Computer Science & AI · Level 3 · Advanced
Human-algorithm teaming in face recognition: How algorithm outcomes cognitively bias human decision-making
Howard JJ, Rabbitt LR, Sirotin YB (2020). PLOS ONE. doi:10.1371/journal.pone.0237855 · CC BY 4.0
Step 1 of 2Read the paper
When a computer or a person has already labelled two face photos as 'same person' or 'different people', does that label change your own judgement? Volunteers rated the similarity of difficult face pairs on a 7-point scale, either with no label or with labels said to come from a human or a computer, and the authors used signal detection theory to separate ability from bias.
How to read it
- Study Fig 2 and the survey design paragraph: three survey variants, with 'same'/'different' labels nested inside two of them.
- Learn the two SDT ideas before Results: sensitivity (d′) is ability; the criterion is the bias to say 'same'.
- Compare Tables 4, 5 and 6: which measures change with the label, and which stay the same?
- In the Conclusion, note how the authors' explanation differs from the earlier attention-based explanation.
Step 2 of 2Take the test
15 questions · about 50 minutes including reading · pick an answer to see straight away if it’s right. Nothing is saved.
Loading progress…
Paper: Howard JJ, Rabbitt LR, Sirotin YB, PLOS ONE (2020), licensed CC BY 4.0. Summary and questions are Research Nest’s own; read the original for the full study.