Professional Documents
Culture Documents
Kassahun, My Notes Edited
Kassahun, My Notes Edited
counseling psychologist:
Plan and develop therapeutic plan
Supervision
Advise clients how they can better be helped by counselors
Data analysis (interviews, tests, session notes)
Consult with other professional, agents or organization
Evaluate the effectiveness of counseling techniques
Refer clients including non-counseling organizations
Training
Research
Administration
Assessment:
Assessment is the process of information gathering, also includes integrating the information gathered,
analyzing, synthesis and interpretations of the information.
Assessment is comprehensive, gather information through diverse methods, and diverse sources,
psychological laboratory tests (specially designed procedure), and then includes evaluation and decision.
Assessment is flexible, we don't use the same procedure, frequently uses tests, doesn't involved defined
steps or procedures it is context based.
Assessment is not usually done for its own purpose, even though it can be beneficial for self-awareness.
The point of assessment is not for just the assessment, it's is either for diagnosis or classification
The focus of group therapy is still the individual, so the assessment in a group is still for the participant
individuals in the group.
Anecdotal report is a subjective observation of the individual at a specific time, it's informal
Questionnaire is a self-report method, we give the questions in writing and they answer in writing. E.g.
basic background information. It is different from the standardized psychological tests.
Self-report techniques: interviews, personal essays and autobiographies, journals (it's more than a diary,
it includes thoughts, improvements, challenges and reflections and is more of longitudinal),
Class room tests are not standardized. The standardized one has wide coverage, and administered
similarly.
Aptitude and intelligence test measure their potential, future performance. Also, a cognitive test. They
are validated and standardized.
Projective: we give ambiguous stimuli like images, incomplete sentence. It is based on psychoanalytic
belief that they are going to project their subconscious thought, feeling, and needs. Its score is difficult,
and might be subjective. E.g. Rorschach inkblot test, thematic apperception test. It also needs the
subjective interpretation of the counselor.
Attitude scale: has components of attitude. But only for a specific subject. Thought, feeling and behavior
domain. To show his agreement and disagreement, using Likert scale. Attitude can also be tested in
observation, projection,
Whole person: different aspects of the individuals, career, family, and social relationship
Referral questions: who is assessed, why, what is assessed and what are the goals of assessment
Can transfer the information with the consent of the client to the third person if they have the expertise
to manage and process the information, and that individual will benefit from it.
Battery means many characters are tested with one test
Intelligence test: is a capacity, ability, and multifaceted. It is the potential we have not what we did. It
manifests in different ways.
Francis Galton (England): the first person that taught thought about intelligence in a most scientific way.
He thought it at sensory level not at cognitive level. he tried to develop a rudimentary level of
intelligence tests.
Alfred Binet, from France: took the meaning from sensory to cognitive level. He is the first person that
developed the intelligence test ever. That was well developed and at cognitive level. In 1900s. He was
the chair of a committee that tried to include the intellectually disabled in the educational system. So, he
developed a test to identify those individuals. Binet Simmon test of intelligence, the first intelligence test
ever. He is guided by theoretical beliefs and knowledge
Stanford Binet: test adopted by Stanford for the us population, it is much popular than the Binet one.
David Wechler: near 1960s, his understanding was acting purposely, think rationally, deal effectively with
his environment. No significant difference with Alfred Binet.
Piaget: intelligence with biological adaptation, evolutionary understanding: two processes, these are,
assimilation and accommodation
The earlier ones believed intelligence was unitary or single like Binet.
There are some that say it has domains or divisions but overall give one intelligence. But recently there
are saying that there are independent intelligence types.
Thurstone: he believes intelligence is not one, rather many primary mental abilities. Criticism: these
factors are correlated and he did not bring a new idea, they are related to previous believes.
John Carroll: it added on the fluid and crystallized intelligence from Cattell and added.
But he has a different idea: these different types of intelligence have strata, general, broad and specific
abilities. These 8 are broad intelligence and each have specific abilities. When each specific abilities are
aggregated, they will give us the broad abilities.
CHC model, recently, Cattell, Horn (added on Cattell), Carroll (he brought strata idea, was student of
Horn). Broad ability, strata.
CHC model is a Carroll idea, 10 broad, and 70 specifics, these 10 broad abilities give a G. The 70s are
included in the 10 broad ability scores
Good Johnson is a CHC model type of cognitive ability test.
All these are a believe that there is one general intelligence but has components.
Howard Gardner: besides the domain, there is a believe that there are multiple independent
intelligences that cannot be grouped into one. He believes both nature and nurture.
He says intelligences, he didn't say there are 7 intelligences, he said there are multiple but what he found
are 7, there could be more.
