177

Developing Reading Comprehension Skills in EFL

University Level Students
K.M. Roebl and Connie Shiue, St.John's University, Taiwan

Abstract: This paper explores the development of reading comprehension in

university level EFL students. Many of these students appear to be able to read
with relative fluency, but a closer examination shows that they do not
understand the text. Comprehension skills are essential if learners are to
assimilate the meaning of the text and understand the content. Reading
comprehension is essential to learning in the modern academic situation.
Students can be assisted in the development of comprehension skills in various
ways, including the development of background knowledge, helping them to
ask pertinent questions of the text, making predictions, interpreting charts and
other illustrations contained in the text, as well as the way language is used in
the text. In this research, students are tested on reading material as a pre-test
and then after a treatment period of three months, in which the development of
their comprehension skills is emphasised, they are retested. A control group is
taught the same material using routine methods and without any emphasis on
comprehension, is given the same test at the same time as the experimental
group. The test results are analysed using the Standard Deviation and the Z
value and represented graphically before interpretation.

Introduction
Reading comprehension skills are important for students to become effective
readers (Grabe & Stoller, 2002). Reading begins with the decoding letters,
letter groups and the sounding out of words. Later, learners begin to read words,
sentences, picture books, short stories and other texts. Reading aloud helps
learners to develop their decoding skills which can be a valuable diagnostic
aid. This process concentrates on the development of fluency. The movement
from passive to active reading involves the development of reading
comprehension skills (Machado, 2010).

Reading comprehension is the ability to understand what we read where words

have context and texts have meaning. Reading comprehension skills allow us
to read proficiently, learn effectively and to conceptualize. These skills are,
basically, based on earlier stages of reading development, including oral
reading and reading fluency. Without developing these earlier reading skills,
students must continually focus on decoding letters and words, rather than
progressing to meaning and understanding (Grabe & Stoller, 2002). The key to
developing proficient reading skills in the early years of education is an even
earlier foundation in underlying language learning skills (Brewster & Ellis,
2002). Therefore, strong reading comprehension skills are viewed as being
 178

dependent on the strength of the cognitive strategies established in the early

years.

Literature Review
Concepts of reading comprehension have changed dramatically over the
decades. Theories of language learning have again shifted dramatically during
the latter part of the 20th century (Crain, 2011). We have moved from a
behavioural perspective, which dominated the field from the turn of the century
to the seventies and eighties to a holistic or interactive approach, which began
in the late eighties, and continues to shape our thinking about reading
comprehension today. Some researchers view reading as a cognitive,
developmental, and socially constructed task that goes beyond understanding
the words on a page (Hedgcock and Ferris, 2009). In the past, reading was
considered a relatively static activity. General meaning was imbedded in the
text, and the readers job was to understand what was being transmitted via the
words on the page. Current research views reading as a more dynamic process
in which the reader constructs meaning based on information s/he gathers
from the text. Katherine Maria (1990) defines reading comprehension as:
holistic process of constructing meaning from written text through the
interaction of (1) the knowledge the reader brings to the text, i.e. word
recognition ability, word knowledge, and knowledge of linguistic conventions;
(2) the readers interpretation of the language that the writer used in
constructing the text; and (3) the situation in which the text is read. (p. 14-15)
University-level reading is much more sophisticated than at high school, and in
a special course load, students may encounter and face many more literary
genres than ever before. They may be asked to read, comprehend, and apply
them in a meaningful way. Understanding these texts are essential for academic
success, yet in an average class, attention will not be given to reading strategy
training which may be important for the language learning tasks.

Metacognition involves several elements in the reading tasks. They are: (1) the
ability to recognize errors or contradiction in text, (2) the understanding of
different strategies to use with different kinds of text, and (3) the ability to
distinguish important ideas from unimportant ones (Nist and Mealey, 1991).
While research suggests that many university level students lack metacognitive
skills (Baker, 1985), intervention studies also show that university students can
try to learn and understand their level of text comprehension by using different
strategies. Studies also reveal that university-age students are highly motivated
to use different strategies than younger, less experienced students. Older
students seem better able to regulate and control their understanding than do
younger children as children become older, their capacity to use
metacognitive skills increases, and their reasons for not using these skills
change (Nist and Mealey, 1991). There are a number of reading strategies,
which can help university students to improve both comprehension and
metacognition.
 179

According to reading specialist John McNeil (1992), schemata are the readers
concepts, beliefs, expectations, processes virtually everything from their
past experiences are used in making sense of reading. In reading, schemata are
used to make sense of text; the printed word evokes the readers experiences,
as well as past and potential relationships (p.20).

