Why do people with schizophrenia struggle to tell faces apart?
People with schizophrenia needed more contrast to identify faces because their visual processing was noisier, not because they used facial information less efficiently.
Source
Noise as a mechanism of anomalous face processing among persons with Schizophrenia
Study at a glance
- Design
- Case-control — Patients vs healthy controls on a two-alternative face match-to-sample task; contrast thresholds measured by staircase for upright and inverted faces at low and high external noise, fitted with a linear amplifier model
- N
- N=47 · 23 people with schizophrenia-spectrum diagnoses and 24 controls; internal-noise analyses used subsets of 20 per group (upright) and 36 participants in total (inverted)
- Population
- Medicated adults with schizophrenia, schizoaffective or schizophreniform disorder and healthy controls matched on age, education and estimated IQ, in Hamilton, Canada
- Outcome
- Face identification contrast threshold; estimated equivalent internal noise and calculation efficiency
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Twenty-three people with schizophrenia and 24 healthy controls saw a face briefly, then had to pick it out from two faces shown side by side. The contrast of the test faces was adjusted trial by trial to find the level needed for 79% correct, both for upright and upside-down faces and with or without visual noise added to the faces. Comparing thresholds with and without external noise let the researchers separate two possible causes of poor performance: more random noise inside the observer's visual system, or less efficient use of the useful information in the face.
What they found
People with schizophrenia had higher thresholds overall, and everyone did worse with inverted faces and with added noise; the patient deficit was similar across these conditions. Estimated internal noise was significantly higher in patients for upright faces and showed a trend in the same direction for inverted faces. Calculation efficiency did not differ between groups for either orientation, although it was lower for inverted faces in both groups.
The limits
What it doesn't show
Internal noise could only be estimated for participants whose threshold rose with added noise, so several people per group were dropped, leaving 20 per group for upright faces and fewer for inverted faces. Only two noise levels were used, which gives rough estimates of the masking function. All patients were medicated and relatively high-functioning (matched to controls on IQ and education), so medication effects cannot be ruled out and results may not generalise; nor does the study show whether higher internal noise is specific to schizophrenia or to faces. The source of the extra noise (for example, eye jitter or neural instability) was not measured.
Key terms
- Internal noise
- Random variability within the observer's own sensory and neural processing that limits how well a signal can be detected.
- Calculation efficiency
- How well an observer uses the available information in a stimulus compared with an ideal observer; low efficiency means relying on less informative features.
- Noise masking
- Adding controlled external visual noise to a stimulus and measuring how thresholds change, which lets researchers estimate internal noise and efficiency.
- Contrast threshold
- The minimum contrast at which a person reaches a set level of accuracy, here 79% correct.
- Face inversion effect
- The finding that upside-down faces are much harder to recognise than upright ones, taken as a sign of specialised face processing.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What was the main conclusion about the face-processing deficit in schizophrenia?
Common questions
How can adding noise reveal what is going on inside the observer?
If thresholds rise at a certain rate with external noise, the point where external noise starts to matter reflects internal noise, and the rate of rise reflects efficiency. Two groups with different internal noise give parallel lines with shifted intercepts; different efficiency gives different slopes.
Why does this matter for therapy?
Training that tells patients to look at the eyes targets efficiency, which was normal here. If the problem is noise, giving more time to accumulate information might help more, as the authors suggest.
Is this deficit specific to faces?
Unknown. The authors suggest increased internal noise may underlie other visual deficits in schizophrenia, but this study only tested faces.
More on Face processing