Assessment for Diagnosis

EDUC 602Week 1

1Benedictine University

Assessment for Diagnosis

• Assessment should… – Inform instruction– Be realistic/authentic

• The best assessments are those closest to skills & strategies being used in classrooms

– Be interactive• Measures both top-down and bottom-up factors

– Be multidimensional: • All major areas are assessed - general cognitive abilities, hearing, spelling, writing

– Be dynamic: • Measure what the student can do now and predict

the student’s potential for change– Cover many categories

Benedictine University2

Principles of Effective Assessment

• Five major factors must be considered when assessing students:

Principles of Assessment

Benedictine University3

5. Situational context in which reading or writing is performed

4. Reading or writing tasks

3. Techniques being used

1. Reader

2. Text

All of the “five interlocking aspects of reading” should be considered because they influence assessment results

• Assessments fall into the following categories:1. Screening: Identify students who appear to be “at risk” and require

additional assessment2. Diagnostic: Determine areas of need3. Monitoring: Are students making adequate progress?

• Formative Assessment4. Outcome: Administered at end of semester or year

• Summative Assessment

• Think about assessments you give your students• Into what categories do these assessments fit?• Prepare to explain an assessment(s), the category(s) and why the

assessment best fits into the category(s) during Session 3

Benedictine University4

Assessment Categories

Assessment

5

• Read Gunning pages 66 – 72 and Bar pages 322 – 329 • As you read pages 66-72 in Assessing and Correcting Reading and Writing

Difficulties and pages 322-329 in Reading Diagnosis for Teachers: An Instructional Approach, compare and contrast the following assessment types:– Norm-referenced vs. Criterion-Referenced – Survey vs. Diagnostic tools – Formal vs. Informal

• Be prepared to respond to the Short Answer Assignment on the following slide

• Portfolio Artifact Case Study Information: You will use one norm-referenced test that we will provide and one criterion-referenced test that you will choose in the case study that is included in your Professional Portfolio Artifact – Begin to consider a criterion-referenced test you might use with your

student

Assessment Types

Benedictine University

EDUC 602Week 1

EVALUATING ASSESSMENT DEVICES BASED ON:- RELIABILITY- VALIDITY- STANDARD ERROR OF MEASUREMENT

Unless stated otherwise the content of this section is based on Chapter 3 – Gunning, T.G. (2010) Assessing and correcting reading and writing difficulties. Boston, MA.: Pearson, Education, Inc

• Assessments must be reliable and valid to make useful educational decisions about and for students

• Therefore, as a Reading Specialist making educational decisions, you must carefully consider the following areas before using assessment data:– Reliability – Validity– Standard Error of Measurement (SEM)

Benedictine University7

Evaluating Assessment Devices

Gunning, T.G. (2010) Assessing and correcting reading and writing difficulties. Boston, MA.: Pearson, Education, Inc. (pp. 73 - 75)

• Reliability is the consistency of an assessment device– Test reliability is high if you give the same test to

the same or matched students on two different occasions and the test shows similar results

Benedictine University8

Evaluating Assessment Devices

• Validity is the degree to which an assessment device measures what it intends to measure and the degree to which it can be used to make educational decisions– Construct Validity is the degree to which a test measures a theoretical

trait, like critical reading or learning ability– Content or Curricular Validity is the degree to which the content of a

test reflects reading or tasks as they are taught in schools• For commercially-produced tests, information about their

reliability and validity is usually found in the test administration manual

9

Evaluating Assessment Devices

Benedictine University

• Standard Error of Measurement (SEM) is an estimate of the difference between the obtained score and what the score would be if the test were perfect– Example:

• If a norm-referenced test has a SEM of 5 percentile points, and a student has a score of 50

• He or she is likely to score between 45-55 if they took the test again

• SEM is a measure of reliability

Benedictine University

Evaluating Assessment Devices

Gunning, T.G. (2010) Assessing and correcting reading and writing difficulties. Boston, MA.: Pearson, Education, Inc. (p. 75)

• Fairness can be impacted by biases caused by differences in background knowledge

• Tests can be biased based on:• Geography• Gender• Socioeconomic status• Ethnicity• Race

Benedictine University

Evaluating Assessment Devices

12

• Read pages 73-76 in Assessing and Correcting Reading Difficulties to learn more about how validity, reliability, and standard error of measurement (SEM) are expressed

• Then, click the following link and choose the latest version of the Illinois State Assessments Technical Manual

• The Illinois Standards Achievement Test Manual is also located in the course resource folder

• Use the Table of Contents to locate information on reliability andSEM data for the ISAT test

Benedictine University

Evaluating Assessment Devices