Open Journal of Modern Linguistics
2013. Vol.3, No.1, 30-39
Published Online March 2013 in SciRes (http://www.scirp.org/journal/ojml) http://dx.doi.org/10.4236/ojml.2013.31004
30
The Effect of Task Characteristics on IELTS Reading
Performance
Behnam Behfrouz, Elham Nahvi
Faculty of Persian Literature and Foreign Languages, Islamic Azad University, Tehran, Iran
Received November 14th, 2012; revised December 17th, 2012; accepted December 25th, 2012
One of the skills that involve thinking processes is reading. Reading is a skill which allows students to
become familiar with other people’s ideas; compare and contrast different ideas; examine and evaluate
arguments, interpretations, beliefs, or theories; make inferences, predictions, or interpretations; and ex-
plore implications and consequences. In essence, reading is a skill that compels students to think critical-
ly. Thus, the present study aims to investigate the effects of task characteristics on the learners’ perform-
ance in the reading section of IELTS. To accomplish the purpose of the research, a sample of 50 male and
female students participated in this study. In order to ensure the homogeneity of the participants and de-
termine the participants’ reading proficiency level, a reading part of the IELTS test was conducted at the
outset of the study. Based upon the results, 30 participants were selected to be included within this study.
Next, five IELTS tests were given to the participants. The subsequent data was then analyzed using a
one-way ANOVA and the Sheffe test. Data revealed that participants performed differently on the IELTS
test due to different task types given as a treatment. The results suggest that task characteristics have a
significant effect on IELTS reading test performance.
Keywords: Task Characteristics; Performance; Reading Section; IELTS
Introduction
Reading materials have been considered to be one of the pri-
mary sources of acquiring comprehensible input and therefore
reading has always been a significant aspect of language learn-
ing (Chastain, 1988). Krashen (1988) believes readers recreate
the message while reading. According to Rumelhart (1977;
cited in Chastain, 1988), the reader, the text, and the interaction
between the reader and the text are involved in reading. In this
interactive process of reading, meaningfulness plays a very
important role in better comprehension of the texts. From
the 1960s most cognitive psychologists have focused on the role
of meaningful learning and the organization of background
knowledge.
Reading comprehension skill has been the home of choice
for language testers and has attracted a lot of attention as a
prime source of input for the ELT researchers and teachers
(Amiri & Maftoon, 2010). Reading is the most available and
foremost source of information and necessary input for the EFL
students (Bachman & Palmer, 2010) and mostly second or for-
eign language learners find it difficult to comprehend the texts
by reading them just once or without looking up the words in
a dictionary. IELTS examinees’ performance in the reading
section could be affected by variety of factors, one of which
is “task characteristics”. The characteristics of the used items
and the nature of the test in general, affect learners’ answers.
The present study is an attempt to focusing on the specific as-
sessment tasks employed in the reading section of IELTS and
their possible impact on the examinees’ (learners’) perform-
ance.
Theoretical Background
Reading
One of the skills that involve thinking processes is reading.
Reading is a skill which makes students familiar with other
people’s ideas; compare and contrast different ideas; examine
and evaluate arguments, interpretations, beliefs, or theories; make
inferences, predictions, or interpretations; explore implications
and consequences, and in short it is a skill which makes stu-
dents to think critically. According to Hudelson (1994) cited in
Edigar (2001: p. 154), in reading “an individual constructs
meaning through a transaction with written text that has been
created by symbols that represent language. The transaction
involves the reader’s acting on or interpreting the text and the
interpretation is influenced by reader’s past experiences, lan-
guage background, and cultural framework, as well as the
reader’s purpose for reading”.
There are different definitions for reading comprehension.
However, the nature of the process of reading is not known
exactly. Some of the common definitions are as follow; in very
simple terms reading is “perceiving a text to understand its
contents” (Richards & Schmidt, 2002: p. 443).
According to Chastain (1988), “reading is a process involving
the activation of relevant knowledge and related language skills
to accomplish an exchange of information from one person to
another. Reading requires that the reader focus attention on the
reading materials and integrate previously acquired knowledge
and skills to comprehend what someone else has written” (p.
216).
Also Bowen et al. (1985) claimed that reading is a problem

B. BEHFROUZ, E. NAHVI
solving activity that involves the reader in the process of
deriving the meaning from the text. According to Carrell (1991)
reading comprehension is a complex behavior, which involves
conscious and unconscious use of various strategies in order to
get the intention of the writer. He claimed that they decode
meaning, respond to what they read and select for particular
aspect for consideration.
Reading is the activation of related knowledge and language
skills to perform an exchange of information from one person
to another. Reading is a receptive skill; it means that the reader
is receiving a message from a writer. Reading is a complementary
skill for language learning. Second language learners are re-
quired to learn to read for communication and to read greater
number of authentic materials (Chastain, 1988). Nuttall (1996)
considered reading as convey of meaning from mind to mind;
the movement of message from writer to reader.
