730596

research-article20172017
SGOXXX10.1177/2158244017730596SAGE OpenTahmasbi and Farvardin

SAGE Open - Original Manuscript

SAGE Open

Probing the Effects of Task Types on

July-September 2017: 1­–10
© The Author(s) 2017
DOI: 10.1177/2158244017730596
https://doi.org/10.1177/2158244017730596

EFL Learners’ Receptive and Productive journals.sagepub.com/home/sgo

Vocabulary Knowledge: The Case of

Involvement Load Hypothesis

Maryam Tahmasbi1,2 and Mohammad Taghi Farvardin2

Abstract
This study examined the effects of task types on English as a foreign language (EFL) learners’ receptive and productive
vocabulary knowledge. To this end, 130 (70 female and 60 male) EFL learners were randomly assigned to one of six tasks of
learning 30 target words. The design of the tasks was based on the involvement load hypothesis (ILH) arguing that learning of
unfamiliar words to be contingent on the amount of task induced involvement. The components of involvement in ILH include
need (N), search (S), and evaluation (E). In this study, the tasks induced the same or different involvement loads regarding the
presence and strength of each component: paragraph writing (+N, +S, ++E), sentence writing (+N, +S, +E), combining (+N,
–S, +E), fill in the blank (+N, –S, +E), translation (+N, –S, +E), and control (–N, –S, –E). After the last treatment session, both
receptive and productive knowledge of the target words were measured. Moreover, a delayed posttest was administered 1
month later. The results revealed that all output tasks were more effective than the control task in enhancing the participants’
receptive and productive vocabulary knowledge. Moreover, paragraph writing task was found to be the most effective task.

Keywords
EFL learners, involvement load hypothesis, output task, receptive vocabulary knowledge, productive vocabulary knowledge

Introduction development in field of L2 vocabulary research, has received

Vocabulary is viewed as the essential building block of sec-
ond language (L2) learning (McCarty, 2005). Vocabulary
knowledge is usually classified into productive and receptive
knowledge (Nation & Meara, 2002). Productive vocabulary
knowledge is referred to the learners’ ability to comprehend
something they hear or read and express their ideas by using
proper vocabulary through writing or speaking (Nation,
2003). Receptive vocabulary knowledge, however, is
referred to recalling the words through listening and reading
(Schmitt, 2000).
As vocabulary knowledge plays a crucial role in L2 learn-
ing, it is necessary for teachers to select and use proper
vocabulary learning tasks (Bao, 2015). Moreover, language
researchers should address the reasons for superiority of
some tasks over others in L2 vocabulary learning. In this
line, it has been argued that learning and retention of words
in an L2 depend on the extent of involvement with a task,
which Laufer and Hulstijn (2001) called the involvement
load hypothesis (ILH).
The ILH argues that the greater demand a vocabulary task
places on an L2 learner, the more likely the target words will
be acquired (Laufer & Hulstijn, 2001). The ILH, as a major
a great deal of attention as it is clear, precise, and can be
operationalized (Bao, 2015; Keating, 2008; Tang & Treffers-
Daller, 2016; Zou, 2017). The ILH consists of three compo-
nents: need, search, and evaluation. Need is a noncognitive
but motivational factor. Search is the relationship between
form and meaning of unknown words. Evaluation includes
making a decision on the appropriate word with its related
meaning in context. These elements are authorized regarding
their distinction. If a component is absent (–) the score is 0.
If the component is moderate (+), the score is 1 and the com-
ponent gains 2 if the involvement is strong. It is believed that
the best result in learning new vocabulary is obtained through
a task with the highest degree of involvement load (Marmol
& Sanchez-Lafunte, 2013).
1
Department of ELT, Khouzestan Science and Research Branch, Islamic
Azad University, Ahvaz, Iran
²Department of ELT, Ahvaz Branch, Islamic Azad University, Ahvaz, Iran
Corresponding Author:
Mohammad Taghi Farvardin, Department of English Language Teaching,
Ahvaz Branch, Islamic Azad University, Ahvaz 6134937333, Iran.
Email: farvardin@iauahvaz.ac.ir

Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License
(http://www.creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of
the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages
(https://us.sagepub.com/en-us/nam/open-access-at-sage).
2 SAGE Open
There are two reasons behind this study. First, as the ILH
claims, the presence of an involvement component leads to
more vocabulary learning than its absence. However, little
research has been done to examine this issue (e.g., Bao,
2015). Second, the ILH proposes that the same amount of
word learning is obtained when there is the same involve-
ment component regardless of type of vocabulary learning
task. However, little research has been conducted to probe
this issue, either (i.e., Bao, 2015; Kim, 2008). Therefore, this
study examined the effect of different output tasks on English
as a foreign language (EFL) learners’ vocabulary
knowledge.
Literature Review
The ILH argues that tasks with three constructs of need (N),
search (S), and evaluation (E) have more effectiveness on
vocabulary learning than tasks with lower involvement loads
(Laufer & Hulstijn, 2001). Need is a motivational construct
dealing with the “need to achieve” (Laufer & Hulstijn, 2001,
p. 14), while search and evaluation are cognitive constructs
having to do with paying attention to form–meaning
relationship.
Tasks may induce these involvement elements to three
possible degrees: none, moderate, and strong. A task involve-
ment load is referred to as the combination of these involve-
ment elements, which can be absent or present, moderate or
strong (Bao, 2015). A moderate involvement is given an
index of 1 and a strong involvement receives an index of 2.
It is argued that “the higher the scores of need, search, and
evaluation are, the greater the involvement load in learning
an unknown word is” (Bao, 2015, p. 85). In assessing the
ILH, L2 reading passages have been used as stimuli in most
previous studies, following Laufer and Hulstijn (2001).
Laufer and Hulstijn (2001) compared three incidental
vocabulary learning tasks: reading with marginal glosses
with index of 1 (+N, –S, –E), fill in the blank with index of 2
(+N, –S, +E), and composition writing with index of 3 (+N,
–S, ++E) with no time controlling. The study investigated the
effect of task types on the retention of 10 target words by
EFL learners. The participants were 97 advanced EFL uni-
versity learners in the Netherlands and 128 in Israel. They
were six intact groups consisting of three parallel groups in
the Netherlands and three parallel groups in Israel. The find-
ings showed that writing task had a better result than two
others. Likewise, reading plus fill-in-the-blank task was
more effective than reading task. This experiment provided
strong support for the ILH. Therefore, it was claimed that
retention of unknown words depends on the amount of
involving learners with tasks including different degrees of
need, search, and evaluation. However, their study lacked a
control group and there was no control for word type.
Moreover, substantiating or rejecting the ILH to some extent
relies on the design of tasks and handling the task time
(Webb, 2005). It should be also noted that the type of
vocabulary knowledge can reject or accept the ILH (Keating,
2008). Most follow-up studies have provided inconclusive
evidence for the ILH.
Keating (2008), for example, investigated the effects of
three tasks with different involvement loads on vocabulary
retention of 79 Spanish learners. To this end, three tasks with
different involvement loads were selected: reading compre-
hension task (involvement load of 1), reading plus fill-in task
(involvement load of 2), and sentence writing task (involve-
ment load of 3). The results showed that the participants
could learn more words in sentence writing task with the
highest involvement load. Nevertheless, Keating (2008) did
not include any control group in his study.
In another study, Marmol and Sanchez-Lafunte (2013)
studied the effects of four types of tasks on EFL vocabulary
learning. The participants were 28 primary school English as
a second language (ESL) learners in Spain. Eighteen words
including six nouns, six adjectives, and six verbs, were
selected randomly from a short story. The participants were
assigned to four different tasks with different involvement
loads: reading comprehension with marginal glosses, read-
ing comprehension and gap-filling, writing with marginal
glosses, and writing and dictionary use. All the participants
took a receptive and a productive vocabulary test. The results
showed that doing a task with the highest involvement led to
the best result in L2 vocabulary learning. Their study, how-
ever, lacked a control group, and the sample size was small.
Some researchers (e.g., Folse, 2006; Webb, 2005; Zou,
2017) have laid emphasis on writing tasks as effective ones
in vocabulary learning. In the same vein, Feng (2014)
included sentence writing task in his study. Feng (2014)
examined the effects of three translation tasks on EFL learn-
ers’ vocabulary learning based on the ILH. In this study, 30
verbs were selected from business documents to be taught to
60 EFL learners via three different translation tasks: transla-
tion-only mode, translation plus fill-in exercises, and transla-
tion plus sentence writing. The result indicated sentence
writing could significantly improve passive and active word
learning and retention, whereas translation-only task had the
lowest effect. However, Feng’s (2014) study lacked a control
group, did not control the word type, and sufficient informa-
tion for power analyses was not provided.
More recently, Bao (2015) investigated how task type
affects EFL learners’ vocabulary knowledge. To this end,
153 Chinese EFL learners (144 females and 9 males) were
selected. The participants learned 18 target words through
five tasks: control (involvement load of 0), definition
(involvement load of 2), combining (involvement load of 2),
translation (involvement load of 2), and sentence writing
(involvement load of 3). The results revealed that all tasks
were significantly better than the control task. Although Bao
(2015) included a control group and the sentence writing
task, he did not examine the effects of different writing tasks
such as paragraph writing on EFL learners’ vocabulary
knowledge.
Tahmasbi and Farvardin
Tang and Treffers-Daller (2016) have recently examined
the effects of different tasks on L2 incidental vocabulary
learning based on the predictions of ILH. To this end, 230
Chinese EFL learners whose proficiency was at A2 level on
the Common European Framework of Reference for lan-
guages (CEFR) were selected. Six different tasks with differ-
ent involvement loads were designed. The results showed
that tasks with a higher involvement load were significantly
better than tasks with a lower involvement load both in the
immediate and the delayed posttests. However, like most
previous studies, this study suffers from lack of a control
group and insufficient information for power analyses.
In Iranian context, few studies have been conducted on
the effect of task types and involvement indices on EFL
learners’ vocabulary learning. Yaqubi, Rayati, and Allemzade
Gorgi (2010), for instance, randomly assigned 60 EFL learn-
ers to three groups: Group 1 completed an input-oriented
task with an involvement load of 3, Group 2 was given the
same type of task but with an involvement load of 2, and
Group 3 completed an output-oriented task with the involve-
ment load of 3. The results were contrary to the prediction of
the ILH, that is, Task 2 was superior to Task 1, which had a
higher index. Moreover, the learners who had completed
Task 3 significantly did better than those who did Task 1,
despite their index equivalency. However, this study lacked a
control group, had a small sample size, provided no informa-
tion for power analyses, and did not measure the participants’
productive knowledge.
In the same vein, Soleimani and Rahmanian (2015) ran-
domly assigned 33 Iranian EFL learners to three groups: fill
in the blank (involvement load of 1), reading comprehension
(involvement load of 2), and sentence writing (involvement
load of 3). The results showed that sentence writing task was
significantly better than the other two tasks. The results sup-
ported the ILH assumptions. Nevertheless, this study suffers
from some limitations such as no control for word type,
small sample size, no measurement of the productive knowl-
edge, and insufficient information for power analyses.
The present study also attempted to overcome the limita-
tions of the previous studies such as having no control group,
no measurements of both receptive and productive vocabulary
knowledge, and no distinction between input and output orien-
tation. In addition, other tasks like paragraph writing and writ-
ing a word definition have been barely used in the literature.
Finally, little research has been conducted to examine the
hypothesis that the same presence of an involvement compo-
nent leads to the same amount of word learning. Therefore,
this study made an attempt to overcome the deficiencies in
previous studies. Moreover, this study focused on the output
tasks and their effectiveness on L2 vocabulary learning based
on the predictions of the ILH. To fulfill the objectives of the
study, the following research questions were raised:
Research Question 1: Are different output tasks (i.e.,
paragraph writing, sentence writing, translation,
3
combining, and fill in the blank) all conducive to EFL
receptive vocabulary knowledge?
Research Question 2: Are different output tasks (i.e.,
paragraph writing, sentence writing, translation, combin-
ing, and fill in the blank) all conducive to EFL productive
vocabulary knowledge?
Research Question 3: Which output task (i.e., paragraph
writing, sentence writing, translation, combining, and fill
in the blank) will be more effective in EFL receptive
vocabulary knowledge?
Research Question 4: Which output task (i.e., paragraph
writing, sentence writing, translation, combining, and fill
in the blank) will be more effective in EFL productive
vocabulary knowledge?
Method
This experimental study adopted a pretest/posttest control
group design. This research was based on five different tasks,
including paragraph writing, sentence writing, combining,
fill in the blank, and translation. The between-subjects factor
was task type.
Participants
First, a total of 144 (78 female and 66 male) junior high
school students in the third grade participated in this study.
They were selected from two high schools in Ahvaz, Iran.
The participants’ age ranged between 14 and 16 (M = 14.7,
SD = 0.35). They had not studied English abroad or in an
English institute, which meant their English background was
just restricted to school. The participants were equally
assigned to six vocabulary learning tasks (n = 24 each)
through simple random sampling. However, after the delayed
posttest, the number of the participants was reduced to 130
(70 female and 60 male). Finally, the number of participants
at each vocabulary learning task was as follows: paragraph
writing (n = 20), sentence writing (n = 22), translation (n =
24), fill in the blank (n = 21), combining (n = 23), and control
(n = 20).
