International Journal of Educational Research 72 (2015) 80–88

Contents lists available at ScienceDirect

International Journal of Educational Research

journal homepage: www.elsevier.com/locate/ijedures

Fostering reading comprehension of expository texts

through the activation of readers’ prior knowledge and
inference-making skills
Christian Tarchi *
Department of Education and Psychology, University of Florence, via di San Salvi, 12, Complesso di San Salvi Padiglione 26, 50135 Florence,
Italy

A R T I C L E I N F O A B S T R A C T

Article history: The importance of being able to critically read expository texts cannot be overstated. It is,
Received 3 January 2015 therefore, fundamental to identify evidence-based classroom practices that can help one
Received in revised form 12 April 2015 with this. Among higher-order skills, prior knowledge and inferences are considered the
Accepted 27 April 2015 most important predictors of reading comprehension. Unfortunately, however, few
Available online 11 June 2015 reading comprehension interventions specifically address the multidimensional nature of
prior knowledge and inference-making skills. In this study, an intervention designed to
Keywords: activate prior knowledge during reading comprehension was used to test inference-
Reading comprehension
making skills and metacognition. One hundred and sixty-six secondary schools students
Exposition
participated in the study. Both control and experimental groups had improved reading
Prior knowledge
Inferences
comprehension, however the intervention based on prior knowledge activation resulted in
Metacognition better reading comprehension, metacognitive and inference-making processes.
Cooperative learning ß 2015 Elsevier Ltd. All rights reserved.

The educational systems of several countries are concerned that a significant amount of students may not have the
necessary literacy knowledge and skills to benefit sufficiently from their educational opportunities. Therefore, it is important
to explore the reading comprehension process. In particular, students often encounter difficulties in expository texts, as their
structure is more difficult to extract, they often contain specific jargon, and discuss difficult concepts as compared to
narrative texts (Best, Floyd, & McNamara, 2008). In this study, an expository text reading comprehension intervention is
presented that was designed to activation students’ prior knowledge. Prior knowledge has been widely studied and often
applied to reading comprehension, but its multi-dimensional nature has rarely been addressed. The efficacy of the
intervention to activate several sub-dimensions of prior knowledge was measured.

1. Reading comprehension: Higher-order skills

Text comprehension is considered a hierarchy of skills from basic-level (or bottom-up), such as working memory and
decoding, to high-level (or top-down) reading skills, such as inference-making skills and metacognition (Rapp, van den
Broek, McMaster, Kendeou, & Espin, 2007). Basic and higher-order skills develop simultaneously and independently,
suggesting that instruction in high-order reading comprehension should be a specific educational target, and treated
independently from basic skills instruction (Rapp et al., 2007). This claim is reinforced by the consideration that certain

* Tel.: +39 055 2055811; fax: +39 055 6236047.

E-mail address: christian.tarchi@gmail.com

http://dx.doi.org/10.1016/j.ijer.2015.04.013
0883-0355/ß 2015 Elsevier Ltd. All rights reserved.
 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88 81

students struggle in reading, even when basic skills are solid (Alvermann, Fitzgerald, & Simpson, 2006). Several studies have
analyzed the influence of basic-level skills like decoding and working memory on reading comprehension, while several
researchers argue for the need for more research on the influence of top-down processes on reading comprehension (Rapp
et al., 2007). The research presented in this article analyzes how to maximize higher order reading skills.

1.1. Prior knowledge

Among higher-order skills, current research on reading comprehension proposes that prior knowledge is the strongest
reading comprehension predictor (Alexander & Jetton, 2000; Dochy, 1994; Gurlitt & Renkl, 2010; Hailikari, Katajavuori, &
Lindblom-Ylanne, 2008; Rapp et al., 2007). Prior knowledge is defined as the whole of a person’s actual knowledge, available
before a certain learning task, structured in schemata, declarative and procedural, partly explicit and partly tacit, and
dynamic in nature (Dochy, 1994). Hailikari, Nevgi, and Lindblom-Ylänne (2007) claimed that it is important to consider what
knowledge type is being evaluated. Many scholars distinguish between declarative, i.e., knowledge of facts and concepts, and
procedural knowledge, i.e., know-how (Hailikari et al., 2007). Thus, prior knowledge is a multidimensional and hierarchical
construct in breadth, depth, and quality (Alexander & Jetton, 2000). Prior knowledge has been assessed in several studies, but
generally through multiple-choice questionnaires inquiring what readers knew about a certain topic. However, not all prior
knowledge types have similar relevance in relation to students’ achievement. For instance, some scholars distinguished
between domain and topic knowledge (Alexander & Jetton, 2000; Dochy, 1994; Tarchi, 2010, 2012). Domain knowledge
represents what a student knows about a certain discipline, e.g., history, science, geography. In other words, domain
knowledge is the breadth of knowledge within a field (Alexander & Jetton, 2000). Instead, topic knowledge is a more focused
knowledge on a specific topic, e.g., The French Revolution, the digestive system, the geography of a specific country. Topic
knowledge, therefore, represents the depth of knowledge in a certain topic (Alexander & Jetton, 2000). Topic knowledge can
be further differentiated from knowledge of facts. Knowledge of facts is a lower level of declarative knowledge that can be
assessed by simply asking the student to enumerate essential concepts on a certain topic, and knowledge of meanings, i.e., a
higher level of declarative knowledge that requires the student to understand the meaning of a specific concept and correctly
define it (Hailikari et al., 2007). Previous studies confirmed that these dimensions of prior knowledge contribute differently
to reading comprehension performances such that topic knowledge of facts and meanings influenced reading
comprehension more than domain knowledge did (Tarchi, 2010, 2012).

