[Music] hi everyone and welcome to kind of our four first topic on measurement and how to take measurement so this is going to be continuous measurement continuous measurement are measurement systems that measure all instances of behavior during an observation period it's different than discontinuous measurement systems which we'll do next topic which give you a snapshot of the behavior within the observation period so anything continuous is always better um but it's harder to do it takes a lot more energy but they are the most reliable measurement systems so it's important if you're able to do continuous you choose to do continuous but if you're not able to do continu it's just it's too hard to capture every second of the observation period Then you would utilize a discontinuous measurement system and a lot of times your supervisor will be the one choosing this for you if you when you utilize these so the first and easiest type of measurement system is a count it's simply a count of how much Behavior occurred you usually do it with a tally count um you use when Behavior has a clear ending or beginning and it occurs at a medium or low rate really high rates are kind of hard to do um higher rate behavior that may lend itself to frequency would be better on a discontinuous measurement system it might just be too difficult to accurately get the count but sometimes people do do it with um clickers like so you just click how many times like hitting is occurring and you just keep your hand on your L clicker and you keep clicking and it keeps count for you that is one way to do those higher frequency behaviors but what's this is usually best for medium to low rate and um clear beginning and hand um and and you can see this is talking out of turn without raising hand and they're just taking a frequency count of that and so on 9:30 it occurred three times on 710 it occurred three times on 7:14 it occurred eight times and you can see you can just keep a easy tally the advantage of frequency data is it's really straightforward and common like everybody you'll everybody will use it it's probably what you kind of thought of when I started saying collect data for an observation everybody thinks of frequency the disadvantage of frequency data is it's difficult to collect when the rate of behavior is really high um or when the behavior doesn't have a clear ending or beginning because you don't know when to say it happened once it happened twice it happened three times um or if the RBT has other duties to perform while they're collecting it can be very difficult so some frequency examples that are good are like raising their hand in class asking for a desired item most verbal behaviors are great for frequency hitting a sibling um biting nails these are all behaviors that have clear beginning and ends um and you can take a tally count of them occurring so rate and we're going to talk about rate a second time in this module but I'll introduce it here so rate is sometimes when you collect data on a learner's Behavior the amount of time You observe might change from day to day or session to session if you only look at frequency does not take into account the time so say You observe for 20 minutes how much time someone hit and then the next day You observe for 30 minutes how much time someone hit if you say if you only without reporting the minutes you observed you said oh they hit four times in the 20 minutes and eight times in the 30 minutes if you say one day they obser they hit four times and the next day they hit eight times that's not really a clear picture because what really what always matters is the length of time You observe that's part of it because the longer you observe the more opportunities you have to observe the behavior so the behavior should increase The More You observe it but so we have to take that into account and so we do that with a calculation and this calculation is you put the frequency count above the amount of time so for example if you got eight counts of aggression or hitting um over a 2hour period what you do is whatever period you observed most of the time we're taking that to one so if You observe 20 minutes we're going to 1 minute if You observe two hours we're trying to get to one hour so eight accounts they eight hit they hit eight times in two hours but we make that into four accounts of aggression per hour so if they had three Tantrums in two hours you would put the three on top and the two on the bottom and you'd go 1.5 Tantrums per hour they're having it's kind of like an almost like an average per hour mans for attention and we know mans is a demand so they were verbalizing they wanted something and you saw 10 mans per in a 2hour period you observed for two hours and you saw 10 mans you would say they're you know on average they're making about five mans per hour and then um if you had five mans a man for a break so that's I need a break so they five times they told you I need a break in two hours you'd say they're doing about 2.5 mans per hour so here's another example of rate because I know it's a harder concept If You observe for one hour on Monday and saw five instances of hitting and two hours on Tuesday and saw five instances of hitting but only graphed only wrote five and five said oh I observe five times this time and five times that time everyone would think that the rate of hitting did not change it was the same throughout the day they have cons they consistently hit about five times but that's not true because you observe double the amount of time on Tuesday so it actually went down quite a bit on Tuesday for Monday they were hitting a lot more on Monday um so rate takes us into account so in the example above Monday they hit five times per hour and on Tuesday they hit 2.5 times per hour and just taking that example you may like we would want to think about that and go well what changed Monday to Tuesday that's a big drop so we can start like better understanding the environment and what's going on with the child and why they're hitting you can report that per hour per minute or per day so you can report rate a couple ways but it's just taking into account that time element um and so again you it's really important because in that example if you didn't do that you might think it was stayed the same Monday and Tuesday and it did not okay so rate is kind of part of frequency I wouldn't call it its own a measurement system it's just a way that we make it so we can understand frequency better so duration is the length of a behavior it's another way to take data so to measure duration you start the stopwatch when the behavior occurs and you stop the stopwatch when the behavior ends and this can be done with a stopwatch computer or phone and you can see here tantrum so the first tantrum was 60 seconds the next tantrum was 120 seconds the next tantrum was um 90 seconds then it was 27 seconds and then it was 113 seconds so every time that child tantrum the person started the stopwatch and they started it and then they stopped it they started it and they stopped and they wrote down those times and then how much time so the whole observation was 12 to 3 so it was 3 hours and in that 3 hours um they they um sorry um in that time they a tantrum a total of 600 seconds which is 10 minutes um so that's an example of how you take duration data so duration examples like good behaviors that are good for duration are kind of the longer behaviors so how long a tantrum last the length of social play like how much someone's engaging in Social play how long does it take for a child to take for a child to um get dressed so you always want the anytime you might take duration over frequency it's because the behavior is kind of longer and that length of time is interesting to us in