As Oxford Placement Test (OPT) is easy to administer
and has a well-established reliability and validity (Allan,
2004), it was adopted in this study. OPT includes two sec-
tions, grammar and listening, each of which consists of 100
items. The required time to complete the test is 60 min. Each
correct item received 1 point. Therefore, the maximum pos-
sible score was 200. The participants’ scores ranged from
105 to 118, suggesting that they were at the elementary level
of English proficiency (M = 111.36, SD = 2.75).
Instruments and Materials
Target words. In this study, first, 42 English target words,
with equal parts of speech, were randomly selected from the
504 Absolutely Essential Words (Bromberg, Lieb, & Traiger,
4 SAGE Open
2011). This book includes 42 lessons and one word was ran-
domly selected from each lesson. The reasons behind select-
ing equal parts of speech were controlling the possible
confusion of part of speech with task type and enhancing the
generalizability of the findings. Moreover, the target words
were selected from 504 Absolutely Essential Words whose
words are a little beyond the participants’ proficiency level.
In other words, this book is appropriate for intermediate-
level EFL learners. The reason behind this was to ensure that
the target words would be unknown to all participants. How-
ever, the sentences in which the target words used in the
main study were ensured to be appropriate for elementary-
level EFL learners by four experienced EFL teachers.
To select the target words, a test was also administered to
60 nonparticipants at another high school. The testees were
asked to write the Persian equivalents of the words. The non-
participants were also at the elementary level as the partici-
pants in the experiment. After administering the test, 30
words (i.e., 10 nouns, 10 verbs, and 10 adjectives) unknown
to all the testees were selected as the target words (see
Appendix A).
EFL reading task. In this study, 30 sentences were adapted
from Cambridge Advanced Learner’s Dictionary (Walter,
Woodford, & Good, 2008) in which each target word was
embedded (see Appendix B). The Persian equivalents were
written from The Aryanpur Progressive English–Persian
Dictionary (Aryanpur Kashani & Aryanpur, 2008). The
teacher-researchers and two other EFL teachers checked the
appropriacy of vocabulary and syntax of the reading sen-
tences for the participants. The 30 sentences were randomly
divided into three sets; each set was presented in one session.
Each set included 10 sentences and 10 target words. After
each sentence, the corresponding gloss for the target word
was provided in the brackets. All the groups received the sen-
tences in the same random order. A sample sentence is pre-
sented below:
Her talent for music showed at an early age. (talent: n.
‫ ذوق‬،‫)استعداد‬
In this example, the target word is talent and its gloss
provides its part of speech (a noun) and Persian translation
(‫ ذوق‬،‫)استعداد‬.
Vocabulary knowledge test. This study adopted Min’s (2008)
four-item Vocabulary Knowledge Scale (VKS) in which the
unknown and known word categories are separated (see
Appendix C). “Categories of VKS offer no clues to the target
words and thus can more accurately reflect the students’
knowledge about the target words” (Min, 2008, p. 85). In this
study, both receptive and productive vocabulary knowledge
of participants were measured. As a result, Categories III and
IV in the VKS measured the EFL receptive and productive
vocabulary knowledge, respectively.
Procedures
This study followed the procedure of Bao’s (2015) study.
First, the participants were randomly assigned to five experi-
mental groups receiving a vocabulary learning task (i.e.,
paragraph writing, sentence writing, translation, fill in the
blank, and combining) and a control task. The same or differ-
ent involvement loads were utilized for all tasks except for
the control task to test the contribution of each task type to
vocabulary learning. The paragraph writing task induced
involvement load index of 4 (+N, +S, ++E), sentence writing
task induced involvement load index of 3 (+N, +S, +E), and
combining, fill in the blank, and translation tasks all induced
an involvement load index of 2 (+N, –S, +E).
Two weeks before the study, all participants took the
VKS. In the main study, the experimental groups were taught
the target words by the teacher-researchers in three sessions.
In each session, 10 target words were taught through a
vocabulary learning task presented immediately following a
reading task in which the target words appeared.
For the control group, the task included meaning match-
ing exercises in which the test words were not the target
words. For each set of exercise, 10 test words had to be
matched with 12 definitions. The number of the definitions
was more than the test words to decrease the participants’
guessing.
For the sentence and paragraph writing tasks, the partici-
pants were required to write semantically acceptable and
grammatically correct sentences or paragraphs in 15 min.
The teacher-researchers were available to answer any ques-
tions and the participants had access to both monolingual and
bilingual dictionaries. In both groups, the participants did the
tasks individually.
For the combining task, each sentence was segmented into
separate parts. The combining group was required to combine
all segments into a grammatically correct sentence.
The participants in translation group had to translate the
same sentences from L2 to L1. The participants could use
English–Persian dictionary during each task. For the fill-in-
the-blank task, the participants were to fill in the blanks with
appropriate given (target) words.
At the end of the last treatment session, the immediate
posttest was administered in 20 min. To assess the partici-
pants’ long-term retention of the target words, a delayed post-
test was also administered 1 month after the experiment.
Scoring and Data Analysis
In this study, Min’s (2008) scoring was followed to measure
the participants’ receptive and productive vocabulary knowl-
edge of the target words. For Categories III and IV, 1 point or
0 point was given. Category III measured the receptive
knowledge of a target word. If the given synonym or transla-
tion was improper, or no response was given, 0 point was
considered. A correct synonym or translation also received 1
Tahmasbi and Farvardin
Table 1. Results of Mixed ANOVA on Receptive Vocabulary Tests.
Source Type III SS MS
Between-subject
Task type 1,561.792 312.358
Error 2,256.682 18.199
Within-subject
Time 15,793.102 7,896.551
Time × Task type 848.790 84.879
Error (time) 1,135.620 4.579
Note. SS = sum of squares; MS = mean square.
point. Category IV measured the productive knowledge of a
target word. If no point was given to Category III, 0 point
was given to Category IV. If a target word’s meaning was
inappropriate or ungrammatical in the sentence context, 0
point was given. One point was given when a target word
was both semantically and grammatically correct in the sen-
tence albeit other parts of the sentence had errors.
The participants’ responses were scored by two experi-
enced EFL teachers. Each participant’s responses were inde-
pendently scored. In the pretest, the Cohen’s Kappa interrater
reliability indices for the receptive and productive tests were
0.98 and 0.95, respectively. Then, the two raters discussed all
discrepancies in scoring until they reached a unanimous
agreement. Hence, the interrater agreement was 100% in the
immediate and delayed posttests.
To address the research questions, a mixed 3 × 6 ANOVA,
one-way ANOVA, and Tukey post hoc tests were run on
receptive and productive test scores separately. Significance
level was set at .05.
Results
First, normality of data was measured through Shapiro–Wilk
test. All significant values in the test were above the signifi-
cance level of .05, implying normal distribution of data. The
results also revealed that the paragraph writing group had the
highest mean scores on both receptive and productive mea-
sures of vocabulary knowledge in immediate and delayed
posttests.
Then, a mixed 3 × 6 ANOVA, with time (i.e., pretest,
immediate posttest, and delayed posttest) and task type (i.e.,
paragraph writing, sentence writing, translation, fill in the
blank, combining, and control) as two main factors, was
done. A one-way ANOVA along with post hoc Tukey tests
was also conducted for the overall comparison of the six
groups on all tests. Results of mixed ANOVA on the recep-
tive tests are displayed in Table 1.
The results show significant main effects of task type
(F2, 124 = 17.163, p = .000, partial η2 = 0.409). There were
also a significant effect for time (F2, 248 = 1724.471, p = .000,
partial, η2 = 0.933) and a significant interaction between time
and task type (F10, 248 = 18.536, p = .000, partial η2 = 0.428).
5
df F Significance Partial η2
5 17.163 .000 .409
124
2 1,724.471 .000 .933
10 18.536 .000 .428
248
The effect size for mixed ANOVA is calculated by the partial
eta squared (Pallant, 2007). Moreover, it is recommended
that L2 researchers adopt the following benchmarks to inter-
pret the practical significance of L2 research effects more
precisely: small = 0.40, medium = 0.70, and large = 1.00
(Plonsky & Oswald, 2014). Therefore, the results suggest an
almost large effect size for time and small effect size for task
type.
To further examine the differences between the groups, a
one-way ANOVA was performed for each receptive test. No
significant difference was found in the pretest (F5, 124 = 0.117,
p = 0.988). However, in the immediate posttest (F5, 124 =
24.546, p = .000) and the delayed posttest (F5, 124 = 15.629, p
= .000), significant differences were found. Table 2 illus-
trates the findings of post hoc analyses on the posttests.
Table 2 shows that on the immediate posttest, the para-
graph writing group did significantly better than all groups
but the sentence writing group. Moreover, the sentence
writing group outperformed the combining and control
groups. All groups also significantly outperformed the
control group. Post hoc tests also showed that paragraph
writing group obtained significantly higher scores than the
combining and control groups on the delayed posttest (p =
.000). Moreover, all groups significantly outperformed the
control on the delayed posttest (p = 000) except the com-
bining group.
A mixed 3 × 6 ANOVA, with time (i.e., pretest, immediate
posttest, and delayed posttest) and task type (i.e., paragraph
writing, sentence writing, translation, fill in the blank, com-
bining, and control) as two main factors, was also conducted
on the productive vocabulary tests. In addition, a one-way
ANOVA along with post hoc Tukey tests were run for the
overall comparison of the six groups on all tests. Results of
mixed ANOVA on the productive tests are shown in Table 3.
The results showed significant main effects of task type
(F2, 124 = 9.985, p = .000, partial η2 = 0.287). There were also
a significant effect for time (F2, 248 = 726.611, p = .000, par-
tial η2 = 0.854) and a significant interaction between time
and task type (F10, 248 = 14.227, p = .000, partial η2 = 0.365).
The results suggest a medium effect size for time and small
effect sizes for both task type and the interaction between
time and task type.
6 SAGE Open