1.2. Inference-making skills

Inference-making is another higher-order skill that has a central role in reading comprehension (Rapp et al., 2007). When
readers are required to integrate information provided by the text but located in different sentences, or when they need to
incorporate information outside the text, they need to make an inference. The contribution of inferences to reading
comprehension is independent from prior knowledge, however these two components work in synergy to fill the gaps found
in the information of the text (Best, Rowe, Ozuru, & McNamara, 2005; Cromley & Azevedo, 2007; Tarchi, 2010, 2012).
Inferences are what make the reader move from a mere interpretation of individual sentences to a global meaning that
integrates multiple sentences (Best et al., 2005). Oakhill (1983) pointed out that poor comprehenders have difficulties in
dealing with constructive inferences (that is, creating links within the text) as well as understanding the meaning of a word
based on context. While addressing the relevance of these two inference types for reading comprehension, Tressoldi and
Zamperlin (2007) proposed a distinction between lexical and semantic inferences. In order to activate the correct and more
appropriate meaning, it is necessary to make a lexical inference, to adapt one of the possible meanings of a term within a
sentence. A semantic inference, in contrast, is based on a wider knowledge of the world. In this study, both inference types
were assessed to determine the specific efficacy of the intervention (see Section 3 for more details).
Cromley and Azevedo (2007) tested the combined effect of prior knowledge and inferences on the reading
comprehension process in their direct and inferential mediation model of reading comprehension. They found that the
reader’s prior knowledge and his/her word reading have both a direct and indirect influence on reading comprehension. Via
the indirect path, the effect of prior knowledge is mediated by metacognition and inference-making skills. Tarchi (2010)
refined this model by including the multidimensional nature of prior knowledge higher-order skill that both the direct and
indirect prior knowledge effects on text comprehension depended on the subject-area and the text content. In this article, the
reader’s prior knowledge of key concept meanings was hypothesized to directly and indirectly influence the comprehension
of a science text, while prior knowledge of facts was hypothesized to directly and indirectly influence comprehension of a
history text with both being mediated by inferences.
Prior research has confirmed the importance of addressing the multidimensional nature of prior knowledge and its
interaction with inference-making skills in designing interventions that foster students’ reading comprehension processes.
Prior knowledge and inference-making skills have often been treated in isolation. Several authors have included prior
knowledge in their reading comprehension interventions (Taboada & Guthrie, 2006). Indeed, some authors have created
activities to facilitate the retrieval of relevant knowledge for comprehending a text (Rapp & Kendeou, 2007) or question-and-
answer generation tasks (Darch & Gersten, 1986). Palincsar and Brown’s (1984) Reciprocal Teaching had students ask
relevant questions of the text, which theoretically activates inferences, and predict the content of the following paragraph
(prior knowledge activation). Pressley et al.’s (1994) Students Achieving Independent Learning program also activated
82 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88

students’ prior knowledge and inferences by including the following reading strategies: predicting, visualizing, questioning,
clarifying, making associations, and summarizing. Regarding inference-making skills, research has demonstrated the
efficacy of strategy-based instruction (Simmonds, 1992) or questions eliciting coherence-based inferences (van den Broek,
Lorch, Linderholm, & Gustafson, 2001) to enhance reading comprehension. A recent review of such studies can be found in
Rapp et al. (2007). The main limitation affecting reading comprehension interventions based on prior knowledge activation,
is that they address this construct generically, and do not attempt to explicitly activate each dimension of the reader’s prior
knowledge: domain and topic (meanings and facts).

2. Research questions and hypotheses

In this study, the efficacy of a reading comprehension program based on the activation of relevant dimensions of prior
knowledge was analyzed. The intervention was designed within the context of peer-assisted education. According to our
hypotheses, an intervention based on the activation of specific dimensions of prior knowledge, would be more effective than
an intervention based on the activation of generic prior knowledge. Intervention influence on the reader’s approach to the
text was analyzed via participants’ performance on standardized tests. The intervention was expected to be effective in
enhancing students’ reading comprehension process by activating relevant prior knowledge. In addition, through the
intervention, students learn to control their reading comprehension process, and, therefore, develop a more metacognitive
approach to the text. Finally, as the activation of relevant prior knowledge creates a basis on which inferences can take place,
we expected that students develop better performances also in this skill.

