Reliability & Validity

Question 1

How should classification and diagnosis of schizophrenia be?

Answer 1

• Relatively easy - use a ‘checklist’ that reduces the problem of misdiagnosis
• However isn’t always easy - clinician’s subjective opinion
• Different too objective methods used in medicine e.g. diagnosing diabetes by testing blood-glucose level

Question 2

What is reliability?

Answer 2

• Consistency of symptom measurement across clinicians
• Affects diagnosis in 2 ways:
  1) Test-retest reliability = extent the same clinician makes same diagnosis with same info
  2) Inter-rater reliability = extent different clinicians agree on same diagnosis
• Measured by Kappa score, 1 = perfect inter-rater agreement, 0 = no agreement, 0.7+ = generally ‘reliable’

Question 3

What is validity?

Answer 3

• Assesses what it claims to measure - patients diagnosed actually have the condition and its separate from all other disorders’ symptoms

Question 4

What are the 4 main factors which lead to an unreliable and invalid diagnosis of schizophrenia?

Answer 4

• Co-morbidity = people diagnosed with one disorder simultaneously showing symptoms of another -> may be confusion of which actual disorder is being diagnosed.
• Gender Bias = accuracy of diagnosis depends on gender -> Longenecker revived studies -> since 80s men have been diagnosed more often.
• Symptom Overlap = of schizophrenia and other conditions e.g. bipolar disorder also involves delusions and avolition
• Cultural Bias = different countries have different systems - USA use DSM, Britain ICD -> may be diagnosed in one place but not in another.

Question 5

What are 2 research studies demonstrating problems with reliability and validity?

Answer 5

Research support illustrating problems with reliability and validity:
- Rosenhan + Seligman ‘s study - to see if psychiatrists can tell the difference between sane and insane
- 8 people with no history turned up at hospitals complaining of auditory hallucinations.
- 7/8 were classified as schizophrenic
- Participants started to say they felt fine - took from 7-52 days to be discharged
- Not one of the professionals detected a pseudo-patient
- Estimated a total of 2,100 tablets were given to pseudo patients, although only 2 were swallowed (rest pocketed or flushed)
This means psychiatrists may not be able to distinguish people with a disorder from those with no such disorder. If diagnosed, everything they say or do may be taken as a symptom of the disorder.

Research support for co-morbidity:
- Buckley et al estimate depression occurs in 50% patients
- 47% suffer substance abuse
- This means its imp for clinician to make dual diagnosis for appropriate treatment for both disorders
This demonstrates the problem with making a valid diagnosis as its difficult to ascertain if patients displaying a symptom as a result of that condition, another, or both.

Question 6

What is a limitation of reliability and validity?

Answer 6

Problem with misdiagnosis - wrongly prescribed medication:
- Typically given antipsychotics - can have harmful side effects and are sedatives e.g. tardive dyskinesia
- Real patients may dispose of medication, behaviour may go unnoticed as long as they are cooperative
- Suggests drug treatment acts as a straitjacket to control their behaviours.
This shows how doctors have to be certain of their diagnosis, which increases the importance of a valid and reliable system.