some way so Tantrums usually are a duration um but like other things that are really interesting to us is like social play how long it takes takes to get dressed okay our third third type of data collection is inner response time um so this is the time between responses so to take IRT data you begin the stopwatch when a behavior ends and you stop the stopwatch when the behavior begins so it's like the opposite of duration at the end of the observation you take the average in response time so when you're interested in the length of time between different responses that is when you might utilize this it's not utilized as much as duration so this little graph I love it it like totally it held me the most when I was learning this okay so this is dog barking Behavior right that's what it's showing so when the dog is a little bit higher and this is time so as time goes on so you can see the dog bark three times so when the dog is placed a little higher he's barking and that's the length of times he's barking So He barked three times the middle time He barked the longest um and it looked like the two times on the outside were about the same now all the other time in the observation period the dog was not barking and you can see he's a little lower and his mouth is not open and there's four of those so four times he didn't bark three times He barked so those four times where he didn't bark is your IR right there it's the time between the responses when behavior is not occurring and then the times he was barking the time a little the little higher dog is your duration that's the time he was barking so duration is when the behavior is occurring and IRT is when the behavior is not occurring what's cool about this is if you get good duration or you get good IR you can fi figure out the other one so even if we go back to our other example with those Tantrums I knew all the time that the Tantrums occurred if I know the length of the behavior I can figure out how the total IRT how much inner response time there was how much did the behavior not occur um and so if you add duration and IRT together you get your whole observation period so you can't you know you wouldn't know exactly where the irts are but you could get total IR or toer duration if you have one or the other so things um you might want to take IR from or inner response time is Time Lapse between doing math problems time lapse between spontaneous communication and then timelapse between pro-social behaviors so those are possible things that IR is important for in a response time but a lot of times that might just be duration duration data this is just not as commonly used as duration are usually as useful as duration okay so latency latency is our last one latency is the time it takes for a behavior to start and typically the time from prompt to respon so someone give someone prompt you start the stopwatch when the prompt ends and you stop it that when the behavior begins it's the time to behavior um and then you take the average of the data you collect it over an observation period so if for example you're taking data on the time it takes for someone to follow the direction open your book you know You' be hanging out in the classroom and every time the teacher gives a directive like open your book you'd start your stopwatch when your child or your client opens their book you'd stop it you'd write that down and then you know they do they do their work all that and then the teachers be like put away that book we're taking out our science book so please take out your science book and so then you take the data again you would take the average of these and you'd write something like the average latency from prompt to um taking out his book is you know 20 seconds or whatever it is and so most of the time we're doing this because we want to shorten the latency there's a problem the latency is too much um people are like hey get dressed and they take 10 minutes to get dressed so you want to take that data and we're going to try to shorten that um because it's going to help the child within the environment so time from instruction to get dressed to get response time from instruction to begin table work to response time lapse between peer asking a favorite movie asking about a favorite movie to client response so time lapse between someone asking them a social question to answering that's a really important one I've taken that a couple times because because um if a peers have if one of your clients is having you know like problems socializing sometimes that's a big part of the problem people take some kids or some clients may need a lot of processing time so hey um you know what's your favorite movie and the client's processing they got to think about it's going through their brain doing all that and they're coming up with an answer and it's just taking them a really long time and it's just kind of you know it's hard and that Pier gets bored because he's also a kid and he walks away because he's like oh he's not answering me so that could really cause some social problems for your client and so you might want to work on that and so you take that data and then you'd start working on that okay so there's this is kind of separate so I'll say that this I just think kind of helps people operationalize this stuff and understand it better there's three measurable dimensions of behavior so behavior um you know there's reasons why it's measurable first Behavior repeats two it has temporal extent and three it has temporal Locus I'm going to talk about each one of those so repeatability behavior Can repeat throughout time so it can occur once and then it can Cur occur again and because it's repeatable we have count and frequency are the same thing so we have frequency and we have rate the reason we are able to take that data is because behavior is repeatable so temporal extent means Behavior occurs at some amount of time so it takes up time on the time line of Our Lives or of um a classroom or whatever so it exists within time and because of that because we know that it we have duration that's the reason we have duration and then um Behavior occurs at some point in time like it's a set time it starts in a set time it ends and that's what temporal Locust means and that's why we have inner response time and that's why we have latency okay so that little last part is hugely important for those who might be going on to become bcbas um that is not um on the RBT exam I just think it kind of if you start looking into those things it will kind of help you understand these Concepts a little bit more understanding repeatability temporal extent and temporal Locus so repeatability means behavior Can repeat temporal extent means it takes up time like it exists with there's time between the start of the behavior and the end and then typ temporal Locus means that time is set is on a timeline so you know we can say it started at 10:00 and it ended at 10:10 so it kind of shows you how we're able to take the data and the point of all that is any one of those Data Systems we just talked about they're only really taking showing a little bit of the obser the ability to observe the behavior that we have so like frequency is leaving out duration it's leaving out temporal Locus and duration is leaving out temporal extent to a certain to a um sorry frequency is not showing you temporal extent or temporal Locus so you're losing some of the things we can observe of behavior and then temporal extent is not taking into account temporal Locust you can get repeatability from temporal extent so you know it's that we're not seeing the whole Behavior they all have their weaknesses and their strengths okay thank you I will see you next um PowerPoint which we'll talk about discontinuous measurement