Table 2. Results of Tukey Post Hoc Tests on Receptive Posttests.

Dependent variable (I) Group (J) Group Mean difference (I-J) SE Significance
Immediate posttest Paragraph writing Sentence writing 1.077 1.047 .907
Paragraph writing Translation 3.225 1.026 .025
Paragraph writing Fill in the blank 3.540 1.058 .014
Paragraph writing Combining 6.089 1.036 .000
Paragraph writing Control 10.250 1.071 .000
Sentence writing Translation 2.148 1.000 .270
Sentence writing Fill in the blank 2.463 1.033 .170
Sentence writing Combining 5.012 1.010 .000
Sentence writing Control 9.173 1.047 .000
Translation Fill in the blank 0.315 1.012 1.00
Translation Combining 2.864 0.988 .50
Translation Control 7.025 1.026 .000
Fill in the blank Combining 2.549 1.022 .134
Fill in the blank Control 6.710 1.058 .000
Combining Control 4.164 1.036 .001
Delayed posttest Paragraph writing Sentence writing 0.895 1.105 .965
Paragraph writing Translation 3.058 1.083 .060
Paragraph writing Fill in the blank 2.564 1.118 .204
Paragraph writing Combining 5.720 1.094 .000
Paragraph writing Control 8.350 1.131 .000
Sentence writing Translation 2.163 1.056 .322
Sentence writing Fill in the blank 1.669 1.092 .646
Sentence writing Combining 4.824 1.067 .000
Sentence writing Control 7.455 1.105 .000
Translation Fill in the blank –0.494 1.069 .997
Translation Combining 2.661 1.044 .118
Translation Control 5.292 1.083 .000
Fill in the blank Combining 3.155 1.080 .000
Fill in the blank Control 5.786 1.118 .000
Combining Control 2.630 1.094 .163

