standards that provide information about where a score of a psychological test ranks in relation to other scores on that test. refers to the average score of a standardised sample. possible through standardisation. used to compare with age mates in terms of a percentile rank
wechsler scores std dev & percentile ranks
100=50th percentile (average). STD DEV 0
85= 16th percentile, STD 1
130= 98th percentile, STD DEV 2
percentile scores & meanings (levels maladaptiveness)
STD error of measurement
STD DEV score of the amount of error that can be expected for test scores. can be used to determine the probability that a given score will fall within a certain range or band (CI)
the longer the test...
the more reliable it will be
which has higher reliability? overall or subtests, multichoice or T/F
overall test score
preferred for stable traits/ abilities, making predictions. not good for states.Can be effected by occarances within the intervals of testing. can also be affected by memory & practise eg; WAIS digit symbol & arithmetic.
age equivalent scores (norming procedures)
raw age data is plotted & age equivalent score is read from a smoothed curved line fitted through points.
pos: easily understood, avaliable for Vineland domain & subdomain, WAIS-R, K-ABC & KAIT
neg: when uneven/skewed does not produce full scale continuum which is necessary for deriving scaled scores
descriptive categories (norming procedures)
compare frequency of behaviour with peers. report maladaptive levels (descriptive categories). scoring of maladaptive levels;
-nonsignificant: 50th percentile & below
-intermediate: 51st-84th percentile
- significant: 85th percentile
alternate/equivalent forms reliability
the relative accuracy of a test at a given time. useful for assessing effects of a treatment programme or changes over time. eliminates recall of previous responses. gives LOWEST estimate of reliability of a test
split half reliability
the internal consistency of the items (odd-even, first-second). preferred for characteristics that are highly subject to fluctuations. can be effected by; warm up, fatigue, boredom in doing halves + if don't have equal items in halves then wider variability & less stabilization for scores around the mean. gives HIGHEST estimate of the reliability.
the degree of agreement between two examiners. used for test with subjective scoring (ie rorschach).
intraclass correlation coefficient vs product moment correlation
--product-moment: measures similarity in orderings of test scores made by evaluators rather than total IQ scores (extent scores go together no whether they are close in absolute terms)
1) can distinguish b/w interclass examiners scores and intraclass.
2) Distinguished between scores ranked in the same order & those ranked + in low, mod or complete agreement with each other
3) corrects for test-retest or inter-examiner agreement based in chance
intraclass can be no higher than product moment and will be lower to the extent that two lots of scores vary from each other.
the extent to which a test measures what it is intended to measure. Need validity for reliability
the extent to which an assessment instrument correlates with an independent criterion of the trait/behaviour assessed by the test. types:
-concurrent: measure at same time
-predictive: outside measurements taken after scoring & scores are used to identify which category of psychiatric populations would be most likely to benefit from specific treatment approaches
criterion validity table meanings.
-sensitivity: % cases indicated pos by lab test pos by clinician- correct classification of impaired cases. (a/(a+c) )
- specificity:% correct classification of normal cases (d/(b+d))
-positive prediction power: % cases that are truly impaired out of those with pos test score (backed up by lab). (a(a+b))
-negative predictive value: the extent to which a negative test score is confirmed by lab test (d/(c+d)
-accuracy: extent to which all tested cases produce the correct correct diagnosis (a+d)/N
square of correlation coefficient can be used to estimate the extent to which a test accounts for a trait being measured
the extent to which a test measures a theoretical construct/ trait. can include data from content & criterion validity, FA & internal consistency. test should correlate highly with theoretically similar variables & diverge from theoretically dissimilar. If it correlated to highly with another test without giving an advantage my be unnecessary.
internal consistency interpretations
inter-rater reliability interpretations
improvement in validity by increasing numbers of data scores. Need to decide if time is better spent doing this or on treatment, administration, scoring and interpreting. clinical and projective test often poor expect MMPI combined with social hist, EEG & physical/neuro exam.
self ass. vs formal ass + unstructured
-self: time & cost effective + facilitate relationship
-formal: social desirability, attributional errors, distortions due to poor judgements + relative self awareness of client
-unstructured: provide flexibility, focus on the uniqueness of a person & are ideographically rich
factors in perceiving others
intelligence, artistic interests, social detachment, good emotional adjustment, similarity in race & cultural backgrounds. NOT confidence in the accuracy of judgements.