3. Method

3.1. Participants

One hundred sixty-six students attending grade 7 and grade 8 in three schools in Florence, Italy, participated in this study
(76 boys and 90 girls; Mean age: 12.86  .81). All schools were located in middle-high socio-economical areas. Students with
learning disabilities, foreign born, with significant language problems, and deemed poor decoders by their teachers were excluded
from the sample. No statistically significant differences between males and females were found in terms of cognitive
performances at the pre-test and post-test assessments.
Students were divided into two groups, an experimental (85 students), and a control group (81 students). The
experimental group underwent 4 weeks of intervention, 1 session per week, and each time on a different text, based on the
activation and sharing of prior knowledge. The control group worked on the same texts as the experimental group and
followed a reciprocal teaching approach (Palincsar & Brown, 1984). Using this teaching approach for the control group
enabled the controlling of a potential confounding effect of cooperative learning methodology in the experimental group. For
each school one class was assigned to the experimental group, and another class to the control group, in order to control a
potentially confounding effect related to schools characteristics. Moreover, the pre-test did not demonstrate differences in
cognitive performances between the two groups, so the experimental and the control groups could be assumed as
equivalent. The two groups did not differ in terms of age and/or socio-economic status. This last data was obtained following
the procedure by Ganzeboom, Graaf, Treiman, and de Leeuw (1992). Finally, before the intervention, students were given a
multiple-choice prior knowledge test on the topics covered by the texts used in the session, and no statistically significant
differences resulted from the comparison of the two groups’ performances.

3.2. Materials and measures

3.2.1. Intervention
In order to activate students’ prior knowledge on different domains, four expository texts of different disciplinary
domains were used in the two experimental interventions: history (The Postwar Period, 1206 words); geography (A World of
Contrasts, 1616 words); sciences (Energy and its Sources, 1296 words); and a expository newspaper article (Earth with three
degrees more: so our planet will be shocked, 884 words). The first three readings were derived from textbooks for grade 8,
whereas the article was derived by an Italian newspaper, La Repubblica.

3.2.2. Assessment

3.2.2.1. Reading comprehension of expository texts. The reading comprehension performance was assessed through a
standardized battery developed by Cornoldi and Colpo (1995) that is currently used in Italy for the assessment and diagnosis
of learning disabilities related to reading comprehension. Two parallel versions were used, one for the pre-test and the other
for the post-test. Students had to read a piece of writing and then answer 10 pre-test and 14 post-test multiple-choice
questions. The number of correct answers represented each student’s final score. The alpha coefficient was .87 for the pre-
test and .83 for the post-test. Here is a translated sample of the questions that were used in this particular component:
From the text. An impressive example of the decrease of the Earth’s capacity to produce nourishment, due to an
excessive ecological stress, comes from the zone located on the south of the Sahara desert. Question: What does the
 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88 83

author mean by ‘‘ecological stress?’’ A. the preoccupation for ecologists’ ideas; B. the excessive exploitation of the
environment; C. the effort and the tension of people working the soil; D. the study of the hard relationship between
men and animals.

3.2.2.2. Semantic inferences. A semantic inference is based on a wider knowledge of the world. For example, to infer the
season in which the event described in the sentence, ‘‘during a sultry afternoon in which you could have heard a pin-drop,
Gimmy arrived at our place’’ takes place, it is necessary to access our knowledge on the meteorological conditions of the
different seasons at our latitude. This component was assessed through a test developed by Tressoldi and Zamperlin (2007).
Two versions were used, one for the pre-test and the other for the post-test. The test consisted of a text followed by 10
multiple-choice questions asking the students to derive inferential information that was not explicitly stated in the text. The
alpha coefficient of this instrument was .75 for the initial version and .78 for the final one. Here is a translated sample of the
questions that were used in this particular component:
From the text: On the 22nd of December I left to visit again my friends; I only had with me 8–10 people who were
guiding me and carrying what I needed for a few days. Question: Why was Piaggia travelling with other 8–10 people?
A. because he did not trust just one person; B. because they were at the same time guides and bearers and he had a lot
of stuff to carry; C. because he was afraid of an assault; C. because he felt affection towards them.

3.2.2.3. Lexical inferences. An example of lexical inference consists in deriving the meaning of ‘‘step’’ and ‘‘space’’ from the
sentence: ‘‘the landing on the moon is considered an important step for the exploration of the space.’’ Knowing the many
meanings of ‘‘step’’ and ‘‘space’’ is not sufficient to choose the pertinent meaning to the sentence containing these two words.
To assess this component a test developed by Tressoldi and Zamperlin (2007) was used. Two versions were given to the
students, one in the pre-assessment and the other in the post-assessment. Students had to read a text and then answer 10
multiple-choice questions to individuate the target-words’ meaning from the passage. The alpha coefficient of this
instrument was .77 for the initial version and .82 for the final one. Here is a translated sample of the questions that were used
in this particular component:
From text: The bee is a social being that lives in community, and the tide cohabitation in the swarm of the bees shows
the existence of a certain exchange of information. Question: Cohabitation means. A. lending him-/herself to someone
else’s interests; B. living tidily; C. living on his/her own; D. living together.

