Introduction
After decades of neglect (Meara, 1980),
acquisition of vocabulary, a critical
component of L2 proficiency (Hafiz &
Tudor, 1990; Horst, Cobb, & Meara, 1998;
Pitts, White, & Krashen, 1989), has received
perceivable attention in the field of second
language acquisition research. This growth
of interest in the domain of vocabulary
development centers on different themes
including incidental versus intentional
vocabulary learning (Ellis & He, 1999).
Despite a bulk of research existing on L2
incidental word learning (e.g., Brown,
Waring, & Donkaewbua, 2008; Cain,
Oakhill, & Bryant, 2004; Daskalovska,
2011; Day, Omura, & Hiramatsu, 1991),
relatively few researchers have explored the
impact of reading purpose on incidentally
acquiring vocabulary knowledge. It is not
yet clear how the variability of reasons
behind reading activity might affect the
amount and kind of vocabulary knowledge
the learners acquire.
A core theoretical assumption behind
constructivism is the centrality of reader’s
purposes or goals in the learning situations
(Cambourne, 2002). Reading researchers see
reading an intentional act (Kulikowich &
Alexander, 2010) during which strategic
readers process text to meet their reading
goals (Burton & Daneman, 2007; Rapp &
The effect of reading purpose on incidental vocabulary learning 65
Kendeou, 2007). In the realm of vocabulary
development through reading, empirical
evidence has proved direct relationship
between the readers’ cognitive processing of
texts and the particular purpose behind the
reading activity which in turn influences the
rate and the amount of information retrieval
and recall (Cerdán & Vidal-Abarca, 2008;
McCrudden, Maglianob, & Schraw, 2010).
van den Broek, Lorch, Linderholm, and
Gustafson (2001), while assigning college-aged participants to a read for entertainment
purposes condition or to a read for study
purposes condition, found that the type of
inferences generated during reading was
greatly influenced by readers’ goals.
Whereas the former group generated more
free associations and more evaluative
comments on the writing or interest value of
the text, readers of the latter group generated
more coherence-building inferences which
in turn resulted in a better information
retrieval and retention.
A study by Linderholm and van den Broek,
(2002) examined the extent to which low-
and high-WMC (working memory capacity)
readers alter their cognitive processes to fit
the reading purpose under the same two
conditions of entertainment or study. One of
the results was an indication of low-WMC
readers’ least demanding processes in the
reading for the study group. The results also
indicated that all readers adjusted cognitive
processes to fit the reading purpose.
Linderholm and Wilde (2010), while
investigating college students’ beliefs about
comprehension when reading for different
purposes, found that the students’ actual test
performance did not differ between readings
for entertainment or study purposes. They
found that although readers engaged in
different strategies when reading in a study
purpose situation versus an entertainment-focused reading situation, the rate of
information recall did not seem to be
significant for the two groups.
Bråten and Samuelstuen (2004) reported a
direct relationship between students’ level of
topic knowledge and the effect of reading
purpose on reported use of memorization
and elaboration strategies just for readers
who read the text for the purpose of
discussing text content and not for those
who read for the purposes of test taking or
summary writing.
Smith (1967) found that while both good
and poor readers read for the two purposes
of reading for details and reading for
general impressions with equal success,
good readers could make more adjustments
to reading purpose than poor readers.
Swanborn and de Glopper’s study (2002)
revealed significant differences between the
three groups of free reading, reading for text
comprehension, and reading to learn about
the topic among the students of Grade 6.
Their study also indicated that good readers
were more successful in incidental word
learning.
The previously cited works mostly explored
the effect of learners’ reading purpose on
incidental vocabulary gain while the learners
just received input tasks. However, what
distinguishes the present study from the
prior ones is its examination of the effect of
reading purpose in both input and written
output tasks simultaneously. Moreover, this
study addresses the learning and retention of
elementary level learners in an EFL context
where reading is usually the chief source of
receiving language input, while the
aforementioned studies did not focus on
these learners in this special kind of context.
Research questions
This study explored learners’ incidental
vocabulary learning and retention in an EFL
context (i.e., elementary Iranian learners)
under three conditions: a) learners who read
the text for comprehension, b) learners who
had a free reading of the text, and c) learners
who read and summarized the text. Two
broad research questions guided the study:
1. Does reading purpose have any
impacts on incidental vocabulary
acquisition?
2. Does reading purpose have any
impacts on incidental vocabulary
retention?
Method
Participants
The participants were 75 female high school
fourth graders with the age range of 17-19
years. To select these participants, the
Oxford Placement Test (reference) was
administered to 99 students who had learned
English only in the formal system of
education at schools and had no experience
of taking part in language learning classes or
institutes. The results of the placement test
showed that the majority of participants
(n=75) were at the elementary level who
were randomly assigned to three groups of
free reading, reading comprehension and
reading to summarize.
