An updated print version of the Low Performer's Manual, now named Strategies for Teaching Students with Severe and Low-Incidence Disabilities, is available for purchase through NIFDI. Click here to go to NIFDI Store.

Prologue to
Low Performers' Manual

Over the years, a lot of people have asked me, "What do you do if you have learners that are too low in language for the beginning language program Language for Learning?"

The answer is bittersweet. Yes, there are tight, specific and effective procedures for teaching very low performers. We wrote up those procedures back in the 1970s and called the work the Low Performers' Manual. However, the chances of someone being able to execute the techniques without training is slim. Even if you're a pretty good teacher, you'll probably need a lot of demonstrations and a lot of practice before you can run the routines effectively. Exquisite timing is essential. So are precise, effective corrections. And possibly the biggest ingredient is responding appropriately to the responses the learner makes. This game is not like following a script and providing corrections when the learners produce a wrong answer or have a weak response to a task that has one correct response. The game is to understand your options for reaching a goal and be prepared to respond to whatever the learner does—either with respect to behavior or to the correctness of the answer.

You have to make up the example sets, following very tight specifications of how to do it, and you have to be able to bring the learner to mastery on everything you present. Mastery definitely will not occur the first time you teach something new or often the fifth or eighth time. But the central rule about mastery is that if you keep bringing the student to mastery, the student will learn related material faster and faster. The reason is that your initial work established a foundation, which means that far less learning is needed for the student to master anything that rests on that foundation.

The Low Performers' Manual has never been submitted for publication. The commercial programs we develop will work for the ordinary teacher who receives appropriate training. The Low Performers' Manual is not like commercial programs because the directions for teaching the material can't be packaged the way they can for a program like Language for Learning.

The overriding rule about low performers is that the lower the performance of the learner, the more you have to know about the subtle details of teaching. The very low performer will do lots of things that are basically "unexpected" by most teachers. You have to know how to respond to all of the behaviors, bring the learner back to the task, and teach the learner to mastery.

That does not mean aimless repetition of items or rituals that are not designed to induce mastery. Nor does it mean assembling lists of items from IEPs or standards and spending the school year trying vainly to teach them. The learner will learn if the teacher teaches well. But you have to know how to diagnose an incoming learner, not with tests that document what you would observe in a few moments, but with tasks that reveal both what the learner doesn't know and what kind of behavioral games the learner plays. What's his best trick to control people who work with him? How much is that trick going to interfere with teaching him? How much work would you estimate will be needed before the learner lets go of his tricks and plays your games? Answer these questions and you'll know how you'll start working on the learner's behavior.

In the same way, the diagnosis of knowledge is relatively quick and simple. But the most important rule, and possibly the most difficult one to teach teachers, is that you have to start as close as possible to where the learner performs, and you have to teach to mastery. You can't achieve mastery if you introduce tasks that are far beyond the learner's ability.

The worst teaching I have ever seen is with low performing children. I've seen teachers prompt the answers by moving their lips or by always saying the answers with the learners. And many teachers have a strong tendency to see what they want to see. In the process, they become completely fooled about both what the learner is learning and what the learner is capable of doing.

Example: Strategies of some echolalic learners

In most cases the learner who responds by saying what the teacher just said doesn't understand yes-no, but the teacher doesn't know it. The teacher repeatedly presents yes-no questions and the learner answers them by repeating the last part of the question the teacher just said. The teacher does not identify this strategy and thinks the learner is answering her questions.

"Would you like a chocolate cookie?"
"Chocolate cookie."
"Of course. Here you are."

Even if the teacher says something that contains a relative pronoun, the teacher thinks that the learner is simply a little mixed up when he responds.

"Would you like to sit on my lap?"
"On my lap."
"No, honey. I'm too big to sit on your lap. You come here and sit on my lap."

"Would you like to walk with me?"
"Walk with me."
"Yes, we'll walk together, won't we?"

In most cases the learner who responds by saying what the teacher just said doesn't understand yes-no, but the teacher doesn't know it. The teacher repeatedly presents yes-no questions and the learner answers them by repeating the last part of the question the teacher just said. The teacher does not identify this strategy and thinks the learner is answering her questions.

I've seen aides who can get learners to respond to their routines when they are in this cubbyhole. But when we take the learner and aide anywhere else, the learner doesn't respond to the same routine. Nor does the learner respond to anyone else who tries to work with him in the cubbyhole. Why? Because the teacher did not teach these generalizations. From the learner’s standpoint, there is one person and one place for these behaviors.

Just as you can expect high performers to generalize what they learn to different people and different settings, you can’t expect low performers to respond to other people or other settings unless the instruction is designed to obviate this problem by systematically having different people run routines as soon as the aide establishes them and to regularly run these routines in different places (lunchroom, playground etc.).