3.2.2.4. Metacognition. Metacognition was assessed through a standardized test originally developed by Pazzaglia, De Beni,
and Cristante (1994) and drawn on by Tressoldi and Zamperlin (2007). It assesses the four areas individuated by Brown
(1978) in her metacognitive model of comprehension: awareness on reading aims, knowledge of strategies to improve
comprehension, monitoring during comprehension, and sensitivity to the text. This test uses different question modalities:
multiple-choice questions with three options, multiple-choice questions with two options, yes/no questions, open
questions, underlining the mistake. The same version was given to students in both assessments. The alpha coefficient of this
instrument was, respectively, .75 and .77.
Here are a few translated sample questions from this test:
Awareness on reading aims
Please describe all the reasons why it is important to read.
Knowledge of strategies to improve comprehension
According to you, a good reader:
A. Reads everything with the same accuracy.
B. Changes his/her way of reading according to text difficulties.
C. Always reads aloud.
Monitoring during comprehension
Some of the following sentences have some mistakes. Find and underline them.
‘‘The student tried to answer to the teacher’s question, but he did not succeed because he had studied very much’’.
Sensitivity to the text
Please, consider the following texts and write down which genre they belong to.
Text nr. 1. Once upon a time there was a prince who was able to wake a princess up from her enchanted sleep. And they lived
happily ever after. This text is . . .. . .. . .. . .. . .

3.3. Research design and procedure

To test this study’s hypotheses, a quasi-experimental research design with pre-test and post-test comparisons between
nonequivalent groups was carried out. The interventions took place over four consecutive weeks, once a week for an hour.
84 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88

Each session was carried out with the following schedule. First, students were given the personal booklets in which they had
to write their personal code which allowed for anonymous performance sessions and assessments. As soon as they received
their booklets, the students were divided into small groups of 5 to 6 students per group and those groups were maintained
throughout all 4 intervention sessions. Group composition was decided by the teacher, who had to ensure that reading
comprehension skills were heterogeneous amongst the members of each group, thus abiding by cooperative learning
principles (Moreno, 2009). Students were also given an instruction sheet, so that they could autonomously follow each
intervention step.

3.3.1. Prior knowledge activation

The ‘‘prior knowledge activation’’ intervention consisted of the following steps. In each session, groups worked
independently, with the researcher (the ‘‘expert’’) acting as a facilitator.

3.3.1.1. Reading the opening question. The first page of the booklet featured an opening question, which was meant to foster
discussion on the topic tackled by the text. For instance, for history: ‘‘What are the consequences of a war?’’ Students opened
the first page of their booklet and read the opening question. The group had to make sure that everyone understood the
question. Starting the reading activity with a prior knowledge activating question allows to situate the learning in a context
that the reader feels as real and authentic (Billett, 1996; Lave & Wenger, 1991; Scardamalia & Bereiter, 2006).

3.3.1.2. First discussion. Each group discussed the opening question. The expert let them discuss it freely, according to their
opinions, beliefs, and experiences. This first discussion allowed readers to activate relevant prior knowledge and to share it.
In this way, the students cooperate in creating common background knowledge to use as a reference to construct the
meaning of the text (Alexander & Jetton, 2000; Dochy, 1994; Hailikari et al., 2008; Rapp et al., 2007).

3.3.1.3. Reading the title and the headings. After the discussion had allowed each student to express his or her ideas and
develop some topics in collaboration with other students, the expert asked the students to go to their booklet’s next page.
Students were asked to read the main title and each subheading and make sure that everyone understood them. Socio-
constructivist research on reading comprehension argues that this process derives from the interaction between the reader
and the text (Armand, 2001; Best et al., 2005). This step supports such an interaction as it helps the reader to focus on the text
characteristics and in particular on its advance organizers (Ausubel, 1960; Boscolo, 1997).

3.3.1.4. Second discussion. Each group, with some guidance by the expert, discussed the title and the subheadings of the text.
This encouraged students to figure out the main idea expressed by the text and discuss basic facts and meanings. Students
had to figure out what the text was about, which facts had to be known in order to understand the text, which key words
were likely to be found in the text, and what the meaning of these key words was. Research on prior knowledge considers it
as multidimensional (Alexander & Jetton, 2000; Tarchi, 2010). In this step, the reader is asked to activate text-specific prior
knowledge of facts and meanings about the topic discussed in the reading assigned.

3.3.1.5. Individual Reading. After the students discussed each point, the expert asked them to individually read the text.

3.3.2. Reciprocal teaching

As this study specifically focused on prior knowledge, also the control group worked on the same texts with a cooperative
learning methodology, reciprocal teaching (Palincsar & Brown, 1984). Students were divided into small groups of 5 or 6
members. Each group had to cooperatively apply four reading strategies:

 questioning, asking relevant questions on the text;

 clarifying, finding out the meaning of unclear words or sentences;
 summarizing, making a synthesis of the passage they had just read;
 and predicting, trying to guess what would the text talk about in the subsequent paragraphs.

Each group worked independently. Students had to take turn assuming the role of the teacher in leading the dialogue
about what the group was reading. The researcher acted as facilitator.

3.3.3. Similarities and differences between prior knowledge activation and reciprocal teaching
Both reading comprehension applied a cooperative learning approach, in which students had to interact to co-construct
the meaning of a text, but had also to independently comprehend the text. Indeed, while the text was read and commented
together, each student was given a reading comprehension test about the text they had just read. Both interventions acted at
a metacognitive level, as students were exposed to other students’ perspectives, which differ from their own, and had to
negotiate the meaning to determine which perspective fitted better the text. Both activate prior knowledge, as reciprocal
teaching includes a predicting strategy, and students need to activate pre-existing schemes to predict what could be the
following paragraph about. However, the most important difference lies in the way prior knowledge is activated. In the prior
 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88 85

Table 1
Descriptive statistics.

