reliability and validity of classification and diagnosis

• reliability is the extent to which findings are consistent across all studies within the same phenomena
• relating to schizophrenia, reliability refers to the extent to which psychiatrists can agree on the same diagnosis when independently assessing patients (inter rater reliability)
• for the classification system of schizophrenia to be reliable - the same diagnosis should be made each time - different psychiatrists should be reaching the same decision
• inter rater reliability is measured by statistic (kappa score) - falls between 0 and 1 (1 being perfect score)
• kappa score of 0.7 or above is seen as good reliability
• kappa score for diagnosis of schizophrenia - 0.46 (means there is inconsistency on how schizophrenia is diagnosed by clinicians) - patients may be inaccurately diagnosed or misdiagnosed
• issue affecting reliability of schizo diagnosis - culture

  • research suggests variation between countries when diagnosing schizophrenia
  • example - description of patient given to 134 US and 194 british psychiatrists - 69% of US and 2% of british diagnosed patient - suggests symptoms of ethnic minorities often misinterpreted
  • findings question reliability of diagnosis - suggests patients display same symptoms and receive different diagnosis
  • patient experiences are different - Ghanda + India had positive hallucinations but US had more violent - environmental exposure can alter diagnosis
• issues affecting reliability of schizophrenia diagnosis

  • culture
• evaluation points for reliability of diagnosis for schizophrenia

  • unreliable symptoms (-)
  • cultural differences (-)
  • rosenhan research into schizophrenia (-) - support unreliability
• unreliable symptoms as eval point for reliability

  • when diagnosing schizo - only one characteristic symptom is required (is delusions are bizarre)
  • minimal requirements - creates problems for reliability of diagnosis
  • example - 50 US psychiatrists were asked to differentiate between 'bizarre' and 'non bizarre' delsuions - lacked consistency (correlation of 0.4)
  • concludes requirements in DSM for schizo lack reliability for it to be classed as an effective reliable method of distinguishing schizo from other disorders
• cultural differences as eval point for reliability

  • positive symptoms (hallucinations and hearing voices) seen as acceptable in some cultures (Africa) - beliefs in communication with ancestors - not seen as a symptom within their culture
  • however, when assessed by psychiatrist form different cultural background the experience of hearing voices can be seen as bizarre and irrational - psychiatrists is culturally biassed towards what's deemed as 'normal' within their own culture (ethnocentrism)
  • therefore deviation from what is seen as normal in some cultures can be misinterpreted by others
• rosenhan research into schizophrenia as eval point for reliability

  • further research supports unreliability of diagnosis for schizo
  • rosenhan conducted study were sane confederates went into psychiatric hospital and told health professionals they were experiencing hallucinations
  • confederates were consistently given same incorrect diagnosis of schizo - found once truth of study had be revealed - new pseudopatients werent admitted despite presenting same symptoms
  • suggests diagnosis of schizophrenia cannot be used reliability to distinguish between sane and insane individuals accurately
• validity is the extent to which what we are measuring is what we intent to measure
• in schizophrenia context - extent to which classification of schizophrenia is a true reflection of the illness a patient is suffering from
• issues affecting validity of classification for schizophrenia

  • co morbidity
  • symptom overlap
  • gender bias
• co morbidity being an issue affecting validity

  • co morbidity is when more than 1 disorder exists alongside a primary diagnoses (schizophrenia and depression)
  • when 2 conditions are diagnosed together it questions the validity of the classification of both illnesses - could be better off as one single condition
  • research support - concluded 1/2 patients with schizo diagnosis also have depression (50%) or substance abuse (47%) - highlights schizo as being co morbid
• symptom overlap being issue affecting validity

  • symptoms of disorder are not unique to 1 disorder - can be found in other disorders - accurate diagnosis difficult
  • example - schizophrenia + bipolar share same positive symptom of delusions and negative symptom of avolition - questions validity of classification
  • most schizophrenics receive at least one other diagnosis
  • difficult to find out what disorder one has because its under more than one symptom category
  • research support - pointed out people with DID have more schizophrenia symptoms than those with schizo
• gender bias being issue affecting validity

  • critics of DSM diagnostic criteria argue some diagnostic categories are biassed towards one gender
  • example - researcher found US clinicians identified a mentally healthy 'adult' with mentally healthy ' male' behaviours (androcentrism) - resulting in women being perceived as less mentally healthy when not showing 'male' behaviour
  • clinicians also biassed due to own beliefs towards genders
• evaluation points for validity of classification for schizophrenia

  • gender bias (-)
  • little predictive validity (-)
• gender bias as eval point for validity of schizophrenia

  • found enhanced gender bias upon diagnosing schizophrenia
  • researcher selected 290 male + female psychiatrists who then read 2 case articles of patients behaviour and then asked them to offer judgement on patients using standard diagnostic criteria
  • patients described as 'males' or not given gender - 56% given diagnosis
  • 'females' - 20% diagnosed
  • found gender bias was not evident amongst female psychiatrists
  • findings suggest diagnosis of individual is not only influences by patients gender but also gender of clinician
• little predictive validity as eval point for validity of schizophrenia

  • prognosis for patients diagnosed with schizo varies -10% achieving lasting improvement, 30% showing improvement but relapse
  • therefore diagnosis of schizo has little predictive validity as patients dont share the same prognosis
  • different patient outcomes mean its impossible to predict the certainty whether someone will recover or not
  • when diagnose we cant accurately state how likely patient will recover due to multiple influences affecting outcome such as family tolerance to schizo behaviour and social skills