 Interval by interval, interobserver agreement. All right, again, I'm gonna say this every time. It is not inter-rater reliability, it's inter-observer agreement because it speaks to validity, really speaks to believability, not validity. Oh my gosh, that was very valid. So you're probably not, wee bit and now you're not gonna believe me at all. So no, interval by interval IOA, right? Really simple. This is used when you do like a partial recording method or a whole interval recording method. All you do is compare the intervals to each other. So you've got two people making recordings about the same interval time, right? So did the behavior happen at all during the first two minutes? Yes, no, whoop, they don't agree. How about this one, did it happen at all? Yes, yes, did it happen at all? Yes, yes. Now we have three intervals that you just saw, picking nose or screaming or getting out of turn, well not picking nose, that's discrete. So continuous type behaviors, right? So they're talking out of turn, whatever. Did the behavior happen at all during that time period? We're using a partial interval method, right? So then we go, we compare check boxes. If they agreed, and so in this case we had three intervals, so did they agree two out of three times? So 66% roughly of the 66% agreement because we take the total number of agreeing intervals divided by the total number of intervals overall, right? So it's a pretty good method. It's not quite as good as exact agreement, but it's not bad at all. And pretty strict for, in fact the strictest one you can have for continuous type behaviors.