Dr. Laura Hart Dr. Teresa Petty

Slides:



Advertisements
Similar presentations
Developing an Outcomes Assessment Plan. Part One: Asking a Meaningful Question OA is not hard science as we are doing it. Data that you collect is only.
Advertisements

Assessment Early Years PE MPESA Fall Conference September 18-20, 2013.
TaskStream Training Presented by the Committee on Learning Assessment 2015.
Pre-Conference Workshop – June 2007 BUILDING A NATIONAL TEAM: Theatre Education Assessment Models Robert A. Southworth, Jr., Ed.D. TCG Assessment Models.
CONNECT WITH CAEP | | Three-Year-Out Review of Assessments (Pending Accreditation Council and CAEP.
TWS Aids for Student Teachers & Interns Overview of TWS.
Day 2: Learning and Teaching Session 4: Dynamic Process NYSED Principal Evaluation Training Program.
Health and Food Technology CPD Presentation Objectives This presentation will help you to : understand the purpose of Verification prepare for Verification.
This video is for UNC Charlotte faculty developing a response form as part of the content validity protocol. The response form is what expert reviewers.
Performance-Based Accreditation
Module 6: Coaching System
Implementing edTPA An Overview.
National Board Process
Laura Hart Joyce Frazier
Refresher Course on Assessment
EVALUATING EPP-CREATED ASSESSMENTS
NCATE Unit Standards 1 and 2
Tenure and Recontracting August 29, 2017
Direct vs Indirect Assessment of Student Learning: An Introduction
A community of learners improving our world
Polices, procedures & protocols
Candidate Support for edTPA: Formative and Summative Models
Phyllis Lynch, PhD Director, Instruction, Assessment and Curriculum
Office of Field and Clinical Partnerships and Outreach: Updates
Benchmarks of Quality (BOQ) Training
Elayne Colón and Tom Dana
EdTPA, Pro-Cert, National Boards
North Carolina coordinators February 1, 2017
ASSESSMENT OF STUDENT LEARNING
Cornell Notes.
Transforming Grading Robert Marzano
Honors Level Course Implementation Guide Q & A Session Social Studies
Steps for Curriculum Mapping
CILIP Professional Registration & Portfolio Building
Subject line: Teacher Evaluation Feedback Form - Deanery ____
Directions for Expert Review Panel
Research on Using Observation Systems with Special Educators
New Student Experience
Building Engagement in Course Evaluation
Level 4 Diploma in Dance Teaching
Connecticut Core Standards for Mathematics
EdTPA 101 TEI June, 2017 UNC Charlotte.
BLT Kick-off Training Pam Lange Marilyn Hofer Barb Rowenhorst
The purposes of grading student work
Office of Education Improvement and Innovation
Tenure and Recontracting February 7, 2018
Tenure and Recontracting August 27, 2018
Principles of Assessment & Criteria of good assessment
Component 4 Effective and Reflective Practitioner
9th Grade Literature & Composition
Tenure and Recontracting February 6, 2018
Connecticut Core Standards for Mathematics
Rick McGee, PhD and Bill Lowe, MD Faculty Affairs and NUCATS
Tenure and Recontracting October 6, 2017
ISTE Workshop Research Methods in Educational Technology
Connecticut Core Standards for Mathematics
Sarah Lucchesi Learning Services Librarian
Implementing Race to the Top
Parent-Teacher Partnerships for Student Success
The Heart of Student Success
Assessing Academic Programs at IPFW
Clinical Educator and Teacher Candidate Training Chapter for the Candidate Preservice Assessment for Student Teachers (CPAST) Form Developed by the VARI-EPP*
SUPPORTING THE Progress Report in MATH
Jeanie Behrend, FAST Coordinator Janine Quisenberry, FAST Assistant
Validity and Reliability II: The Basics
Tenure and Recontracting February 26, 2019
Community Builder Activity 3 min-2 min
National Board Certification
Reviewing organizational policies with an equity lens
Cooperating Teacher and Student Teacher Training Chapter for the Candidate Preservice Assessment for Student Teachers (CPAST) Form Developed by the VARI-EPP*
Presentation transcript:

Taming the tiger: Developing a Valid and Reliable Assessment System in Partnership with Faculty Dr. Laura Hart Dr. Teresa Petty University of North Carolina at Charlotte

Two parts to our presentation: Establishing our content validity protocol Beginning our reliability work Content Validity Protocol is available http://edassessment.uncc.edu

Developing Content Validity Protocol Setting the stage … So … you can’t take a tiger from the jungle and put him right into a cage and expect him to be happy about it, right? Need to get him used to the idea first … so that’s how we started, by laying the groundwork over a period of time.

Stop and Share: What have you done to build this capacity at your institution for validity work? (turn and share with a colleague: 1 min.)

Setting the Stage Primarily with advanced programs where we had our “homegrown” rubrics Shared the message early and often: “It’s coming!” (6-8 months; spring + summer) Dealing with researchers  use research to make the case CAEP compliance was incidental  framed in terms of “best practice” Used expert panel approach -- simplicity Provided one-page summary of why we need this, including sources, etc.

Using the Right Tools You don’t expect tigers to train themselves … having the right tools is essential to their success so we made it as easy as possible for them.