Text and the tests
The input text ‘A good night’s work’ was
selected from a graded reader series Reading
Comprehension 4 by Louis Fidge. Eight
readability formulae scored the text as ‘very
easy to read’ and suitable for grade 4.
Nearly all difficult words were put into a
questionnaire and administered to ten
experienced English teachers to check the
degree of familiarity of learners with the
words. Twenty words which were labeled by
the English teachers as ‘the students have
never seen the word’ were chosen for the
study. To engage the learners in noticing,
and to assist incidental vocabulary learning,
the text was enhanced by bolding and
Persian glossing of all difficult words. The
post-test questionnaire was a mixture of both
Vocabulary Knowledge Scale (Paribakht &
Wesche, cited in Paribakht, 2005) and its
modified version by Brown (2008). As
mentioned previously, twenty words which
the teachers believed the learners were never
familiar with, were selected for the test. The
test intended to evaluate the rate of
vocabulary learning as the byproduct of
reading for different purposes. The same test
but with a different order of items was used
as the delayed post-test for checking gains
of retention rate after two weeks.
Procedures
The students were divided into three groups:
a) reading comprehension, b) free reading
and c) summarizing. In order to determine
the level of the participants, an Oxford
Placement Test was administered. Based on
the tests’ interpretations, the three groups’
level was determined as elementary. The
normality of the scores of the three groups
was also checked through different
procedures on SPSS. Performing a one-way
ANOVA, it was ensured that no significant
differences existed between the three groups
prior to the treatment.
Next, the text was administered to the three
groups of learners. The learners were not
forewarned about the two vocabulary gain
tests that were to be administered shortly
afterward and with a time interval of two
weeks. One group was asked to read the
enhanced text as they liked. The reading for
comprehension group was asked to read the
same text and answer the questions which
followed the text. It is worth mentioning that
the questions were only in multiple-choice
and true/false formats. No essay questions
were given to the learners so that the effect
of writing and generation would be more
distinctive for the summarizing group. The
summarizing group read the text and
summarized the story on the blank sheets
which were available to them. Immediately
after finishing the reading phase, the
vocabulary test was administered to the
learners. The test was made up of twenty
vocabularies (six nouns, thirteen verbs and
one adjective) from the bolded and glossed
words. The students were tested both for
recognition and production of words. Using
the target word in the sentence with the true
part of speech was the most preferred
situation.
In scoring the test, choice A (I know what
this word/phrase means and I can use it in a
sentence) received a value of 3, choice B
received a value 2 if the correct meaning of
the word was given by the learner and a
value of 1 if the meaning was not correct.
Choice C was given a value of 1 and choice
D was given a value of 0. Since word
learning is a gradual and incremental
process, a second test was administered to
ensure the retention of the learned words.
The participants’ vocabulary retention gains
were tested by the same test but with a
different order within a time interval of two
weeks.
Results
Having obtained the immediate posttest
data, they were first screened and extreme
scores were discarded. Table 1 shows the
descriptive statistics for the screened data of
immediate (learning) post-test.
FR: Free Reading, RC: Reading Comprehension,
Sum: Summarizing, S: Skewness, SE: Standard Error,
SR: Skewness ratio
Ensuring the normality of the data through
Shapiro-wilk test and other procedures on
SPSS, the parametric test of one-way
ANOVA (Table 2) was used to test H01
which states reading purpose has no
statistically significant effect on incidental
vocabulary learning.
The probability level of the ANOVA in
Table 2 rejects H01 (P<.05). That is, the
ANOVA test shows that reading purpose
had a statistically significant effect on
incidental vocabulary learning. The effect
size or strength of association of 0.17,
according to Dörnyei (2005), indicates a
large effect size which means that 17% of
the between group variance is due to the
difference in the reading purpose.
The ANOVA test showed the significant
effect of reading purpose. Therefore, to
understand where exactly the difference
existed, a post hoc test of Scheffe was
performed
As the data in Table 3 show, free reading
group is significantly different from both
reading comprehension and summarizing
groups (P<.05). However, the difference
between reading comprehension and
summarizing groups is not significant
(P>.05).
The same procedure followed for immediate
posttest was followed to ensure the
appropriateness and normality of the data.
That is, the data were explored to discard
possible outliers. Table 4 shows the
descriptive statistics for the screened data of
the delayed (retention) posttest.
Ensuring the normality of the data through
Shapiro-wilk test on SPSS, the parametric
test of one-way ANOVA (Table 2) was used
to test H02. Table 5 displays the result of the
One-way ANOVA to test the second null
hypothesis.
