Reliability and Validity in Classification and Diagnosis

HideShow resource information
  • Created by: imanilara
  • Created on: 29-02-16 21:55


Two classification systems-plus furhter published diagnostic criteria-affects reliability between clinicians

Reliability-likelihood of clinicians reaching same diagnosis for same patient

-No objective test like a physical illness-eg to test for a bacteria or x-ray
-In psychiatric medicine, no objective tests for detecting abnormal behaviours or thought processes

-Diagnosis by matching presenting symptoms against a list in a manual:
List of criteria are published in a manual such as DSM IV (Diagnostic and statistical manual) or ICD (international classification system for diseases)
Criteria include symptoms such as hallucinations/delusions/how long dysfunction present/to which degree normal functioning affected

1 of 7


Inter rater reliability:
Whaley 2001-inter rater reliability for diagnosis of schizophrenia as low as 0.11%

Divergent Prevalence Rates: 
Prior to the 1970s there was a significant difference in the prevalence rates of Sz in different countries+over time 
Copeland et al-description of patient to 134 US and British psychiatrists
69% of US diagnosed Sz
2% Brit diagnosed Sz
Suggests the diagnostic criteria were problematic-lack of reliability in diagnosis
Another example is that in the USA 20% of psychiatric patients in 1930, rose to 80% in 1950s- during this same time, maudsley hosp remained at 20%- USA has broader definitions 

2 of 7


Still a difference in the time in which symptoms must be present: ICD (1 month) DSm IV (6months)
Other criteria, eg scheider criteria

The use of other criteria can help improve the reliability of diagnosis. However the fact that other diagnostic criteria are used suggests that there are concerns about the validity of the manuals
Efforts have been made to bring ICD+DSM in line w eachother-
-more operatioalised, standardised interview techniques used to assess patients
-Farmer et al (1988) high levs of reliability using PSE (present state examination), however it depends on clinicians using same criteria on same diagnosis, when realistically this will not happen.

However existence of other criteria means there are issues of reliability and validity

3 of 7


For some of these, subjective decisions required

In the DSM criteria, one of the characteristic symptoms are "if delusions are bizarre" but what one clinician thinks is bizarre can differ to another- it is subjective and therefore leads to different decisions being made-lack of reliability

Rosenhan: 1973
Aim-distinguish whether staff in psych hospitals could distinguish between normal and abonormal mental states

1: Eight normal ppl went to admission departemnts claiming they could hear voices-7 were admitted w diagnosis of Sz-nobody recgonised
2:Hospital warned some pseudopatients would be admitted-none were-but 10% of real patients were suspected of being fakes

-They cannot reliably tell the difference
-peudopatients felt powerless and everything was a 'symptom'-diagnosis is a self fulfillinf prohecy

4 of 7


Cultural differences:

Psychs in China, India use different criteria, eg CCMD (chinese class of mental disorders)
Similar to ICD but not the same
Eg, neurasthenia or weakness of the nerves which covers up depression/Sz-avoid stigma
-prev of Sz is lower in these countries but criteria is unrealiable

5 of 7


If Sz is a valid concept it should allow us to describe the nature of the disease and predict the outcome of the disease


does it allow us to describe the nature of the disease?

Schneider 1959- proposed first rank symptoms: halluncinations/delusions
But many of these symptoms also occured in other disorders, Ellason found that people with DID had more Sz symptoms than people diagnosed as Sz
Hard to define the boundaries-mood disorders it is difficult, eg Sz+depression are comorbid, which symptoms belong to which, eg avolition
They tried to sort this by addressing the problem and creating mixed disorder categories-schizo-affective disorder-isnt this more confusing?
Wider issue of imposing classification system on to a phenomenon as complex sa mental illness-isnt eveyr case unique and so isnt diagnosis and class. doomed from the start? However, diagnosis does have pos. outcomes

6 of 7


-predict outcome, ie the course of disease+treatment?

Prognosis for patients diagnosed as suffering from Sz varies:
-20% recover their previous level of functioning
-10% achieve lasting improvement
-30% show some improvement with some relapses

Diagnosis has little predictive validity

Factors that affect outcome:
Gender-women have mroe positive outcomes
Psychosocial afctors-social skills and fam tolerance of behaviour

Gender diff influenced by psychosocial-women do better in all these areas

7 of 7


No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Schizophrenia resources »