Introduction to ArtificiaI Intelligence in Higher Education
Smart brain training in education
1. Dr. Sarah Cassidy
B.S. (Hons) Psychology, PhD Psych Science
Cert. CBT
Professional Member APA, ACAMH (Ireland), NEPS (SCPA Panel)
Educational Psychologist
Behaviour Therapist
Lecturer in Educational Psychology and Child Development
www.raiseyouriq.com
3. What is SMART?What is SMART?
SMART is a scientifically developed
educational brain training programme
that has been proven in published
scientific research to raise intelligence
levels (measured using standardised IQ
tests) by 20 - 30 IQ points. That makes
SMART training a breakthrough in
psychological science.
4. What is SMART?What is SMART?
SMART is based on the new scientific
discovery that "relational skills" underlie
most forms of intelligence. Our
computer based online brain training
programme helps people to enhance their
relational skills and their intellectual skills
go up as a result.
Let’s see if you have got good relational
skills....
6. Some background information....Some background information....
What is Intelligence?What is Intelligence?
• “The ability to learn, understand and make judgments or have
opinions that are based on reason”
-Cambridge Advance Learner’s Dictionary (2006)
• “Ability to adapt oneself adequately to relatively new situations in
life.” -R. Pinter, quoted in Sterberg (2000)
• “Intelligence measures an agent’s ability to achieve goals in a wide
range of environments.” -S. Legg and M. Hutter (2006)
• “Intelligence is a general factor that runs through all types of
performance.” -A. Jensen
7. What is Intelligence??What is Intelligence??
“Viewed narrowly, there seem to be
almost as many definitions of intelligence
as there were experts asked to define it.”
--RJ Sternberg
In a nutshell, intelligence is a summary of
a set of skills that allows us to interact
successfully with our environments.
8. What is IQ?What is IQ?
An Intelligence Quotient is the measure
of your performance on an IQ test
RELATIVE to people of your own age and
sex.
11. Problems with intelligence testsProblems with intelligence tests
While the WISC-IV
does measure g, it
fails to take other
possible intelligences
into account
14. More recent evidenceMore recent evidence
IQ scores are often stable across time
◦ i.e., if two tests are taken a few months /
years apart the results are likely to be quite
similar particularly for older children and
adolescents.
Test-retest correlations typically range
from .70 - .85
We also know that IQ scores are good
predictors of academic success, career
success, and even health and happiness.
15. Some factors that that are linked toSome factors that that are linked to
increases in IQincreases in IQ
Quality parent-child interactions:
◦ Emotionally responsive
◦ Language-rich
◦ Complex, stimulating environment
◦ High expectations
◦ Operate in children's ZPD
Early intervention e.g. Head Start, Carolina Abecedarian
Project
16. Education: “A year in education most likely increases IQ by
somewhere between 2 and 4 points” (Winship &
Korenman, 1997)
Nutrition (e.g., Lynn)
Enriched environments -- Brain plasticity (e.g.
Rosenzweig ). Studies have shown that intellectual
stimulation in the form of “cognitive training” can help
to slow down cognitive decline in the elderly and
among those with dementia and Alzheimer’s disease
(Belleville, Gilbert, Fontaine, Gagnon, Menard &
Gauthier, 2006; de la Fuente-Fernandez, 2006; Spector,
et al., 2003; Willis et al., 2006; Wilson, et al., 2002).
17. Brain Plasticity
The big challenge is to show that such
interventions do more than simply improve
people’s skills at the very tasks at which they
practice. What is needed is evidence that
training at one task type generalizes at least to
real-life situations (e.g., involving problem-
solving or remembering) or to other aspects
of cognitive functioning.
19. A behavioral approach assumesA behavioral approach assumes
intelligence to be a malleable skill setintelligence to be a malleable skill set
(Cassidy, Roche, & O’Hora, 2010).(Cassidy, Roche, & O’Hora, 2010).
20. Relational Frame Theory and human
intelligence
A modern behavioral research programe has been
designed specifically to understand how the skills
associated with high intelligence might be
established in educational interventions.
In a nutshell, RFT is a functional-analytic
behavioural theory of human language and
cognition.
RFT claims that the foundational skill for most
intellectual abilities is Derived Relational
Responding.
21. RFT and human intelligence
What is Derived Relational Responding??
DRR is the skill of relating objects to each
other in accordance with a small family of
mathematical relationships (e.g. symmetry,
equivalence, opposition, more-than, less-
than).
Equivalence is the most basic of the relations
that can be trained.
23. Multiple Exemplar Training
Relations are taught by caregivers across
multiple exemplars in normal language
interaction (whether the caregiver
realizes it or not).
Equivalent
24. Multiple Exemplar Training
By giving a child multiple exemplars, a
caregiver is teaching a child skills that become
more and more abstracted over time.