p < .05.

Table 3. Results of Mixed ANOVA on Productive Vocabulary Tests.

Source Type III SS MS df F Significance Partial η2

Between-subject
Task type 315.130 63.026 5 9.985 .000 .287
Error 782.667 6.312 124
Within-subject
Time 1,817.093 908.546 2 726.611 .000 .854
Time × Task type 177.899 17.790 10 14.227 .000 .365
Error (time) 310.096 1.250 248

Note. SS = sum of squares; MS = mean square.

To further examine the differences between the groups,
a one-way ANOVA was performed for each receptive test.
No significant difference was found in the pretest (F5, 124
= 0.072, p = 0.996). However, in the immediate posttest
(F5, 124 = 9.301, p = .000) and the delayed posttest (F5, 124
= 19.562, p = .000) significant differences were found.
Table 4 depicts the findings of post hoc analyses on the
posttests.
Table 4 shows that, on the immediate posttest, para-
graph writing significantly did better than all groups
except the sentence writing group. Moreover, the sentence
writing group outperformed the combining and control
groups. Also, it was found that the translation and the fill
in the blank groups were superior to the control group.
Post hoc tests also showed that the paragraph writing
group obtained significantly higher scores than all groups
Tahmasbi and Farvardin 7

Table 4. Results of Tukey Post Hoc Tests on Productive Posttests.

