2 Methods in Psychology · 2.1 Methods in psychology
2.1.1 Scientific method, variables, hypotheses and sampling
Psychological research follows the scientific method: pose a problem, state a hypothesis about how an independent variable affects a dependent variable, control extraneous variables, collect data from a sample that represents the population, and draw conclusions others can repeat.
- APPSC Paper I
- Standard level
- 12 min read
- Free
Core facts
- Scientific study = the systematic study of facts by a reliable and correct method; its marks are being empirical, objective, systematic, replicable and self-correcting.
- Four goals of science (NCERT): description, prediction, explanation, control (changing behaviour, e.g. through therapy or training).
- Steps (NCERT): (1) conceptualising a problem, (2) collecting data, (3) drawing conclusions, (4) revising conclusions. Textbook long form: observe → state the problem → hypothesis → test / collect data → analyse → conclude → report and replicate.
- Four types of psychological data (NCERT): demographic (name, age, sex, education, occupation, caste, religion), physical (rural or urban, house size, room, climate), physiological (heart rate, blood pressure, reaction time, EEG, fatigue), psychological (intelligence, personality, emotions, attitudes, values).
- Quantitative data = numbers (scores, counts); qualitative data = non-numerical (words, narratives, observations). The main difference between the two methods is the type of data they collect.
- Variable: any event, characteristic or behaviour that varies (takes more than one value), e.g. age, noise, anxiety.
- Independent variable (IV): the one the researcher manipulates (the cause). Dependent variable (DV): the one measured (the effect). These are the two main types.
- Extraneous variable: any other factor that can affect the DV; if it changes along with the IV it becomes a confounding variable.
- Operational definition: defining a variable by the procedure used to measure or produce it (anxiety = score on an anxiety scale).
- Hypothesis: a tentative, testable statement about the relationship between two or more variables.
- Null hypothesis (H₀): no difference or relationship. Alternative / research hypothesis (H₁): there is one.
- Population (universe): all the members of a defined group. Sample: a subset chosen to represent it.
- Probability (random) sampling: simple random, systematic, stratified, cluster, multistage. Non-probability: convenience (accidental), purposive (judgement), quota, snowball. Mechanics and sampling errors: see STA-DS-02.
Explained
1. Characteristics of psychological studies (the syllabus wording).
- Empirical: based on observation and measurement, not opinion.
- Objective: two researchers looking at the same behaviour should record it the same way.
- Systematic: follows planned steps.
- Replicable: another researcher using the same procedure should get similar results.
- Self-correcting: wrong conclusions are revised when new data come.
Psychological data are, however, relative and context-bound: behaviour depends on the person, the time and the setting, and the data are often about subjective experience reported by the person. So findings are statements of probability, not exact laws. Methods that study it: observation, experiment, correlational research, survey, case study, psychological tests (see EDU-PS-02 for each method).
2. The goals in a probation setting.
| Goal | Question | Example |
|---|---|---|
| Description | What happens? | How many juveniles in a home reoffend |
| Prediction | When will it happen? | Peer delinquency predicts reoffending |
| Explanation | Why does it happen? | Weak family supervision leads to delinquent peers |
| Control | How can it be changed? | Family counselling reduces reoffending |
3. Variables: kinds and examples. Experiment: "Does noise affect memory?"
- IV: noise (present / absent), set by the researcher.
- DV: number of words recalled, measured.
- Extraneous: time of day, practice, the subjects' intelligence and fatigue.
- Levels: the values the IV takes (e.g. 0, 60 and 90 decibels: three levels).
Other names you will see:
- Organismic (subject) variables: age, sex, intelligence, personality: traits of the participant, which cannot be manipulated, only selected.
- Situational variables: light, temperature, noise, instructions.
- Sequence variables: practice and fatigue when the same person does several tasks (controlled by counterbalancing).
- Intervening (mediating) variable: an inner process between IV and DV that cannot be seen (hunger, motivation, learning); term from Tolman.
- Moderator variable: changes the strength of the IV–DV link (stress harms performance more in anxious people).
- Continuous (height, reaction time) vs discrete / categorical (sex, religion, number of children).
4. Operational definitions. "Aggression" is vague; "number of times a child hits the Bobo doll in 10 minutes" is operational. Operational definitions make studies measurable and repeatable. The idea comes from physicist P.W. Bridgman (operationism, 1927).
5. Hypotheses.
| Type | Says | Example |
|---|---|---|
| Null (H₀) | No effect or no relationship | Noise has no effect on recall |
| Alternative / research (H₁) | There is an effect | Noise affects recall |
| Directional (one-tailed) | States the direction | Noise reduces recall |
| Non-directional (two-tailed) | Effect, direction not stated | Noise changes recall |
A good hypothesis is testable, falsifiable (could be shown wrong), specific (names the variables), simple and grounded in earlier theory or observation. Statistics test the null hypothesis: we reject H₀ or fail to reject it; we never "prove" H₁.
