Saral Shiksha Yojna
Courses/Behavioral Research: Statistical Methods

Behavioral Research: Statistical Methods

CG3.402
Vinoo Alluri•Monsoon 2025-26•4 credits
Sample Papers/Mock Paper 1 — All Units Mix

Mock Paper 1 — All Units Mix

Duration: 120 min • Max marks: 100

Section A — MCQ (20 × 1 = 20)

20 marks
  1. 1.Reaction time of 400 ms is twice as fast as 800 ms. The statement is valid because RT is on which scale? (a) Nominal (b) Ordinal (c) Interval (d) Ratio1 m
  2. 2.Correct frequentist interpretation of a 95% CI: (a) 95% probability the parameter lies in this interval (b) 95% of data lie in this interval (c) If we repeated the procedure many times, 95% of such intervals would contain the parameter (d) The parameter is in this interval 95% of the time1 m
  3. 3.A researcher reports p = 0.03. Which statement is correct? (a) P(H₀ true) = 3% (b) Assuming H₀ true, 3% chance of data this extreme or more (c) 3% chance the finding is a fluke (d) 97% chance H₁ is true1 m
  4. 4.Three independent groups, normality severely violated. Test? (a) One-way ANOVA (b) Repeated-measures ANOVA (c) Kruskal-Wallis (d) Friedman1 m
  5. 5.CLT says: (a) Any large sample is Normal (b) Sampling distribution of sample mean approaches Normal as n grows (c) Population must be Normal (d) Outliers cancel in large samples1 m
  6. 6.Mauchly's test of sphericity is relevant for: (a) Chi-square (b) Repeated-measures ANOVA (c) Independent t-test (d) Pearson r1 m
  7. 7.A scale consistently reads 5 kg above true weight: (a) Reliable but not valid (b) Valid but not reliable (c) Both (d) Neither1 m
  8. 8.Bonferroni with m = 10 tests at family-wise α = 0.05 sets each test's α to: (a) 0.05 (b) 0.005 (c) 0.5 (d) 1 − 0.95¹⁰1 m
  9. 9.R command for P(≤ 6 heads in 10 fair flips): (a) dbinom(6,10,0.5) (b) pbinom(6,10,0.5) (c) qbinom (d) rbinom1 m
  10. 10.Bayesian posterior is proportional to: (a) prior × evidence (b) prior × likelihood (c) likelihood ÷ prior (d) evidence ÷ prior1 m
  11. 11.Pearson r = 0 between two variables. The correct conclusion: (a) They are independent (b) No linear association; nonlinear possible (c) Identical means (d) One causes the other inversely1 m
  12. 12.In a 2×3 contingency table, df for χ² independence test: (a) 5 (b) 6 (c) 2 (d) 11 m
  13. 13.Tiny effect, p < 0.001, n = 50,000. Conclude: (a) Very important effect (b) Real but practically meaningless (c) p-value wrong (d) H₀ definitely false1 m
  14. 14.NOT a measure of central tendency: (a) Mean (b) Median (c) Standard deviation (d) Mode1 m
  15. 15.Regression with R² = 0.4 means: (a) 40% predictions correct (b) Correlation is 0.4 (c) 40% of variance in Y explained by model (d) Slope is 0.41 m
  16. 16.Sphericity violation in RM-ANOVA is corrected by: (a) Welch (b) Greenhouse-Geisser (c) Bonferroni (d) Tukey HSD1 m
  17. 17.Bayes Factor BF₁₀ = 8 indicates: (a) Strong for H₀ (b) Anecdotal for H₁ (c) Moderate for H₁ (d) No evidence1 m
  18. 18.Best sampling method to ensure rare subgroups are represented: (a) Simple random (b) Convenience (c) Stratified (d) Snowball1 m
  19. 19.A scree plot is used in: (a) Hypothesis testing (b) Factor analysis / PCA to choose # factors (c) χ² tests (d) Regression diagnostics1 m
  20. 20.Failing to reject H₀ when H₀ is false: (a) Type I (b) Type II (c) Sampling (d) Measurement1 m