Using the Right Tools Started with CAEP Assessment Rubric / Standard 5 Distilled it to a Rubric Review Checklist (“yes/no”) Anything that got a “no”  fix it Provided interactive discussion groups for faculty to ask questions – multiple dates and times Provided examples of before and after Asked “which version gives you the best data?” Asked “which version is clearer to students?” Created a new page on the website Created a video to explain it all Big item here was the idea of construct to measure: what are these items supposed to be measuring? Some faculty struggled with that. Led to some questioning what they had included in the first place. Some of the rubrics had been around for a while! This forced us to review them in a really thoughtful way.

The “Big Moment” Every trainer has this moment where you think “ok, this is how I know the tiger loves me, is happy, is okay to continue.” For us, it was the response form, the actual document we were sending to external experts for them to review. Once we got rubrics where we wanted them, we needed to have experts confirm that the rubrics were measuring what we wanted them to measure.

The “Big Moment” – creating the response form Example: Concept we want to measure: Content Knowledge Level 1 Level 2 Level 3 K2a: Demonstrates knowledge of content Exhibits lapses in content knowledge Exhibits growth beyond basic content knowledge Exhibits advanced content knowledge K2b:Implements interdisciplinary approaches and multiple perspectives for teaching content Seldom encourages students to integrate knowledge from other areas Teaches lessons that encourage students to integrate 21st century skills and apply knowledge from several subject areas Frequently implements lessons that encourage students to integrate 21st century skills and apply knowledge in creative ways from several subject areas Had to stop and make sure we were all understanding the common idea here … looking at validity: does it measure what we want it to measure? (review the example). Some faculty struggled with understanding that they weren’t looking at the indicators – just the constructs to make sure they measured what we wanted them to.

The “Big Moment” – creating the response form Could create an electronic version or use pencil and paper Drafted a letter to use/include to introduce it Rated each item 1-4 (4 being highest) on Representativeness of item Importance of item in measuring the construct Clarity of item Open ended responses to allow additional info

Talking to Other Tigers We knew it might be cantankerous, but interacting with experts (other tigers) was going to help us know our rubrics could stand up to scrutiny. So we found experts and sent our rubrics for their review.

Talking to Other Tigers (experts) Minimum of 7 (recommendation from lit review) 3 internal 4 external (including at least 3 community practitioners from field) Mixture of IHE Faculty (i.e., content experts) and B12 school or community practitioners (lay experts). Minimal credentials for each expert should be established by consensus from program faculty; credentials should bear up to reasonable external scrutiny (Davis, 1992).

Compiling the Results (seeing the final product)

Compiling the Results Submitted results to shared folder Generated a Content Validity Index (CVI) (calculated based on recommendations by Rubio et. al. (2003), Davis (1992), and Lynn (1986)): The number of experts who rated the item as 3 or 4 The number of total experts A CVI score of .80 or higher will be considered acceptable. Working now to get the results posted online and tied to SACS reports Did this for all our internally generated rubrics in 2015-16.

Stop and Share: Based on what you’ve heard, what can you take back and use at your EPP? (Turn and talk: 1 minute)

Beginning Reliability Work Similar strategies as with Validity: “logical next step” Started with edTPA (key program assessment):

Focused on outcomes CAEP  incidental Answering programmatic questions became the focus: Do the planned formative tasks and feedback loop across programs support students to pass their summative portfolios? Are there varying degrees within those supports (e.g., are some supports more effective than others)? Are there patterns in the data that can help our programs better meet the needs of our students and faculty? Are faculty scoring candidates reliably across courses and sections of a course?

Background: Building edTPA skills and knowledge into Coursework Identified upper-level program courses that aligned with domains of edTPA (Planning, Implementation, Assessment) Embedded “practice tasks” into these courses Becomes part of course grade Data are recorded through TaskStream assessment system; compared later to final results Program wide support and accountability (faculty identified what “fit” into their course regarding major concepts within edTPA even if not practice task)

Data Sources Descriptive Data Feedback Scores from Formative edTPA tasks scored by UNC Charlotte faculty Scores from summative edTPA data (Pearson) Feedback Survey data from ELED faculty The idea here was to use the data not just to check a box but to better inform program faculty, and therefore, create better experiences for our students.

Examination of the edTPA Data Statistically significant differences between our raters in means and variances by task Low correlations between our scores and Pearson scores Variability between our raters in their agreement with Pearson scores Compared Pass and Fail Students on our Practice Scores Created models to predict scores based on demographics Like to add: nothing here was particularly surprising … most of us might predict similar results with our data before reliability training.

Feedback from faculty to inform results – next steps Survey data Idea for the survey was to help inform data – how are we scoring these, how can we get better at it?

Considerations in data examination Not a “gotcha” for faculty but informative about scoring practices (too strict, too variable, not variable) Common guidance for what is “quality” for feedback (e.g., in a formative task that can be time consuming to grade drafts, final products, meet with students about submissions, etc., how much is “enough?”) Identify effective supports for faculty (e.g., should we expect reliability without Task-alike conversations or opportunities to score common tasks?)

Faculty PD opportunity 1 ½ day common scoring opportunity Review criteria, reviewed common work sample Debriefed in groups Rescored a different sample after training Results indicate faculty were much better aligned Will analyze 2017-18 results next year Scoring common work sample will be built into the faculty PD schedule each year

So to wrap up …

Questions?? Laura Hart Teresa Petty Director of Office of Assessment and Accreditation for COED Laura.Hart@uncc.edu Teresa Petty Associate Dean tmpetty@uncc.edu