There have been a number of tightly controlled experimental studies, both on RFT and ACT processes. There are resources within the RFT section of the site that will help you with your own studies, but there are a few that are currently posted to the ACT section of the site that you might find useful as well.

Please visit the Component Studies Information page for information on features of experimental analogs and sample files (visual basic command code, videos, instructions, etc.) from an ACT component study.

Welcome to the Ghent-Odysseus Implicit Relational Assessment Procedure (GO-IRAP)! 

• Training and Traditional Testing IRAPs
• Traditional (Label and Target) and Natural Language (Sentence) IRAP formats 
• The easy input of complex languages and characters for stimuli and instructions 
• The use of text or images as labels, targets, and response options
• Block level or trial-type level application of mastery criteria
• Label-Target-Response Option stimuli locking 
• Incorrect and Correct responding feedback options
• One easily readable .txt data and experimental file 

The 2 program files (32 bit and 64 bit versions) and the PDF of the manual are attached for all ACBS members to download. 

We are also attaching a folder named 'Training IRAP' which was recently used at a workshop presented at the Division of Behavior Analysis Conference in Dublin to demonstrate how the Training IRAP can easily be used in ABA settings to teach both basic and complex relational skills (Barnes-Holmes, McEnteggart, & Kavanagh, 2017). This folder also contains a paper entitled 'Teaching Important Relational Skills for Children with Autism Spectrum Disorder and Intellectual Disability Using Freely Available (GO-IRAP) Software'  which followed on from this workshop (Murphy & Barnes-Holmes, 2017).

Note: Bug Testing. We have been using the GO-IRAP in our research lab without encountering any problems, however, bug testing of the GO-IRAP is ongoing, so no guarantees can be made that no bugs or errors can be found within the program.

Note: Source Code. Because the GO-IRAP is still in the testing phase and there are additional features that we plan to add to the final version, we cannot make the source code available at this time, however we intend to make the source code open access in late 2020.

Also see the https://go-rft.com/ website!

An introduction to the Mixed Trial Implicit Relational Assessment Procedure (MT-IRAP)

We designed this program to hopefully provide a more flexible and sensitive tool for examining implicit relational responding. In particular, we are interested in developing/refining an implicit measure that is sensitive and reliable for measuring implicit relational responding at the level of individual stimuli and individual participants (as opposed to implicit effects at the level of stimulus categories and participant groups).

Preliminary data suggests the MT-IRAP can be used to measure implicit relational responding at both the stimulus category and individual stimulus level (Levin, Hayes & Waltz, 2010). It is unclear whether this measure is sensitive and reliable at the level of individual participants and whether it is more or less effective than the standard IRAP. Further research is needed to determine the reliability and validity of this measure, particularly with more ambiguous stimuli (e.g., where participants may be unsure of what is “truth” and what is a “lie”).

At this point, the exact configuration of the MT-IRAP is somewhat of a moving target and depends on the specific research questions being asked. Due to this, the program we have created provides a wide variety of configuration options. The flexibility in this program will hopefully be helpful to you as a researcher, though it may also require more time figuring out how to use it and the most appropriate configuration for a given study.

Please log in and download the MT-IRAP instructions (attached to the bottom of this page), this document contains a great deal more information about the procedures strengths and limitations.

Also, for further information about the MT-IRAP, we’d recommend you read the following article:
Levin, M.E., Hayes, S.C. & Waltz, T. (2010). Creating an implicit measure of cognition more suited to applied research: A test of the Mixed Trial – Implicit Relational Assessment Procedure (MT-IRAP). International Journal of Behavioral Consultation and Therapy, 6, 245-262.

Disclaimer: This program is still being developed so you may run into errors while using it. Although we have tried to make it user friendly when possible, due to limited resources and the complexity of the program, it does require some time to get used to using.

The Implicit Relational Assessment Procedures (IRAP)

This software has been used extensively to assess implicit relational responding, or responding that is outside of conscious awareness. This program was developed as an alternative to the Implicit Assessment Test (IAT) which was developed with social psychology as a way to assess implicit biases. While the IAT requires participants to categorize lists of words, the IRAP is designed to assess relational responses using contextual cues of interest. There are numerous studies that have used the IRAP, compared its results to that of (explicit) self-reported beliefs, and some that have compared it directly to the IAT. See the Empirical Support page for a detailed list of studies.