6. Population and sample. A parameter describes the population (population mean μ); a statistic describes the sample (sample mean). The sampling frame is the list from which the sample is drawn. A sample is useful only if it is representative: large enough and chosen without bias. Sampling follows the scientific approach: results from a well-drawn set of units are generalised to the whole population (APPSC 2023).
7. Sampling methods in psychological research.
| Method | How | Psychology example |
|---|---|---|
| Simple random | Every member has an equal chance (lottery, random numbers) | 100 names drawn from a list of all inmates |
| Systematic | Every k-th name after a random start | Every 10th case file |
| Stratified | Divide into strata, sample from each | Boys and girls in proportion to their numbers |
| Cluster | Pick whole groups at random | Choose 5 observation homes, test every child |
| Convenience | Whoever is easily available | Students in one's own college |
| Purposive | Researcher picks typical or expert cases | Only probation officers with 10+ years' service |
| Quota | Fill set numbers per category, non-randomly | 50 men and 50 women, whoever comes first |
| Snowball | Each participant refers others | Drug users, sex workers: hard-to-reach groups |
Random sampling (who is chosen for the study) is not the same as random assignment (who goes to the experimental or control group). Sampling protects generalisation; assignment protects cause-and-effect conclusions.
Common traps
- ✗ "The dependent variable is the one the researcher changes." ✓ The researcher changes the independent variable; the dependent variable is measured.
- ✗ "The two main types of variables are field and case variables." ✓ Independent and dependent variables.
- ✗ "A hypothesis is a proven fact." ✓ A tentative, testable statement about the relationship between variables.
- ✗ "The null hypothesis states the expected difference." ✓ It states no difference; the alternative states the difference.
- ✗ "Quantitative and qualitative research differ in the type of hypotheses." ✓ They differ in the type of data they collect (numbers vs words).
- ✗ "Making a stereotype or a table is a step of the scientific method." ✓ The steps begin with making an observation / conceptualising a problem (APPSC 2023 keyed "make an observation"; drawing a conclusion is also a step, so read such options carefully).
- ✗ "Quota sampling is a random method." ✓ Quota, convenience, purposive and snowball are non-probability methods; stratified is the random one that uses categories.
- ✗ "Random sampling and random assignment are the same." ✓ Sampling picks who is studied; assignment decides which group each joins.
- ✗ "Type I error = accepting a false null." ✓ Type I = rejecting a true null; Type II = accepting a false null.
Quick revision
- Goals: describe, predict, explain, control.
- Steps (NCERT): conceptualise → collect data → conclude → revise.
- Data types: demographic, physical, physiological, psychological.
- Quantitative = numbers; qualitative = non-numerical.
- IV manipulated (cause); DV measured (effect); extraneous → confound.
- Operational definition = defined by how it is measured.
- Hypothesis = testable statement relating two or more variables.
- H₀ = no effect; H₁ = effect; directional = one-tailed.
- Population = all; sample = representative subset.
- Random: simple, systematic, stratified, cluster; non-random: convenience, purposive, quota, snowball.
Go deeper: Deeper (beyond what APPSC Paper I usually asks)
Deeper Advanced
1. Errors in testing hypotheses.
- Type I error (α): rejecting a true null hypothesis (a false alarm: finding an effect that is not there). The level of significance (0.05 or 0.01) is the risk of this error we accept.
- Type II error (β): failing to reject a false null hypothesis (missing a real effect). Larger samples reduce it; power = 1 − β.
2. Control of extraneous variables. Elimination, constancy, matching, randomisation and counterbalancing (detail in EDU-PS-02). Two biases need special controls: experimenter bias (the researcher's expectations shape results; controlled by a double-blind design) and demand characteristics or the placebo effect (subjects guess the aim or expect a benefit; controlled by a single-blind design, where subjects do not know their condition, and a placebo group). A double-blind design controls both biases at once.
3. Internal vs external validity. Internal validity: the change in the DV is due to the IV, not a confound (strong in laboratory experiments). External validity: findings generalise to other people and settings (strong in field studies and with representative samples). Tight control usually trades one for the other.
4. Correlation is not causation. A correlational study shows that two variables move together (school failure and delinquency) but not which causes which, or whether a third variable (poverty) causes both. Only an experiment with manipulation and random assignment supports a causal claim.
5. Ethics in psychological research (NCERT). Voluntary participation, informed consent, debriefing after the study (especially if deception was used), confidentiality of data, and sharing results with participants. Children and people in custody need extra protection: consent from a guardian or the authority, and no pressure linked to release or benefits.
Practise this topic. Questions on Scientific method, variables, hypotheses and sampling, with answers explained.
All Methods in psychology notes · APPSC Paper I syllabus · APPSC Paper I Mock 1 with answers · Start free
Medha AI writes its own notes and checks them against the sources listed. Medha AI is an independent preparation service. It is not affiliated with, endorsed by or connected to APPSC, TSPSC, MPSC, UPSC or any other commission or government body.