Italian poor comprehenders: are they overloaded by irrelevant information?

1
Italian poor comprehenders are they overloaded
by irrelevant information?

• Paola Palladino
• University of Pavia, Italy

paola.palladino_at_unipv.it
2

• Italian reading decoding
• Italian poor comprehenders
• Italian assessment of reading comprehension
• Working memory in poor comprehenders
• Intrusion errors and evidence of less efficient
  activation control

3

• Italian is a transparent language
• Shallow orthography
• Regular print to sound translation

Example
G
U
F
GUFO
O
OWL
U
M
R
O
WALL
MURO
Most cases NO lexical and semantic look-up
4

• Therefore
• learning to read Italian
• Is mainly a question of speed (even dyslexics
  make almost no errors from the fourth-fifth
  grade)
• And in normal children the speed increases of
  half a syllable a year
• then, after an initial phase, decoding and
  comprehension result in partly independent
  skills. Chilosi et al. 2003, 80 dyslexic are
  good comprehenders

5
Reading speed of a passage normal vs. dyslexic
children
Sill/sec
grades
Tressoldi et al. 1996
6
poor comprehenders-cattivi lettori are children
who fail in reading comprehension but they can
read at a normal speed and correctness (Yuill
Oakhill, 1991)
They are about from 3 to 7 when non-verbal
intelligence is controlled for and reading
decoding is normal
7
Palladino et al. 1995 589 children in 6th
grade -1 s.d. 35 poor comprehenders (6) De Beni
et al. 1998 3103 college students -1.5 s.d. 95
poor comprehenders (3) Palladino et al.
2001 265 college students -2 s.d. 18 poor
comprehenders (6.8)
8
Reading comprehension in Italy is generally
measured with MT test a battery where short
texts are followed by inferential questions

• No time constrain
• Free re-reading of both text and questions
  (inferential Q)
• Multiple choices
• Collective administration

9
MT reading comprehension test Example of
inferential question

• The passage describes a child who encounters a
  dolphin and refers to it as a black triangle
  rising from the water
• Example of a question
• What is the black triangle?

10
MT reading comprehension test Example

1. A black spot on the back of the fish
2. The fin typical of dolphins
3. A snake in the sea
4. The head of the dolphin

11
Who are Italian poor Comprehenders?
Average non-verbal skills gtmean Adequate
decoding speed gtmean Poor reading
comprehension performance Score of 1-2 s.d.
below the mean
12
The Working Memory deficit De Beni et al.
1998 The exam of poor comprehenders memory
performance showed Poor verbal working
memory the Listening Span Test requiring sentence
processing and last word recall Average
performances in forward and backward digit
span Average in the LST when all words where to
be recalled
13
De Beni et al. 1998 Intrusion errors processed
but not to be recalled words more frequent in
poor comprehenders recall When words activation
was manipulated poor comprehenders made more
intrusions of more activated irrelevant
information 30 of highly processed words (tap
the hand with an animal noun) 14 of other words
14
What is going on doing a working memory task?
In which of the many processes taking place in a
wm task can we find the source of individual
differences in wm efficiency?
One possibility is that the efficiency in
reducing activation of irrelevant information
does matter.
Intrusions could be a clue but
15
What is going on doing a working memory task?
Hypothesis People with poor comprehension have
also a poor wm as they are not able to control
activation-interference of processed but no
longer relevant information
Gernsbacher, 1990 reading comprehension
requires suppression of irrelevant information
16
What is going on doing a working memory task?

• Doing the LST many information are processed
• Only a minimum part has to be recalled (last
  word)
• The rest of the words have to be suppressed

17
Rationale of the Experiments
A way of testing the inhibitory hypothesis could
be to measure the activation of no longer
relevant information during a wm task as for
example with a lexical decision
18
Rationale of the Experiments
If participants are suppressing activation of no
longer relevant information they should not show
any positive priming (advantage) for them
compared to completely new items Whereas
participants with poor wm should manifest a
certain amount of positive priming as a
consequence of a difficulty in reducing
activation of no longer relevant information
19
Method
Palladino 2006 QJEP
Method

• Participants
• From a sample of about 150 third graders three
  groups of
• 26 control children,
• 11 children at risk of ADD
• 10 children at risk of ADD and RD have been
  selected.
• Children with ADD and RD were poor in reading
  decoding and comprehension, vocabulary, verbal
  reasoning besides being at risk for the Attention
  Deficit Disorder (ADD).

20
Material Categorization Working Memory
Span Thing Mother Dog Word Night Year Champion
Gas Location Match Woman Doe Head Visit
Giraffe Tap the hand when animal noun (dog, doe
and giraffe) remember the final words (night,
match and giraffe). Lexical Decision
task Champion , Visit Old NoAnimal Dog, Doe
Old Animal Statue, Answer new Non Animal Cat,
Cow New Animal
21
Categorization Working Memory Span number of
words recalled
22
Lexical Decision Task RT
23
Lexical Decision Task gain-cost for old
words(RT new- RT old)
24
Data seem to indicate that in doing a working
memory task control participants are controlling
the activation of no longer relevant information
showing no advantage for just presented material
when it becomes no longer relevant.
25
It could be seen a quite paradoxical result the
group with poor working memory that has slow RT
shows a greater advantage for no longer relevant
information than the control group Therefore a
poor recall is associated with a higher
activation in memory of irrelevant information
and not simply with a general poor memory of any
information presented.
26
Method
Palladino 2008
Aim examine a group with poor comprehension
without ADD

• Participants
• From a sample of about 200 third graders, 8 years
  and five months old,
• 73 control children,
• 14 children poor comprehenders
• 11 children poor decoders
• 23 children poor calculators
• 16 children at risk of ADDLD

Material The task was similar to the previous
one
27
Categorization Working Memory Span number of
words recalled
28
Lexical decision gain-cost of old items
(correct new-correct old)
29
DISCUSSION
Again results seem to indicate that in doing a
working memory task control but also poor
decoders or calculators children are controlling
the activation of no longer relevant information
showing no advantage for just presented material
when it becomes no longer relevant.
30
DISCUSSION
Previous results are replicated with poor
comprehenders poor working memory is associated
with a greater advantage for no longer relevant
information Therefore a poor recall is
associated with a higher activation in memory of
irrelevant information and not simply with a
general poor memory of any information presented.
31
What is going on doing a wm task?
People with poor comprehension have also a poor
wm as they are not able to reduce
activation-interference of processed but no
longer relevant information
reading comprehension similarly to working
memory requires suppression of
irrelevant information (Gernsbacher, 1990)
32
What is going on doing a reading comprehension
task?
Further research could examine whether poor
comprehenders are less able to reduce activation
of irrelevant information when reading a passage
De Beni Palladino 2000 Poor comprehenders free
recalling a passage where as able as good in
recalling less relevant information but poorer
in recalling relevant one
33
Thank you for your attention!!