Variable N Mean Skewness Kurtosis

Reading_comprehension1 162 5.79  1.90 .47  .19 .29  .38

Lexical_inferences1 166 6.94  1.74 .79  .19 .44  .38
Semantic_inferences1 146 6.43  1.86 .97  .20 .75  .40
Metacognition1 161 8.84  2.12 .50  .19 .43  .38
Reading_comprehension2 152 7.31  1.50 .77  .20 1.09  .39
Lexical_inferences2 166 6.74  1.49 .33  .19 .29  .38
Semantic_inferences2 137 5.73  1.94 .33  .21 .08  .41
Metacognition2 154 9.36  1.66 .40  .20 .09  .39

knowledge activation intervention, each component of readers’ prior knowledge is specifically targeted, and guided by the
interpretation of the text’s advance organizers (title, sub-headings, bolded words, and the like). Students are asked to
activate fact and meanings that they consider relevant to the text. Another difference is that in the prior knowledge
intervention, prior knowledge is activated and shared at the beginning, and then children individually read the text, which is
how usually reading comprehension is performed in school. Instead, in the reciprocal teaching intervention, prior knowledge
is recursively activated at the beginning of each paragraph.

3.4. Data analysis

A missing data analysis was carried out in order to decide how to substitute missing values. Each variable’s extreme
outliers have been individuated and eliminated by observing the relative box-plots. The normality of each dependent
variable’s probability distribution was explored: in those cases in which a variable distribution was not similar to a Gauss
curve, the appropriate monotone increasing transformations were applied before carrying out the statistical inferential
analysis. In order to test the hypotheses and the efficacy of an intervention on reading comprehension based on prior
knowledge activation, a repeated measures multivariate analysis of variance was performed, with group (experimental
versus control) as a between-subject variable and time (performances between the pre-test and at the post-test assessment)
as a within-subject variable.

4. Results

Descriptive analyses (see Table 1) and correlation analyses (see Table 2) are presented.
Most of the measures correlate with other measures, both at the pre-test, at the post-test, and across time. This data
confirms the multicomponential nature of reading comprehension: the reader needs to be aware whether he/she is
comprehending and developing sensitivity to the text (metacognition), and to infer unstated information (inference-
making), in order to deeply comprehend a written text.
According to the repeated-measures multivariate analysis of variance, both interventions contributed to enhance
students’ performances in reading comprehension, as the main effect of time was statistically significant (L = .64; F1,
2
146 = 82.63, p < .01, h = .36). Also the interaction between group and time produced an effect, with the prior knowledge
activation group scoring a higher improvement in reading comprehension skills than the reciprocal teaching group (L = .91;
F1, 146 = 14.64, p < .01, h2 = .09). The semantic inferences test scored a statistically significant interaction effect between
group and time (L = .88; F1, 118 = 15.53, p < .01, h2 = .12). Indeed, while the control group’s performances got worse, the
students in the prior knowledge activation group improved their performances at the end of the intervention. The prior
knowledge activation group improved more than the reciprocal teaching group also in the metacognition task (L = .96; F1,
2
147 = 5.83, p < .05, h = .04). Finally, the two groups did not differ in their performances in the lexical inferences test. Marginal
means for all analysis are reported in Table 3.

Table 2
Correlation matrix of dependant variables (Pearson’s r).

RC1 LI1 SI1 M1 RC2 LI2 SI2 M2

RC1 1
LI1 .58** 1
SI1 .55** .34** 1
M1 .37** .37** .41** 1
RC2 .33** .33** .28** .42** 1
LI2 .30** .31** .30** .33** .35** 1
SI2 .01 .05 .08 .09 .35** .22** 1
M2 .22** .24** .17 .39** .32** .21** .21* 1

Note. RC: reading comprehension; LI: lexical inferences; SI: semantic inferences; M: metacognition; 1: pre-test; 2: post-test.
* p < .05.
** p < .01
86 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88

Table 3
Marginal means for all dependant variables analyzed with the repeated-measures analysis of variance.

Dependent variable Pre-test assessment Post-test assessment

Prior knowledge activation group Control group Prior knowledge activation group Control group

Reading comprehension 5.88  1.77 5.89  1.98 7.90  1.35 6.72  1.43
Lexical inferences 7.04  1.85 6.83  1.63 6.78  1.58 6.70  1.40
Semantic inferences 6.33  2.09 6.60  1.65 6.53  1.55 5.08  1.83
Metacognition 9.23  2.04 8.56  1.98 10.01  1.24 8.54  1.63