Another technique that generates incredibly inappropriate behavior is “shaping” responses. . For example, the learner produces no identifiable words. However, the learner produces many verbalizations that begin with m and f, "Mumufufee fumumafufoo." You're a behaviorist, so you believe that you'll start with words that begin with m and f and shape the learner's repertoire by reinforcing closer "approximations" of saying the words correctly. You reinforce with treats. You assume that the learner is trying to communicate by naming things. So if the object is mustard and the child said something that begins with "mu", you would reinforce the child. If the object is fence, and the child said something that begins with "fe", you would reinforce the child. If the child later said "femu", you could reinforce the learner more lavishly because this response is a closer approximation of fence.

You work with the learner possibly 400 trials. The net result of the program is that the child produces “reinforceable responses” at a much higher rate. What the learner has learned, however, is not that these utterances refer to specific things or features, but that babbling at a high rate leads to a higher rate of reinforcement. If you present a flag and tell the child the name, the child does not tend to produce a higher percentage of responses that begin with f. The child just produces more verbalizations of the random variety. In other words, you have now reinforced the learner to say the same verbalizations he always did, except at a higher rate.

To teach this child anything about language or speech, you have to begin with a sound that is not part of his superstitious rituals—any sound. If you present a ball and teach any kind of stable response that does not have an m sound or an f sound such as "ubu", you've got something on which you can build. But without a differential response that would not be produced in any situation, you've got nothing.


What you do, how you test the learner, what kind of standards you use for measuring performance, and the general formats that you follow are spelled out in detail in the Low Performers' Manual. These details are based on extensive work with many learners who were labeled, "unteachable", "profoundly retarded", "language delayed", "autistic", and so forth. I worked with Jerry Silbert, Peter Lormier, and Jeff Sherman on teaching hundreds of seriously autistic children. I worked primarily with Linda Olin with seriously brain-damaged teens and adults.

In Eugene, Oregon, we ran what had to be the most awesome preschool for seriously low learners that probably existed anywhere ever. Gary Davis ran it. The teachers were 10s and Gary was a 10 at training and supervising them. We worked primarily with the seriously "developmentally-delayed" preschoolers from the Eugene school district. Some learners could not follow any basic directions. For these learners, you could present a ball and say, "Touch this." They might not perform correctly after 100 trials and 100 corrections.

We accelerated the performance of these learners many times the rate that they would have performed if we had used traditional techniques. A high percentage of them who entered unable to follow basically any verbal directions would go through the low performer sequence and then complete DISTAR Language 1 within two years. A fair percentage of them were integrated into regular classes. Unfortunately, a lot of them were betrayed in school because they went into traditional special-ed classes where there was no attempt to systematically build on what they had been taught.

In the preschool Deborah Loschiavo, Kim Weiherman, or Gary Davis would evaluate and place a new learner in possibly 15 minutes. They would start by giving the learner a task of low difficulty. To test the learner's best behavioral tricks they would behave in different ways, first with poor pacing and the demeanor of a doting parent or teacher, then briskly as they repeated the same material. As part of what they did, they would "punish" a misbehavior, with voice or possibly physical prompting. The responses of the learner would reveal diagnostic information about how much the child's misbehavior will escalate when it is put on extinction. These procedures are spelled out in the Low Performers' Manual.

Understand that the more a child has relied on particular behaviors that control others, the more the misbehavior will escalate when it is put on extinction (is no longer rewarded or permitted). This fact is very important for changing those inappropriate behavior patterns that preempt the learner from learning. When typical teachers attempt to be strict and not allow misbehavior, the behavior will escalate and most teachers back off, thinking that what they are doing is not working. By backing off, however, they are simply strengthening the behavior they're trying to eliminate. They are providing the learner with further evidence that the favored trick works in controlling people.

I'm not blaming teachers for not knowing effective practices for working with very low performers. They were never taught these practices, and what they were taught is inadequate, greatly misleading, and often presents the opposite of what needs to be done to be effective. Some of the worst teaching I've ever seen was conducted by behaviorists who believed they were following behavioral principles.

Although the mistakes that people make in working with low performers are not their fault, the fact is that most teachers of low performers do many things wrong. For example, they use words after the fact, not as predictors of behavior, and the result is that some learners never learn how language functions. For instance, the teacher guides the learner's hand to the object and then says, "Touch, touch, touch." Wrong. The directive comes first. The demonstration follows immediately, making it clear that the verbalization is supposed to predict the behavior.

In summary, it can be done, but only by the best teachers. It requires a lot of thinking, learning, and practice. Sometimes, it requires letting go of most of what you think you know about working with seriously low performers. So don't expect it to be easy.


The Low Performers' Manual describes how to:

1. teach learners all the basic skills that they need to go into Language for Learning;
2. teach them how to follow directions;
3. bring unacceptable behavior under instructional control.

