How to Build Valid Tests & Assessments

How to Build Valid Tests & Assessments

Using  Generative AI

A 3-Part Certificate Series: 

Fridays 10a.m. to 12:30p.m. Central Time

April 10 to April 24

Intended Audience:  This program is for internal and external consultants wanting to gain the expertise required to develop valid tests and assessments and deliver quality outputs in significantly less time by leveraging generative AI.  You will learn what is required to create a valid test, the value and limitations of different testing methods, how to fairly score performance-tests, how to select and qualify judges, and how to leverage generative AI to accelerate the process.   You will see real examples of AI prompts and the outputs: response supplied (multiple-choice tests), response not-supplied (performance, portfolios), and rubrics.   You will get job aids with practical guidelines that you can apply to your work.  

The Test and Assessment series is part of Hale Center's 2026 theme on Truth, Innovation, & Technology.   It is designed to strengthen your technical skills in assessment and measurement.   When you complete the 3-part program you will receive a certificate in Testing.  

Topics in the series are:

Part A:  Basics

1. Test Basics:  learn what makes something a test and what is required for a test to be valid. 
2. How to Decide What to Test, Why, and How:  learn the value of  developing operational definitions and doing a criticality analysis to determine what and how to test.  You will get guidelines and tools for developing an operational definition and doing a criticality analysis.  

Part B: Multiple-Choice

1. How to Develop Multiple-Choice Questions that Test Higher Level Skills:  find out how to use the work of David Merrill and Robert Bloom to create forced-choice questions (multiple-choice) that measure memory and reasoning.  Learn how generative AI can significantly accelerate the process allowing you to make smart use of your subject matter experts. You will receive a toolkit with guidelines for developing effective multiple-choice tests 
2. How to Analyze Test Results:  learn how to do an item analysis to identify which questions and answer choices are effective and ineffective.  You will also learn how to compare tests to confirm they are equivalent, that is they measure the same knowledge and are equivalent in difficulty. 

1. How to Develop Performance Tests: learn about the different types of performance tests, what differentiates them in terms of fidelity (how life-like they are), and what is required of each.  You will receive a set of guidelines for developing performance tests, and how to accelerate the process by using generative AI.   You will see examples. 
2. How to Develop Checklists and Rubrics:  learn how to more quickly design and develop checklists and rubrics to score performance tests using generative AI .  Learn to train and qualify judges in how to score performance tests and control inter-rater reliability.  You will receive templates and guidelines. 

If you want to register for specific topics only, please contact Judy@HaleCenter.org and she will complete your registration.  Please note, one paid $444 registration allows up to six people in your organization to attend at no additional charge.   Just send Judy the names. 

Materials:  Each topic comes with a workbook containing the slides, guidelines, and other supporting materials. 

Cost:  For the series is $444.  A total of six in your organization may attend at no additional cost. Please contact Judy@HaleCenter.org if you want to find out how to arrange for an in-house program.