From RFT perspective - all language for
humans involves being able to use words in
this abstracted way - that is use and respond
to words whose meaning has been derived
through equivalence.
So according to RFT - language and deriving
relations involve the same process.
25. What do we know so far??What do we know so far??
Relational skills are foundational to
language.
Language and intelligence are highly
correlated.
We can train language through MET.
26. Empirical evidence
Relational skills are foundational to language
(Lipkens, Hayes and Hayes, 1993, Devany, Hayes
& Nelson, 1986; Barnes, McCullagh & Keenan,
1990).
There is a correlation between WAIS subtests
and a complex relational skill (O’Hora, Pelaez &
Barnes-Holmes; 2005).
27. Relational skills are foundational to language
O’Toole and Barnes-Holmes (2009) found that
fluency at a complex relational task correlated with
IQ as measured by the Kaufman Brief Intelligence
Test
Gore, Barnes-Holmes and Murphy (2010) found
significant correlations between performance on a
test for deictic relations (i.e., perspective taking
relations involving “I” and “You”, and “here” and
“there” relations) and Full Scale (r = .43), Verbal (r
= .45) and Performance IQ (r = .45; p.12) as
measured by the Wechsler Abbreviated Scale of
Intelligence (WASI; Psychological Corporation,
1999).
28. Related Experimental Work
Barnes-Holmes, Barnes-Holmes, Smeets, Strand
& Friman (2004).
Barnes-Holmes, Barnes-Holmes & Smeets
(2004).
Berens & Hayes (2006).
Gomez, Lopez, Martin, Barnes-Holmes &
Barnes-Holmes (2007).
These studies all showed that MET was an effective means of
developing relational responding skills and these relational responding
skills then successfully generalized to novel stimuli.
29. Implications……
If there is a connection between
language, intelligence and relational skills
(as all of the foregoing research has
shown that there is) then improving
relational skills should improve
intelligence (as measured by commonly
used IQ tests such as WISC, WAIS).
Barring this research body, no other
study to date has attempted to use MET
to improve general cognitive abilities.
30. Cassidy, Roche & Hayes, (2011) researchCassidy, Roche & Hayes, (2011) research
Experiment 1Experiment 1
Found mild and modest correlations between almost
all subtests of WISC-IIIUK
and speed of deriving
symmetrical and transitive relations.
Found that providing intensive relational skillsFound that providing intensive relational skills
training to normally developing children raises theirtraining to normally developing children raises their
IQ by more than one standard deviation and in someIQ by more than one standard deviation and in some
cases by as much as 2 standard deviations.cases by as much as 2 standard deviations.
Also found increases of as much as 30 points in full
scale IQ scores in a small sample of normally
functioning children who participated in a MET
intervention designed to raise IQ score.
31. Cassidy, Roche & Hayes (2011)
Experiment 2
8 children presenting with learning difficulties were
administered a comprehensive psychometric
assessment (IQ and attainment test) and a baseline
test of relational abilities.
All children then took part in an intensive relational
training intervention programme over the course of
an academic year.
All children showed increased relational abilities and
increased performance on an IQ test following the
research programme. Mean IQ rise was 13 points.
Many children moved from the categorisation of
“LD” to the categorisation of “normal range”.
32. Some demographicsSome demographics
Specific learning difficulty in reading
Mild General Learning Disability, borderline general
learning disabilities
Mild to moderate delay in comprehension and expression
of language, mild to moderate delay in vocabulary
development,
Specific learning difficulties in reading, writing and spelling
ADHD
ADHD, Specific learning difficulty in reading and
mathematics
Dyspraxia
33. Involvement with other professionalsInvolvement with other professionals
Psychiatrist
Clinical Psychologist
Educational Psychologist
Speech and Language Therapist
Occupational Therapist
Learning Support Teacher
Resource Teacher
SNA
34.
35. •Children were re-assessed 4 years later andChildren were re-assessed 4 years later and
it was found that no child returned to theirit was found that no child returned to their
previously lower than average IQ score.previously lower than average IQ score.
All IQ rises were largely maintainedAll IQ rises were largely maintained
and in many instances, the IQ scoresand in many instances, the IQ scores
(particularly verbal and reasoning(particularly verbal and reasoning
skills) continued to rise.skills) continued to rise.
36. Read this again pleaseRead this again please
IQ RISES WERE
MAINTAINED FOUR
YEARS LATER!