Reading is a language process. Students should be assisted to
answer the visual symbols which depict the same auditory sig-
nals to which they had answered before (Finocchiaro & Bonomo,
1973).
Barnett, cited in Hadley (2003: p. 163) listed reasons for in-
cluding reading skill development as an important part of sec-
ond language curriculum: 1) reading is still essential in the
teaching of literature, which remains an important goal in many
programs; 2) it is a skill that can be maintained after students
complete formal language study; 3) it fosters the development
and refinement of literary skills.
There are various theories to explain what is involved when
we read. Current research shows that reading is an interactive,
socio-cognitive process which involves a text, a reader and a
social context within which the activity of reading takes place
(Bernhardt, 1991, cited in Edigar, 2001).
Widdowson (1979) cited in Grabe (1988) stated that reading
is the process in which textual information is combined with
the information a reader brings to a text. In this view the read-
ing process is not simply a matter of extracting information
from the text. Rather, it is the process in which a range of
knowledge in the reader’s mind is activated and this knowledge
may be influenced by the new knowledge. Grabe (1991) cited
in Edigar (2001: p. 154) listed six general component skills and
knowledge areas, within the complex process of reading:
1) Automatic recognition skills—a virtually unconscious
ability, ideally requiring little mental processing to recognize
text, especially for word identification;
2) Vocabulary and structural knowledge—a sound under-
standing of language structure and a large recognition vocabu-
lary;
3) Formal discourse structure knowledge—an understanding
of how texts are organized and how information is put together
into various genres of text (e.g., a report, a letter, a narrative);
4) Content/world background knowledge—prior knowledge
of text-related information and a shared understanding of the
cultural information involved in text synthesis and evaluation
skills/strategies—the ability to read and compare information
from multiple sources, to think critically about what one;
5) Read, and to decide what information is relevant or useful
for one’s purpose;
6) Meta-cognitive knowledge and skills monitoring—an
awareness of one’s mental processes and the ability to reflect
on what one is doing and the strategies one is employing while
reading.
He said that when a fluent reader reads, he brings together all
of these components into a complex process.
Reading Purposes
Rivers and Temperly cited in Nunan (1999: p. 251) sug-
gested seven main purposes for reading:
1) To obtain information about some topic;
2) To obtain instructions about doing some task;
3) To act in a play, play a game, do a puzzle;
4) To keep in touch with friends by correspondence or to
understand business letters;
5) To know when or where something will happen or what is
available;
6) To know what is happening or has happened;
7) For enjoyment or excitement.
Chastain (1988) stated that the ultimate reading goal must
include other factors that convert a laborious problem solving
task into a viable skill. Language learners have to learn in order
to dealing with linguistic materials over which they have no
control. They must learn to interact with the reading as a pro-
ductive way to determine meaning even when some of the word
endings and forms are not meaningful.
Task Characterization an d Rea di n g Comprehension
A lot of researchers agreed that reading is rapid, purposeful,
and interactive (Grabe, 1999; Alderson, 2000). At first level,
the different components and the levels of reading passage
might interact with each other. Second, reading is interactive in
that the reader’s background knowledge and other attributes
interact with the content of the text.
With regard to the fact that reading involves interaction be-
tween readers and texts, it can be concluded that characteristics
of both, the reader and the text, can affect the reading process
(Alderson, 2000).
According to Bachman and Palmer (1996, 2010) task char-
acteristics can be synonymous with test method facets. There-
fore, the characteristics evaluation of the reading comprehen-
sion tests may be the optimal means for the assessment of the
learners’ task performance.
Of greater relevance here is the point that text characteristics
are describable in a systematic manner. Bachman and Palmer
(1996) referred to tasks as the activities which can be done in
proper settings for gaining specific objectives. With reference
to their definition of the task, they concluded that tasks are
associated with specific situations and the active participation
of the language users is goal-oriented. In other words, the par-
ticipants perform a specific task in a particular setting in order
to achieve a specific goal. This definition of language use tasks
entails both the specific activity and the setting in which it takes
place. Bachman (1990), and later modified in Bachman and
Palmer (1996), presented a model of language ability which
draws attention to a range of factors which can affect test per-
formance. Bachman and Palmer (1996) also posited the impor-
tance of method facets, which they now term “task characteris-
tics” as follows:
Language use involves complex and multiple interactions
among the various individual characteristics of language
users, on the one hand, and between these characteristics
and the characteristics of the language use or testing situa-
tion, on the other. Because of the complexity of these in-
teractions, we believe that language ability must be con-

B. BEHFROUZ, E. NAHVI
sidered within an interactional framework of language use
(p. 62).
