My name is John Wilson. I'm a special education teacher for the Lenape Regional High School District in Medford, N.J. I currently teach vocational skills to a small group of autistic boys. I'm relatively new to Precision Teaching and charting. I've been lurking on the ListServ for about 6 months. I do not have a chart parent. I'm attempting to learn on my own. As this is my first attempt at charting a non-repeatable behavior (at least I think that's the proper term), I would very much appreciate both positive and corrective feedback!
 

Kevin is a 17 year old autistic boy. During the 2000-2001 school year, he performed volunteer work in his high school's media center. The media center staff was so impressed with hi productivity that they offered him a paid summer job. They assigned him the task of cleaning all of the school's overhead projectors. The chart shows his performance during training.  Initially, I developed a 20-step task analysis and provided Kevin with a written procedural checklist. I taught him to follow the written checklist using procedures outlined in McClannahan and Krantz, 1999.
I included the raw data sheet/task analysis taught him to follow the written checklist using procedures outlined in McClannahan and Krantz, 1999. Kevin actually has the opportunity to clean as many overhead projectors per day as his work schedule permits (he has several other tasks to complete). I decided to chart his first opportunity of the day as a cold probe - test-teach-wait/test again. Duration data was within the supervisor's expectation of 15 minutes. His probes took 12, 12, 11, 9, 9, and 8 minutes. It seemed to me that duration was not an issue. And Kevin's supervisor said that t he quality of his work was much better than he expected.
McClannahan, L.E., and Krantz, P. (1999). Activity schedules for children with autism: Teaching independent behavior. Bethesda, MD: Woodbine House.

ListServer Comments from Alison Moors, 7/11/2001
John- I love this chart. I also appreciated the raw data to show me how you broke down the skill. Did you do any component motor skills training? Like the "big 6"? Let me ask- I couldn't tell if there was a timing interval charted. It looked like the corrects and errors always added up to 20,right??? So- am I right to assume that the data only represent one overhead projector cleaned per day? Now that his accuracy criterion is at 100%, and he no longer needs the schedule (I assume this from the probe), I would love to see you put the entire 20 task activity on a duration timing to see how long it is taking him to be accurate. Does this student have an opportunity to clean more than one projector per day? Also- just a little charting convention, when charting 0 errors, one can put a little question mark just below the timing bar (or in this case, below the one per day as you've done with an X)...this way, the observer can immediately pick out which days had errors (X) and which days were 100% accurate (?).... Great job! I can't believe that you did this without a chart parent. You are a faster learner than I ever was when I started... :-) Ali

From Dennis Edinger, 7/12/2001
John and fellow charters,
This is an outstanding effort, especially without an on site mentor to help. I want to analyze this chart as I see it just as it was posted. For folks not familiar with my terminology or with chart analysis in general the general references I use are available on the web. They are: "Standard Chart Topics -- A Precision Teaching Resource Website" by John W. Eshleman, Ed.D. located at http://members.aol.com/standardcharter/index.html. Another site by John is "The Universal Frequency and Celeration Site" at http://members.aol.com/JWESHLEMAN/. Not available on the web, but absolutely necessary for understanding chart analysis is the "Handbook for the Standard Behavior Chart" by Hank Pennypacker, et. al. which is available from a printer in Gainesville FL, but I can't find the phone or address handy. Perhaps someone on the list can find it.
John, the first thing I would do, is change the N/A on item 15 of your Programmed Instruction. Kevin must check the light one way or another. I think that item was just poorly written. This would give you a total of 20 "task steps".
Next, you should note, as Ali pointed out, that you have two pinpointed movement cycles. As she notes, "task steps correct" would be noted with a . while "task steps incorrect" would be noted with an x. Some folks use a ? for zero. This is because it is quite difficult to directly observe nothing!