Effective teaching of reading comprehension necessitates an understanding and

analysis of its nature and components, including both text and reader variables.
Grabe (1977) put it in this way: The central components of reading processing
include: orthographic processing, phonological coding, word recognition
(lexical access), working memory activation, sentence parsing, propositional
integration, propositional text-model formation, text-model development, and
the development of an appropriate situation model (mental model). (p.9). We
need to understand when and why some readers fail to generate appropriate
situation models of the text they read in spite of adaptation applied to the
linguistic features (Grabe, 1997).

Reading comprehension skills are essential for meaningful and effective

reading. Early reading is grounded in strong cognitive skills, i.e. learning styles
such as auditory analysis, sound blending and segmenting, memory and
visualization. Therefore, the key to improving reading comprehension skills is
to attack weak language learning skills at the foundational level.

Effective reading comprehension requires not only accurate reading skills but
also automatic and fluent reading ability. Many struggling university level
students have difficulty moving to a level of automaticity and fluency that
allows them to comprehend what they are reading. Automaticity is the ability
to identify, at the single word level quickly, accurately and effortlessly. The
speed and accuracy with which single words are identified is taken as a
predictor of text comprehension (Wallace, 2010), however, reading fluency
involves not only automatic word identification, but also the application of
prosodic features such as rhythm, intonation and phrasing. Wood, Flowers, and
Grigorenko (2001) suggest that fluency involves the prediction of what comes
next in the text. They also say that reading speed and practice are not enough to
promote fluency and comprehension. The ability to predict what comes next
improves reading speed and is important for text comprehension.

TESOL research was influenced by Krashens hypotheses on language

acquisition, and the The Schema Theory on reading comprehension.
Increasingly, empirical research attests to the importance of schemata in
reading comprehension. Most of the research involved reading comprehension
in the first language, although the insights were adapted to the needs of second
language reading comprehension studies. Attention is given to interactive
approaches to reading, which suggest that reading comprehension is a
combination of word identification and interpretation. Grabe (1991) lists the
 180

five most important areas of current research: Schema theory, language skills
and automaticity, vocabulary development, comprehension strategy training,
and reading writing relations (p.375).

The principle of reading comprehension is the cognitive tasks involved in

reading as well as the various activities teachers use in teaching reading
comprehension. Current research suggests that lack of automaticity in lower-
level processing (i.e. automatic lexical access through bottom-up process)
leads to poor reading skills. Consequently, most current versions of interactive
approaches to reading have taken a strong bottom-up orientation to the
processing of lower-level linguistic structure through extensive research of eye
movement. Research suggests that most words are recognized before higher-
level (non-automatic context information can be used to influence lexical
access. (ibid: 385)

Methodology

In this study, data collected from two sets of tests were used. The first tests
were carried out during the period from September 2011 to December 2011,
and the second series of tests were carried out from February 2012 to May
2012. Although the work taught progressed, the same test was used to evaluate
the students. The students were evaluated on four occasions, twice for each
period. In this study, the first and last test is used to establish whether the
students progressed over the study period of eight months.

As this experiment uses two different groups of students (the experimental

group and a control group), it is necessary to use a calculation that enables a
comparison of the two groups to be made. The standard deviation and Z-Values
provide such a means where the scores of the four tests taken by the two
groups can be compared. The Z-Values and the Standard Deviation calculations
enable the researcher to take factors such as the distribution of the scores
obtained by the groups into consideration. This enables a more meaningful
comparison to be made than by simply looking at raw percentages. The higher
the Z-Value obtained, the better the individual students performance (Agresti
& Finlay, 2009).