5. Discussion

This study was designed to assess the efficacy of a reading comprehension intervention based on the activation of
relevant prior knowledge types with 7th and 8th grade students. Many reading comprehension interventions focus on
readers’ prior knowledge (Taboada & Guthrie, 2006), but the main innovation brought by this study is that prior knowledge
was not considered as a general stash of information, but as an organized multidimensional construct, in which the
knowledge of facts matter as much as the knowledge of meanings. Indeed, previous research (Cromley & Azevedo, 2007;
Tarchi, 2012) confirmed that the predictive effect of prior knowledge on reading comprehension is boosted when its
multidimensional nature is acknowledged, as well as when its indirect effect, as mediated by inferences and metacognition,
is included.
The first relevant data is that both groups, prior knowledge activation and reciprocal teaching, improved in reading
comprehension over time, as the performances in the post-test were significantly higher than the ones at the pre-test, with a
large effect size (36% of variance explained). This finding confirms a vast set of theories and data included in the socio-
constructivist approach (cooperative learning, Johnson & Johnson, 1975; community of learners, Brown & Campione, 1990;
community of practices, Lave & Wenger, 1991; Knowledge Building, Scardamalia & Bereiter, 2006). These pedagogical
approaches differ in many aspects, but they all share a focus on the learner’s active role in constructing knowledge by
inserting him or her in a learning environment that features the appropriate means and cultural artifacts, such as open-
ended or authentic questions (Soter et al., 2008). In the experimental interventions implemented in this study, the students
were at the center of the meaning-construction process, while the adult constructed and scaffolded a learning environment
that could guide the students when needed and shift the responsibility of the activity to them when they were able to
organize themselves. Reading strategies and activities, such as discussion, served as cultural artifacts, with the role of
supporting the internalization of a higher mental cognitive functioning.
However, this was the only variable in which the main effect of time was statistically significant. Indeed, data
substantially confirmed the research hypothesis, as the prior knowledge group outperformed the control group in almost all
the measures: reading comprehension, semantic inferences, and metacognition. Only the lexical inferences task did not
show any statistically significant difference at the post-test assessment. In addition to the main effect of time, the reading
comprehension task produced an interaction effect (group  time), with the prior knowledge activation group increasing
their performances at a higher rate than the reciprocal teaching group. These results confirmed the fundamental importance
of activating different prior knowledge types before and during the reading comprehension task. The students in the
experimental group learned how to activate relevant background information in terms of facts and meanings and apply them
to information extracted from the text. This was an important step in this study, as it represents a substantial move from a
unidimensional perspective on prior knowledge to a multidimensional one. Such a shift is well present in scientific literature
aiming at verifying predictors of reading comprehension (Alexander & Jetton, 2000), but no evidence could be found in
studies aiming at verifying the efficacy of reading comprehension intervention.
This focus on multiple dimensions of prior knowledge helped students also to improve their semantic inferences skills
(12% of variance explained by the differences in groups), but not the lexical ones. Many studies underlined the fundamental
importance of making inferences according to information provided in the text (Rapp et al., 2007). Previous studies brought
evidence of the mediating role that semantic inferences have on the relationship between prior knowledge and reading
comprehension (Tarchi, 2010). Also, for semantic inferences to be effective, first relevant prior knowledge must be activated
or, if absent, must be acquired from peers through a discussion. Data of this study support Tressoldi and Zamperlin’s (2007)
claim to keep distinct lexical from semantic inferences: the activation and sharing of prior knowledge allowed readers to
better infer missing information in the text by drawing from their background knowledge, whereas their ability to select the
appropriate meaning of a word to the context did not seem to benefit. Indeed, the first discussion prompted by an open
question, and the second discussion prompted by a quick reading of the text title and headings, supported the activation of
general background knowledge. According to these findings, it seems more effective to help students to gain awareness of
their own and each other’s pre-conceptions (‘‘semantic’’ ‘‘knowledge’’), instead of having a contextualized lexical jargon.
The two group discussions offered the possibility to have a metacognitive discussion on the approach to the text they
were reading, which consequently had a positive effect on their overall metacognitive approach to reading. It is important to
notice that also the reciprocal teaching methodology aims at enhancing students’ metacognitive approach to the text, and it
does so through the joint application of reading strategy. According to our data, discussing how our prior knowledge relates
 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88 87

to the text has a stronger effect on metacognition than discussing reading strategies does. Probably this happens because
sharing prior knowledge among students allows them to create and negotiate a situational model of the text and better
control the overall reading comprehension process. As the effect size for this measure was low (4% of variance explained),
data have to be interpreted with caution. In conclusion, as we are living in the so-called ‘‘knowledge age,’’ students’
interaction with written texts is becoming an increasingly urgent matter. The evidence-based potentiality of the prior
knowledge activation intervention in fostering deep elaboration and constructive use of written information should be a
criterion to be used to revise curricula in the Italian school system. Furthermore, these practices, specifically addressing deep
processing of sources of information, should be included in teachers’ training, in the context of an educational psychology
that is specifically characterized by a methodological strictness and, therefore, a strong validity.
This study had the following limitations and guidelines for future research. The prior knowledge activation intervention
included three main components, which could play a potentially confounding role: the initial question, the repetitive use of
discussions, and the implementation of the highlighting strategy. The main hypothesis of this study was that the combined
effect of these components would activate and improve reader’s prior knowledge, thus enhancing their comprehension of a
text. Results from this study could also depend on the unique contribution of one or more of these components. First, the
initial question was designed to activate readers’ prior knowledge before beginning to read. However, the initial question
could also play a role in activating reader’s interest, which plays a fundamental role in reading comprehension (see Boscolo &
Mason, 2004). Second, group discussion could have contributed to reading comprehension in quantity, besides than they did
in quality, i.e., students might have benefitted from discussing the same material several times, in addition to discussing it
from different perspectives. Murphy, Wilkinson, Soter, & Hennessey’s (2009) meta-analysis of the effects of classroom
discussion on students’ comprehension of text demonstrated that group discussion does not necessarily determines an
improvement in reading comprehension. Indeed, they found stronger effects of group discussion on literal and inferential
comprehension than they did for critical and reflective comprehension. Third, highlighting, as well as underlining, is a
cognitive strategy able to promote active learning per se (see Gersten, Fuchs, Willimas and Baker’s review of research on
teaching reading comprehension strategies, 2001). Future studies should replicate this study with different control groups in
order to better isolate the unique contribution of activating specific prior knowledge types.