Bachman and Palmer (1996) test task characteristics model
provides a means for systematically describing various charac-
teristics of tests and testing procedures. Bachman and Palmer
(2010) described the characteristics of the setting for tasks and
discussed that “the setting comprises the circumstances under
which either language use or testing takes place” (p. 68).
The characteristics of the setting based on what Bachman
and Palmer (2010) presented, included the physical characteris-
tics, the participants, and the time of task.
Bachman (1990) classified test method facets into five cate-
gories: 1) testing environment; 2) test rubrics; 3) the nature of
the input; 4) the nature of the expected response; and 5) the
interaction between the input and the response. According to
Bachman (1990), these factors can affect test performance; it is
important for testers to be aware of their influences and, if pos-
sible, to minimize them.
Based on the concept of task characteristics, Carr (2006)
used a new methodological approach to describe variation in
test task characteristics and explored how differences in these
characteristics might relate to examinees’ performance. He con-
structed an expanded test task characteristics instrument and
added a large number of syntactic features to this instrument
(Celce-Murcia & Larsen-Freeman, 1999). Ratings and numerical
counts were gathered for three forms of the Test of English as a
Foreign Language (TOEFL) Reading Comprehension Section.
The items were the objects of the measurement and based on
them, the results were used in a series of exploratory and
confirmatory factor analyses, together with IRT (item response
theory) parameter estimates for the items in question.
The most important finding of this study was that key sen-
tence variables were the most important determining factors in
the measures of difficulty of reading items. Key sentence vari-
ables accounted for 15 of the 20 variables in the final task
characteristics model. Carr (2006) asserted that the findings are
particularly significant because other studies have generally
ignored key sentences as a unit of analysis.
Bachman and Palmer’s model (1996) is initially intended as
a tool for designing and constructing language tests, but be-
cause of the detailed descriptions of tasks, the model also pro-
vides a useful research model for empirical studies and other
research. A number of language testing studies have made use
of this model (Clapham, 1996) or other systems of describing
test task characteristics (Freedle & Kostin, 1993).
Task Characteristics and Learners’ Performance
The characteristics of tasks could be taken into account as
highly influential factors affecting the learners’ performance on
the test items. Task characteristics, reading comprehension na-
ture, and task difficulty are among the concepts which require
clarification before beginning to conduct a research on the ef-
fect of task characteristics on the performance of learners or
examinees on the tests. IELTS as the high-stake proficiency test
taken into consideration in the present study also requires some
description.
Bachman (1990) provides a framework of task characteristics.
This framework includes a set of features that describe five
aspects of tasks: setting, test rubrics, input, expected response,
and relationship between input and response.
Characteristics of the setting refer to all physical conditions
under which testing takes place. This includes the physical set-
ting, participants, and the time of task. The second aspect to
mention here is called the characteristics of the test rubric. Test
rubric refers to those features that show how the test takers
should proceed during the test to accomplish the tasks. The
characteristic of rubric include: the organization (structure) of
the test, instructions, the duration of the test as a whole and of
the individual parts, and how the language that is used is evalu-
ated and scored. In a test task the instructions should be as ex-
plicit and clear as possible because students should know ex-
actly what to do and just in this way the teachers can make safe
decisions based on their performance on the test. Structure of
the test task shows how the different part of the test are put
together and presented to the test takers. This aspect deals with
the number of parts or tasks, the salience of the parts/tasks
which shows how the different parts are clearly distinguished
from each other, the order or sequence of parts/tasks, relative
importance of tasks, and the number of items in each part. Time
allotment is the amount of time for each part of test and the
entire test. Scoring method specifies how numbers will be as-
signed to test takers’ performance.
The third aspect of test task characteristics which is utilized
in this study is characteristic of the input. Input consists of the
material contained in a given test task which the test takers are
going to process in some way and to which they are expected to
respond. Input can be studied from different aspects namely
format and language. By format Palmer means the way in which
the input is presented. Format includes channel, form, language,
length, type, degree of speediness, and vehicle. Channel deter-
mines the way of presenting the input which can be aural, vis-
ual, or both. Also input can take the form of language, non-
language, or both. If the form is language, it can be the testees’
native language or target language. In terms of length, input can
be just a single word, a phrase, a sentence, a paragraph, or an
extended discourse. Type of input may be either an item or a
prompt. The purpose of an item is to elicit either a selected or a
limited response. An example of a test item is the familiar mul-
tiple-choice question. The purpose of a prompt is to elicit an
extended production response. A test task as summary writing
contains input of this type. Degree of speediness refers to the
rate at which the test taker has to process the information. Ve-
hicle is the means by which input is delivered which can be live
or reproduced.
Characteristics of the expected response are another aspect of
test task features. Here again two factors are important: format
and type of response. The explanation of format is the same as
what was said about input. Genesee and Upshur (1996) state
that when selecting language tests, the response characteristics
or demands of the test task should be taken into consideration.
