One-page CSIR NET Part A research aptitude revision: study types, hypotheses, variables, controls, sampling, statistics, ethics, authorship and metrics.
Pre-written messageResearch Aptitude Formula Sheet and Short Notes — notes, formulas and practice, free. csirnetnotes.com/general-aptitude/research-aptitude/quick-revision
Per the NTA information bulletin for December 2026, the same for all five subjects:
Given
Attempt
Correct
Wrong
Maximum
20
15
+2
−0.5
30
Only the first 15 answered questions are evaluated. Research aptitude is named on CSIR-HRDG’s subject pages, not in the bulletin, and has no official topic list. Details on the General Aptitude overview.
Types of research and study designs
Axis
Types
Purpose
Basic (understanding) vs applied (practical problem); action research (practitioners improve their own practice)
Depth
Exploratory → descriptive → explanatory
Data
Quantitative vs qualitative vs mixed
Control
Experimental (researcher assigns) vs observational
Time
Cross-sectional (one point) vs longitudinal (followed over time)
Logic
Deductive (theory → test) vs inductive (observations → generalisation)
Cohort: exposure → forward to outcome. Case–control: outcome → back to exposure. Cross-sectional: both at once.
Only random assignment breaks the link between treatment and everything else.
Blinding: single (subjects), double (subjects + assessors).
Look-alike effects: placebo, Hawthorne, regression to the mean, observer bias.
Sampling and bias
Method
Rule
Simple random
Every sample of size n equally likely
Systematic
Random start, every k-th, \( k = N/n \)
Stratified
Random units from every stratum; proportional allocation \( n_i = n \frac{N_i}{N} \)
Cluster
All units from randomly chosen clusters
Convenience, purposive, quota, snowball
Non-probability; weak for generalisation
Bias: selection, non-response, survivorship, recall, measurement, confirmation, publication. A bigger sample cuts random error, not bias.
Measurement and statistics
Scale
Zero
Example
Nominal
—
Blood group
Ordinal
—
Disease grade
Interval
Arbitrary
°C
Ratio
True
Mass, K, concentration
Accuracy = near the true value (systematic error spoils it; fix by calibration). Precision = near each other (random error spoils it; repeating helps).
Reliability ≈ consistency; validity ≈ measuring the right thing.