Below is a zip file containing the Ole Miss IRAP (version 3.5). The program is easy to learn and readily adapted for your own research interests. The pdf file is the current manual for installing and using the program. I recommend reading this manual before anything else (it's only six pages). In addition to installation instructions and explanations of the program features, it also contains some useful information about handling participants and accessing data.

While the Ole Miss IRAP is copyrighted, this was done to keep it freely available. I encourage researchers to use, distribute, and modify the program as they see fit. Feel free to contact me if you encounter technical problems installing or using the program, or even if you would just like to bounce a study idea off of me. I'm all about the research and will do what I can to help.

Have fun!

The lab at the University of Mississippi has made software available for the traditional matching-to-sample procedure. Below is an attachment containing a Visual Basic.NET program for a three-member, three class procedure. You will need to be logged into the ACBS website as a member in order to see and download the attachment.

PsyScope: Easy-to-Use Software for Running Computer-Controlled Experiments in Psychology

PsyScope is for the Macintosh Platform only but several FREE Apple Macintosh emulators are available for download on line. PsyScope (Cohen, MacWhinney, Flatt, & Provost, 1993) is a user-friendly freeware Macintosh application which undergoes on-going development by psychologists for psychologists. It is the easiest to use and most versatile experiment generation software available, and it’s free.

PsyScope allows researchers to design complex psychology experiments without the need for programming skills. As the user constructs spider diagrams in a graphic interface using a limited number of graphic tools, PsyScope writes a scripting file which can be accessed directly by users familiar with scripting.

An alternative to match-to-sample procedures, the Relational Completion Procedure is available below. 

The article is available in the publications section of the site, and by clicking directly here. 

The "Relational Completion Procedure.zip" file below is the more recent of the two files. Updates may be available at Simon Dymond's website as well.

Understanding and accounting for the ability to produce and understand completely novel sentences--accounting for the generativity of language--is critical to any account of language development (Malott, 2003), and to the creation of programs for teaching flexible and fully functional language repertoires.

RFT explains linguistic generativity in terms of learned contextually controlled relational responding referred to as relational framing. Typically developing children learn relational framing through natural language interactions during which they are exposed to contingencies that establish these response patterns (e.g., Lipkens, Hayes & Hayes, 1993; Luciano, Gómez & Rodríguez, 2007). However, children with autism spectrum disorders (ASD) do not easily learn this key form of responding (e.g., Rehfeldt, Dillen, Ziomek, & Kowalchuk, 2007). Nonetheless, they can benefit from training of this repertoire (e.g., Murphy & Barnes-Holmes, 2009).

The Training & Assessment of Relational Precursors & Abilities (TARPA) is a recently developed computer-based protocol for the assessment of a progression of key domains of responding critical to the development of generative language. The TARPA is comprised of ten stages as follows: (i) basic discrimination; (ii) conditional discrimination involving similarity; (iii) conditional discrimination involving non-similarity (2 comparisons); (iv) conditional discriminations involving non-similarity (3 comparisons); (v) mutually entailed relational responding [e.g., deriving the symmetrical relation B --> A from the trained relation A --> B] (2 comparisons); (vi) mutually entailed relational responding (3 comparisons); (vii) combinatorial entailed relational responding [e.g., deriving the combinatorial relations A --> C and C --> A when trained with A--> B and B--> C] (2 comparisons); (viii) transfer of function [responding to a stimulus in a new and appropriate way based on it’s participation in a derived sameness relation] (2 comparisons); (ix) combinatorial entailed relational responding (3 comparisons); (x) transfer of function (3 comparisons). Each stage is further subdivided into multiple levels, and in the stages assessing derived relations (i.e., Stages 5-10), levels are subdivided into training sections and derivation sections.

A preliminary version of the TARPA has been correlated with the Vineland Adaptive Behavior Scale (VABS; Sparrow, Cicchetti & Balla, 2005). Currently ongoing research is using the most up-to-date version to assess the emergence of relational responding with typically developing children and children with autism in order to correlate performance on this protocol with level of functioning as assessed using standardized measures of language and cognition (e.g., PLS-4; Zimmerman, Steiner & Pond, 2002) as well as to gain some insight into the hierarchical structuring and other features of the protocol to aid its further development and refinement. For access to the TARPA and the TARPA manual, please email Siri Ming at siri@siriming.com

REFERENCES

• Lipkens, R., Hayes, S.C., & Hayes, L.J. (1993). Longitudinal study of the development of derived relations in an infant. Journal of Experimental Child Psychology, 56, 201-239.

