Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
28 Cards in this Set
- Front
- Back
Reliability
|
Consistency; Ranges from 0 to 1. R = .9 (excellent), .7 (acceptable), .6 (ok for research)
Interpreted as the proportion of variability that is "true score variability." A coefficient of .84 means that 84% of the variance is due to true score differences with remaining 16% due to error. 1. Test-retest = two administrations of same test to sample people. Primary source of measurement error is related to passage of time. 2. Alternate Forms = two or more versions of the same test are administered. Primary source of error is content; an examinee's knowledge base being tapped slightly differently on different tests. If with delay then time also contributes error. 3. Internal Consistency: useful for tests measuring single characteristics NOT speed tests (if speed tests use alternate forms); (1) Split-half = equal halves of one test are correlated to each other, tends to underestimate a test's true reliability because r decreases with decrease in items (2) cronbach's alpha = formula used to determine the average degree of inter-item consistency. Average reliability that would be obtained from all possible splits of the tests. Tend to be conservative. 4. Inter-rater = the % agreement between raters. Sources of error include rater motivation and characteristics of measuring device. Factors Affecting: 1. Test length = larger is more reliable 2. Range of test scores = if range of scores is unrestricted it helps reliability, having moderate item difficulty helps create greater variability 3. Guessing = all else equal, T/F will have lower reliability than multiple choice which has less than free recall |
|
Validity
|
Accuracy, measures what it is intended to measure; must be reliable in order to be valid
0. Face = not a real measure, but does the measure appear to measure what it is supposed to? Increases trust/motivation. May lead to social desirability. 1. Content = extent to which it samples the content/behavior that it is designed to measure. Most often associated with achievement tests (e.g., stats test asks stats questions). Is it comprehensive and representative? Relies on experts. 2. Criterion = bring in external criteria that are proven to measure the same constructs and compare. (1) concurrent criterion - is the comparison test data used at the same time (GPA and SAT) or (2) predictive validity - external criteria is examined in the future (GRE and Grad School); assessed using correlation coefficients, SEE provides a confidence interval for a more accurate range for predictive validity 3. Construct = theory based, can be abstract or empirically supported (1) convergent - correlate one test with another assessing same concept (2) divergent - want to demonstrate that a new test will not be related to another construct (3) discriminant - 2 different constructs measured by 2 different tests 4. Incremental = degree to which assessment contributes to decision making, what added variance is explained by this test? |
|
Normal Curve Percentages
|
68.26; 95.44; 99.72
34.13; 13.59; 2.14 |
|
SEM and SEE
|
SEM = SD x sqrt (1-r) :: CI = obtained score +/- (Zcrit) (SEM) [estimate of error to use in interpreting test score since test score is combo of true score and error]
SEE = SDcrit x sqrt (1-Rsquared) [used to make prediction with correlations] |
|
Z Scores
|
Way to convert a raw score into an interpretable standard score, based on its position on the normal distribution curve
Mean = 0, SD = 1 Z = (X - M)/SD |
|
T Scores
|
Similar to Z scores but M = 50, SD = 10
Arbitrarily moves mean up so all numbers are positive T = (Z x 10) + 50 |
|
Percentile Rank
|
Indicates the percentage of people who scored below or at the same level as a given point. Demonstrates relationship to other
Look at Z table to find corresponding rank |
|
WISC-IV
|
Individual Administration
Published in 2003 Age range 6-16:11, normed for every 4 months (200 children each year) Normative Sample = 2200 children, stratified 10 subtests to obtain FSIQ; 4 index scores |
|
WISC Subtests
|
VCI (SI, VC, CO, IN, WR) - the ability to work with abstract semantic info
PRI (BD, PCn, MR, PC) - the ability to integrate perceptual stimuli, ability to evaluate spatial info WMI (DS, LNS, AR) - concentration, attention, and manipulation of info; sequencing, mental flexibility PSI (CD, SS, CN) - information processing, planning and organizing, motor control, motivation, visuo-motor coordination FSIQ GAI New subtests = PCn, MR, WR |
|
WIAT-III
|
Revised 2009
Age range 4-50:11, pre-k to adult 8 composite scores from 16 subtests |
|
WIAT Areas of Achievement
|
Total Reading
Basic Reading Comprehension and Fluency Oral Language Written Expression Math Calculation Math Fluency Total Achievement |
|
CHC Theory of Intelligence
|
CHC Theory is the theory of intelligence that has received the greatest empirical support
Cattell-Horn proposed the Gf-Gc Model (1941-1989) - structural, developmental, genetic, outcome, and neurocognitive evidence supports Broad Abilities Broad Abilities: Crystallized (Gc) Quantitative (Gq) Reading/Writing (Grw) Fluid (Gf) Visual-Spatial (Gv) Auditory Processing (Ga) Long-Term Retrieval (Glr) Short-Term Retrieval (Gsm) Processing Speed (Gs) Acquired Knowledge (Gc, Gq, Grw) Thinking Abilities (Gf, Gv, Ga, Glr) Cognitive Efficiency (Gsm, Gs) |
|
WJ-III COG & ACH
|
Based on CHC Theory and Structure
Age 2-90+ 3 Broad Abilities and 7 Ability factors General Intellectual Ability Very Superior - Very Low W score = equal interval, centered at 500, equal to 10 yr old performance Grade or Age Equivalent Standard scores Relative Proficiency Index (PRI) - 60/90 means subject is 60% successful on tanks that peer reference group completes with 90% success WJ-COG Verbal Ability (Gc) Thinking Ability (Glr, Gv, Ga, Gf) Cognitive Efficiency (Gs, Gsm) WJ-ACH Reading (Grw) Oral Language (Gc) Mathematics (Gq) Written Language (Grw) Knowledge (Gc) |
