Reliability of ADA- CAT for ADA Compliance Denis Anson Director of Research and Development Assistive Technology Research Institute Denis Anson Director.

Slides:



Advertisements
Similar presentations
NETS-S Curriculum Review
Advertisements

You can use this presentation to: Gain an overall understanding of the purpose of the revised tool Learn about the changes that have been made Find advice.
Design of Experiments Lecture I
What is a CAT?. Introduction COMPUTER ADAPTIVE TEST + performance task.
The Research Consumer Evaluates Measurement Reliability and Validity
Plan Evaluation/Progress Monitoring Problem Identification What is the problem? Problem Analysis Why is it happening? Progress Monitoring Did it work?
© Cambridge International Examinations 2013 Component/Paper 1.
Consistency/Reliability
Linear Regression and Correlation Analysis
18 January Writing a Functional Spec. Administrivia How many teams will want departmental web space vs links to your own space? Please send me your CS.
Personality, 9e Jerry M. Burger
Correlational Designs
Classroom Assessment A Practical Guide for Educators by Craig A
Introduction to GREAT for ELs Office of Student Assessment Wisconsin Department of Public Instruction (608)
But What Does It All Mean? Key Concepts for Getting the Most Out of Your Assessments Emily Moiduddin.
Electronic Communication and Web Accessibility Workshop.
ANCOVA Lecture 9 Andrew Ainsworth. What is ANCOVA?
 Basic information on the essential elements of a Measurable Behavioral Objective (MBO).  A Template to guide you in the creation of an individualized,
Topic 6.1 Statistical Analysis. Lesson 1: Mean and Range.
LECTURE 06B BEGINS HERE THIS IS WHERE MATERIAL FOR EXAM 3 BEGINS.
Analyzing Reliability and Validity in Outcomes Assessment (Part 1) Robert W. Lingard and Deborah K. van Alphen California State University, Northridge.
A MULTIMEDIA APPLICATION FOR THE TEACHING OF THE MODULE "WATER" Anna Thysiadou 1, Sofoklis Christoforidis 2, Panagiotis Jannakoudakis 1 1 Aristotle University.
Copyright © 2005 Brooks/Cole, a division of Thomson Learning, Inc. Chapter 2 The Data Analysis Process and Collecting Data Sensibly.
Classroom Assessment A Practical Guide for Educators by Craig A
Slide 13-1 Copyright © 2004 Pearson Education, Inc.
L 1 Chapter 12 Correlational Designs EDUC 640 Dr. William M. Bauer.
Tests and Measurements Intersession 2006.
No criminal on the run The concept of test of significance FETP India.
An Examination of Science. What is Science Is a systematic approach for analyzing and organizing knowledge. Used by all scientists regardless of the field.
Power Point Slides by Ronald J. Shope in collaboration with John W. Creswell Chapter 12 Correlational Designs.
Creswell, Educational Research: Planning, Conducting, and Evaluating Quantitative and Qualitative Research, 4e © 2012, 2008, 2005, 2002 Pearson Education,
Statistical analysis Outline that error bars are a graphical representation of the variability of data. The knowledge that any individual measurement.
SINTEF Telecom and Informatics EuroSPI’99 Workshop on Data Analysis Popular Pitfalls of Data Analysis Tore Dybå, M.Sc. Research Scientist, SINTEF.
Objectives 2.1Scatterplots  Scatterplots  Explanatory and response variables  Interpreting scatterplots  Outliers Adapted from authors’ slides © 2012.
Research in Communicative Disorders1 Research Design & Measurement Considerations (chap 3) Group Research Design Single Subject Design External Validity.
Appraisal and Its Application to Counseling COUN 550 Saint Joseph College For Class # 3 Copyright © 2005 by R. Halstead. All rights reserved.
Analyze Design Develop AssessmentImplement Evaluate.
Introduction to Web Authoring Ellen Cushman our syllabus
Data Mining BY JEMINI ISLAM. Data Mining Outline: What is data mining? Why use data mining? How does data mining work The process of data mining Tools.
1. DEVELOP THE PROJECT QUESTION/PURPOSE Find a relevant topic of interest Write a question to be answered (How, What, When, Which, or Why?) Write down.
A Content-Based Approach to Collaborative Filtering Brandon Douthit-Wood CS 470 – Final Presentation.
Research Methodology and Methods of Social Inquiry Nov 8, 2011 Assessing Measurement Reliability & Validity.
Regression Analysis: Part 2 Inference Dummies / Interactions Multicollinearity / Heteroscedasticity Residual Analysis / Outliers.
Nurhayati, M.Pd Indraprasta University Jakarta.  Validity : Does it measure what it is supposed to measure?  Reliability: How the representative is.
Ensemble Methods Construct a set of classifiers from the training data Predict class label of previously unseen records by aggregating predictions made.
Measurement Experiment - effect of IV on DV. Independent Variable (2 or more levels) MANIPULATED a) situational - features in the environment b) task.
1 Evaluating the User Experience in CAA Environments: What affects User Satisfaction? Gavin Sim Janet C Read Phil Holifield.
LISA A. KELLER UNIVERSITY OF MASSACHUSETTS AMHERST Statistical Issues in Growth Modeling.
IT323 - Software Engineering 2 1 Tutorial 3.  Suggest ways in which the user interface to an e-commerce system such as an online stores might be adapted.
1 Collecting and Interpreting Quantitative Data Deborah K. van Alphen and Robert W. Lingard California State University, Northridge.
Slide Slide 1 Chapter 10 Correlation and Regression 10-1 Overview 10-2 Correlation 10-3 Regression 10-4 Variation and Prediction Intervals 10-5 Multiple.
Introduction Many problems in Engineering, Management, Health Sciences and other Sciences involve exploring the relationships between two or more variables.
Session 5: How Search Engines Work. Focusing Questions How do search engines work? Is one search engine better than another?
Stats Methods at IC Lecture 3: Regression.
Web Accessibility John Rochford Rich Caloggero UMMS Shriver Center
You Can’t Afford to be Late!
Review of Assessment Toolkit
Writing a sound proposal
Software Quality Control and Quality Assurance: Introduction
Unit 1: Matter, Measurement, and unit conversions
Classroom Assessment A Practical Guide for Educators by Craig A
Collecting Data Sensibly
Analyzing Reliability and Validity in Outcomes Assessment Part 1
Student Satisfaction Results
Nature 2018 Summer Camp Hypothesis and Product Testing
An Introduction to Correlational Research
Collecting Data Sensibly
Collecting and Interpreting Quantitative Data
Presentation transcript:

