IPPCR 2018: Measurement Part 1: Reliability

>> David Luckenbaugh: Hello, my name is David Luckenbaugh. I'm a statistician with the Office of Equity, Diversity, and Inclusion here at NIH in the Office of the Director. I want to welcome you to a conversation about reliability and measurement in clinical trials. The first thing we need to do as we're talking here

is just define our objective. Our objective is to understand reliability of measurement as it fits into a clinical trial. In order to understand reliability, the first thing we need to do is to talk about what a construct is. So, a "construct" is a theoretical concept. It can be a concept like depression or mental health;

it could be something like a biological measure. Or it could be something much more -- much more theoretical. A construct is necessary to understand the measurement process because we need to measure something. The construct is the something that we're going to measure. So, "measurement" is a system of defining the level of a construct.

Often, we talk about in clinical research using an operational definition. An "operational definition" is a method used for examining some kind of domain. Have you ever considered when you go to do something the first time it's a unique experience, but then when you go to do it again it's not quite the same?

Well, when we talk about whether something is consistent from one point to the next we talk about reliability. So, reliability can come up in any kind of different field. It could be in depression where you have multiple different kinds of rating scales. It could be in other fields that study tremors where you might have something that's rated by a judge