References

Alexander, P. A., & Jetton, T. L. (2000). Learning from text: A multidimensional and developmental perspective. In M. L. Kamil, P. B. Mosenthal, P. D. Pearson, & R.
Barr (Eds.), Handbook of reading research (vol. III, pp. 285–310). Mahwah, NJ: Erlbaum.
Alvermann, D. E., Fitzgerald, J., & Simpson, M. (2006). Teaching and learning in reading. In P. A. Alexander & P. H. Winne (Eds.), Handbook of educational psychology
(2nd ed., pp. 427–455). Mahwah, NJ: Lawrence Erlbaum Associates.
Armand, F. (2001). Learning from expository texts: Effects of the interaction of prior knowledge and text structure on responses to different question types.
European Journal of Psychology of Education, 16(1), 67–86. http://dx.doi.org/10.1007/BF03172995
Ausubel, D. P. (1960). The use of advanced organizers in the learning and retention of meaningful material. Journal of Educational Psychology, 51, 267–272. http://
dx.doi.org/10.1037/h0046669
Best, R. M., Floyd, R. G., & McNamara, D. S. (2008). Differential competencies contributing to children’s comprehension of narrative and expository texts. Reading
Psychology, 29, 137–164. http://dx.doi.org/10.1080/02702710801963951
Best, R. M., Rowe, M., Ozuru, Y., & McNamara, D. S. (2005). Deep-level comprehension of science texts: The role of the reader and the text. Topics in Language
Disorders, 25(1), 65–83. http://dx.doi.org/10.1097/00011363-200501000-00007
Billett, S. (1996). Situated learning: Bridging sociocultural and cognitive theorizing. Learning and Instruction, 6(3), 263–280. http://dx.doi.org/10.1016/0959-
4752(96)00006-0
Boscolo, P. (1997). Psicologia dell’ Apprendimento Scolastico: Aspetti Cognitivi e Motivazionali. [Psychology of school learning: Cognitive and motivational aspects].
Torino, IT: Utet Libreria.
Boscolo, P., & Mason, L. (2004). Topic knowledge, text coherence, and interest: How they interact in learning from instructional texts. Journal of Experimental
Education, 71, 126–148. http://dx.doi.org/10.1080/00220970309602060
Brown, A. L. (1978). Knowing when where and how to remember: A problem of metacognition. In R. Glaser (Ed.), Advances in instructional psychology (vol. 1, pp.
77–165). Hillsdale: Erlbaum.
Brown, A. L., & Campione, J. C. (1990). Communities of learning and thinking, or a context by any other name. Contributions to Human Development, 21, 108–126.
Cornoldi, C., & Colpo, G. (1995). Nuove Prove MT per la Scuola Media Inferiore. [New MT tests for secondary school]. Firenze, IT: Giunti Organizzazioni Speciali.
Cromley, J. G., & Azevedo, R. (2007). Testing and refining the direct and inferential mediation model of reading comprehension. Journal of Educational Psychology,
99(2), 311–325. http://dx.doi.org/10.1037/0022-0663.99.2.311
Darch, C., & Gersten, R. (1986). Direction-setting activities in reading comprehension: A comparison of two approaches. Learning Disability Quarterly, 9, 235–243.
http://dx.doi.org/10.2307/1510469
Dochy, F. (1994). Prior knowledge and learning. In T. Husen & T. N. Postlethwaite (Eds.), International encyclopedia of education (2nd ed., pp. 4698–4702). Oxford/
New York, NY: Pergamon Press.
Ganzeboom, H. B. G., De Graaf, P., Treiman, D. J., & De Leeuw, J. (1992). A standard international socio-economic index of occupational status. Social Science
Research, 21(1), 1–56. http://dx.doi.org/10.1016/0049-089X(92)90017-B
Gersten, R., Fuchs, L. S., Willimas, J. P., & Baker, S. (2001). Teaching reading comprehension strategies to students with learning disabilities: A review of research.
Review of Educational Research, 71, 279–320. http://dx.doi.org/10.3102/00346543071002279
Gurlitt, J., & Renkl, A. (2010). Prior knowledge activation: How different concept mapping tasks lead to substantial differences in cognitive processes, learning
outcomes, and perceived self-efficacy. Instructional Science, 38, 417–433. http://dx.doi.org/10.1007/s11251-008-9090-5
Hailikari, T., Nevgi, A., & Lindblom-Ylänne, S. (2007). Exploring alternative ways of assessing prior knowledge, its components and their relation to student
achievement: A mathematics-based case study. Studies in Educational Evaluation, 33, 320–337. http://dx.doi.org/10.1016/j.stueduc.2007.07.007
Hailikari, T., Katajavuori, N., & Lindblom-Ylanne, S. (2008). The relevance of prior knowledge in learning and instructional design. American Journal of
Pharmaceutical Education, 72(5), 1–8. http://dx.doi.org/10.5688/aj7205113
Johnson, D. W., & Johnson, R. (1975). Learning together and alone: Cooperation, competition and individualization. Englewood Cliffs, US: Prentice-Hall.
Lave, J., & Wenger, E. (1991). Situated learning—Legitimate peripheral participation. Cambridge, US: Cambridge University Press.
Moreno, R. (2009). Constructing knowledge with an agent-based instructional program: A comparison of cooperative and individual meaning. Learning and
Instruction, 19, 433–444. http://dx.doi.org/10.1016/j.learninstruc.2009.02.018
88 C. Tarchi / International Journal of Educational Research 72 (2015) 80–88

