Classification Flashcards
what is the purpose of classification?
- theoretical development
- guides assessment practices
- guides treatment practices
what does a good set of criteria allow?
basis of comparative study and when research is successful then can be useful for informing assessment practices
- treatment is individual based
what does classifying into meaningful groups help with?
identification of between-group differences
why should you not classify?
- individual differences may be obscured (within group variability)
> knowledge base concerning may be irrelevant to patient - preempt more meaningful diagnostic exploration
- categorization often seen as pejorative (gut feeling, treat people not disorders, replace identity with illness/diagnosis)
-> used in legal defences: redeeming feature you failed to look for after diagnosing
what do humans have a strong tendency to do?
categorize
- learn via associations
- we acknowledge we do this and proceed systematically
- recognize commonalities among group members
weak system characteristic
devoid of any organizing principle
-> impressionistic sys - poor psychometric capability so poor chances of survival, not practical
- individual may belong to multiple groups
- DSM-5 based on impression and processors
what are some issues where classification is relevant?
- level of risk
- institutional housing decisions
- degree of supervision vs privileges
- offender needs
- treatment inclusion -> D/A, anger management, meds
are between=group differences theoretically informative?
- developmental trajectories/rf
- circumstantial risk management post-release -> specific treatment plans for individuals
Chaiken study
8 dimensions based on past crimes
- 256 combos
- 10 accounted for 59% inmates
- is the difference for ex between certain crimes meaningful? no rational
what did Megargee describe as a good system?
- comprehensive - every offender
- unambiguous - clearly defined categories
- statistically reliable assignment procedures
- type distinction are valid - what separates one category from another
- sensitive to individual changes - rehab with expectation of lowering risk
- treatment relevant - deliver appropriate treatment for offender
- economical to apply
what were the old systems based on?
impressions have borne little fruit - term more used despite lack of precision
- cohesive personality/criminogenic theory (driven to certain type of behavior)
- empirically driven systems - more testable and most accurate - cluster and factor analysis
what is the impressionistic system?
- individual or group of indiviauls think they recognize pattern or type of some members of clinical/institutional sample
- not reliable
- no theory or systematic method of measurement guides impression
- essential features identified after fact
- lots of problem with reliability -> prognostic validity very poor
** utility of classification system
what is now a diagnostic valid tool?
dsm-5
what were problems with the original dsm-5?
personality disorder with cluster b features
- many NOS categories
- systematic disagreements - geographically diagnosing differences
- definitions are highly behavioral and predictions are circular
- assumes discontinuity amount personality disorders
-> discrete taxons
circular preditcions
- diagnosis/symptoms talk about people being more inclined to breaking rules
- some behave in way that draw them to police
-> APD and BPD dont far behind B cluster
what system stems from developmental or personality theory?
theoretically derived system
-> testing utility of related classification scheme puts whole theory on line
-> related specific attributes or points of maturation to antisocial behavior
-> appeal to empathic capacity, social reasoning, intelligence
-> measurement instruments draw heavily from concepts central to theory -> significant bias (narroed perspective_
what did Hunt and Hardt create?
conceptual level
- ability to adopt other perspectives
- attempt to tap reasoning process more than outcome
- assessed via incomplete sentences task
what do both system represent?
efforts to match treatment to offenders need/level - responsivity principle
-> lower conceptual = more behaviorally-oriented
-> higher conceptual level kids paired with staff more sensitive to manipulation attempts
-> favorable outcomes over general assignment
what are the empirically derived classifications based on?
- cluster analysis
- factor analysis
cluster analysis
- numerical taxonomy
- group subjects according to level of similarity as measured by variables selected by the researcher
- not including useful variable so analysis is useless
factor analysis
correlational
- reduce large number of variables to smaller number
- underlying dimensions
- most predictor variables are highly correlated with one another
what did Megargee and Bohn do?
cluster analysis
- MMPI based scheme (MOST pop)
- clinical personality measure
- clinical interpretation based on individual scale configurations as well as profile analysis - gut feeling
- unlike type 1 error, results have been replication but predictive accuracy is in question
problems with mmpi
high scale overlap
smaller number is discriminating
- profile derivation atheoretical -> results may be unqiue and not theoretically informative on own
what did Hewitt and jenkins do?
behavioral data from 500 delinquent children
- based on unsocialized aggressive behavior -> poor prognosis
- socialized delinquency -> better prognosis
- over-inhibited -> more likely to be mentally ill
- physical deficiency -> less interpretable