The probability level of the ANOVA in
Table 5 rejects H022 (P<.05). That is, the
ANOVA test shows that reading purpose
had a statistically significant effect on
incidental vocabulary retention. The
magnitude of partial eta squared in Table 5
shows the strength of association or the
effect size is .40 which is a much greater
than what Dörnyei (2005) regard a large
effect size. The unadjusted effect sizes for
the study are also .68 for the delayed and .21
for the immediate posttest.
In order to understand where exactly the
difference between groups existed, a post
hoc test of Scheffe was performed
As the data in Table 6 show, free reading
group is significantly different from both
reading comprehension and summarizing
groups (P<.05). However, the difference
between reading comprehension and
summarizing groups is not significant
(P>.05).
Discussion
The first research question addressed the
rate of vocabulary learning based on the
reading purpose. As it was mentioned in the
result section, the findings of the study did
reveal statistically significant differences
between the three groups (Table 2). In other
words, the result of this study did indicate
some impacts of reading purpose on
incidental vocabulary learning. Concerning
the second research question which
investigated whether the learners’ reading
purpose had any impacts on their lexical
retention, the result of the study also
indicated a statistically significant effect
(Table 5).
As the results of Scheffe tests (Tables 3 & 6)
in the result section indicated the two groups
of reading comprehension and summarizing
yielded a better outcome than free reading in
both immediate and delayed posttests. There
are several reasons which might have led to
this outcome. This outperformance can be
due to deeper processing (Craik & Lockhart,
1972) of the text, the more involvement load
(Laufer & Hulstijn, 2001) and the more
amount of time which the participants in
these two groups spent while reading and
answering the post-reading-comprehension
questions and summarizing. The cognitive
psychologist, Eysenck (cited in Mahdavy,
2011), agrees that processing new lexical
information more elaborately results in a
better retention than processing it less
elaborately.
It seemed that the factor of need also played
an important role here. Two groups of
reading comprehension and summarizing
had to re-read the text in order to answer the
comprehension questions, or summarize the
text. They had to re-read to confirm
understandings, and to clarify details, what
free reading group felt no need for. The
summarizing group had the opportunity to
fill the gaps and the holes in their
interlanguage (IL) while reconstructing the
text (Swain & Lapkin, 1995) and generating
sentences. Besides, while answering
comprehension questions learners of reading
comprehension group might have had to
guess the meaning of new words from
context or their background knowledge;
what subjects in free reading group seemed
not to have the chance for, neither might
they have felt the need to do so because they
know they were reading for their own
enjoyment. As noted above, this elaboration
probably increased the chances that the word
and its meaning would be available for use
at a later time. In a nutshell, in both
immediate and delayed post-tests, free
reading group had the lowest vocabulary
acquisition rate which is in line with
Swanborn, and de Glopper’s (2002) finding.
Contrary to expectations, summarizing did
not lead to higher incidental word learning
gains than reading for comprehension.
While this finding might lend partial support
to Laufer and Hulstijn’s (2001) involvement
load hypothesis (i.e. the effectiveness of a
task is determined by the involvement load
it induces irrespective of whether the task is
input or output oriented), it might also be
due to the fact that, in the Iranian
educational system, students are mostly
familiar and somehow skillful in reading for
comprehension and weak in summarizing.
However, the study did indicate a
considerable difference between the
recognition and production scores of the
summarizing group and the other two groups
of free reading and reading for
comprehension group in both posttests.
This group outperformed the other two
groups in the production scores. This group
also used acquired vocabularies within more
complex and longer sentences. It seems that
because of the effect of writing and
production this group had a better chance
and bravery to produce and generate
sentences instead of just giving the Persian
translation of the word, a finding which is in
line with Griffin and Harley (1996),
Mondria and Wiersma (2004), and Waring
(1997) who suggest receptive learning is
more effective in contributing to receptive
knowledge, whereas productive learning
may yield better outcome in increasing
productive knowledge, but it contrasts with
Hashemi Shahrakia, and Kassaian’s (2011)
finding. It seems that productive learning
due to the deeper processing brings a higher
awareness of vocabulary use to the learners
of the language. Following Brown and
Payne (1994) who see conversion of
Applied Research on English Language: 3(1) 61
receptive vocabulary into productive
vocabulary as the final stage of vocabulary
learning, one can feel the importance of a
writing course and more specifically
summarizing in an educational setting.
Furthermore, the study has also found that it
is more difficult for foreign language
learners to develop productive vocabulary
than receptive one.
The highest rate of incidental word learning
was found for the word ‘detective’. This
could be attributed to the higher frequency
of this particular word form as compared to
those of the others, suggesting that the
frequency of a target feature (Tekmen &
Daloglu, 2006) in L2 input could play a
crucial role in learners’ attentional
allocation. In other words the probability of
learning a word from context increases
substantially with additional occurrences of
the word.