They state that based on the response characteristics, test me-
thods can be described as closed-ended, limited, and open
ended. For describing the type of response other measurement
specialists use three categories: selected, limited production,
and extended production. In a selected response no production
is needed. The test taker just chooses or selects one response
from among two or more that are provided. A typical example
of selected response is a multiple-choice test task. A limited
production response includes a single word or phrase, and may
be as long as a single sentence or utterance. In short completion
items and cloze test this type of response is required. If the test
taker has to provide a response that is longer than a single
32

B. BEHFROUZ, E. NAHVI
sentence and its range is from two sentences to a whole compo-
sition, then the response is named an extended production one.
The summary writing test task used in this study requires an
extended response and the test takers are to provide a summary
of about 100 words. Other features relating to response are the
degree of speediness and the language of expected response.
Task characteristics model presented by Bachman and Palmer
(1996) and modified as test task characteristics (Bachman and
Palmer, 2010), was developed based on “test method facets
model” presented by Bachman (1990). This notion provides a
means for systematically describing various characteristics of
tests and testing procedures. It is primarily intended as a tool
for designing and constructing language tests, but because of
the detailed descriptions of tasks, the model also provides a
useful research paradigm for empirical validation studies and
other research. A number of language testing studies have made
use of this model (e.g. Bachman et al., 1996; Clapham, 1996)
or other systems of describing test task characteristics (e.g.
Brown, 1989; Freedle & Kostin, 1993; Perkins et al., 1995).
Most researchers agree that reading is rapid, purposeful, and
interactive (Alderson, 2000). Specifically, this interactivity oc-
curs at two levels. First, regardless of the exact components or
levels that they posit, models of the reading process depict
these components or levels as interacting with each other. Sec-
ond, reading is interactive in that the reader’s background
knowledge and other attributes interact with the content of the
text. Given that reading involves interaction between readers
and texts, it logically follows that characteristics of both the
reader and the text will affect the reading process (Alderson,
2000). Reader characteristics have been examined in previous
research, in terms of both readers’ background characteristics
and their abilities (Bachman, 1990). Of greater relevance here
is the point that text characteristics are describable in a system-
atic manner, using frameworks such as the Bachman and Palmer
(1996) model. This has been demonstrated in a number of pre-
vious studies, some of which have also examined interactions
between these characteristics and examinee performance as
indicated by item difficulty and/or discrimination (e.g. Bach-
man et al., 1996; Clapham, 1996).
Method
Participants
To accomplish the purpose of the research, a sample of 50
male and female students participated in this study. The sub-
jects were between 19 and 28 years old. These participants
were studying English as a foreign language. After giving a
reading part of the IELTS test, 30 students whose scores were
between one standard deviation above and below the mean
were chosen as the subjects of this study. They were under-
graduate students from Azad universities of Tehran, the major
field of study was English translation and English language
teaching.
Instruments
First of all, a proficiency test on the basis of IELTS reading
section was held in order to homogenize the participants’ levels
of reading proficiency.
Five original tests of IELTS were utilized to achieve the
purpose of this study. These tests were administered in 2009
and 2011 (Cambridge IELTS 8, 2011; Cambridge IELTS 8,
2007). IELTS is a standard test which is provided to measure
the students English proficiency level. It consisted of four sec-
tions: 1) listening (40 items); 2) reading comprehension (40
items); 3) writing; and 4) speaking. The total time for this test
was 90 minutes. Since listening, writing and speaking sections,
not being the focus of the present study, were deliberately
omitted for practical reasons.
Procedures
To collect appropriate data for this study several steps were
taken. In order to ensure the homogeneity of the participants
and determine the participants’ reading proficiency level a
reading part of IELTS test was conducted at the outset of the
study. Those participants whose scores were one standard de-
viation above and below the mean were selected as a homoge-
nized group.
After homogenizing the students 30 of them were selected to
be included in this study. During five sessions, five IELTS tests
were taken from the participants. The purpose of this study was
to compare the performance of the participants on different
tasks of the IELTS.
Three kinds of tasks out of twelve different tasks were se-
lected and included in the study: 1) Matching heading; 2) True/
false/not given; 3) Multiple-choice. These tasks were separated
and all the participants were scored based on their tasks. The
statistics of the distribution of different tasks in the tests are as
follows:
Task Number of Questions
Matching heading 32
True/false/not given 27
Multiple-choice 23
Data Analysis
In order to test the Research Hypothesis, the following statis-
tical techniques were utilized:
1) Descriptive statistics and Standard deviation were utilized
to homogenize the participants based on their reading profi-
ciency.
2) A one-way ANOVA was run to compare the mean scores
of the groups, to see the effect of the task characteristics on the
IELTS tests.
3) Multiple-comparisons were conducted using Sheffe test to
distinguish the exact difference of the groups on the IELTS
tests.