sources of measurement error
1) Lasting general characteristics of the individual (ability, skills in test-taking techniques, ab to understand instructions)
2) lasting specific characteristics of indivi (level of ab on traits required on specific test, skills related to test Q, knowledge of fact
3) temp. gen. characteristics of indivi (health, fatigue, motivation, emotional strain, external conditions, test-wiseness, understanding how test work)
4) temp. gen. characteristics indivi (understanding specific task, specific coping techniques, practice, memory fluctuations, carelessness)
matching clinical samples with control on demographics- able to show KAIT had applicability in a variety of clinical and neurological problems
extent to which measure correlates with earlier instrument measuring same/similar construct. factors; what the new test proposes to measure in relation to the old one, the extent to which societal & cultural changes since the development of a new test may have necessitated major item changes & extent to which earlier instruments have been inappropriately standardised or normed
extent to which subdomains & their respective items loaded on their intended domains
innate, brain efficiency & intactness- sensitive to BD. nonverbal, independent of culture/ed. measures prob solving ab. increases till 14, levels 20 then grad decline. MATRIX REASONING & VISUAL PUZZLES
dev from interaction b/w fluid intelligence & environmental factors (culture, ed.) permanent, less sensitive to BD. measures accumulated knowledge & is what most IQ tests measure. VOCABULARY & INFORMATION.
typical ways of defining intelligence
- ab to cope with envi through learning
- ab to judge, comprehend & reason
- ab to understand + deal with people, objects & symbols
- ab to act purposefully, think rationally & deal effectively with the envi
general qualitative analysis of WAIS & WMS
1)observe individuals interactions with examiner
2) study CONTENT of test responses particularly information, vocabulary, comprehension and similarities
3) info from PATTERN of subtest scores
IQ scores influenced by...
- understanding of English
full scale IQ & global/general ability index
difference between the two shows the extent to which brain/age factors are lowering overall functioning. The difference becomes less meaningful if greater then 23
culture free indexes
perceptual reasoning, processing speed
Indexes sensitive to DB, mental disability & learning diff
- working memory
-processing speed (most_
index that begins to decrease earliest
processing speed. Also the lowest in gifted children
WM index is lower for
ADHD, learning disabilities- reading + math
which WAIS subtest can be used as a predictor for prognosis
similarities- good introspection
Similarities sensitive to what?
-L-hem BDparticularly temporal/ frontal
- schizophrenics, rigid thinkers, senile
high vocab scores...
May be compulsive or intellectualising defence mechanisms
subtests that are stable & resistant to disturbances
Vocab + Information
what kind of individuals may do poorly on arithmetic
antisocial or histrionic personalities
Block design subtest can be used to...
-ob. impulsivity, motor co, speed of info processing, frustration tolerance, feedback
-R-hem BD- parietal
-early alzheimers & disting Alzheimers from dep
culture free subtests
matrix reasoning, block design,
ANX or ADHD effect with what subtests
- digit span, arithmetic, coding, letter-number sequencing
In comprehension who does not perform well?
what would make an individual poor at pic completion
digit span lower for people...
anxious, mentally disabled, learning diff
coding subtest is sensitive to
organic and functional impairment
tests that should have reduced emphasis if time is an consideration
coding, symbol search, cancellation
poor test motivation decreases scores on
digit span, arithmetic, coding, symbol search, letter-number sequencing
fast processing ability can be informed by what subtests
cancellation, symbol search, coding
what subtests deteriorate due to situational variables
arithmetic, coding, block design
if sequencing is an issue low scores on..
digit span, arithmetic, coding, letter-number sequencing
what to large diff on DSF, DSB & DSS mean?
cog deterioration due to poor attn, sequencing, chunking + visualising no
low DSB and coding scores might suggest
solvent caused DB
what subtests can be considered in qualitative analysis
information, vocabulary, comprehension, similarities
Brief cognitive screening exam scores
average 25%-100%, low average 10-24%, borderline 5-9%, low 2-4%, very low >2%
what WMS index is considered measure of global memory
delayed memory index
contrasting index scores: auditory vs visual
>7= visual mem is low compared to auditory
13+= auditory mem is a strength compared to visual
contrasting index scores: visual WM vs visual mem
>7= visual mem low compared to VWM
13+= high visual mem compared to VWM
WAIS GAI comapred to WMS index
>7= mem index weakness compared to gen ab.
what to WAIS vs WMS compairsons are impt?
- GAI vs visual mem is the most sensitive to impairemnt
- GAI vs delayed mem b/c delayed is considered a global measure of mem