|
LD
|
IDEA - 8 areas of specific learning disabilities
Oral expression Listening comprehension Written expression Basic reading skills Reading comprehension Reading fluency Math calculation Math problem solving Based on referral question and clinical interview, generate hypothesis Use either AAD or PSW approach Corroborate among 3 sources or methods |
|
Ability-Achievement Discrepancy Analysis
|
Predicted difference method - regression-based discrepancy method
Simple difference method Necessary but not sufficient condition, used as supplementary evidence |
|
Patterns of Strengths and Weakness Discrepancy Analysis
|
Separate analysis could be conducted for each identified achievement deficit
Common cut off for achievement deficit = 85 Select an achievement composite or subtest corresponding to an IDEA area Select ability representing processing weakness (should be related to achievement deficit area) Select ability representing processing strength (should be unrelated to achievement deficit; don't use PSI, WMI, STM, LTM) Measures must have same SD Use Standard Error of Difference (SED) to evaluate the following 3 conditions Significant difference between processing strength and processing weakness Significant difference between processing strength and achievement weakness Consistency between processing weakness and achievement weakness (no significant difference) SED = SD * SqRt (2 - reliability of X - reliability of Y) 1.96 * SED - formula used to determine significance, use this formula for each side of the triangle |
|
ADHD assessment
|
Subtypes:
Inattentive Hyperactive-Impulsive Combined Must have 6+ out of 9 symptoms; these symptoms must have persisted for 6 months in social and academic setting 2:1 male to females, 5% in children, 2.5% in adults Clinical interview - parents and teachers input for minors School and learning difficulties, socio-behavioral difficulties Self-Report Scale (CAARS or Brown ADHD scale) Cognitive abilities/memory test (pair WISC or WAIS w/ WMS) WISC alone is not enough for diagnosis TOVA - measures attention, impulsivity, and adaptability |
|
ADHD Hypothesis
|
Brain structure hypothesis = those with ADHD have a slightly different brain structure
Dopamine hypothesis = those with ADHD have malfunctioning or decreased functioning of dopamine systems Executive functioning = people with ADHD have weakened executive functioning, inability to stay on task and ignore distraction, initiating testing planning sequencing and regulating goal-directed behaviors/emotions, cognitive flexibility and set-shifting, synthesizing and integrating details into a coherent whole to solve novel problems |
|
Neuroanatomy of memory
|
Temporal lobe = LTM
Hippocampus = episodic memory Perirhinal cortex = semantic memory Frontal lobe = encoding, retrieval/learning strategies, screens for relevant info Parietal lobe = attention, working memory, episodic memory Left hemisphere = verbal Right hemisphere = perceptual and visual material, emotion, big picture things Thalamus = pathway Basal ganglia = similar in function to frontal lobe |
|
Baddeley Working Memory Model
|
Proposed more than just a central executive, phonological loop and visuospatial sketchpad
Added in episodic buffer - if a memory was processed with the episodic buffer, it went directly to immediate episodic LTM |
|
WMS-IV
|
Norm - 1400 participants, 100 in each of 14 age groups, stratified
16-69 - 900 65-90 - 500 10 subtests 5 index scores - Auditory, Visual, Visual Working Memory, Immediate, Delayed |
|
Age and IQ
|
IQ increases in adolescence, peaks in early-mid 20s, declines after
VCI increases until 40s-50s PRI starts to decline around 40 PSI steeply declines in mid 20s WMI decreases after mid 20s |
|
Gender and IQ
|
Myth - females do not score significantly lower on math
Females score statistically significantly lower than males on spatial tasks, significantly higher on processing speed Females and males only have difference in VCI at younger age |
|
Race and IQ
|
Asians have highest FSIQ (107), then Whites (103), Hispanic (92), then Blacks (89)
Differences could be because values or SES Education - when education controlled for, differences disappear SES, value of education, home environment, parents time with kids, teacher reinforcement, school system all impact IQ scores |
|
Adverse Impact
|
The use of test results consistently leads to a disadvantage for a particular group
Disparate treatment - treating people differently because they belong to different groups Disparate impact - using tests and interviews to hire, but biased impact 4/5th rule of EEOC - hiring rate of a given ethnic group cannot be lower than 8-% of the top ethnic group's hiring rate (ex. hiring rate of whites is 8% so .08*.8=.064; minority required hiring rate of 6.4%) |
|
Griggs vs. Duke Power Company
|
Employers must justify use of standardized testing in making hiring decision, must be closely related to job responsibilities (adverse impact)
|
|
Larry vs. Riles
|
Claimed that state's method (California) for classifying students with LD was culturally biased against black students
Set the precedent that tests must be validated for use with a given population before they can be administered to minority children |
|
Gratz vs. Bollinger, Grutter vs. Bollinger
|
Gratz = undergrad minority applicants to U of Michigan were given bonus points for being minority - ruled as unconstitutional because it did not assess diversity contributions of each individual (special consideration is ok, but fixed formula is not)
Grutter = white female not admitted after learning minority groups were favored in admission decision; ruling was upheld because the school considered race meaningfully alongside all other factors |