Skip to main content

Research Design: Meaning, Function, and Basic Terminology

Overview​

Every research study needs a blueprint — a detailed plan that guides the investigator from the first research question to the final statistical conclusion. This blueprint is the research design. A well-crafted research design answers two fundamental questions: How will we find answers to our research problem? and How will we control the sources of variation that could distort our findings? This unit introduces the foundational concepts of research design before focusing on the widely used Single Factor Design.

Source PDFs:

1.2 Meaning of Research Design​

Research design is formally defined as:

"The plan, structure, and strategy of investigation conceived so as to obtain answers to research questions and to control variance." — Kerlinger (2007)

Let us break this three-part definition:

ComponentWhat it MeansPractical Example
PlanThe overall scheme/program of researchDeciding to study the effect of sleep deprivation on memory
StructureOutline of the research design; the paradigm of variable operationAssigning 60 participants to 3 sleep-deprivation conditions
StrategyMethods to gather and analyse data; how objectives will be reachedUsing recall tests scored by blind raters; analysing with ANOVA

Thyer (1993) adds a practical dimension: a research design specifies (a) how variables will be operationalised, (b) how a sample will be selected, (c) how data will be collected for hypothesis testing, and (d) how results will be analysed.

In brief: a research design is a plan adopted by the researcher to answer a research question.

Why Does Design Matter?​

Without a sound design:

  • Results may be attributable to extraneous variables rather than the independent variable
  • Findings may lack internal validity (did our manipulation cause the effect?) and external validity (do findings generalise?)
  • Statistical conclusions may be misleading due to uncontrolled error variance

A poorly designed study wastes resources and may produce conclusions that mislead policy, clinical practice, or theory development.

1.3 Function of Research Design​

Kerlinger (2007) identified two basic purposes of research design:

  1. Provide answers to research questions — The design specifies exactly what data to collect and how to analyse it to test hypotheses.
  2. Control variance — The design maximises sensitivity by managing different sources of variation.

The MAXMINCON Principle​

The statistical heart of research design is the MAXMINCON Principle, which governs three types of variance:

Variance TypeDefinitionControl Strategy
Systematic VarianceVariability in DV due to the IV manipulation — the signal we wantMaximise by designing strong, meaningful treatment differences
Extraneous VarianceVariability due to variables other than the IV that may influence the DVControl via randomisation, elimination, or matching
Error VarianceVariability from uncontrollable factors (measurement error, individual differences not accounted for)Minimise through standardisation and reliable instruments

Indian Context: The MAXMINCON principle is directly applicable in NIMHANS research protocols. For instance, a study on yoga-based interventions for anxiety must maximise the yoga vs. no-yoga difference (clear protocol), control for prior meditation experience (matching), and minimise measurement noise (standardised questionnaires like STAI administered under identical conditions).

1.4 Basic Terminology in Research Design​

Before studying specific designs, mastering the core vocabulary is essential. These terms appear constantly in research reports, IGNOU exam questions, and journal articles.

i) Factor​

The independent variable of an experiment is known as a factor. An experiment always has at least one factor.

  • In a study of the effect of caffeine on alertness → caffeine is the factor
  • In a study of the effect of teaching method on achievement → teaching method is the factor

ii) Level​

A level is a particular value of an independent variable. An independent variable has at least two levels (otherwise, there is no comparison to make).

Example: If the factor is reward, it might have two levels:

  • Level 1 → Reward present
  • Level 2 → Reward absent (control)

If the factor is study strategy, it might have three levels:

  • Level 1 → Rereading
  • Level 2 → Highlighting
  • Level 3 → Self-testing

iii) Treatment​

A treatment refers to a particular set of experimental conditions. In a 2 × 2 factorial experiment, subjects are assigned to 4 treatments (all combinations of 2 factors × 2 levels each).

Quick Summary Table:

TermDefinitionExample
FactorIndependent variableTeaching method
LevelSpecific value of the IVMethod A, Method B, Method C
TreatmentA specific experimental conditionStudents taught by Method B

Memory Aids 🧠​

MAXMINCON Mnemonic​

"MAX the Signal, MIN the Noise, CONtrol the Confounds"

Or remember it as a research mantra:

  • MAXimise → Make your IV strong enough to create real differences
  • MINimise → Reduce random fluctuations (error) through standardisation
  • CONtrol → Eliminate/neutralise competing explanations

Factor-Level-Treatment Chain​

"FLT = Factory Line Test"

  • Factory = Factor (the IV — what you're testing)
  • Line = Level (specific values of the factory setting)
  • Test = Treatment (the actual condition each subject experiences)

🎓 Self-Assessment Questions​

Level 1 — Recall

  1. Write the formal definition of research design as given by Kerlinger (2007). What are the three components (plan, structure, strategy)?
  2. What does MAXMINCON stand for, and what are the three types of variance it addresses?
  3. Define factor, level, and treatment. Give a single example that illustrates all three.

Level 2 — Application 4. A researcher studies the effect of noise level (quiet, moderate, loud) on reading comprehension. Identify: (a) the factor, (b) the levels, (c) a possible extraneous variable, and (d) how the MAXMINCON principle should guide the design. 5. Two students argue about research design: Student A says "Just collect lots of data and analyse it." Student B says "You need a design before you collect a single data point." Who is right, and why?

Level 3 — Analysis (IGNOU Exam Style) 6. A researcher claims their study proves that mindfulness reduces depression. However, they only measured depression after the mindfulness program began (no pre-test), and participants chose whether to join the program. Identify the design flaws in terms of variance control using MAXMINCON.

📚 External Resources​

Essential Reading​

Educational Videos​

Academic Resources​

  • 📖 Kerlinger, F.N. (2007). Foundations of Behavioural Research (10th reprint). Delhi: Surjeet Publications. — Primary textbook for this unit
  • 📖 Thyer, B.A. (1993). Single-System research design. In R.M. Grinnell (ed.), Social Work, Research and Evaluation (4th ed.). Itasca, Illinois: F.E. Peacock Publishers.

Recent Research (2020–2025)​

  • 🔬 Lakens, D. (2022). Sample size justification. Collabra: Psychology, 8(1), 33267. — Modern perspective on designing studies with adequate power
  • 🔬 Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2021). Pre-registration: Why and how. Journal of Consumer Psychology, 31(1), 151–162. — How pre-registration strengthens research design

Indian Context Resources​

🔗 Cross-References​

Research design is not bureaucratic paperwork — it is the intellectual architecture that transforms curiosity into reliable knowledge. A strong design is the single most important decision a researcher makes.