Apart from the word ‘detective’, words such
as ‘handcuffs’, ‘torch’, and ‘two-way-radio’
are among the most learned and retained
vocabularies. One possible reason might be
the place and position of these nouns in the
text. These words were introduced in the
first lines of the passage, where the mind is
still fresh. Therefore, as VanPatten,
Williams, and Rott (2004) argue there are
some input features and learner factors
which determine noticing and recalling of a
specific word form. Input factors include the
salience of the target form (Schmidt, 2001);
the ratio of known to unknown words
(Laufer, 1992); meaningfulness and
communicative value of the word
(VanPatten, 1990); the quality of
information processing (Laufer & Hulstijn,
2001); the number of occurrences of the
unknown word, the importance of the
unknown word to text comprehension
(Paribakht, 2005), and the importance and
significance of the word to the learner
herself; and learner factors include learners’
language proficiency; their communicative
need (Williams, 2001); and individual
interests and differences (Ellis, 2004) both
in terms of “abilities” and “propensities”
(i.e. learning style, motivation, anxiety,
personality, ..).
This study led to the discovery, however,
that many errors are not traceable to the
structure of the first language, but are the
result of intralingual factors, such as
differentiating the correct forms of the
words. The word ‘fair’ has been wrongly
translated and used 32 times in the
immediate post-test. This large proportion of
wrong answers was due to students
misunderstanding of the word form. This
word was mistaken for 18 times with the
word ‘afraid’, 10 with ‘failure’, three with
‘fire’, and one with ‘near’. None of these
errors shows any traces of learners’ first
language, i.e. Persian language. First
language influence just appeared to be
strongest in complex word order and in
word-for-word translations of phrases such
as ‘two-way-radio’ in two groups of reading
comprehension and free reading. The
summarizing group mostly used the exact
text phrases or sentences with greatest
accuracy in grammaticality of sentences.
Nevertheless, the large effect sizes of .21
and .68 for the immediate and delayed
posttests denote that purposeful reading led
to one-fifth and more than one-half of a
standard deviation improvement in outcome,
respectively. These figures also denote that
59% and 73% of the control group would be
below average person in experimental
group, correspondingly (Coe, 2002).These
large effect sizes of both immediate and
delayed post-tests extremely suggest the
impact of having and setting a purpose of
reading on incidental vocabulary learning
and retention.
Conclusion
Theoretically, the findings of the present
study generally provide positive support of
the role of readers’ purpose on incidental
gain of vocabulary knowledge. Creating
purpose in the classroom reading situation
will heighten motivation and enhance
readers’ interest and performance.
Furthermore, having a purpose for reading
will assist students to choose the most
appropriate method of reading. In practice, it
means by providing stimulating and
appealing reading tasks and materials,
educationalists and course book designers
can increase students’ motivation and
interest. They can push students toward
autonomous and self-regulated learning, and
make them better metacognitive strategic
readers. There seems to be an urgent need
for including pre-reading activities with
authentic texts or other reading selections,
and for adequately cueing readers to the
purpose (i.e. given intentions, Graesser,
Singer, & Trabasso, 1994; van den Broek,
Risden, & Husebye-Hartmann, 1995) for
reading a particular text in order to better
assist the learners’ mind activation.
Moreover, the findings of the study imply
that successful teachers can ask students
some carefully-selected questions before
approaching the reading text so that students
are urged to think, talk, or even argue, and
finally answer. Teachers need to remind
students of the importance of the big
question of "why do I read this text?"
Furthermore, the low rate of incidental
vocabulary learning and retention in English
as a foreign language context (in the best
situation, the mean of the reading
comprehension learning and summarizing
groups are 25 and 24 out of total score of
60) extremely suggests the explicit teaching
of new lexicon on the part of teachers along
with encouraging the learners toward
extensive reading. Narrow reading
(Krashen, 2004) as well as spaced repetition
and exposure of the new lexicon can be of
significance in committing words to
memory. Summarizing and writing tasks (e.
g. reading logs, diaries and portfolios) can
equip students with a deeper level of text
processing and as a result, a more durable
and long-lasting learning.
Although these results are promising, and
they cast a new light on the question of
incidental teaching effectiveness in the
context of L2 acquisition, considering the
situated nature of L2 learning, it would be
intriguing to see whether similar results are
obtained in different situations
complemented by a more comprehensive
examination of the TL input and with more
students involved. It would also be crucial
for future studies to examine how individual
differences interact with reading goals to
influence the rate of vocabulary learning.
Finally, it should be noted that since the
participants in the study were female
elementary language learners, the results
may face problems of generalizability.