Dependent variable (I) Group (J) Group Mean difference (I-J) SE Significance
Immediate posttest Paragraph writing Sentence writing 1.655 .612 .081
Paragraph writing Translation 2.158 .599 .006
Paragraph writing Fill in the blank 2.271 .618 .005
Paragraph writing Combining 2.570 .605 .001
Paragraph writing Control 4.150 .626 .000
Sentence writing Translation 0.504 .584 .955
Sentence writing Fill in the blank 0.617 .604 .910
Sentence writing Combining 0.915 .590 .633
Sentence writing Control 2.495 .612 .001
Translation Fill in the blank 0.113 .591 1.00
Translation Combining 0.411 .578 .980
Translation Control 1.992 .599 .015
Fill in the blank Combining 0.298 .597 .996
Fill in the blank Control 1.879 .618 .034
Combining Control 1.580 .605 .102
Delayed posttest Paragraph writing Sentence writing 1.891 .550 .010
Paragraph writing Translation 3.217 .539 .000
Paragraph writing Fill in the blank 3.610 .556 .000
Paragraph writing Combining 3.800 .545 .000
Paragraph writing Control 5.000 .563 .000
Sentence writing Translation 1.326 .526 .126
Sentence writing Fill in the blank 1.719 .543 .024
Sentence writing Combining 1.909 .531 .006
Sentence writing Control 3.109 .550 .000
Translation Fill in the blank .393 .532 .977
Translation Combining .583 .520 .871
Translation Control 1.783 .539 .015
Fill in the blank Combining –.190 .538 .999
Fill in the blank Control 1.390 .556 .132
Combining Control 1.200 .545 .243

p < .05.