So you now have a single chart with two pinpoints configured as an Learning Pair. It is important to note that you used the DAILY chart, not the MINUTE chart. For this kind of problem this choice is correct for this pinpoint as stated and the counts as observed. We will see that a DAILY chart may be needed later. Since both charts are STANDARD CHARTS the analysis below would be the same regardless of the chart you chose. This is a unique advantage of precision teaching over any other form of instruction. Kevin will tell us what is appropriate with his plotted movements, not some expert. The precision teaching principle here is: "The Kid Knows Best" (over the years that may have be politically corrected to "Child" or "Person", I'm not sure).
First I want to note that you filled out all the nominal data--the blanks that request information. Often when I look at a chart this is the first thing omitted. You correctly made it your first task, and that's GREAT precision teaching. As I noted above, you do need to note by "Counted" that you are counting two movement cycles, correct and incorrect, indicating that the person reading the chart will know that a "Learning Pair" or "Accuracy Pair" is on the chart.
When I drew Celeration lines through the correct and incorrect, I immediately noted that in week one Kevin's tasks correct and incorrect present a classic "JAWS" pattern: http://members.aol.com/standardcharter/jaws.html. This tells us that the program as written is working and that we need to make no interventions or changes in instruction. I didn't replot your data on a full chart so my celerations may be a bit off, but I find Kevin to be accelerating tasks correct at x3.3 movements per day and decelerating tasks incorrect at /10 movements per week. If you have Pennypackers "Handbook..." you will find procedures for calculating an "Accuracy Ratio" as well as an "Improvement Index", two different ways for showing accuracy. These are far better than just calculating % correct and they can be done directly on the chart. If one MUST deal with % that also can be calculated on the chart with a "Frequency Finder" or by hand. I'm probably in the minority, but I like the Improvement Index because it uses Celeration as it's base. By my rough analysis I see that Kevin has an Improvement Index (noted II) of X10 in week one. He is really improving fast!
Apparently your gut or genius told you not to intervene after week one. Given the II of x10 that is clearly the correct choice. In week two I calculate Kevin's celeration for "tasks correct" at x1.0 movements per day and his "tasks incorrect" at /1.5 movements per day or so. This is a classic "DIVE" learning picture: http://members.aol.com/standardcharter/dive.html where correct moves to a flat x1 or so and incorrect continues to decelerate toward zero. I calculate the Improvement Index at about X20. Kevin is really flying as he gets ready for summer vacation.
On week five after summer vacation you have correctly refrained from any additional intervention as the program is working fine by itself. I calculate Kevin's "tasks correct" accelerating at about X1.2 and his "tasks incorrect" decelerating at /2.0 or so to zero at the end of the week. Again I get an Improvement Index of X20. Kevin rules!
On week six I see that the learning picture has again changed. He is now showing a classic MID LEVEL (sometimes called by kids as RAILROAD TRACKS): http://members.aol.com/standardcharter/midlevel.html. John, it shows that you were really reading the chart correctly when you saw an "Accuracy Ratio" (in Pennypacker, et. al.) of x20 the same as his last weeks II of x 20 as you chose to "Probe".
Now that Kevin is performing with perfect accuracy, you may want to see if he can clean the overhead even faster. To do this, you will have to change the chart to a "Count per Day" chart with the same Sync Days at the top so any reader will immediately see the continuous charting. You will have to determine a Record Floor or Timing Floor for this chart. You didn't note how many minutes it is taking him to do this task, but that will give you a good idea of where this floor lies by examining the times for the last three days of Kevin's cleaning. The movement cycle will change however. Now that you have finished the 20 task program, you no longer need it. I prefer to take a week and just look to see where 1/minutes taken for "cleans overhead correct" and "cleans overhead incorrect" will lie. Other precision teachers will just say they have enough data from week six and jump directly into some change in the program.
Should you elect to continue to teach this as a Free Operant, you can Arrange (which you hope will be a Schedule of Reinforcement or Contingency) and some Subsequent Event (which you hope will become a Reinforcer or Consequence) for either the correct or incorrect movement or for both simultaneously. I'd opt out of targeting the incorrect completely given the data shown and concentrate on "Cleans Overhead Correctly". Over time your experience will tell you. With your description of Kevin, I'd keep it simple.
