20
particles of the gases are in constant movement". The total score is the sum of the correct
answers based on a correction template. Thus, the maximum score is 30. The reliability
analysis comprised these 30 items and Cronbach's alpha reported an acceptable internal
consistency (α = .74).
Text.
This

is

about

Atmospheric

Pressure.

It

contains

1126

words

and

20
paragraphs distributed in five sections preceded by a subtitle. Eleven of these paragraphs
contain information relevant to the AOQ and SE task.
Twelve self-explanations vs. twelve open-ended questions. Both the SE and AOQ
are different in procedure. While the
AOQ students had to read the text first and t hen
answer 12 open-ended questions; SE students did not do an initial reading of the text, but
self-explained 12 target-sentences (in bold) while they were reading.
The
twelve
AOQ
questions
were
designed
according
to
two
levels
of
representation: six TB questions and six SM questions. An example of TB question is the
following: ‘What causes atmospheric pressure?’ His answer corresponded to the idea of
the first self-explaining target-sentence (fourth paragraph). On the other hand, an example
of a SM question is as follows: ‘If you tried to replicate Torricelli's experiment at the top
of an eight thousand meters mountain, instead of at sea level where the original was made,
why would more mercury come out of the tube into the bucket, going down more than 760
mm in height?’. To answer this question, the student must have understood Torricelli's
experiment explained in the text. Moreover, he should relate the ideas of the tenth and
eleventh paragraphs and integrate his prior knowledge.
The target-information that SE students had to self-explain
corresponded to the
relevant information needed to answer all the questions. To determine these phrases, a
pilot study was conducted with 20 first-year college students. They had to highlight the
relevant information in the text to answer the questions.
All
self-explanations

were

preceded

by

the

following

statement:

‘Explain

the
highlighted fragment in bold. You can use all the information in the text that you consider
necessary’. In this way, an attempt was made to encourage the student not to limit to the
target-sentence and to make the appropriate inferences between textual information and
prior knowledge.
This text has been widely used by the ERI-Lectura group in relation to the AOQ
task. Their reliability analysis was performed with the 12 questions and Cronbach's alpha
reported an acceptable internal consistency (α = .71). Likewise, the processing strategies
of
all

the

AOQ

and

SE

responses

were

categorized

by

inter-rater

agreement.

Cohen's
Kappa (k) reported almost perfect agreement (k = .88) and substantial agreement (k = .75),
respectively. Then, disagreements were discussed and resolved.
Software: Read&Learn (Vidal-Abarca et al., 2018). It is a software designed by
ERI-Lectura (University of Valencia). It does not require the previous installation of any
program. The AOQ condition is characterized by the use of two screens, one for the