• Luciano, C., Gomez Becerra, I. & Rodriguez Valverde, M. (2007). The role of multiple- exemplar training and naming in establishing derived equivalence in an infant. Journal of the Experimental Analysis of Behavior, 87, 349-365.

• Malott, R. W. (2003). Behavior analysis and linguistic productivity. Analysis of Verbal Behavior, 19, 11-18.

• Murphy, C. & Barnes-Holmes, D. (2009). Derived more-less relational mands in children diagnosed with autism. Journal of Applied Behavior Analysis, 42, 253–268.

• Rehfeldt, R. A., Dillen, J. E., Ziomek, M. M., & Kowalchuk, R. E. (2007). Assessing relational learning deficits in perspective-taking in children with high functioning autism spectrum disorder. The Psychological Record, 57, 23-47.

• Sparrow S. S., Cicchetti D.V & Balla D. A. (2005). Vineland II Adaptive Behavior Scales. (2nd ed.) American Guidance Service, Inc., Circle Pines, MN.

• Zimmerman, I. L., Steiner, V. G., & Pond, R. E. (2002). Preschool language scale-4. San Antonio, TX: Harcourt Assessment.

Below and attached includes a listserv discussion on training relational operants. 

THIS IS A DISCUSSION FROM A POST BY DARIN CAIRNS ON THE RFT LISTSERVE (I HAVE ATTACHED IT AS A WORD DOCUMENT AS WELL IN CASE THE FORMATTING IS TOUGH TO READ HERE.)

Hi there all I’m currently putting a curriculum together for working with children with Autism and have been getting remarkable success using exemplar based relational frame procedures.
Currently we are;
1) Using exemplar training (altering certain elements to highlight distinctions etc and using familiar labels etc – eg for Dad we’ll use child’s Dad’s name first then moving to less familiar names)
2) targeting specific frames - mainly hierarchical and didactic
3) using specified entailment and transformation of functions as performance goals
4) Using mastery criteria is still hard to pin but functionality is the goal anyway so we’re using RESA(A); Retention, Endurance, Stability and Application (we view the Entailment as Adductions) – for those familiar this from the Precision Teaching guys (sans the response rate at this stage – but very soon will have that).
We are messing around with prompting strats, errorless learning preparations, different ways to present the exemplars to vary the ease of acquisition of the response and then moving to framing (not sure if this is right as over selectivity may actually be enhanced and framing hindered), also looking at how to present exemplars in different orders to make the relational aspect more apparent etc etc etc….
My questions for the list are;
a) Other than coordination which seems to have to come first; do people have any thoughts on the order frames should be taught in?
b) Has anyone studied how one frame impacts upon another frame?
c) Is anyone aware of what sort of exemplars are better for teaching relational frames
d) Has anyone studied how the acquisition of a set of frames (or just a specific one) facilitates a learners ability to move through a curriculum or other tests eg how the acquisition of condionality and causality impacts on problem solving in general)
e) Has anyone studied how to ‘pull part’ relational networks (ie to identify how a person is making reasoning errors from an RFT perspective we could identify the required frames and then assess if they are functioning for the learner/performer)
Lots of questions. Any thoughts, references, ideas would be much appreciated.
All the best (and congrats on the Time article Steve – I think it will resonate with many by being presented the way it was, and engender curiosity even from those who thought it was ‘kooky’)
D
Darin Cairns
Clinical Psychologist Supervisor
Country Autism Services
Country Services Coordination
Hi Darin -- the attached three articles are relevant to questions a, b, c,
and e (nothing on d yet). Ian Stewart, John McElwee, Yvonne and I have
been looking at broadly similar issues -as those below, but in the context
of producing relevant educational software. Ian and John ran a workshop at
last year's ABA.

D.

Darin,

As Dermot mentioned in his post, myself, John McElwee, Dermot and Yvonne
put together a protocol for training up arbitarily applicable relational
responding in learning disabled children, which we've been refining for
the last year or so. This protocol is computer-based, and, drawing on a
database of electronic pictures and sounds, it allows a teacher to lead
a child through various stages which train and test mutual and
combinatorial entailment via multiple exemplars.

