Producing Knowledge in Psychology
What is psychology trying to understand?
The study of mind and behaviour
Why people think, feel, and act as they do, covering everything from basic brain functions, emotions, and personality to complex social interactions, development, and mental health, with the ultimate goals of describing, explaining, predicting, and improving human life and well-being.
What counts as knowledge in psychology?
Controlled experiments - 𝜟s one thing on purpose (the independent variable) to see what effect it has on something else (the dependent variable)
How they produce knowledge
They allow researchers to test cause-and-effect relationships.
Standardised procedures make the data more scientific and objective.
How reliable the knowledge is
Strengths:
High internal validity because of tight control.
Easy to replicate, increasing reliability.
Weaknesses:
Often low ecological validity (the setting is artificial).
Participants may show demand characteristics.
Brain scans - neuroimaging methods such as fMRI, PET, EEG, or CT that show brain structure or activity while a person performs tasks
How they produce knowledge
Provide biological data about which brain areas are linked to certain behaviours.
Support the biological approach by showing brain–behaviour relationships.
How reliable is the knowledge?
Strengths:
Highly objective data.
Non-invasive and scientifically precise.
Weaknesses:
Brain activation shows correlation, not causation.
Expensive → smaller sample sizes.
Behaviour patterns - researchers watch and record behaviour in natural or controlled settings
How they produce knowledge
Show how people behave naturally.
Help identify patterns, social behaviours, and interactions.
How reliable is the knowledge?
Strengths:
High ecological validity (especially naturalistic).
Useful when behaviour cannot be manipulated.
Weaknesses:
Observer bias can reduce reliability.
Difficult to establish cause-and-effect.
Survey data - questionnaires or structured interviews where people report their own thoughts, feelings, or behaviour
How they produce knowledge
Gather large amounts of data quickly.
Allow access to psychological constructs (e.g., stress levels, attitudes) that cannot be directly observed.
How reliable is the knowledge?
Strengths:
Large samples increase generalisability.
Standardised questions improve reliability.
Weaknesses:
Social desirability → people may not answer honestly.
Memory can be inaccurate.
Rosenhan’s Experiment (1973)
“On Being Sane in Insane Places” 🏥🧠
Aim 🎯
To investigate whether psychiatrists can distinguish between “sane” and “insane” people in psychiatric hospitals.
Procedure 🔍
Rosenhan sent 8 “pseudopatients” (mentally healthy people) to 12 psychiatric hospitals in the US.
They all said they heard a single voice saying “empty,” “hollow,” “thud.”
Once admitted, they stopped pretending and behaved normally.
Their task was to get discharged by acting sane.
Findings 📊
All pseudopatients were admitted, most with a diagnosis of schizophrenia.
None were detected as healthy by hospital staff.
Average stay was 19 days (range: 7–52 days).
Normal behaviour (like writing notes) was interpreted as a sign of disorder.
Staff often ignored patients; depersonalisation was common.
After the study, a hospital challenged Rosenhan to send pseudopatients.
The hospital flagged 41 out of 193 patients as “fake.”
Rosenhan had actually sent zero pseudopatients.
→ Shows the system makes both false positives and false negatives.
Conclusion 🧩
Psychiatric diagnosis at the time was unreliable.
Hospitals often could not distinguish healthy from sick individuals.
The labels (e.g., “schizophrenia”) shaped how all behaviour was interpreted.
The environment itself created loss of identity and power for patients.
Strengths 💪
Powerful real-world evidence (high ecological validity).
Sparked major reforms in psychiatric diagnosis and treatment.
Demonstrated the impact of labels and context on behaviour.
Limitations ⚠
Ethical issues: hospitals and staff were not informed → deception.
Could have taken up beds needed by real patients.
Pseudopatients were highly trained and may not represent typical patients.
Interpretations may have been biased against psychiatry.