Reliability of ADA- CAT for ADA Compliance Denis Anson Director of Research and Development Assistive Technology Research Institute Denis Anson Director of Research and Development Assistive Technology Research Institute

What is ADA-CAT? ADA-CAT is a combination of a website and set of physical tools ADA-CAT is a screening tool for accessibility of the environment, based on ADA standards ADA-CAT was designed to allow non- engineers and non-architects to perform reliable and accurate assessments of compliance with ADA requirements

The Website The ADA-CAT website provides assessment checklists (Audits) for features of the environment

The Website Each item in an audit includes: The ADA (or other) requirement, in understandable language A link to the ADA-ABA requirement on the web An explanation of why this feature is important Just-In-Time Training

Just-in-Time Training Training you get today will be forgotten when you need it ADA-CAT allows you to review how to assess an item as you are assessing This includes both text and pictures of the process The training is based on the ADA-CAT toolkit

The ADA-CAT Toolkit The toolkit includes eleven tools in a custom bag The tools include three off-the-shelf tools 25-foot tape measure 24 inch spirit level Stopwatch

Three Off-The-Shelf Tools The tools are easily found in sports and hardware stores The assumption might be that everyone knows how to use them Watching novices attempt to use these common indicates that this isn’t the case

The ADA-CAT Toolkit The toolkit contains three hard to find, but off-the-shelf tools Light meter reading in Lux A sound meter reading decibels A force gauge

Three Hard-To-Find Tools It’s unlikely that many have ever used tools like this Most people have little understanding of the concepts being measured Inverse Squared Rule? Although the tools are “simple” to use, if not used correctly, they won’t give good results

The ADA-CAT Toolkit The toolkit contains five custom tools not available anywhere else Magic Slope Block MultiTool Key Torque Tool The Font Guide The StoryStick

Five Custom Tools These tools were created specifically for use in ADA-CAT assessments Although the principles are hidden from the user, they can be complex No-one who has not used ADA-CAT will have any experience with these tools.

Little or No Training Required? The assertion is that ADA-CAT, the combination of the website and the tools, will produce reliable results when used by non-architects This is a bold claim, and requires support It is also testable

What is Reliability? There are two types of reliability that are commonly recognized in tests and measurements Test-Retest Reliability Inter-Rater Reliability

Test-Retest Reliability Test-Retest reliability means that when a measurement is taken repeatedly, the outcome is the same When you measure the width of a door, you should always get the same width for the same door Tools that have large subjective components will not show this reliability

Inter-Rater Reliability Formally, this means that the results of a measurement are not dependent on the person doing the measurement If I measure the weight of a glass of water, then you do, we should get the same result In accessibility assessment, the “I was/wasn’t able to do it” test depends on who “I” am

Testing ADA-CAT Reliability A group of four OT research students undertook this task The students received an introduction to the tools of the toolkit, but no observed trial in the real world The students used the Elevator Audit as their sample test

Why Elevators? The Elevator Audit uses almost all of the tools of the toolkit We didn’t use the Font Guide, but did use all the other tools Elevators are commonly available There is a wide range of accessibility in elevators in daily use

Method Identify eight elevators to be assessed Each researcher measures each elevator After at least three weeks, each researcher measures each elevator again

Test-Retest Reliability Each researcher’s measurements can be compared with their measurements three weeks later. Assuming no damage or repair to the elevators, there should be good agreement between the results

Inter-Rater Reliablity The measurements of each researcher can be compared to show the inter-rater reliability When the scores for all four researchers were compared, the inter-rater reliability was.6 Statistically significant, but not exciting Why?

Inter-Rater Reliability

Examining the data shows that, in most cases, three of the researchers had closely related data, and one was different. Removing this individual increased the inter-rater reliability score to above.9 So why was this researcher’s scores different? We’ll look at that in a minute.

Inter-Rater Reliability (Filtered)

Test-Retest Reliability If the scores of the researchers on first measurement are compared with the measurements 3 weeks later, the correlation is.97

Test-Retest Reliability

So, What’s Wrong with Inter-Rater? Individual items were compared between researchers where there was not good agreement The differences clustered around two features Sound levels Braille

Sound Levels The Elevator Audit includes standards for the relationship between elevator signals and background noise, and for background noise levels The outlier researcher did her evaluations during times when there was a lot of surrounding activities During a basketball game for the athletic center elevator

Sound Levels The background sound levels at busy times represent the people in the area, not the elevator, and cannot be controlled by the manufacturer The sound measurements should be taken at quiet times, as the other three researchers did

Braille? Visually dependent people have a very difficult time determining if Braille is appropriate In some of the older, and more abused elevators used in this study, the “raised” component of Braille was lost due to damage and wear This was interpreted differently by different researchers, resulting in some variability

Conclusions Some additional instructions are needed on when to take sound measurements Additional examples and counter- examples of Braille need to be added to the assessments

Conclusions Overall, the assertion that ADA-CAT reliably measures compliance with accessibility standards is well supported by this project