The purpose of the present study was to determine the effect
of task characteristics on the learners’ performance in the read-
ing section of IELTS. Considering the purpose and based on the
problem specified above, the following research question was
raised:
Do task characteristics affect the learners’ performance in the
reading section of IELTS?
Considering this question the following null hypothesis was
formulated:
H0: Task characteristics DO NOT affect the learners’ per-
formance in the reading section of IELTS?
To test the above hypothesis, the researcher conducted a se-
ries of statistical procedures which are elaborated in the fol-
lowing parts of this section.

B. BEHFROUZ, E. NAHVI
In order to validate the IELTS test with Iranian students, the
test was administered to 30 subjects the same as our sample.
The reliability coefficient of this test is in Table 1.
To homogenize the group, they were pre-tested through the
reading section of IELTS. The descriptive statistics for the pre-
test is presented in Table 2. The mean score is 25.44, and the
SD is 3.27.
Figure 1 presents the bar graph of the scores with normal
curve in pretest.
After pretest the group took five different IELTS reading
tests. The intended task types were separated and scored dif-
ferently. The results were calculated based on the different task
types. The statistics of the raw scores of task types are as fol-
lows.
The descriptive statistics for the raw scores of matching tasks
is presented in Table 3 the mean score for the matching task is
24.16, and the SD is 2.64.
The bar graph of Figure 2 represents the scores with normal
curve in matching task.
The descriptive statistics for the raw scores of T/F/NG tasks
is presented in Table 4 the mean score for the matching task is
21, and the SD is 2.62.
The following bar graph, i.e. Figure 3 deals with the raw
scores of participants in T/F/NG task.
The descriptive statistics for the raw scores of multiple-
choice task is presented in Table 5 the mean score for the mul-
tiple-choice task is 15.83, and the SD is 3.16.
In Figure 4, some information about the raw scores of M/C
task has been stated.
After selecting task types they were scored. Since the number
of items for each task is not the same, all the scores were con-
verted to percentage.
The descriptive statistics for the percentage of the scores of
matching task is presented in Table 6 the mean score for the
matching task is 75.52, and the SD is 8.25.
The following Figure 5 is about the percentage of scores
with normal curve in matching task.
The descriptive statistics for the percentage of scores of
T/F/NG task is presented in Table 7 the mean score for the
T/F/NG task is 77.77, and the SD is 9.72.
Figure 6 is the bar graph of the percentage of the scores with
normal curve of T/F/NG task.
The descriptive statistics for the percentage of the scores of
multiple-choice task is presented in Table 8 the mean score for
the multiple-choice task is 68.84, and the SD is 13.75.
Figure 7 also stated the percentage of the scores in M/C task
as follow:
A one-way ANOVA was run to compare the mean scores of
the percentage of the scores. The descriptive statistics for
one-way ANOVA is presented in Table 9. The mean scores for
the matching, T/F/NG and multiple-choice tasks are 75.52,
77.77, and 68.84 respectively. As it can be figured out the par-
ticipants did better in T/F/NG then matching and multi-
ple-choice.
As displayed in Table 10, the F-observed value is 5.52. This
amount of F-value at 2 and 87 degrees of freedom is higher
than the critical value of F, i.e. 3.11. As it is shown the prob-
ability level is lower than .05 i.e. .006.
Therefore it can be concluded that task type has a significant
effect on the participants’ reading skill. So our null hypothesis
is rejected. We can conclude that task type has a significant
effect on the Iranian EFL learners’ reading skill.
Table 1.
Reliability coefficient of the pretest.
NMinimum Maximum Mean Std. Deviation Reliability
Coefficient
IELTS 3025 35 30.46 4.77 .83
Table 2.
Descriptive statistics of the pretest.
Valid 50
N Missing 0
Mean 25.44
Std. Error of Mean .46
Median 25
Mode 25
Std. Deviation 3.27
Variance 10.7
Skewness .39
Std. Error of Skewness .33
Kurtosis −.05
Std. Error of Kurtosis .66
Range 13
Minimum 20
Maximum 33
Sum 1272
PRETEST
34.032.030.028.026.024.022.020.0
Frequency
16
14
12
10
8
6
4
2
0
Std. Dev = 3.27
Mean = 25.4
N = 50.00
Figure 1.
Bar graph of the pretest.
In order to distinguish the difference among the performance
of the group on different task types, multiple comparisons are
conducted using Sheffe test. As it is represented on Table 11
the main difference is between multiple-choice group and the
other groups.
Discussion
As mentioned previously, the present study aimed at provid-
ing answers to the addressed research question:
“Do task characteristics affect the learners’ performance in
34

B. BEHFROUZ, E. NAHVI
Table 3.
Descriptive statistics of raw scores of matching.