37. Cassidy et al. (2011, Experiment 2)...
Baseline Post-Intervention Four-year Follow-up
Mean SD Range Mean SD Range Mean SD Range
Full Scale IQ 82.9 8.3 70-92 95.9 10.6 76-111 102.57 12.19 91-126
Verbal
Comprehension 82.3 7.3 73-93 92.4 9.2 83-110 100.86 6.91 90-112
Perceptual
Reasoning 82.1 10.3 65-96 94.5 6.7 84-106 103.14 16.12 83-123
Working
Memory 94.9 16.6 59-116 97.5 12.3 77-116 88.14 14.78 65-103
Processing Speed 91.0 9.8 83-109 107 15.6 78-121 98.86 10.33 86-114
✔
✔
✔
✖
See Roche, Cassidy & Stewart, 2013 for more on four year follow-up
✖
38. Recent Study in a SchoolRecent Study in a School
Replication of Cassidy et al (2011), using anReplication of Cassidy et al (2011), using an
online relational skills training intervention asonline relational skills training intervention as
offered at RaiseYourIQ.comoffered at RaiseYourIQ.com
Fifteen mainstream school children (normal IQs)Fifteen mainstream school children (normal IQs)
Administered the WISC-IV UK, standardAdministered the WISC-IV UK, standard
scholastic tests, and a specially designedscholastic tests, and a specially designed
Relational Abilities IndexRelational Abilities Index (RAI).(RAI).
Twice weekly training sessions in Same,Twice weekly training sessions in Same,
Opposite, More-than/Less-than relationalOpposite, More-than/Less-than relational
responding.responding.
Standardised tests re-administered followingStandardised tests re-administered following
trainingtraining
39. Recent StudyRecent Study
55 training and testing stages55 training and testing stages
100% accuracy and high speed100% accuracy and high speed
required on each block of training andrequired on each block of training and
testingtesting
16-20 weeks to complete16-20 weeks to complete
All stimuli were nonsense syllablesAll stimuli were nonsense syllables
Every training and testing trial wasEvery training and testing trial was
unique – only the FORMAT of derivedunique – only the FORMAT of derived
relational responding was taughtrelational responding was taught
40. 40
Sample trial from one level ofSample trial from one level of
training….training….
Relational
Statements
Relational Question
30s to Respond
41.
42.
43. Q: Which of these two coins is bigger?
The size relation between these coins is non-arbitrarynon-arbitrary
A: The 5c.
44. Which of these two coins is more valuable?
“is greater than”
In the context of value the relation
between the coins is ARBITRARY.
Kids must learn the ABSTRACT nature of “More than”
45. RAI ScoresRAI Scores
Increase in relational abilities was significant (t=7.235, df =14, p<0.000).Increase in relational abilities was significant (t=7.235, df =14, p<0.000).
Effect size (Cohen’sEffect size (Cohen’s dd) = 1.89 (very large).) = 1.89 (very large).
46. 46
Full Scale IQ ScoresFull Scale IQ Scores
The increase in IQ from baseline to follow up was significant (t=19.18, df =14, p<0.000).
Effect size = 4.96 (Cohen’s d, very large)
48. IQ Rises by Diagnostic CategoryIQ Rises by Diagnostic Category
49. Sigma T Sten ScoresSigma T Sten Scores
Sigma TSigma T test measures standard mathematical ability in Irishtest measures standard mathematical ability in Irish
SchoolsSchools.
49
Increase in Mathematical ability was significant (t=1.87, df=14, p<0.05, one-tailed)
The Effect Size (Cohen’s d) was 0.5 (medium).
50. Micra T Sten ScoresMicra T Sten Scores
Micra TMicra T test measures standard reading ability in Irish Schools.test measures standard reading ability in Irish Schools.
50
Increase in Reading ability was not significant (t=1.45, df=14, p>0.05)
The Effect Size (Cohen’s d) was 0.37 (medium).
52. If a woman were to marry her own
adopted Son…
What relation would she be to herself?
53. School Principal has noted....School Principal has noted....
“The kids enjoyed the training. They were
challenged by the varying sets of tasks and
seem to have benefited from the programme in
terms of their general intelligence. There were
marked improvements in IQ scores, but also in
terms of their ability to concentrate in school, in
their reading and in their mathematics. We
would definitely be interested in seeing all kids
at the school benefit from this product.”
54. Resource Teacher has noted....Resource Teacher has noted....
“Having worked with some of the kids who have
completed SMART, the improvements I have
observed as a result of using this programme
include increased confidence, increased ability to
problem-solve, and greater speed and accuracy in
completing tasks. I believe the repeated practice
of a set of skills in a staged manner enables
a student to gain these improvements. It is a type
of 'brain training' that we, both adults and children,
all can avail of to improve our cognitive abilities.”
55. Class Teacher has noted....Class Teacher has noted....
“I have been working with many of these
kids since they were in Junior Infants. I
have seen the results of this programme
in the last month and it is powerful.
Not only have the children's IQ's risen,
but many of them have improved
significantly in their reading, spelling and
mathematics.”