John and myself have been involved in workshops on Early Intensive
Behavioral Intervention and RFT the previous few ABAs and this summer
we're giving a similar one with the same theme - demonstrating to EIBI
people the effectiveness of RFT-type interventions for training language
generativity in learning disabled children. This summers workshop we'll
be discussing the extension of the original computer protocol via the
creation of a protocol that trains and tests relational pre-cursors from
simple dicriminations, through non arbitrary relational conditional
discriminations, through arbitrary relational conditional
discriminations. We'll also be discussing how this work relates to the
ABLA, and the ABLLS early learning protocols, and more specifically, how
RFT predicts the type of patterns that are seen in terms of the
correlations between ABLA and stimulus equivalence and language
performance. We are currently writing a paper on the latter and applying
for funding to test specific RFT-based predictions with regard to the
difficulty of various stages of the ABLA.

Your questions are very important ones. In our RFT-based analysis of the
ABLSS curriculum we are stressing the importance of the order of
training particular frames and how particular types of frames are indeed
necessary precursors of others. For example, flexible contextual control
over same/different responding may be an important precursor of
hierarchy. If you or others are interested I can send you past ABA
workshop .ppt presentations on EIBI and RFT.

Best,

Ian.

Hey Darin;

I think the work you are doing is important. In terms of the
sequence of which frames should be taught first, I am not sure the
data is there to guide us on this issue. Thus, put on your
clinicians hat and monkey around. Choose a frame that will have the
greatest impact for the child in his or her environment. In other
words, what frames will the child's environment immediately support
and what frames, once established will be of most use for that child
to more effective participate and operate in his/her environemnt.
This is a logical exercise. But, I suspect that a) the logic will
steer you right in many cases & b) you will hit walls that will
require troubleshooting. These successes and temporary walls are
perhaps the most important bits of information for the field. Keep
us posted so the basic minded folks can help work out the issues
that will assuredly crop up.

Question: Do have non-arbitrary tasks built into your training
sequences? Whether these are absolutely necessary pre-requisites is
still uncertain. Several studies they have definitely shown to
facilitative or the development of arbitrarily applicable derived
relational responding. But I suspect when working with language
delayed indivduals non-arbitrary tasks will be particularily useful
in making the targeted relation more salient. There are a lot of
good old studies (50'-70's) showing how MET can facilitate what's
been labeled transposition. Which is, crudely put, responding in
terms of a relation and not an absolute value.

TONY;
I agree the mistakes may be very enlightening. Again, I think the
early transposition literature sheds light on the issue of these
mistakes and implicitly (not explicity because authors often
relegate these mistakes and sucesses to the realm of schemas or
language as mediator) demonstrates ways to program training
materials such as to promote derived relational responding.

I agree that errorless procedures are often times great remedies. I
found it VERY difficult, during my pilot work for my thesis that
when dealing with derived relational responding, it is very hard to
get kids ("typically developing") to "get" the target relation and
not some other absolute property of the stimulus conditions.

FINAL NOTE (I half-heartedly promise): Would it be alright if I
posted these contributions to this thread on the RFT list to the
education section of contextualpsychology.com

Hi Nick hows it going?
I think it would be fine to put this on the education list. The more ideas and insight the better.
Choose a frame that has greatest impact for the child;
1) To me Baer’s behavioural cusp concept applies here. But it’s interesting to consider how one identifies a required frame. We could do an example of cusp analysis with frames.
2) Cusps also apply to ‘bang for your buck’ in terms of what the curriculum sequence will benefit from the most. Ie the PT guys look at tool skills to promote generative learning. I wonder what the ‘tool’ frames are in a given curriculum context?
Do we have non arbitrary tasks built in to help language delayed individuals target relation more salient?
1) We’re messing around with it but this is definitely something that interests me. If I’m interpreting you correctly, what I can feedback at this point is that the non-arbitrary tasks have proven the fastest road to relational framing for the learner. We kinda think of it as a sort of ‘within stimulus’ prompt as the answer lay in explicit previous learning and then we lead them to more purely relational situations where they must entail. When we don’t prep this way the children become anxious and start guessing.
2) Can you throw me some refs? Interested in the MTS examples you are referring to.
3) The issue of protecting the learner from over-selectivity pops up at this point too as the non arb stuff can play right into the hands of non relational behaviour being reinforced at the expense of relational…
Mistakes…
1) Have been fascinating to watch and have directed our approach with the learner quite a lot.
2) I wonder if IRAP procedures can be used here to give us insights into the relevant networks and construct the instructional design from there…?
3) What errorless learning strats (ie prompts) did you try in your thesis Nick? It’s a real struggle here too. Non arb being the only real strat we have (I’m wary of visuals). We have tried to conmstruct a hierarchy based on how previous learning will support the correct answer (ie non arb), not impact directly on framing (ie we use hypothetical examples), to impacting against framing relationally (Ie use examples where the entailment is logically true but experientially wrong – ie the child has to ‘trust’ the entailment not his explicit learning history)
Here’s a question. If we forget the order issue what is the process (I have avoided the word mechanism ) by which a frame effects another frame? What is necessary for a frame to ‘bump’ into another frame and then change it? This will go a long way to helping us understand generative learning and adductions I’d think. Real bang for your buck instructional designing.
D
Darin Cairns
Clinical Psychologist Supervisor
Country Services Coordination