Valid 30
N
Missing 20
Mean 24.16
Std. Error of Mean .48
Median 24
Mode 25
Std. Deviation 2.64
Variance 6.97
Skewness .1
Std. Error of Skewness .42
Kurtosis −.13
Std. Error of Kurtosis .83
Range 11
Minimum 19
Maximum 30
Sum 725
Matching
30.028.026.024.022.020.0
Frequency
10
8
6
4
2
0
Std. Dev = 2.64
Me an = 24.2
N = 30.00
Figure 2.
Bar graph of the raw scores of matching.
the reading section of IELTS?”
The comparison made between matching and T/F/NG had
revealed that there is not a significant difference between the
performances of the participants on the test in these two groups,
the probability level is .723. And the comparison made between
matching and multiple-choice showed that the performance of
the participants on the test is not significant too, the probability
level is .063. The difference between the performance of the
participants of the test, in T/F/NG and multiple-choice is sig-
nificant i.e. .008.
As the results showed, task characteristics have a significant
effect on the learners’ performance in the IELTS reading tests.
All teachers and test developers ought to pay attention to test
characteristics especially those which were utilized in this study,
i.e. the degree of being communicative, and the difference in
Table 4.
Descriptive statistics of raw scores of T/F/NG.
Valid 30
N
Missing 62
Mean 21
Std. Error of Mean .47
Median 21
Mode 21
Std. Deviation 2.62
Variance 6.89
Skewness −.18
Std. Error of Skewness .42
Kurtosis −.94
Std. Error of Kurtosis .83
Range 9
Minimum 16
Maximum 25
Sum 630
T/F/NG
26.024.022.020.018.016.0
Frequency
10
8
6
4
2
0
Std. Dev = 2.63
Mean = 21 .0
N = 30.00
Figure 3.
Bar graph of the raw scores of T/F/NG.
expected response. Teachers should be aware of task character-
istics and try to use the test which best suit the students’ needs.
The participants did better in T/F/NG then matching and
multiple-choice.
The participants on the T/F/NG and matching did better than
the multiple-choice group. It can be inferred that multiple-
choice task is a kind of traditional test that students need more
practice in mastering the strategies to answer.
Conclusion
Based on the results, it can be concluded that the participants
performed differently on the IELTS test due to the different
task types utilized as the treatment; therefore, the null hypothe-
sis of the study is rejected.

B. BEHFROUZ, E. NAHVI
Table 5.
Descriptive statistics of raw scores of M/C.
Valid 30
N
Missing 62
Mean 15.83
Std. Error of Mean .57
Median 16
Mode 20
Std. Deviation 3.16
Variance 10
Skewness −.12
Std. Error of Skewness .42
Kurtosis −1.38
Std. Error of Kurtosis .83
Range 9
Minimum 11
Maximum 20
Sum 475
M/C
20.018.016.014.012.0
Frequency
10
8
6
4
2
0
Std. Dev = 3.16
Mean = 15.8
N = 30.00
Figure 4.
Bar graph of the raw scores of M/C.
“Task characteristics affect the learners’ performance in the
reading section of IELTS?”
Implications and Applications
The findings confirm the previous research on the effect of
task characteristics on the reading ability.
These findings have some implications and applications:
Testing characteristics have a significant effect on the per-
formance of the learners on the test. Task types should be in
accordance to the class procedures and the specified goals. In a
well-organized curriculum all the teaching materials and pro-
cedures should be predefined and based on the goals of the
course the appropriate task type should be selected. In choosing
a suitable test, some of the characteristics of tasks which are
setting, input and expected response should be considered.
Table 6.
Descriptive statistics of the percentage of the scores of matching.
Valid 30
N
Missing 62
Mean 75.52
Std. Error of Mean 1.5
Median 75
Mode 78.13
Std. Deviation 8.25
Variance 68.07
Skewness .1
Std. Error of Skewness .42
Kurtosis −.13
Std. Error of Kurtosis .83
Range 34.38
Minimum 59.38
Maximum 93.75
Sum 2265.63
MATCHING
95.090.085.080.075.070.065.060.0
Frequency
10
8
6
4
2
0
Std. Dev = 8.25
Mean = 75.5
N = 30.00
Figure 5.
Bar graph of the percentage of scores of matching.
The results of the current study can be of use in all educa-
tional centers. It has direct or indirect implications and applica-
tions in teaching, learning, test development, syllabus design,
and material development. Considering the important decisions
which are made based upon the tests and their possible influ-
ence on students’ fates, there is no doubt that some task types
which are formative, motivating, and anxiety-reducing methods,
should be incorporated into the syllabuses and lesson plans of
the educational centers. Since in recent years, the focus has
shifted from the products of instruction to the processes, the
type of task used in the classes should be considered. Hence,
considering the instructional as well as evaluation value of task
and their positive effects, course designers should plan for their
proper use in instruction.
