7 Teaching Procedures to Smash ITT

Table work, ITT (intensive table teaching), DTT (discrete trial training), are all ways of talking about working at the table.


The following teaching procedures are taken from the excellent Carbone et al (2010) paper. These teaching procedures will make table sessions, and pretty much all teaching, more effective. Most importantly, they get your learner to learn because they want to, not because they have to!

1. Pairing and Manding

When beginning a table session, pairing and manding is your first priority. You should present an array of reinforcing items for ‘free’ (no requesting necessary). Then it’s your job to follow your learner’s motivation, see what they are most motivated for among the items that are presented. To be sure that the item will function as a reinforcer, you should require them to mand for it, if they are willing to ask for it, then is most likely a reinforcer. This process helps you identify effective reinforcement for your table session, which is essential to promote good responses, and also decreasing the likelihood that your child will engage in escape motivated problem behaviour.


2. Stimulus Fading

Another method to prevent escape motivated problem behaviour (crying, whining, flopping etc when demands are placed) is to fade in the amount of demands. This will be relative to your child’s VR. A Variable Ratio schedule of reinforcement has been found to be the best schedule to maintain steady rates of responding (sorry got carried away there, but it is a juicy science). A VR is basically how many demands you can place before your learner loses interest. You should start at the lower end of your learners’ VR, for example if your learner has a VR of 2, you should start with 1 demand then reinforce, and increase the amount of demands each time until you reach the higher end of the VR which would be no more than 4. If the learner’s VR is 10, stick between 5 and 20 (half below, double above). It’s not just about the amount of responses; you should also fade in the effort and difficulty of responses (don’t probe acquisition skills (skills you’re teaching) too soon).


3. Differential reinforcement

If your learner responds well (not making errors, or getting a ‘yes’ on the probe (the first time you ask them)) then you should reinforce more. You can do this in various ways, either longer duration of an activity (e.g. giving your learner longer on the iPad), more than one reinforcer (e.g. iPad, slinky, and bubbles), or a higher quantity of a reinforcer (e.g. 3 crisps instead of 1). Equivalently, if your learner responds poorly (e.g. errors frequently on mastered targets, or gets a ‘no’ on a probe) then you should deliver less reinforcement. This process is referred to as differential reinforcement. Think of it as ‘performance related pay’.


4. Errorless Teaching

Throughout a table session you should minimise errors (your learner responding incorrectly) as much as possible. Frequent errors increase the likelihood of escape-motivated behaviour.

If your child errors this is the error correction procedure you should follow: –


Re-state the SD (the demand)

Prompt response


Re-state the SD

Fade on your prompt


Distracter (between 1 and 3 previously mastered skills)


Re-state the SD

Fade again on prompt if needed/let child respond independently


Effective prompting will also help minimise errors. You should follow the prompt schedule of most to least (go in with a higher prompt and fade as needed). Your prompts should be the most effective and least intrusive you can do. Remember to prompt as much as necessary to ensure a correct response while not over prompting when not needed. Each trial you run will be different. Don’t fade if you think an error is likely.


5. Pace of Instruction

Another teaching procedure to consider when at the table is the pace of instruction. A fast pace of instruction is important as it prevents the likelihood of escape motivated problem behavior. Using short ITI’s (inter trial intervals – the time between the learners last response and your next demand) gives less opportunity for problem behavior to occur, and also provides socially mediated reinforcement for your learner quicker. It is also worth paying attention to your learner’s latency (how long it takes your learner to respond to your demand) to responding, the longest time between a demand and a response should be 2 seconds, and anything longer should be error corrected. Basically, don’t hang about, get it done well and in a timely fashion.


6. Intersperse Instructions

When teaching a target at the table, you should be using a master pile; these are skills that have previously been mastered. The mastered targets are regarded as ‘easy’ tasks, and acquisition targets are regarded as ‘difficult’ tasks. Difficult tasks have been found to be associated with a worsening set of conditions due to higher errors, more effort, and less reinforcement (basically new targets are harder). It’s important to intersperse difficult tasks within the easy tasks – 80% easy, 20% difficult to help prevent this. This (hopefully) ensures loads of success.


7. Mix and Vary Tasks

This is short and simple. Research has shown if you repeat the same task over and over again, it’s boring! This can lead to an increase in escape-motivated behaviour. So mix demands across different verbal operants; listener responding, imitation, tact, intraverbal, visual etc.


I hope this all makes sense. I’m hoping to get a video of this to post it in action.

Smash ITT.

Reflexive MO article – Carbone + Tirri

Leg it to the Table!

So, with some learners, it’s appropriate to be running ‘table sessions’. A table session is an intensive teaching period of many tasks to provide a lot of opportunities to teach targets. As a general rule, I wouldn’t run table sessions with pre school age children, as most of those programmes are based around natural environment teaching and play.


It’s important to generalise all teaching in to the natural environment, but learners (and therapists) often respond well to table sessions, as it’s very structured, and (should be) reinforcement rich.


First things first; set up the table as a fun place to be. You want your learners to run to the table, not run away. If you are working with a learner who has a history of problem behaviour related to table teaching, forget demands, just pair. Build a stronger history of reinforcement at the table than problem behaviour. Just go there for fun! Deliver all of the learners favourite things, and play with their favourite toys, and the only requirement is that they stay at the table. 


Once your learner is ready for table sessions (no longer having problem behaviour when asked to sit at the table), make sure you prepare well. Have a range of possible reinforcers ready, all necessary teaching materials (targets on acquisition – being taught, and mastered skills), and probe data sheets needed.


For most learners, a 15-minute schedule is appropriate, (15 minutes NET, 15 minutes’ table, 15 minutes NET etc.). For younger learners it may be less, and older learners (particularly secondary age and over) it would be more, but of course this should be totally tailored to the individual learner.


Version 2

Master Pile

Version 2

Tact Picture

Version 2


Version 2



















The next post is going to be on awesome teaching procedures to use at the table, and in general life!