Warning: Some of the techniques described in the Low Performers' Manual would not be permitted in some districts. So make sure you get approval before you institute any of them. A good idea is to either be monitored or taped, particularly when you work with out-of-control, seriously noncompliant learners.

The rules in Eugene, Oregon about "touching" a learner or the like are so absurd that I will not work with autistic learners or learners who are out of control, even though most of the ones up to age 7 or 8 could be turned around the fast way. Legislators and district administrators don't understand much about punishment and how it is different from punition. Punishment works in changing behavior, and works fast because it provides the learner with very clear information about the consequences of behavior. But it must be done properly, infrequently, and in a way that provides the learner with information about what leads to positive reinforcement. In other words, punishment is simply a tool used to get the learner under control so that the learner will receive high rates of positive reinforcement.

The manual addresses some of these issues, but be very clear, I am not warranting that you will be able to use all the techniques. Check it out before you do something that will get you in trouble.

You should be able to use all the cognitive stuff. Also, you should be able to use the key paradigms—the hard-task paradigm, the four-level memory paradigm, the procedures that address content that is taught, like yes-no, prepositions, interactions that involve relative pronouns, etc.

Following the Program

The routines specify what you must vary from trial to trial and how you must vary it. There's a reason for every detail in the routines, but the reasons may not be obvious. For example, a teacher who has received no training presents a simple task like pointing to a piece of paper on the table and saying, "Touch." The learner doesn't do it but is compliant. So the teacher presents repeated trials, models the behavior of touching, and assists the learner in touching it until the child does it without physical assistance. The teacher says, "Touch," and the learner touches the paper. This takes part of 3 sessions, each with about 40 trials. For the next session, the teacher puts the paper on the table, says, "Touch," and the learner does nothing. The teacher assists the learner in touching. Then she returns his hand and says, "Touch." The learner touches. The teacher reinforces the learner and presents 15 more trials, and the learner performs on every one.

At this point, the teacher thinks she has taught something, but she's actually taught something else because she created practice that's ambiguous. Here's how she could test what she actually taught the learner. Present a couple of more trials. Then move the paper, point to it and say, "Touch." The odds are about at least 2 to 1 that the child will touch the same place he had been touching. The touch was not referenced to the paper. The learner was just doing the same motor response she had done on the previous trials—touching the same spot.

What you need to do to teach touching is to move the paper to a new position after each trial. You need to do this from the beginning, and you need to change the objects so the child learns a generalized rule about touching. Next you start working at discriminated touching by teaching "Touch paper." You put other objects in the set that serve as distracters. After some trials, you move one of the distracters, not the paper. Each task you present, however, is "Touch paper."

You need good pacing and pauses in the right places. And you need to follow the rule that if the learner touches anything before you complete the direction, "Touch paper," it's wrong, and you correct it by varying the pacing.

Once you've established this discriminated touching, you have an important window for reliably observing what the learner can and can't do. Now you can teach other objects, such as "Touch pencil." The learner's performance gives you ongoing information of the misconceptions the learner has, which implies exactly what you have to do to correct each misconception. You're teaching receptive concepts, so you don't have to rely on the learner's verbal performance to discover what the learner knows about concepts like over, in front of, red, or whatever you choose to test. Just present a group of objects and direct the learner to "Touch the ball," or "Touch the red circle."

The Future

One of the reasons I'm putting The Low Performers' Manual online is that I think it is important for at least some practitioners out there to have the skills and knowledge needed to teach very low performers effectively. I think the manual is an important part of the technology of effective teaching. I also think it's important for those kids who are currently going nowhere and who have very unhappy lives because they are not being taught properly. Certainly, you're not going to turn most of them into "normal children", but you'll sure make life a lot easier for moms and a lot more rewarding for the kids.

So I'll make a deal with you. If you're serious and want to become a super star with very low performers, and if you have permission from your district to use the program, I'll give you explicit directions about what to do when you get hung up.

You have to provide me with detailed information about the exact nature of the problem. Record yourself so it clearly shows what you are doing and what the learner is doing. Send it to me, and I'll critique it and tell you exactly what's wrong and what you have to do to correct it.

There's only a handful of people who are really good at this game of teaching very low performers, and I would hate to see all knowledge of it die when they quit. (Most of them are no longer working with super low performers or difficult preschoolers.) So if you're serious and already know how to teach DI programs well, I'll be glad to work with you.

(541) 485-1163



Featured Video

Kindergarteners Showing Off Their Math Skills 1966 Uncut demonstration of at-risk children who were taught math by Zig Engelmann as four year olds and five year olds. The session was filmed in front of a class of college students in August with no rehearsal. Children work addition, subtraction, multiplication, division problems, basic algebra problems, fraction problems, area problems, factoring, and simple simultaneous equations.

Watercolors by Zig

Oregon coast, south of Coos Bay

© 2012 Siegfried Engelmann. All rights reserved.

website statistics