Reliability And Validity
- Test-retest reliability: extent to which ppl being observed perform taking the same test twice,
sometimes weeks apart.
- Alternate-form Reliability: patients are tested twice but with 2 different types of tests, in case
they aim for consistency and try to mimic the same answers as their previous test when given
- Inter Consistency Reliability: Asses whether the items on a test relate to one another. A person
with anxiety who says yes to dry mouth should also say yes to muscle tension.
Generally related to whether a measure fulfills its intended purpose.
Ex: Does a questionnaire intended to measure hostility really do that?
Note that validity is related to reliability: Unreliable measures will not have good validity.
-Content Validity: Refers to whether the measure adequately samples the domain of
interest. Ex: Axis I Diagnostic test has high content validity because it asks questions
about all the symptoms involved in Axis I diagnosis.
-Criterion Validity: Evaluated by determining whether a measure is associated in an
expected way with some other measure (the criterion). Ex: In a test that measures
depression, depressed people should score higher than non-depressed people. If these
relationships are measured at the same time, it is known as Concurrent Validity. Also,
you can have Predictive Validity which is the ability of a measure to predict some other
variable that is measured in the future. Ex: IQ tests were made to predict future school
-Construct Validity: The effect to which scores on an assessment relate to some hypothesis
or theory. A construct is an inferred attribute such as anxiousness. Ex: if people with an
anxiety disorder and those without take an anxiety test, if the people with the disorder
come up with a higher score, the test has construct validity. Also, when people’s self-
reports on tests match up with observations of their clinicians, the construct validity
increases. [I DON’T REEEEAAAALY GET THIS]
Psychological assessment techniques are designed to determine cognitive, emotional, personality
and behavioural factors in psychopathological functioning.
-Interviews are ways in which one person uses language as the means of finding out about
Characteristics of Clinical Interviews
-The interviewer notes not only the responses of the patient but the emotions in the patient
when he/she responds.
-The paradigm in which the interview works influences the type of information sought and
how it is obtained/interpreted.
-Clinicians operate from vauge outlines, how the information is actually obtained is up to
the interviewer and the responsiveness/responses of the interviewee.
-Reliability in clinical interviews is low, as 2 different interviews can lead to 2 different
interpretations. Sine it is confidential, it has been hard to obtained reliability or validity
through systematic research.
-Used for professionals that need to collect standardized information to make diagnostic
judgments based on the DSM-IV.
-Interview in which questions are set out in a prescribed fashion for interviewer.
-Ex: The SCID (Structured Clinical Interview Diagnosis) is branching, meaning the
clients response to 1 questions will lead to the next question being asked.
-measure a persons performace on a particular task or to assess his or her personality.
-Statistical norms for these tests can be established as soon as sufficient data has been
collected. This is known as the process of Standardization
-Help find conditions that the patient may not be willing to admit in clinical interviews,
such as presense of dilusions that point to schizophrenia.
-This is a Self Report Personality Inventory that the person is asked to complete which
is a questionnaire that says statements and asks how much it applies to that person.
-Most common is the MMPI – Minnesota Multiphasic Personality Inventory. This was
developed in the early 1940’s and was meant to serve as an inexpensive means of
-Called Multiphasic because it was designed to detect a number of psychological
-In the development of this test, clinicians made statements that they considered
indicative of various mental problems, and these items were then rated as self-descriptive
or not by patients. This test was taken by a large group of individuals considered
“normal” and by patients with various mental disorders. Items that discriminated among
patients was retained, and if a person taking this test had behaviour that resembled those
scores of a certain group of ppl (those that were Schizo’s) then they would be classified
into that particular diagnostic group.
-Revised version of this test compensated for the difference in racial minorities, sexist
wording and outmoded idioms.
-Test also contains “validity” scales used to detect fake responses. Ex: The lie scale has a
series of statements that set a trap for a person who is trying to look “too good” (I read
editorials every night)
Projective Personality Tests
-Psychological assessment in which abstract stimuli are presented to the patients and they
are asked to describe it.
-Projective Hypothesis: is the belief that since the pictures ate unstructured, the patients
response will show his/her unconscious attitudes, motivation and modes of behaviour
-The Rorschach Inkblot test is a technique in which a person is shown 10 inkblots and 1
at a time ask what they see. Content of a persons responses are called “symbolic” Ex: a
guy seeing an inkblot as buttocks may be judged homo. Other forms of analzyation such
as on the form of the persons responses are also used.
-The Thematic Appreciation Test shows patients a picture and asks them to describe a
story about it. Since it is abstract, it may show patients inner thoughts.
-Initially used to specify which children needed “special help” in schooling
-IQ tests are now standard for measuring a persons current mental ability
-IQ tests are also used: In conjunction with achievement tests to diagnose disabilities and
find area of strength, to help determine mental retardation, to identify gifted children, and
as part of neuropsychological evaluations.
-100 is the mean score, 65% of ppl get scores between 85 and 115. People with scores < 70
are considered to have “significant subaverage general intellectual functioning” and those
with > 70 are “intellectually gifted”
Behavioural and Cognitive Assessment
-These clinicians assess four sets of variables, referred to as SORC
-S: Stimuli. Environmental situations that precede (happen before) the problem.
-O: Organismic. Both physiological and psychological factors assumed to be operating
beneath the surface. Ex. Fatigue caused by use of alcohol
-R: overt Responses. Clinicians give this the most attention. They must determine what
problems are problematic and the frequency, intensity and form.
-C: Consequent Variables: events that seem to be reinforcing of punishing the behaviour in
question. Ex: when person avoids a feared situation, does their spouse provide sympathy
and thus keep the person from facing his fears?
Direct Observation of Behaviour
-Behaviour therapists pay considerable attention to observable overt behaviour, but they do
so in a way that works within a framework that is consistent with their points of view.
-An important point of behavioural assessment is it’s link to intervention
-Although much of this research is conducted within an operant framework, it can also be
applied within a framework by the use of mediators. Ex: Gordon Paul’s technique where
ppl with fear of public speaking were asked to speak in front of a group and certain
members of the group were trained to rate the behaviour of the patient every 30 seconds.
Using this, he arrived at a behavioural index of anxiety.
Test-retest reliability: extent to which ppl being observed perform taking the same test twice, sometimes weeks apart. Alternate-form reliability: patients are tested twice but with 2 different types of tests, in case they aim for consistency and try to mimic the same answers as their previous test when given test-retest reliability. Inter consistency reliability: asses whether the items on a test relate to one another. A person with anxiety who says yes to dry mouth should also say yes to muscle tension. Generally related to whether a measure fulfills its intended purpose. Note that validity is related to reliability: unreliable measures will not have good validity. Content validity: refers to whether the measure adequately samples the domain of interest. Ex: axis i diagnostic test has high content validity because it asks questions about all the symptoms involved in axis i diagnosis.