56. What kids have said about SMART...What kids have said about SMART...
P1/11 year old boy. I really liked the SMART training
programme. It has helped me in school because I
notice that it is easier to answer questions in maths.
I’m also quicker to understand new stuff we haven’t
learned before.
P2/12 year old boy. Maths are easier for me now
and I am quicker at figuring things out in all my
subjects. I feel like I am faster at everything in school
now. I have definitely benefitted from this IQ
programme. My reading and writing have improved
too.
57. What parents have said about SMARTWhat parents have said about SMART
I have noticed that my child’s problem solving skills are stronger since
using the SMART programme. He seems to be using his head more.
He is thinking more practically and approaching things more sensibly.
For example, he is using concrete strategies to figure out why the SKY
box wasn’t working. He thought of things that had never even crossed
my mind. He also quickly rearranged some furniture in the office for
me so that I could plug my phone charger in and use it at my desk
while I worked. It was a silly thing that I should have probably thought
of myself, but I didn’t and he did! He was not that type of child before.
In fact, I would have described him as “flighty”. That’s why these things
surprised me so much. He is also more settled into his work and he
seems to be taking it more seriously now. His concentration and
attitude toward his homework have improved. He also really enjoyed
working at the programme so that was a relief to me because I really
wanted him to do it.
58. Why Should This Matter toWhy Should This Matter to
You??You??
SMART training helps people to....
Reach Intellectual Potential
Improve Problem Solving Skills
Enhance Creative Thinking
Increase Mental Processing Speed
Improve Ability to Learn
59. Why Should This Matter toWhy Should This Matter to
You?You?
The relational skills we train form the
very building blocks of intelligence and
teaching these skills will increase
student’s abilities to access the
curriculum in all subject areas, incl.
mathematics, reading and spelling,
science, etc
THIS WILL MAKE YOUR JOB EASIER!!!
60.
61. Implications
Now we can!
Perhaps for the first time in the history
of psychology we have an educational
technology with which we can literally
nurture genius as well as supporting the
students that most need our help, but for
whom we very rarely have enough
resources to properly support their
needs.
62.
63.
64. Get a free two week trial for your school here:
https://secure.raiseyouriq.com/schools
See some of our media coverage here:
https://secure.raiseyouriq.com/media-coverage
Like us on Facebook here and listen to radio
interviews:
https://www.facebook.com/RaiseYourIQ
Sign up for our newsletter here:
https://secure.raiseyouriq.com/users/sign_up
Notes de l'éditeur
Average = 100. SD = 15.
67% of the population will obtain a score that is one Standard deviation of the average score
96% will obtain ascore that is within 2 standard deviations of the average.
Hypothetical curve: In reality, the curve is negatively skewed...with modern medicine etc...
Winship & Korenman (1997) Chapter 10 in Intelligence, genes and Success: Scientists respond to the Bell Curve
Given a long enough social history of multiple exemplar training for deriving relations, a child taught to match these stimuli as illustrated by the solid lines, can derive the stimulus relations represented by these hashed lines without reinforcement. This is stimulus equivalence - the simplest form of derived relational responding.
Imagine a child is taught that the word “dog” goes with an actual dog and then in another context, perhaps in a different house by a different person, the child learns that “bow wow” also goes with “dog”, in other words, both words refer to the same thing. So, in 2 different contexts, a child learns that 2 different words refer to the same thing. Then in another context a child hears a caregiver or peer use the terms interchangeably or is told explicitly that one word means the other. This child has effectively been shown how to derive an equivalence relation.
On each training trial a task like this one in presented, consisting of several relational statements and a relational question. The participant has to derive a stimulus-stimulus relation in each case, from the arbitrary relations represented by the statements. Other levels of training involve other relation types and mixtures of relation types. Test blocks consist of 16 novel tasks presented without feedback.
Here is another example involving Opposite relational statements, and a question in terms of a frame of Sameness or coordination.
Here are some puzzles that require relational skills the read puzzles slowly
The average IQ score at the baseline was 97 (average intellectual range). The average IQ score at the follow-up was 120 (superior intellectual range). This represents an average IQ rise of 23 points.
The range of IQ rises was from 14 points to 32 points. The increase in IQ was significant and the Effect size was very large.
Narrative: The Sigma T test measures standard mathematical ability in Irish Schools. This graph represents STEN scores for the Sigma T for each of the 15 participants. The sten scores rose on average from 6.06 to 6.66. This increase in Mathematical ability was significant, with a medium effect size.
Narrative: The Micra T test measures standard reading ability in Irish Schools. This graph represents STEN scores for the Micra T for each of the 15 participants. The sten scores rose on average from 5 to 5.46. This increase in reading ability was NOT significant, BUT there was a medium effect size.
Summary – read it all out in Plain English ad-libbing – or at least the highlights