In healthcare conversations, the "popular" topic of predictive analytics is often discussed, but infrequently is it understood. Without doubt, significant value can be realized using predictive analytics, but only if all the essential components are present. One of these critical components is measurement. Creating a measure that is calibrated to measure what you say it measures (it is accurate) requires three types of psychometric testing. Dr. John Nelson explains this in layman’s terms, explaining how testing your measure for validity ensures it measures what you say it does; for reliability to ensure all who are responding to the survey understand the measure the same way; and invariance testing to compare demographics to ensure there is no bias from one group to another. Many of the "solutions" promoted by vendors today are based on measurements that are insufficient or irrelevant to an organization's or unit's context. In this episode, Dr. Nelson shares how you can approach measurement in a manner that will truly benefit you and your organization.