Murphy, P. K., Wilkinson, I. A. G., Soter, A. O., Hennessey, M. N., & Alexander, J. F. (2009). Examining the effects of classroom discussion on students’ comprehension
of text: A meta-analysis. Journal of Educational Psychology, 101, 740–764. http://dx.doi.org/10.1037/a0015576
Oakhill., J. V. (1983). Instantiation in skilled and less skilled comprehenders. Quarterly Journal of Experimental Psychology, 35, 441–450. http://dx.doi.org/10.1080/
14640748308402481
Palincsar, A. S., & Brown, A. L. (1984). Reciprocal teaching of comprehension-fostering and comprehension-monitoring activities. Cognition and Instruction, 1,
117–175. http://dx.doi.org/10.1207/s1532690xci0102_1
Pazzaglia, F., De Beni, R., & Cristante, F. (1994). Prova di metacomprensione—Valutazione dell’abilità strategica, della sensibilità al testo e dell’attività di controllo
implicate nella comprensione della lettura. [Metacomprehension test—Assessment of strategic ability, text sensitivity, and control of activities involved in
reading comprehension]. Firenze, IT: Giunti Organizzazioni Speciali.
Pressley, M., Almasi, J., Schuder, T., Bergman, J., Hite, S., El-Dinary, P. B., et al. (1994). Transactional instruction of comprehension strategies: The Montgomery
County, Maryland, SAIL Program. Reading and Writing Quarterly: Overcoming Learning Difficulties, 10, 5–19. http://dx.doi.org/10.1080/1057356940100102
Rapp, D. N., & Kendeou, P. (2007). Revising what readers know: Updating text representations during narrative comprehension. Memory & Cognition, 35(8),
2019–2032. http://dx.doi.org/10.3758/BF03192934
Rapp, D. N., van den Broek, P., McMaster, K. L., Kendeou, P., & Espin, C. A. (2007). Higher-order comprehension processes in struggling readers: A perspective for
research and intervention. Scientific Studies of Reading, 11, 289–312. http://dx.doi.org/10.1080/10888430701530417
Scardamalia, M., & Bereiter, C. (2006). Knowledge building: Theory, pedagogy, and technology. In K. Sawyer (Ed.), Cambridge handbook of the learning sciences (pp.
97–118). New York, NY: Cambridge University Press.
Simmonds, E. P. M. (1992). The effects of teacher training and implementation of two methods of improving the comprehension skills of students with learning
disabilities. Learning Disabilities Research and Practice, 7, 194–198.
Soter, A. O., Wilkinson, I. A., Murphy, K., Rudge, L., Reninger, K., & Edwards, M. (2008). What the discourse tells us: Talk and indicators of high-level comprehension.
International Journal of Educational Research, 47, 372–391. http://dx.doi.org/10.1016/j.ijer.2009.01.001
Taboada, A., & Guthrie, J. T. (2006). Contributions of student questioning and prior knowledge to construction of knowledge from reading information text. Journal
of Literacy Research, 38, 1–35. http://dx.doi.org/10.1207/s15548430jlr3801_1
Tarchi, C. (2010). Reading comprehension of informative texts in secondary school: A focus on direct and indirect effects of reader’s prior knowledge. Learning and
Individual Differences, 20, 415–420. http://dx.doi.org/10.1016/j.lindif.2010.04.002
Tarchi, C. (2012). La comprensione del testo espositivo. L’interazione tra conoscenze precedenti del lettore e contenuto del testo. Giornale Italiano di Psicologia, 1,
91–114. http://dx.doi.org/10.1421/37064
Tressoldi, P., & Zamperlin, C. (2007). La valutazione della comprensione del testo [Assessment of reading comprehension]. Psicologia Clinica dello Sviluppo, 11,
271–290. http://dx.doi.org/10.1449/24878
van den Broek, P., Lorch, R. F., Jr., Linderholm, T., & Gustafson, M. (2001). The effects of readers’ goals on inference generation and memory for texts. Memory &
Cognition, 29, 1081–1087. http://dx.doi.org/10.3758/BF03206376