Section B — MSQ (10 × 2 = 20)

20 marks
  1. 1.Valid forms of reliability: (a) Test-retest (b) Construct (c) Inter-rater (d) Parallel forms (e) Ecological2 m
  2. 2.A ratio scale supports: (a) Ordering (b) Add/subtract (c) Multiply/divide (d) Meaningful mean (e) All of the above2 m
  3. 3.Threats to internal validity: (a) History (b) Selection bias (c) Practice / testing effects (d) Sampling from one university (e) Experimenter bias2 m
  4. 4.True of the t-distribution: (a) Heavier tails than Normal (b) Used when σ unknown (c) Shape depends on df (d) Symmetric around 0 (e) Approaches Normal as df → ∞2 m
  5. 5.Correct about p-hacking: (a) Optional stopping inflates Type I (b) Trying many analyses and reporting favorable one is p-hacking (c) Pre-registration helps (d) Reporting only the most interesting of many outcomes is fine (e) Selectively dropping participants until p < .05 is p-hacking2 m
  6. 6.Assumptions of OLS regression: (a) Linearity (b) Independence of errors (c) Homoscedasticity (d) Predictors must be normal (e) Normality of residuals2 m
  7. 7.Inflate family-wise Type I error: (a) Many independent t-tests on same data (b) Pre-specifying one hypothesis (c) Testing 50 outcomes (d) Many subgroup analyses (e) Bonferroni correction2 m
  8. 8.Correct Bayes-rule term mappings: (a) P(H) is prior (b) P(D|H) is posterior (c) P(D|H) is likelihood (d) P(H|D) is posterior (e) P(D) is the marginal probability of the data2 m
  9. 9.Non-parametric tests: (a) Mann-Whitney U (b) Independent t-test (c) Kruskal-Wallis (d) Friedman (e) One-way ANOVA2 m
  10. 10.Increase statistical power: (a) ↑ sample size (b) ↑ α from .01 to .05 (c) Within-subjects design (d) Reduce measurement noise (e) ↓ true effect size2 m

Section C — Short descriptive (6 × 5 = 30)

30 marks
  1. 1.A university shows overall bias against admitting women, but every individual department admits women at higher rates than men. Explain how this is possible.5 m
  2. 2.Distinguish reliability and validity. Can a measurement be valid without being reliable? Concrete example.5 m
  3. 3.Diagnostic test has 95% sensitivity and 5% false-positive rate. Disease prevalence is 2%. A patient tests positive. Compute P(disease | positive). Comment.5 m
  4. 4.Differentiate FWER and FDR. When use each? Which is more conservative?5 m
  5. 5.PCA vs Factor Analysis — conceptual differences and when to use each.5 m
  6. 6.A researcher claims that ice-cream sales correlate with drownings in a city. Critically evaluate.5 m

Section D — Long descriptive (3 × 10 = 30)

30 marks
  1. 1.A psychologist tests whether a new mindfulness app reduces anxiety. 60 students, randomly assigned 30 to app and 30 to waitlist; anxiety measured pre and 8 weeks later. (i) IV and DV with scales. (ii) Hypotheses. (iii) Recommend a test with justification. (iv) Assumptions. (v) Additional analyses.10 m
  2. 2.Multiple regression: Income = β₀ + β₁·Education + β₂·Experience + β₃·Age + ε, n = 200. R² = 0.42, adjR² = 0.41, F(3,196) = 47.3, p < .001. β₁ = 4500 (SE 800, p<.001); β₂ = 1200 (SE 400, p<.01); β₃ = 50 (SE 150, p = .74). VIFs 9.8 / 11.3 / 12.1. (i) Interpret R² and adjR². (ii) Interpret each β. (iii) Why is Age non-significant? (iv) What problem do VIFs reveal and what to do?10 m
  3. 3.Define statistical power. List five factors affecting it. A study finds p = 0.18 with n = 30/group and claims 'no effect of the treatment'. Critically evaluate and explain how Bayesian analysis helps.10 m

Track your attempt locally — score and time are recorded in your browser. (Coming soon: timed-attempt mode.)