The cusp concept looks like another way of saying “increases psychological flexibility” (“behavioral cusp” is in a frame of coordination with “increases psychological flexibility”?)
Re: “Here’s a question. If we forget the order issue what is the process (I have avoided the word mechanism ) by which a frame effects another frame? What is necessary for a frame to ‘bump’ into another frame and then change it?”
Along the lines of a comment in Ian’s earlier post, it seems to me that coordination and difference are essential for hierarchy. In broader RFT/ACT terms, once coordination and difference are established – which respectively have the most precision and the broadest scope, hierarchical relations serve to attenuate/balance precision and scope in context - and thus to increase behavioral flexibility. In other words, frames don’t bump into other frames unless/until we relate relations….relate multiple exemplars qua correspondence AND difference and you get hierarchy….
Once you get correspondence, difference, hierarchy, only then are more complex relations possible - deictic, conditional, and beyond…
Seems right
One thing
It seems to me that as soon as you have multiple frames you can
leverage training by demanding several simultaneously.
In other words, you can make them "bump into" each other
usefully
For example
Once a kid has coordination and difference, you can
a) use exclusion to train new material so that it simultaneously
strengthens frames of cooridination and differnce.
For example, saying "where's the gug-gub" in the presence of a
novel object and a ball (that is in a frame of coordination with "ball")
is a good way to have another "ball trial" AND distinguish balls and gub-gubs
AND train a frame of coordination between the novel object and "gub-gub."
b) similarly you can train using deliberate errors. Once the kid relaly knows Mommy, point to
Ma-Ma and say "there's Daddy" and pause. The kid will laugh or otherwise
indicate the error. "That right, That's not Daddy ... that's Mommy!"
You simultaneously apply a frame of coordination twice and a frame of distinction
c) similarly you can early on play with flexible contextual control.
(If Ma Ma was Da Da, where's Ma Ma?). This is cool because it simultaneously
strenghtens two instances of a FoC and a FoD, plus it makes the Crel control
more arbitrary
To do these kinds of things you need multiple frames ... and I think even in
the earliest training of FoC, it should be done along of training others frames.
it makes the relational feature more likely to be the invariant. For example,
why just train "Da Da" when you can also train "not Da Da."
Lots of discrete trail programs seem to ignore the benefits this can provide
- S
Steven C. Hayes

Right on! And... In defense of training just "da da." I think
there are two issues.
1) Thinking that way has not hit the main pipeline just yet. It is
out of the box. I do not believe that training "not da da" is in
most peoples realm of importance let alone possibility.
2) (this one may take me a while). Discrete trials are most commonly
used in the world ABA with autism. Initially, it is often difficult
to get any kind of vocal responding under stimulus control. With
some kids it is a real pain in the @** struggle.
As such, I think that a large part of the field has been or is
quite impressed by the POWER of reinforcement to train just about
behavior (yes yes not sophisticated language behavior like I am
displaying right here before your very eyes) that we just plug on to
see what else we can bring to strength. This is done, without
looking at the broader desired repertiore. Sometimes I feel that we
are engaging in a modern form of S-R psychology. Call it R-S
psychology. By the way this statement is not meant to indeminfy any
individual/s. As I have the utmost respect for the work I just
described.

Nick B