on the delayed posttest. Moreover, the paragraph writing,
the sentence writing, and the translation groups signifi-
cantly outperformed the control group on the delayed
posttest (p < .05).
Discussion
The results revealed that, regardless of vocabulary knowl-
edge type, all output tasks were significantly better than the
control group in receptive vocabulary knowledge test.
Moreover, the findings showed that paragraph writing task
was more effective than all groups except the sentence writ-
ing group. In addition, the sentence writing group outper-
formed the combining and control groups. The relative
efficacy of the tasks can be briefly reported as follows: para-
graph writing (index = 4) ≥ sentence writing (index = 3) ≥
translation (index = 2) = fill in the blank (index = 2) ≥ com-
bining (index = 2). However, some discrepancies were
detected as the patterns for each type of vocabulary knowl-
edge were examined separately.
In the immediate receptive vocabulary test, the paragraph
writing task was better than other tasks except the sentence
writing task, implying that the higher the involvement load a
task has, the more effective the task will be for L2 vocabu-
lary learning. The findings of this study are in line with
Laufer and Hulstijn (2001) who found that the effectiveness
of an output task depends on its involvement load. Fill in the
blank, translation, and combining tasks had similar involve-
ment load indices (+N, –S, +E). Although, the translation
and fill in the blank tasks showed no significant difference,
combining task had the lowest score in immediate and
delayed posttests. However, the findings of this study are not
in line with some previous studies supporting the ILH (e.g.,
Bao, 2015; Keating, 2008; Marmol & Sanchez-Lafunte,
2013). Comparing the groups’ performance in the immediate
and delayed posttests revealed that paragraph writing task
with involvement index of 4 was not superior to sentence
writing task with involvement index of 3. In addition, the
participants who had completed fill in the blank and transla-
tion tasks with involvement index of 2 outperformed those
8 SAGE Open
who did combining task despite its involvement index is 2. In
other words, combining task with the components (+N, –S,
+E) of involvement load had the lowest score in comparison
with fill in the blank and translation with the similar compo-
nents in this study. In combining task, the learners were
required to arrange a string of words in a proper sentence.
Hence, the learners might have no special attention to the
target words. Moreover, in the combining task, the partici-
pants might be too obsessed with combining the sentences
per se to notice the semantic and syntactic aspects of the tar-
get words. The disconnected strings of words given to the
participants might significantly diminish the contextual clue-
ing that led the combining task to be the least effective output
task. The results of delayed receptive vocabulary test, how-
ever, revealed that the paragraph writing task was similar to
the sentence writing, translation, and fill in the blank tasks.
The reason might be attributable to the opportunity the par-
ticipants had to “retrieve more grammatical information
about those target words whose meanings they remembered”
(Bao, 2015, p. 92).
Regarding productive vocabulary knowledge, paragraph
writing was better than other tasks on both immediate and
delayed productive tests. The reason might be attributable to
the fact that it could draw participants’ attention more to both
semantic and grammatical features of target words, which
might help them remember the meaning of words better.
Moreover, it can be argued that when participants wrote a
paragraph with target words they evaluated the appropriacy
of using the target words in context.
The sentence writing task, however, was not superior to
other output tasks on the immediate posttest. The results
revealed that “the same involvement loads did not necessarily
lead to similar vocabulary learning, nor did the higher involve-
ment loads necessarily lead to better word learning” (Bao,
2015, p. 91). This may imply that the writing task might be
superior to other output tasks in productive vocabulary knowl-
edge when a more challenging writing task like writing a para-
graph or essay is required. In this regard, Keating (2008) found
that sentence writing task had no superiority over the reading
plus fill-in task on the immediate recall test. Keating (2008)
argued that “producing connected sentences might involve
more elaborate processing of the target words than discon-
nected sentences” (Bao, 2015, p. 93). The findings of the pres-
ent study confirm Keating’s (2008) claim about the higher
effectiveness of more complex writing tasks such as paragraph
writing on the EFL learners’ productive vocabulary knowl-
edge. The results are also in line with Williams’ (2012) claim
that writing tasks can push learners to evaluate and analyze
their linguistic knowledge and hence improve their L2 vocab-
ulary knowledge. Therefore, it can be concluded that writing
tasks can reinforce and encourage EFL learners to extend their
knowledge of learning new lexical items (McDonough &
Fuentes, 2015; Zou, 2017). In the delayed productive test, the
paragraph writing task was also superior to other tasks in the
retention of the target words.
As the results showed, a decline in the mean scores of all
groups was observed from the first posttest to the second
posttest, implying that some newly learned words are often
forgotten after a few days. Hence, “reinforcement of newly
learned words is still needed if students are to remember
them in the longer term, regardless of the amount of involve-
ment load of the vocabulary learning task” (Tang & Treffers-
Daller, 2016, p. 138). Therefore, it seems necessary to take
other factors, such as attention, contextual clueing, and
learners’ awareness of the target words, into account.
Conclusion
The results of this study confirm the predictions of ILH that
tasks with a higher involvement load better help EFL learn-
ers recall and retain the target words. The findings have also
important implications for EFL practitioners. Teachers can
arrange activities to help students develop their vocabulary
learning through tasks with high involvement load and strong
evaluation. Moreover, the teachers should apply useful
opportunities to involve learners to learn more vocabulary
knowledge and obtain meaningful usage of unknown words
through tasks such as paragraph writing and sentence writ-
ing. Furthermore, teachers can actively increase the effec-
tiveness of vocabulary learning through noticing and
considering the learners’ needs. EFL teachers may some-
times integrate various tasks, for instance, translation and
sentence writing tasks, to use the merits of each task for
vocabulary learning.
The findings of the present study are subject to a number
of limitations. First, this study was conducted in Iran as an
EFL context. Hence, future studies can examine the effec-
tiveness of different vocabulary learning tasks in ESL and
other EFL contexts. Second, the effect of each output task
was studied separately. Thus, researchers can integrate the
output tasks and investigate their effect on EFL learners’ pro-
ductive and receptive vocabulary knowledge. Third, this
study explored the predictions of ILH in vocabulary learning
tasks. However, the role of other factors such as learners’
awareness of target words and the contextual clues in EFL
vocabulary learning were not examined. Therefore, further
studies are needed to investigate other influential factors in
L2 vocabulary learning. Finally, future studies can compare
and contrast ILH with other theoretical frameworks such as
Technique Feature Analysis (Nation & Webb, 2011) to better
examine the predictability power of these frameworks.
Appendix A
Target Words
Nouns: decade, disaster, event, homicide, lecture, menace,
obstacle, panic, talent, vicinity.
Verbs: anticipate, commence, conceal, detect, detest, excel,
perish, pursue, squander, vanish.
Tahmasbi and Farvardin 9

Adjectives: excessive, frigid, massive, mediocre, rural, III. I know this word. It means …………………. (Give
urgent, thorough, vacant, vague, weary. the meaning in English or Persian.)
IV. I can use this word in a sentence. (Write a sentence.)
Appendix B (If you do this section, please also complete III.)