(Note: for those of you who have heard about IS/DOES my language above illustrates it. The Movement Cycle, Arrangement and Subsequent are from the IS side; they are just the Teachers Plan; an instructional sequence planned but not confirmed. When they are demonstrated to be effective instructional interventions, the language changes to the DOES side: Contingency and/or Consequence. It is possible to have an Arranged Consequence, which would tell one to look at the count or time (depending on the arrangement) for tweaking. It is also possible that the Movement Cycle change I suggested is too big for Kevin, in which case we would slice back to the tasks. If the Movement Cycle links up with the Arrangement or Consequence we would, by definition, have a Response or Behavior on the DOES side. In summary, the IS side is a Lesson Plan not yet validated by Kevin. The DOES side will be the charted validation of each instructional component, so it becomes the Performance Evaluation. I hope that is clear. Look at Lindsley's "Direct Prosthesis...." in the 1964 Journal of Education for the original discussion of this critical part of precision teaching.)
Another choice within the precision teaching repertories would be to move to Fluency Trials. I have very little experience with setting up this type of instruction, but you will need to get Timings Charts. Sadly, Timings Charts are not Standard with the Minute and Daily charts, but for the Best Timing you will plot that on a Daily Chart.
I'm certain any of the other list members can make a suggestion as how to proceed should you decide to go to Fluency Trials in some detail.
At this point you can clearly see that Kevin is in the driver's seat. His charted performance tells us on a day by day basis how to help him learn.
John, I must again complement you on a wonderful chart for instructional demonstration and discussion. Ian, thanks for making this available for us.
Looking Forward,
Dennis
[In the beginning there was the Sponse...

From Richard McManus, SCList, 7/12/2001
I am with Ali! I love this chart! Congratulations on doing such a great job with trying to figure out how to do this on a chart!
I suspect that it is time to look at the components so that you can figure out (now that the task is being DONE) how close to mastery Kevin has come. As Ali said, how about charting some of the elements in isolation? And how about figuring out the duration of a skilled projector cleaner to see how close Kevin is to achieving a competitive level skill.
There is a risk in charting a multi-step task this way and calling the resultant combination of steps "20." That is that since these are NOT repeatable movement cycles, they are sort of apples and oranges mixed together.
It has not been a problem SO FAR but you are masking some potential problems. If one of the steps hangs up the others you won't be able to see it on this chart, so you won't be able to diagnose the problem.
I did lots of charting of complex tasks in a vocational setting. We typically called the completion of an entire set of steps "one" so that we had movement cycles that were repeatable. So in that case he has ONE cycle divided by the total time. Say that is 10 minutes, so now we have .1 cycle per minute.
So then, like Ali I wonder if he can clean multiple projectors per day? That way you can chart the result and take a look at the increase of speed. For example, if he could then clean TWO projectors in 10 minutes he will be going UP to .2 per minute--doubling!
The other problem you will encounter is that the above scenario (2 projectors in a very lucky ten minutes) is unlikely. More likely is that the number of minutes changes but the movement cycles stay the same. So now you have one projector cleaned in seven minutes, and that is trickier to chart. Now you are trying to lift the frequencies AND also lift the record floor. So the chart will look a bit ragged. You can't fix the time unless you make it REALLY long, like the entire day, or several hours, because the task is too complex to expect that he will do 100's of projectors in a day.
So I would figure out how rapidly the projector cleaning experts can do the task, then set an aim for the entire twenty steps based upon that standard. You will be looking at both a record floor aim AND a count per minute aim, which will be a fraction of a movement cycle.
CONGRATULATIONS ON A GREAT JOB! Many people would not have the guts to risk school AV equipment and teach a useful skill.
>> Great job! I can't believe that you did this without a chart parent. You are a faster learner than I ever was when I started...
I never tried charting any thing this complex until I had years of easy to chart behavior! I agree with Ali! GREAT POST and thanks for sharing your chart!
Richard