PsychologyOrbit
Sign in Join

Spot the flaw

Past research methods puzzles

Each puzzle is a short fictional study with one flaw. Its page gives the answer and explains the methods concept behind it. Today's puzzle is not here until tomorrow.

  1. #18 Office noise and attention Cognitive · Cardiff, UK · Counterbalancing in repeated-measures designs
  2. #17 Manager praise and performance Work & organisational · Birmingham, UK · Which way does the arrow point?
  3. #16 A classroom behaviour programme Educational · Chicago, US · Blinding outcome assessors
  4. #15 Lessons from long-lasting founders Work & organisational · London, UK · Studying only the successes
  5. #14 Ranking teams by engagement Work & organisational · Bengaluru, India · Reliability limits what a measure can support
  6. #13 What a positive screen means Clinical · Seattle, US · Positive predictive value and prevalence
  7. #12 A sleep-hygiene course Health · Glasgow, UK · Matching the test to the design
  8. #11 Ranking two depression clinics Clinical · Dhaka, Bangladesh · Aggregated data can reverse a trend
  9. #10 Upbeat music and mood Social · Islamabad, Pakistan · Participants guessing the hypothesis
  10. #9 A resilience course at work Work & organisational · Ohio, US · Intention-to-treat and dropout bias
  11. #8 Memory training for older adults Cognitive · Chennai, India · Measures too easy to detect change
  12. #7 Dog owners and depression Health · Leeds, UK · Correlation does not establish causation
  13. #6 Group therapy for insomnia Clinical · Karachi, Pakistan · Statistical power and null results
  14. #5 Favourite colours and personality Social · Boston, US · False positives from many tests
  15. #4 How common is smartphone addiction? Health · Delhi, India · Representative sampling for prevalence estimates
  16. #3 A new phonics programme in first grade Educational · Texas, US · Confounding variables in group comparisons
  17. #2 A self-help leaflet for anxiety Clinical · Manchester, UK · Regression to the mean after extreme scores
  18. #1 A mindfulness app before exams Clinical · Lahore, Pakistan · Why a study needs a control group