Below, a discussion of the two calculations and their meaning is given.

The Standard Deviation and the Z value

The analysis here uses the Standard Deviation which is the distribution round
the mean. It is a widely used statistical measurement in educational research.
The standard deviation is defined as the square root of the variance, which is
the average of the squares minus the square of the mean (Agresti & Finlay,
2009).
 181

The greater the standard deviation, the more widely the scores are scattered;
the smaller the standard deviation, the smaller the deviation of the scores from
the average.

The Z value is used to compare students performance. The Z value is the

distance from the mean in terms of standard deviation units. The higher the Z
value the better a students performance (Agresti & Finlay, 2009).

Participants
The research participants are members of two classes of first year students at a
general university located in the northern part of Taiwan.
Class A is the experimental group and Class B is the control group. Both
classes are at the same university and both classes consist of 61 students.

The Experimental Group

The raw scores of the experimental group indicate improvement over the
period of the study as indicated in graph form in Figure 1.

Figure 1. Experimental Group. Series1 (First test) Series 2 (Final test)

According to the raw score data, there was some improvement, as the
difference between the highest score and the lowest score, narrowed from 60 to
38. These scores also show that the lowest score in the first test was 20% and
the highest was 80%; the mean was 61.73%. The lowest in the second test was
59% while the highest was 97% and the mean was 81.36%. The difference
between the highest and lowest scores diminished, as did the distribution
around the mean. The difference in the first test is 16 and in the last test 38.
This indicates a significant drop in the difference between the highest and
lowest scores, suggesting an improvement in the weaker scoring students. It is
apparent, then that, according to the raw scores, the experimental group
improved and benefited by the work done during the experimental period.
 182

The Control Group

The control group is a similar group of students in terms of size and
composition to the experimental group.

Figure 2. Control Group. Series1 (First test) Series 2 (Final test)

The raw scores indicate a lower set of scores in the first test. The first test
indicated a low of 26% and a high of 72% with a difference of 46 and a mean
of 56.206%. In the final test, the lowest score 45%, with a highest of 84%. The
difference is 39, with the mean 66.81% as shown in Figure 2. As with the
experimental group, the raw scores indicate an apparent improvement.

Data Analysis
After collection, the data was analysed using the standard deviation and Z-
value, as mentioned above. The table below shows the mean Z-Values arranged
in descending order.

Table 1. Z-Value means arranged according to descending order

Test Z-Value
Experiment test four 61.8
Control test one 61.50
Experimental test one 61.27
Control test four 61.005

These values indicate that the experimental group actually improved. The
fourth test gained a mean Z-value of 61.8, while the first test gained a mean Z-
value of 61.27. There is a positive difference of 0.53, indicating an
improvement.

The control group, declined on their Z-values. The fourth test indicates a mean
Z-value of 61.005, while the first test indicates a mean Z-value of 61.50. There
 183

is a negative difference on the Z-value of -0.495, indicating an overall decline

in their average performance, though individual students may have improved.
The Z-value, therefore gives a somewhat different picture regarding the
students actual performance, from test-to-test, to the raw scores.

Findings
The results, using the Z-Value, indicate that the experimental group performed
better than the control group. It is apparent that the control group declined in
their over-all performance, an indication that is cause for concern, especially as
it appears that the students actually lost ground, given that the same test was
used in the pre-test and the post-test.

These findings indicate that the methods used by the researchers in the
experimental group actually benefited the class.

Recommendations
The control group, as shown by the Z-Value, declined in their performance
over the term, while the result for the experimental group indicates an
improvement. This suggests that there is value to be accrued in reading-
comprehension type teaching.

Further research is necessary to show why the control group declined and to
understand the gains made by the experimental group.

Conclusion
As shown in the findings, the Z-Value calculations show that the experimental
group improved over the period of the experiment. The use of reading and
reading comprehension can, therefore, be a valuable aid to language teaching.
Such methods appear to help students gain understanding, as they are able to
see the language in context. This is in keeping with the theoretical review and
recent research, as shown in the literature review.

