1. Dehaene (2001)
The question addressed by the research.
What are the cerebral mechanisms when processing masked
(invisible) words?
The population and equipment used in the study.
fMRI and ERP’s used for measuring brain activity. 37 French
students; 19 to 34 years old.
An overall picture of the (experimental) protocol used (what data
was collected? e.g.neuroimaging method, behavioral task etc.).
Experiment 1: Within the word processing circuit, image areas
activated by masked words. Participants viewed a random series of
masking shapes and blank screens, during which brain activity was
measured with fMRI and ERP’s. (figure below)
Condition 1: blank before and after word, which makes the
word consciously perceivable.
Condition 2: mask before and after word, which makes the
word invisible.
Control: In order to
measure the difference
in brain activity due to
merely seeing the
masking stream and
actually seeing a
masked/unmasked word,
the words were
sometimes replaced by
blanks.
Data collected: fMRI, ERP, and behavioral tests: name the
word out loud (1), decide if word has been shown before (2),
forced choice test between two options (3).
Experiment 2: (in order to
investigate whether masked words
can actually facilitate repetition
priming). A masked word similar or
different from the target word
would be presented in either the
same case or different case (capital
or small letters) than the target
word. The difference in casing was
in order to see whether the
abstract identity of the word was
actually processed, not just the
physical similarity of the letters.
(figure on the right).
, Data collected: fMRI, ERP, and behavioral tests: response time of
classifying the words in “man-made” (train) or “natural” (apple)
words by the use of a response button (1), after that a forced choice
test (2).
The methods used for data analysis (univariate contrasts,
multivariate decoding,etc.).
For fMRI: two random-effect analyses.
For ERP: sample by sample t-tests and ANOVA.
The main results (all experimental factors and dependent measures
are important).
Experiment 1:
fMRI: visible words
activated a left-
lateralized set of
areas (similar to
PET studies of
word reading).
Masked words
activated two
significant clusters
in this same
circuit. (image on
the right).
ERP’s: (image
below)
Behavioral tests:
(figure below)
Experiment 2:
fMRI: the left
fusiform is “case-
independent”,
meaning that the
activity in this area
is reduced to the
same extend
regardless of
whether the prime
word is shown in
the same or in a
different case.
The right extrastriate is “case-specific”, meaning that activity in this
area is extra
reduced when the
, case is similar, and not reduced when the case is different (see
image next page).
Behavioral tests: (see part b in the figure on the previous page)
The question addressed by the research.
What are the cerebral mechanisms when processing masked
(invisible) words?
The population and equipment used in the study.
fMRI and ERP’s used for measuring brain activity. 37 French
students; 19 to 34 years old.
An overall picture of the (experimental) protocol used (what data
was collected? e.g.neuroimaging method, behavioral task etc.).
Experiment 1: Within the word processing circuit, image areas
activated by masked words. Participants viewed a random series of
masking shapes and blank screens, during which brain activity was
measured with fMRI and ERP’s. (figure below)
Condition 1: blank before and after word, which makes the
word consciously perceivable.
Condition 2: mask before and after word, which makes the
word invisible.
Control: In order to
measure the difference
in brain activity due to
merely seeing the
masking stream and
actually seeing a
masked/unmasked word,
the words were
sometimes replaced by
blanks.
Data collected: fMRI, ERP, and behavioral tests: name the
word out loud (1), decide if word has been shown before (2),
forced choice test between two options (3).
Experiment 2: (in order to
investigate whether masked words
can actually facilitate repetition
priming). A masked word similar or
different from the target word
would be presented in either the
same case or different case (capital
or small letters) than the target
word. The difference in casing was
in order to see whether the
abstract identity of the word was
actually processed, not just the
physical similarity of the letters.
(figure on the right).
, Data collected: fMRI, ERP, and behavioral tests: response time of
classifying the words in “man-made” (train) or “natural” (apple)
words by the use of a response button (1), after that a forced choice
test (2).
The methods used for data analysis (univariate contrasts,
multivariate decoding,etc.).
For fMRI: two random-effect analyses.
For ERP: sample by sample t-tests and ANOVA.
The main results (all experimental factors and dependent measures
are important).
Experiment 1:
fMRI: visible words
activated a left-
lateralized set of
areas (similar to
PET studies of
word reading).
Masked words
activated two
significant clusters
in this same
circuit. (image on
the right).
ERP’s: (image
below)
Behavioral tests:
(figure below)
Experiment 2:
fMRI: the left
fusiform is “case-
independent”,
meaning that the
activity in this area
is reduced to the
same extend
regardless of
whether the prime
word is shown in
the same or in a
different case.
The right extrastriate is “case-specific”, meaning that activity in this
area is extra
reduced when the
, case is similar, and not reduced when the case is different (see
image next page).
Behavioral tests: (see part b in the figure on the previous page)
