that the information is given in a small minority of remand productions does not tell you whether the magistrate does not know the duty, has no time, or has nobody to assign.
A mechanism without a number is not persuasive. An account of why one magistrate does not do it is a story until somebody establishes how common it is, chapter 730's point about persuading a ministry.
Quantitative and Qualitative Research
So the ordinary design is: count, then ask why. Or, where nothing is known at all: ask a few people first to find out what should be counted, then count it.
That second order is the more common in law, because the field is so unmeasured that a researcher usually does not yet know what the right questions are. Chapter 1230 on the exploratory design is the same idea from the design side.
Distinctions
| Quantitative | Qualitative | |
|---|---|---|
| Produces | Magnitude, frequency, distribution | Meaning, mechanism, context |
| Instruments | Structured schedule, questionnaire, counts | Interview, observation, case study, documents |
| Sample | Larger, and its selection is critical | Smaller, chosen purposively |
| Can discover the unanticipated | No | Yes |
| Supports a claim about how many | Yes | No |
| Checkable by | Re-analysis of the same data | Reading the researcher's records |
| Cost per respondent | Low | High |
What a student should watch for
Do not present qualitative material as though it were quantitative. Fifteen interviews do not produce a percentage, and a claim that sixty per cent of respondents said something, where the respondents were nine people chosen because they were available, misuses both.
Do not treat a number as self-explanatory. A table needs a sentence saying what it shows and a paragraph saying what it does not.
Record qualitative material properly. Notes made the same day, in the respondent's words where possible, and kept, because the only way a reader can check qualitative work is by inspecting the records.
And observe the ethics in both, chapter 1400, with the additional point that qualitative material is far more identifying: a table of numbers protects a respondent and a quoted account may not.
A worked example
One question, done both ways and then together.
The question. Why do so few people in this taluk use the legal aid system?
Quantitatively. A schedule of six questions to every fifth household in three settlements: have you heard of free legal services; do you know where to go; have you ever gone; if not, why not, with four fixed options and a fifth for other. Result: proportions, and a distribution of reasons across the four options.
Qualitatively. Twenty unstructured interviews with people who said they had not gone. Result: accounts, including two reasons nobody had put on the schedule, that the office hours coincide with work and that a previous applicant in the settlement was turned away and told everybody.
Together, and this is the design to use. Twenty interviews first, to find out what the reasons actually are; then a schedule built from those reasons, administered to a larger sample, to find out how common each is; then a few more interviews to understand the largest category.
Quantitative and Qualitative Research
The finding that would result is both a proportion and a mechanism, which is what a District Authority would need in order to act, chapter 740.
Quick revision
Quantitative research measures numerically and analyses statistically, producing magnitude, frequency and distribution; strong on comparison and checkability, weak in that it measures only what was decided in advance and strips context.
Qualitative research investigates meaning, quality and context through interviews, observation and documents; strong in that it finds what was not anticipated and supplies the mechanism, weak in that it cannot support claims about magnitude and is hard to check.
The usual right answer is both, and in an unmeasured field the order is normally qualitative first to find out what to count, then quantitative to find out how common it is.
Watch for: presenting qualitative material as percentages; treating a table as self-explanatory; failing to keep qualitative records; and the greater identifiability of qualitative material for ethics.
Test yourself
1. State what each kind of research produces and one strength and one weakness of each. Quantitative research produces magnitude, frequency and distribution; its strength is that it supports comparison and can be checked by re-analysis, and its weakness is that it can measure only what was decided in advance, so it discovers nothing unanticipated. Qualitative research produces meaning, mechanism and context; its strength is that it finds what the researcher did not anticipate, and its weakness is that it cannot support a claim about how many.
2. In an unmeasured field, in what order should the two be used, and why? Qualitative first, then quantitative. Where nothing is known, the researcher does not yet know what the right questions are, so a small number of open interviews is needed to discover what should be counted before a structured instrument can be built.
3. What is the commonest misuse of qualitative material? Presenting it as though it were quantitative, for example reporting a percentage of respondents from a handful of people chosen because they were available, which claims a magnitude the method cannot support.
4. Why does qualitative work raise sharper ethical questions? Because its material is far more identifying. A table of numbers conceals the respondent, while a quoted account of what somebody said, in a small community, may identify them to anybody who reads 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.