In our own experience in test development, we have found
that task characteristics are useful for describing the character-
36

B. BEHFROUZ, E. NAHVI
Table 7.
Descriptive statistics of the percentage of scores of T/F/NG.
Valid 30
N
Missing 62
Mean 77.77
Std. Error of Mean 1.77
Median 77.77
Mode 77.78
Std. Deviation 9.72
Variance 94.6
Skewness −.18
Std. Error of Skewness .42
Kurtosis −.94
Std. Error of Kurtosis .83
Range 33.33
Minimum 59.26
Maximum 92.59
Sum 2333.33
TFNG
95.090.085.080.075.070.065.060.0
Frequency
10
8
6
4
2
0
Std. Dev = 9.73
Mean = 77.8
N = 30.00
Figure 6.
Bar graph of the percentage of the scores of T/F/NG.
istics of task types. Language teachers intuitively realize that
the types of tasks that are included in language tests are impor-
tant. Frequently one of the first questions asked in our testing
classes is about our opinions of the best way to test a particular
area of language ability. These teachers may not yet have re-
fined their thinking as to what specific characteristics might
make one test task more appropriate for a given purpose than
another. However, they are clearly aware that the way they test
language ability affects how their students perform on language
tests and hence the quality of the information obtained from
their tests.
There is also considerable research in language testing that
demonstrates the effects of task type on test performance. This
research and language teachers’ intuitions both lead to the same
conclusion: the characteristics of the tasks used are always
likely to affect test scores to some degree, so that there is virtu-
Table 8.
Descriptive statistics of the percentage of the scores of M/C.
Valid 30
N
Missing 62
Mean 68.84
Std. Error of Mean 2.51
Median 69.56
Mode 86.96
Std. Deviation 13.75
Variance 189.14
Skewness −.12
Std. Error of Skewness .42
Kurtosis −1.38
Std. Error of Kurtosis .83
Range 39.13
Minimum 47.83
Maximum 86.96
Sum 2065.22
MC
85.080.075.070.065.060.055.050.0
Frequency
10
8
6
4
2
0
Std. Dev = 13.75
Mean = 68.8
N = 30.00
Figure 7.
Bar graph of the percentage of the scores of M/C.
ally no test that yields only information about the ability we
want to measure. The implication of this conclusion for the
design, development, and use of language tests is equally clear:
since we cannot totally eliminate the effects of task characteris-
tics, we must learn to understand them and to control them so
as to insure that the tests we use will have the qualities we de-
sire and are appropriate for the uses for which they are in-
tended.
When we think about the different types of tasks that are
commonly used for language tests, we realize that they are not
single wholes, but rather collections of characteristics. Consider,
for example, the multiple-choice item, which has often been
used as a kind of test task: multiple-choice test items vary in a
number of ways, such as in their length, syntactic complexity,
level of vocabulary, topical content, and type of response re-
quired, to name but a few. Similarly, the “composition” task

B. BEHFROUZ, E. NAHVI
38
Table 9.
Descriptive statistics of one-way ANOVA.
N Mean Std. Deviation Std. Error Minimum Maximum
Matching 30 75.52 8.25 1.5 59.38 93.75
T/F/NG 30 77.77 9.72 1.77 59.26 92.59
Multiple-choice 30 68.84 13.75 2.51 47.83 86.96
Total 90 74.04 11.36 1.19 47.83 93.75
Table 10.
One-way ANOVA.
Sum of Squares Df Mean Square F Critical F Observed Sig.
Between Groups 1295.93 2 647.96 3.11 5.52 .006
Within Groups 10202.96 87 117.27
Total 11498.89 89
Table 11.
Sheffe test.
(I) GROUPS (J) GROUPS Mean Difference (I-J) Std. Error Sig.
Matching T/F/NG −2.25 2.79 .723
Matching Multiple-choice 6.68 2.79 .063
T/F/NG Multiple-choice 8.93(*) 2.79 .008
Note: *The mean difference is significant at the .05 level.
type encompasses a wide variety of prompts that can differ in
characteristics such as the intended audience, purpose, and
specific organizational pattern requested. It is thus clear that we
cannot characterize test tasks precisely if we think of them only
as holistic types. Therefore we need a descriptive framework of
task characteristics.
This framework also enables test developers to better under-
stand which specific characteristics can be varied, and to sug-
gest ways in which these can be varied, thus providing a valu-
able tool for tailoring tests appropriately for specific groups of
test takers, and so enabling them to perform at their best.
The researcher therefore believes that attempting to control
the test task characteristics by design provides the most useful
and practical means for maximizing the usefulness of our tests
for their intended purposes.
Suggestions for Further Research
The results of the study can help teachers, administrators, test
task writers, curriculum designers and EFL textbook/material
writers in decision making in the classroom teaching, test de-
velopment, curriculum design, and textbook writing. The study
is intended to serve as a guide to orient teachers to task selec-
tion.
