Research Methodology and Medical Statistics अनुसन्धान पद्धति एवं चिकित्सा सांख्यिकी

Research methodology and medical statistics · Third (Final) Professional (Tritiya Varsha) · AyUG-RM

Papers

• Paper I — Research Methodology and Medical Statistics

Units in this subject

Study notes

How to read a paper, design a study and understand a number. The Ayurveda-specific part matters most: classical evidence and clinical evidence are different kinds of claim, and being able to say how is what separates a researcher from an advocate.

Design

• Observational: case report, case series, cross-sectional, case-control, cohort. Interventional: randomised controlled trial.
• What each design can and cannot show. A case series cannot establish efficacy; a cross-sectional study cannot establish sequence.
• Bias: selection, information, recall, observer. Confounding and how randomisation addresses it.
• Blinding, allocation concealment, intention-to-treat.

Sampling and measurement

• Probability sampling: simple random, systematic, stratified, cluster. Non-probability: convenience, purposive, quota, snowball.
• Sample size and power; why an underpowered study that finds nothing has found nothing at all.
• Validity and reliability; scales of measurement — nominal, ordinal, interval, ratio.

Statistics

• Descriptive: mean, median, mode; range, variance, standard deviation; normal distribution.
• Inferential: null hypothesis, p value, confidence interval, type I and type II error.
• Tests: t test, chi-square, ANOVA, correlation, regression — and which to use for which data type, which is the commonest exam question.
• A p value below 0.05 is not proof, and a non-significant result is not proof of no effect. Say so.

Ayurvedic research and its particular problems

• Whole-system versus single-molecule research, and why a standardised extract trial may not test what the tradition uses.
• Standardisation of the drug itself: the same dravya name can mean different species, as Brahmi, Shankhapushpi and Rasna all demonstrate. A trial that does not name its binomial has not stated what it tested.
• Placebo and blinding in procedures like Panchakarma, where blinding is often impossible.
• Ethics: informed consent, ethics committee approval, CTRI registration, and reporting negative results.
• Literature review, referencing styles, plagiarism.

Literature review and referencing

• Types of review: narrative, systematic, meta-analysis, scoping. What each can claim.
• Databases: PubMed, AYUSH Research Portal, DHARA, Google Scholar, and their different coverage of Ayurvedic literature.
• Referencing styles: Vancouver is standard for medical writing; know how to cite a classical text, which needs the sthana, chapter and shloka number rather than a page.
• Plagiarism, self-plagiarism, and authorship criteria.

Data collection and measurement

• Primary versus secondary data; questionnaires, interviews, observation, records.
• Reliability: test-retest, inter-observer, internal consistency. Validity: face, content, construct, criterion.
• Case record forms and the importance of defining every variable before collection begins.
• For Ayurvedic research specifically: how prakriti, agni and bala are operationalised, and why unvalidated assessment tools weaken an otherwise sound study.

Where the texts disagree

• Classical pramana and modern evidence hierarchies are not the same thing, and mapping aptopadesha onto expert opinion is an analogy rather than an equivalence.
• Whether classical observational evidence can be graded on a modern scale is genuinely contested in the Ayurvedic research literature.

Commonly confused

• Incidence (new cases) and Prevalence (all cases).
• Type I error (false positive) and Type II error (false negative).
• Case control (backwards from outcome) and Cohort (forwards from exposure).
• Systematic review (the synthesis) and Meta-analysis (the statistical pooling within it).

Practise it, and prove it