Gardner proposed 7 but currently there are 9 and 2 evidence collected.
The 2 new are, naturalist intelligence, existential (why we live and die) intelligence
Gardner doesn't believe most tests are not context based. He is influential currently.
Gardner has not influenced the testing community as much. Because most of the influential tests are
mostly oriented by previous works.
Individual intelligence tests differences are based on domains included.
Many tests have verbal and nonverbal (performance), there are some that are entirely nonverbal.
Group intelligence testing, idea came during world war, because the test at that time (Stanford Binet was
highly verbal, it was not good for non-English speaking and illiterates at that time, so for military
recruiting).
Aptitude, is ability in one specific area. Aptitude is for prediction and readiness. In that specific area to
be successful.
Specific:
Watson Glaser critical thinking appraisal,
Modern language,
General intelligence can also predict but aptitude is very specific to certain areas.
Achievement: is not potential but success. Can be standardized at regional, local or international. Or can
be non-standardized unlike aptitude and intelligence tests.
Standardized are made by expert teams, have wider and detailed coverage.
Standardized can be made for each subject as well. There are commercially available.
MMPI is the most widely adopted and validated test. It is most widely used test. Unlike other types this
test diagnose psychopathology. There will be reverse coded if negatively stated questions.
Edward: forced choice items, in each item there are two choices you are forced to choose. It is different
because it has forced item choice.
Objective tests are usually based on trait and state theory, you rate yourself against those statement.
MMPI had 3 validity tests, one is to check whether you are faking or not
All are about the test validity as a whole
Projective methods:
Reliability:
Reliability should be understood as a degree not as an all or none
The most reliable test for error is reliability coefficient but that's not possible to determine true score
Since this cannot be done other source of testing used
If reliability coefficient is negative means its error is above 100 % which is not usually possible
We have to select which test of reliability are important but we may not be able to keep the other tests
constant
If the test-retest reliability is changed then tested in short period where you wouldn't expect it to change
The individuals tested are not necessarily the same individuals, but a group
Test-retest reliability is not necessary or useful for achievement tests
Internal consistency: if the domains are mutually exclusive or independent we can do test for each
domain but if the whole test has a general meaning as a whole then the whole question as a test can be
done. So both are different types of approaches in internal consistency.
Coefficient alpha or Cronbach alpha when each item cannot be answered like PHQ it doesn't matter You
start from 1 or 0 as long as it is constant and square bis variance for
Interrater, for observational rating, their rating agreement, correlation is not good, but use percent
agreement
Kappa statics is preferred
Whatever the type of reliability usually when it's 0.7 and above it's acceptable.
Standard error of measure talks about how reliable that is on individual from the test developed from
the group.
Reliability for individual,
SEM is to infer, from that test to its true score, to infer from observed score the true score
The mean would be the true score and the SD is the SEM. If that test is administered for the same
individual many times, it will be normal distribution. But it is difficult to do that practically.
SEM is a type of SD
If administered 100 times for that individuals its true score will be between -1 and 1in 68% of the times,
95 % between -2 and 2, 99.7 between -3 and 3.
SEM is also a type of error so as the reliability increases the SEM decreases.
It tells us about the individual
Validation is the study of validity or cumulating evidence for validity of the test.
There are different types of validity because each test has different purposes,
Face validity: appears, if people with expertise say that test appears valid
But it is not to say it is a valid measure
Content validity: does it reflect the definition of the target construct?
Criterion related validity:
Predictive, the test data is collected early then the criterion data.
Construct validity: the test does the psychological test it wants to test. It is determined by accumulated
result unlike criterion and predictive validity
Inter item if they are close to 0.8 0.9 it means they are the same.
We need to check before adapting or accepting a test we need to check the theory it's developed on.
Adaptation of tests:
Forward translation
Translation as conceptual translation than word for word translation
The translators to back translation should be native language speakers of the original language and
should be able to speak the second language as well and the subject matter translates the other
The back translated are compared to the forward translated single one to form a single questionnaire
The pilot testing is more of a qualitative test. The simplicity, the relevancy, offensive terminology, and
time it takes.
In preliminary psychometric testing to bilingual individuals both the original and the adapted. The same
individuals.
Adaptation allows deletion and adding new questions, but it should not be big enough to totally change
the test.
One validation test might not be enough, many researches might need to be done.
In semantic you label the beginning, the middle and the end.
If inter item is very large like .89 and above, they might be similar so merge
The minimum is 0.3, if below it means they are totally different and should not be in the same construct.
Item variance should not be 0, if so, it means all people responded the same.
After pilot test, you make a table and record all those values from the pilot, the cognitive interview Nd
the like and send it to experts they then. Can keep or delete or modify