Screening

Intervention Planning

Progress Monitoring

Program Evaluation/Accountability

1- educational need (screening; performance discrepancy)

2- educational benefit (progress monitoring; rate of improvement)

Present Level of Performance

write goals using PLOP (to reduce the gap)

T1

T2

T3

Consideration for Special Ed. Eligibility

Special Education

Student, Instruction, Curriculum

(required to assess all these things)

or

Instruction, Curriculum, Environment, Learner

Positive Skew

(kids low, hard test/task, or both)

Negative Skew

(kids high, easy test/task, or both)

Bomodal Distribution

(two groups)

-behavior/skill to measure

-time frame

-PLOP

-CAP

T2/T3- grade level

SPED- could be a lower grade level

 P= E = O

      S

*related to a discrepancy

-Reliable

-Unusual

-Meaningful

Aptitude by Treatment Interaction (Chronbach)

-types of kids, patterns of kids

-matched to treatment

(doesn't work)

1- Inclusionary Components

2-Exclusionary Components

-Determinant Factors

-circle, point, underline

-1/4 chance of getting it right due to luck

Salvia and Ysseldyke Framework

(when; procedural)

-Referral

-Screening

-Eligibility/Entitlement

-Intervention Planning

-Progress Monitoring

-Program Evaluation

Deno Framework

(why; problem solving; philosophy)

-Problem Identification

-Problem Certification (or Validation)

-Exploring Solutions

-Evaluating Solutions

-Problem Solutions

Problem Identification

-a problem is defined by differences between what is expected in gen ed curriculum and how the student performs

*if the difference is large enough, further investigation (assessment) may be warranted)

-How severe is the problem?

*If the severe problem unlikely to be resolved by gen ed and if not, what resources may be needed to reduce the discrepancy?

-intervention planning

-What are the goals for the intervention?

-What is the content of the intervention (what to teach)?

-What is the process of the intervention (how to teach)?

-progress monitoring

-Is the intervention successful? Is the student attaining the established goals?

-If not, does the intervention require modification?

-Student Characteristics and Skills

-Instruction

-Curriculum

*SIC or ICEL

*student usually not modifiable

-collectin samples of behavior under specific conditions

-not exactly the same meaning as assessment (just a type of assessment)

-meaning information is collected, scored, and interpreted in a standard way

-not just "norm-referenced"

-put out by a company or organization, usually for profit

-not synonymous with "good quality"

-"formal" should mean standardized

-"informal" should mean "without technical information" (not for high-stakes)

-information is interpreted relative to a norm group

-not to be confused with "standardized"

-information is interpreted relative to the domain the device is used to sample

-not solely because there is a "criterion"

(ex: reading--> kid should know 90%)

-information is collected at the end of instruction for the purpose of determining who has learned and what they learned

(ex: final exam)

-information is collected during instruction for the purpose of determining whose programs need modification to improve student learning

(while forming knowledge)

-information is collected over time and is interpreted relative to an individual's expected rate of progress

(goals can still come from norms)

-the implicit or explicit standards of conduct or performance "imposed" on all who would be "successful" members of a society

-e.g.- basic literacy skills, social skills

-those activities, proficiencies, and standards of conduct that are valued within a society, but are not required of ALL members

-e.g.- playing an instrument, trigonometry

6 Big Things to Remember about

Educational Assessment

1. Error is always present

2. Tests are samples of behavior

3. Measure as directly as possible

4. Current performance is observed, future performance is an inference

5. Acculturation is similar

6. Not all scores are the same

-NO test is perfectly reliable

Test Score= "True Score" + Error

-Reliable Tests have LESS error

Agreements/(Agreements + Disagreements)

ex: 84/84+6= 84/90= 93.3% Agreement

(reliable!)

-We can never assess ALL of something

-Therefore, we SAMPLE

High Fidelity <----------------------> Low Fidelity

(Lifelike Behavior- Simulated Behavior- Knowledge- Opinion)

-Here and Now

-Predicting the future is a high risk occupation

-when making comparisons, we'll try to use HIGH quality metrics

-yes to percentiles

-no to age and grade equivalent scores

ex. 33 WRC, 15 Out of Seats, 5 Crying Episodes

-typically convey little meaning unless referenced to a standard

-Comparison Standards

*Past Performance: Individually Referenced

  *Criterion Level of Performance-- to a "standard", often arbitrary

  *to a Group-- Norm-Referenced

Types of Data

(Understanding Scores)

Cardinal (some meaning, but not really relational)

-ex.: uniform numbers, bus numbers

Ordinal (reflects the order, but not the amount of a variable index of bigger-smaller, greater less)

-ex.: pain scales, percentiles, class rankings

Interval (like ordinal + equal distances/diffs between the numbers; scale indicated amount, no zero point)

-ex.- most conventional test scores

Ratio (true zero + equal intervals, like the real # system)

-ex.: time, number correct, frequency of behavior

-scores of a relative standing with a set, fixed, predetermined mean and standard deviation

z-score: mean (0); SD (1)

T-score: mean (50); SD (10)

IQ: mean (100); SD (15)

-use the same scale; used to compare scores

-the "go-betweens" that are used to: calculate other standard scores, go from one standard score to another, and estimate percentile ranks

-the number of persons who scored AT or BELOW the score of interest

ex.: Larry's PR (63); 63% of those who took the test scored lower than Larry; Larry scores as well as or better than 63% of those who took the test

-scores of relative standing with lots of baggage

-ordinal data

-GE and AE (determined by matching raw scores with a mean at a grade)

-refers to the stability, consistency, or accuracy of test scores

-accuracy in the sense of freedom from random error

-a standard indez of HOW MUCH ERROR is in a test score

-conceptual (we can never really know true score)

-the extent to which a test measures what it says it measures

-the single most important psychometric characteristic of a test

-the extent to which one's score on a criterion can be estimated from the person's test score

Predictive: the extent to which a test estimated future performance on a criterion measure

Concurrent: the extent to which a test estimated performance on a criterion measure

-ALL types of validity are construct validity

-The IF, THEN proposition