Sentences in the EFL Reading Task Declaration of Conflicting Interests

1. Her talent for music showed at an early age. The author(s) declared no potential conflicts of interest with respect
2. The hospital has no vacant beds. to the research, authorship, and/or publication of this article.
3. I tried to conceal my surprise when she told me her age.
4. Few plants can grow in such a frigid weather. Funding
5. The child vanished while on her way home from The author(s) received no financial support for the research, author-
school. ship, and/or publication of this article.
6. Life in rural areas is simpler and cheaper.
7. Human ear cannot detect some sounds. References
8. They did a thorough search of the area but found Allan, D. (2004). Oxford placement test. Oxford, UK: Oxford
nothing. University Press.
9. The earth is a massive planet. Aryanpur Kashani, M. A., & Aryanpur, M. (2008). The Aryanpur
10. Parents don’t want their children going to mediocre progressive English-Persian dictionary. Tehran, Iran: Amir-
schools. Kabir Publications.
11. There are several hotels in the vicinity of the station. Bao, G. (2015). Task type effects on English as a foreign language
12. Rebecca always excelled in mathematics at school. learners’ acquisition of receptive and productive vocabulary
13. Drunk drivers are a menace to everyone. knowledge. System, 53, 84-95.
Bromberg, M., Lieb, J., & Traiger, A. (2011). 504 absolutely essen-
14. Many people moved out of this city in the last decade.
tial words. Hauppauge, NY: Barron’s Educational Series.
15. Many people are in urgent need of food and water. Feng, T. (2014). Involvement load in translation tasks and EFL
16. He commenced speaking before all the guests had vocabulary learning. The New English Teacher, 9(1), 83-101.
finished eating. Folse, K. S. (2006). The effect of type of written exercise on L2
17. The car was pursued by helicopters. vocabulary retention. TESOL Quarterly, 40, 273-293.
18. Carmel was in a panic about her exam. Keating, G. D. (2008). Task effectiveness and word learning in a
19. Three hundred people perished in the earthquake. second language: The involvement load hypothesis on trial.
20. I think he’s a little weary after his long journey. Language Teaching Research, 12, 365-386.
21. At this stage, we can’t really anticipate what will Kim, Y. (2008). The role of task-induced involvement and learner
happen. proficiency in L2 vocabulary acquisition. Language Learning,
22. The number of homicides in the city has highly 58, 285-325.
Laufer, B., & Hulstijn, J. (2001). Incidental vocabulary acquisition
increased.
in a second language: The construct of task-induced involve-
23. I detest getting up early in the morning. ment. Applied Linguistics, 22, 1-26.
24. Do not squander your money by buying what you Marmol, G. A., & Sanchez-Lafunte, A. A. (2013). The involvement
cannot use. load hypothesis: Its effect on vocabulary learning in primary
25. Excessive exercise can sometimes cause health education. Revista Española de Lingüística Aplicada, 26, 11-24.
problems. McCarty, M. (2005). Discourse analysis for language teachers.
26. It will be a disaster for me if I lose my job. Cambridge, UK: Cambridge University Press.
27. Who’s giving the lecture this afternoon? McDonough, K., & Fuentes, C. G. (2015). The effect of writing
28. I do have a vague memory of meeting her many years task and task conditions on Colombian EFL learners’ language
ago. use. TESL Canada Journal, 32, 67-79.
29. The Olympic Games are the biggest sporting event in Min, H. T. (2008). EFL vocabulary acquisition and retention:
Reading plus vocabulary enhancement activities and narrow
the world.
reading. Language Learning, 58, 73-115.
30. The biggest obstacle in our way was a tree trunk in Nation, P. (2003). Vocabulary. In D. Nunan (Ed.), Practical English
the road. language teaching (pp. 129-152). New York, NY: McGraw-
Hill.
Appendix C Nation, P., & Meara, P. (2002). Vocabulary. In N. Schmitt (Ed.),
An introduction to applied linguistics (pp. 35-54). London,
Modified Vocabulary Knowledge Scale (Min, England: Arnold.
2008) Nation, P., & Webb, S. (2011). Researching and analyzing vocabu-
lary. Boston, MA: Heinle.
I. I don’t remember having seen this word before. Pallant, J. (2007). SPSS survival manual: A step-by-step guide to
II. I have seen this word before, but I don’t know what it data analysis using SPSS version 15. New York, NY: McGraw-
means. Hill.
10
Plonsky, L., & Oswald, F. L. (2014). How big is “big”? Interpreting
effect sizes in L2 research. Language Learning, 64, 878-912.
Schmitt, N. (2000). Vocabulary in language teaching. Cambridge,
UK: Cambridge University Press.
Soleimani, H., & Rahmanian, M. (2015). Vocabulary acquisition
and task effectiveness in involvement load hypothesis: A case
in Iran. International Journal of Applied Linguistics & English
Literature, 4, 198-205.
Tang, C., & Treffers-Daller, J. (2016). Assessing incidental vocabu-
lary learning by Chinese EFL learners: Testing the Involvement
Load Hypothesis. In G. Yu & Y. Yin (Eds.), Assessing Chinese
learners of English: Language constructs, consequences and
conundrums (pp. 121-148). London, UK: Palgrave.
Walter, E., Woodford, K., & Good, M. (Eds.). (2008). Cambridge
advanced learner’s dictionary (3rd ed.). Cambridge, UK:
Cambridge University Press.
Webb, S. (2005). Receptive and productive vocabulary learning:
The effects of reading and writing on word knowledge. Studies
in Second Language Acquisition, 27, 33-52.
Williams, J. (2012). The potential role(s) of writing in second lan-
guage development. Journal of Second Language Writing, 21,
321-331.
SAGE Open
Yaqubi, B., Rayati, R. A., & Allemzade Gorgi, N. (2010). The
involvement load hypothesis and vocabulary learning: The
effect of task types and involvement index on L2 vocabu-
lary acquisition. Journal of Teaching Language Skills, 29,
145-163.
Zou, D. (2017). Vocabulary acquisition through cloze exercises,
sentence-writing and composition-writing: Extending the
evaluation component of the involvement load hypothesis.
Language Teaching Research, 21, 54-75.
Author Biographies
Maryam Tahmasbi holds an MA in TEFL the Islamic Azad
Unicersity of Ahvaz, Iran. She has been teaching English for 20
years. Her research interests include L2 vocabulary learning and
psycholinguistics.
Mohammad Taghi Farvardin is an assistant professor of TEFL at
the Departmet of English Language Teaching, Islamic Azad
University, Ahvaz Branch, Iran. His research interests include L2
vocabulary learning, reading in a second/foreign language, and psy-
cholinguistics. He teaches at both undergraduate and postgraduate
levels, and also supervises postgraduate students.