It is apparent that being able to understand the text (Maria, 1990), as tested in
the two tests, comprises a number of reading skills, including prediction and
schema (McNeil, 1992), as discussed in the literature review, above.
During the tests, the students were given a time limit, which means that they
had to work quickly and as efficiently as possible. Those who had been taught
using reading comprehension methods, were clearly, more able to read and
work efficiently than the control group who were not taught using these
methods. Being able to make inferences and predictions is clearly important in
 184

enabling learners to develop their understanding of the target language. This is

shown in the better scores obtained by the experimental group.

University level students often perceive the great importance of reading as well
as the escape provided by reading. Therefore they develop personal
approaches to reading development, combining useful instruction, individual
and personal persistence for learning and extended practice. In other words,
they develop their own content-based language instruction.

We are encouraged by the fact that many students learn to read well and many
develop a reading habit and read for pleasure. In spite of the fact of the
dilemmas that teachers encounter, the difficulties that learners face and the
many unanswered problems that emerge from research findings, many students
become proficient readers. Thus, it is essential to explore how learning and
teaching can be carried out more effectively and to find ways to make
instruction work better in future.

References
Agresti, A & Finlay, B. (2009). Statistical Methods for the Social Sciences 4th
Edition. New Jersey: Pearson, Prentice Hall.

Baker, L. (1985). Differences in the standards used by college students to

evaluate their comprehension of expository prose. Reading Research
Quarterly, 20, 297-313.

Brewster, J & Ellis, G. (2002). The Primary English Teachers Guide.

Edinburgh Gate: Pearson.

Crain, W. (2011). Theories of Development Concepts and Development (6th

Edition). Boston: Pearson.

Diaz-Rico, L. T. (2004). Teaching English Learners Strategies and Methods.

Boston: Pearson.

Dornyei, Z. (2011). Research Method in Applied Linguistics. Oxford: Oxford

University Press.

Grabe, W. (1997). Reading research and its implications for reading

assessment (Language Teaching Resource Center Paper). Flagstaff:
Northern Arizona University.

Grabe, W. (1991). Current developments in second language reading research.

TESOL Quarterly. 25 (3): 375-406.
 185

Grabe, W. & Stoller, F. L. (2002). Teaching and Researching Reading. Harlow:

Pearson.

Hedgcock, J. S. & Ferris, D. R. (2009) Teaching Readers of English Students,

Texts and Contexts. London: Routledge.

Johnson, K (Ed). (2005). Expertise in Second Language Learning and

Teaching. Basingstoke, Hampshire: Palgrave, Macmillan.

Kern, R. (2000). Literacy and Language Teaching. Oxford, Oxford University

Press.

Machado, J. M. (2010). Early Childhood Experiences in Language Arts

Early Literacy (9th Edition). Belmont, CA: Wadsworth.
Maria, K. (1990). Reading Comprehension Instruction, Issues and Strategies.
Parkton, MD: York Press.

McNeil, J.D. (1992). Reading comprehension new directions for classroom

practice (3rd. Ed.) Los Angles: U. of California.

Nist, S.L. & Mealey, D.L. (1991). Teacher-directed comprehension strategies.

In R. Flippo & D. Caverly (Eds.). Teaching reading and study strategies at
the college level. Newark, D.E.:International Reading Association.

Wiersma, W & Jurs, S. G. (2005). Research Methods in Education an

introduction 8th Edition. Boston: Pearson Ltd.

Wood, F.B., Flowers, L. and Grigorenko, E. (2001). On the functional

neuroanatomy of fluency or why walking is just as important to reading as
talking is. In M. Wolf (Ed.), Dyslexia, fluency, and the brain Timonium,
MD: York Press.
 186