The researcher hopes that the results are actively and care-
fully employed by teachers, and teachers and test task writers
benefit from the results when they choose activities, and it also
gives them insight to what goes on in the learning process, thus
shedding much needed light on some of the dark areas created
in the classroom. It is also very practical in the sense that by
showing the different phases of the process, it gives teachers
and test task writers clues to precisely what the difficulties are,
and, indirectly, what they can do to help students overcome
them and it also may assist test task writers and curriculum
designers when they are trying to gauge the level and impact of
a task and to determine just what difficult or easy means. Ex-
amining task types helps educators get a better idea of how
students approach and complete tasks and what they learn by
doing so.
In conducting any research project, there exist a number of
limitations which can be overcome in future studies. It is sug-
gested that the present study can be replicated with various
methods and techniques of testing, larger samples, different
levels of language proficiency and other language skills. The
followings are some further suggestions:
1) Is there any relationship between consciousness-raising
and task type?
2) Is there any relationship between task characteristics and
students’ personality character?
3) Do task types show the exact weaknesses of the learners’
reading comprehension skill?
4) Is there any relationship between task type and learners’
learning styles?
REFERENCES
Alderson, J. C. (2000). Assessing reading. Cambridge: Cambridge Uni-
versity Press. doi:10.1017/CBO9780511732935
Amiri, M., & Maftoon, P. (2010). Awareness of reading strategies
among Iranian high school students. In EDULEARN10 Proceedings
CD (pp. 6782-6791). Valencia: IATED.
Bachman, L. F. (1990). Fundamental considerations in language test-
ing. Oxford: Oxford University Press.
Bachman, L. F., & Palmer, A. S. (1996). Languag testing in practice.
Oxford: Oxford University Press.
Bachman, L. F., & Palmer, A. S. (2010). Language assessment in prac-
tice. Oxford: Oxford University Press.
Bowen, J. D., Maadsen, H., & Hilferty, A. (1985). TESOL techniques
and procedures. Rowley: Newbury House.
Carr, N. T. (2006). The factor structure of test task characteristics and
examinee performance. Language Testing, 23, 269-289.

B. BEHFROUZ, E. NAHVI
doi:10.1191/0265532206lt328oa
Carrell, L. P. (1991). Second language reading: Reading ability or lan-
guage proficiency? Applied Linguistics, 9, 233-242.
Celce-Murcia, M., & Larsen-Freeman, D. (1999). The grammar book:
An ESL/EFL teacher’s course. Boston, MA: Heinleand Heinle.
Chastain, K. (1988). Developing second-language skills: Theory and
practice (3rd ed.). New York: Harcourt Brace Jovanovich.
Clapham, C. (1996). The development of IELTS: A study of the effect of
background knowledge on reading comprehension. Cambridge: Uni-
versity of Cambridge Local Examinations Syndicate.
Edigar, A. (2001). Teaching children literacy skills in a second lan-
guage. In M. Celce-Murcia (Ed.), Teaching English as a second or
foreign language (pp. 153-169). Boston: Heinle & Heinle.
Finocchiaro, M., & Bonomo, M. (1973). The foreign language learner:
A guide for teachers. NewYork: Regents.
Freedle, R., & Kostin, I. (1993). The prediction of TOEFL reading
item difficulty: Implications for constructvalidity. Language Testing,
10, 133-170.
Grabe, W. (1988). What every EFL teacher should know about the
reading in English. TESOL Journal, 31, 130-148.
Grabe, W. (1999). Developments in reading research and their im-
plications for computer-adaptive reading assessment. In M. Chal-
houb-Deville (Ed.), Issues in computer-adaptive testing of reading
proficiency (pp. 11-48). Cambridge: University of Cambridge Local
Examinations Syndicate.
Hadley, A. O. (2003). Teaching language in context (3rd ed.). Boston:
Heinle & Heinle.
Nunan, D. (1999). Second language teaching & learning. Boston, MA:
Heinle & Heinle.
Nuttall, C. (1996). Teaching reading skills in a foreign language. Ox-
ford: Macmillan.
Perkins, K., Gupta, L., & Tammana, R. (1995). Predicting item diffi-
culty in a reading comprehension test with an artificial neural net-
work. Language Testing, 12, 34-53.
doi:10.1177/026553229501200103
Richards, C. J., & Schmidt, R. (2002). Longman dictionary of lan-
guage teaching & applied linguistics. London: Pearson Education.
Krashen, D. S. (1988). Second language acquisition and second lan-
guage learning. Englewood Cliffs, NJ: Prentice-Hall International.
Genesee, F., & Upshur, J. A. (1996). Classroom-based evaluation in
second language education. Cambridge: Cambridge University
Press.