Appendix

Results of the Four Tests

Experimental Group Control Group

Test 1 Test 4 Test 1 Test 4
Test 1 Z-Value Test 4 Z-Value Test 1 Z-Value Test 4 Z-Value
65 61.0074 87 61.3043 61 61.0029 81 61.0060
72 61.1309 84 61.1 60 61.006 78 61.00087
65 61.0074 71 61.0129 41 61.0199 60 61.00018
54 61.00063 77 61.06 50 61.0018 65 61.1486
52 61.194 74 61.02 49 91.0049 56 61.0045
63 61.057 94 61.00386 56 61.043 71 61.00046
72 61.1309 91 61.012 30 61.000026 56 61.0045
54 61.349 76 61.0478 41 61.000094 78 61.0002
70 61.0332 79 61.1398 51 61.0182 71 61.00046
61 61.000066 78 61.90 46 61.0028 56 61.0045
63 61.057 97 61.093 50 61.0018 73 61.0024
49 61.201 82 61.010 41 61.000094 72 61.0014
78 61.165 87 61.304 52 61.027 66 61.0031
40 61.00162 59 61.092 43 61.000158 52 61.00022
80 61.00004 94 61.00386 72 61.00008 84 61.000017
52 61.194 71 61.0129 49 61.0049 58 61.000096
74 61.044 85 61.0738 66 61.02050 60 61.00018
48 61.022 76 61.609 39 61.000049 52 61.00022
51 61.314 73 61.021 47 61.001088 51 61.000077
52 61.194 76 61.609 54 61.001225 50 61.000170
73 61.25 85 61.0738 60 61.006 76 61.00029
20 61.008 94 61.00386 51 61.0182 63 61.000016
71 61.66 91 61.012 50 61.0018 61 61.000001
58 61.147 79 61.1398 42 61.003 52 61.00022
56 61.000017 84 61.1 62 61 80 61.000002
65 61.00063 85 61.0738 53 61.000154 80 61.000002
70 61.0332 88 61.02941 62 61 74 61.00458
71 61.002895 83 61.01190 63 61.0012 78 61.0002
63 61.057 81 61.016949 61 61.0029 52 61.000222
64 61.35 84 61.1 61 61.0029 75 61.000002
63 61.057 88 61.02941 60 61.006 62 61.000355
54 61.00063 74 61.02 61 61.0029 45 61.000167
60 61.785 76 61.0478 48 61.000158 71 61.00046
51 61.314 72 61.176 47 61.001088 70 61.000002
60 61.785 79 61.1396 48 61.000158 69 61.002785
64 61.35 82 61.010 60 61.006 59 61.000374
53 61.048 70 61.0752 51 61.0182 72 61.0014
75 61.152 86 61.0044 70 61.0315 76 61.000015
56 61.6129 76 61.0478 51 61.0182 73 61.000246
55 61.18 78 61.090 50 61.00187 66 61.00316
55 61.18 70 61.00311 60 61.006 77 61.000065
75 61.152 91 61.012 61 61.0029 70 61.000002
77 61.08 93 61.0062 55 61.0714 80 61.000002
75 61.152 86 61.0044 72 61.0066 72 61.0014
63 61.057 85 61.0738 34 61.0032 69 61.002785
64 61.35 88 61.02941 55 61.0714 73 61.00246
48 61.022 64 61.0193 46 61.00282 50 61.000170
66 61 95 61.0017 69 61.00093 81 61.000013
62 61.36 88 61.02941 58 61.0042 61 61.000001
71 61.39 81 61.016949 50 61.00187 62 61.000355
68 61.05 86 61.0044 51 61.0182 79 61.00002
64 61.35 75 61.036 50 61.00187 52 61.00022
52 61.194 68 61.0045 26 61.000048 54 61
65 61.0074 82 61.010 60 61.006 76 61.000015
71 61.000005 81 61.016949 60 61.006 65 61.1486
53 61.048 69 61.00088 45 61.043 62 61.00035
57 61.032 83 61.01190 52 61.027 75 61.000002
68 61.051 82 61.010 60 61.006 68 61.000615
72 61.011 85 61.0738 41 61.0199 63 61.000016
61 61.000066 83 61.01190 43 61.00015 73 61.00246
62 61.36 72 61.176 52 61.027 70 61.000002
Mean: Mean: Mean: Mean: Mean: Mean: Mean: Mean:
61.73 60.2686 81.36% 61.8498 56.206 61.50172 66.81% 61.005770
Difference: Difference: Difference: Difference:
60 38 46 39