Reading in a Foreign Language April 2014, Volume 26, No.

1
ISSN 1539-0578 pp. 29–48

Pleasure reading and reading rate gains

David Beglar
Temple University, Japan Campus
Japan

Alan Hunt
Kansai University
Japan

Abstract

This study investigated the effects of (a) the amount of pleasure reading completed, (b)
the type of texts read (i.e., simplified or unsimplified books), and (c) the level of
simplified texts read by 14 Japanese university students who made the largest reading
rate gains over one academic year. The findings indicated that the participants who made
the greatest fluency gains read an average of 208,607 standard words and primarily read
simplified texts up to the 1,600-headword level. This study also provides an empirically
supported criterion for the minimum amount learners should read annually (i.e., 200,000
standard words), provides direct evidence that simplified texts are more effective than
unsimplified texts for reading rate development, and is the first study to provide empirical
evidence that reading lower-level simplified texts within learners’ linguistic competence
is effective for developing the reading rates of Japanese learners at a lower-intermediate
reading proficiency level.

Keywords: pleasure reading, extensive reading, graded readers, reading rate, reading fluency

Second language (L2) reading authorities widely acknowledge that reading fluency is an
important aspect of skilled reading, but as Grabe (2009) has noted, “relatively little research on
fluency or fluency training has been conducted with L2 populations” (p. 294). Even though
several characteristics believed to lead to greater reading fluency (e.g., extensive engagement
with meaningful and communicative texts) are present in extensive reading and pleasure
reading,1 little empirical research on fluency development has been conducted with both types of
reading. One exception was a recent paper by Beglar, Hunt, and Kite (2012), who found that
large amounts of pleasure reading, particularly of highly comprehensible simplified graded
readers, resulted in significant reading rate gains. However, a number of important issues were
not examined in that study. Consequently, the primary purposes of this paper are to further
investigate the development of reading rate with highly successful foreign language learners and
to distinguish between the effects of the amount, types, and levels of texts on reading rate gains

http://nflrc.hawaii.edu/rfl
Beglar & Hunt: Pleasure reading and reading rate gains 30

across one academic year. In this article we operationalize reading fluency as increases in
reading rate accompanied by high levels of comprehension. Increased processing rate is a central
characteristic of fluency development and an important aspect of theories of fluency and
automatization2 in cognitive psychology (e.g., Anderson & Lebiere, 1998; Logan, 1997) and in
second language acquisition (e.g., DeKeyser, 2007; Segalowitz, 2010). Thus, we agree with
Breznitz (2006) that “fluency in reading is expressed by performance time” (p. xiii).3

Literature Review

Extensive Reading and Reading Rate Gains

Much of the published research on extensive reading and reading rate gains is plagued by
numerous problems that make the results difficult to interpret: (a) no standard metric, such as
standard words4 (Carver, 1982, 1990), is used across studies to measure the total amounts of
reading; (b) no information about piloting the instruments is reported, (c) the administration of
reading rate tests is not described clearly; (d) comprehension measures are rarely reported, even
though reading rate increases must be accompanied by high levels of comprehension to be
considered as fluency development (e.g., Pikulski & Chard, 2005); (e) little information is
presented about the readability of the reading rate passages themselves or about the students’
reading proficiency levels; and, (f) control groups are often absent (See Beglar, Hunt, and Kite
[2012] for a detailed discussion of these issues).

Bearing these potential limitations in mind, Table 1 provides a summary of empirical studies in
which the effect of extensive reading on reading rate was investigated. Participants in short-term
studies of less than one academic year (i.e., Imamura, 2012; Iwahori, 2008; Lai, 1993; Lao &
Krashen, 2000) read between 12,000 to 330,000 standard words (6-28 books; Iwahori [2008] did
not report how many books her participants read) and showed reading rate gains ranging from
10.16 wpm to 96 wpm. Participants in studies of one academic year or longer (i.e., Beglar, Hunt,
& Kite, 2012; Bell, 2001; Nishino, 2007; Robb & Susser, 1989; Sheu, 2003) read 136,000 to
approximately 400,000 standard words (9-42 books; Bell [2001] and Robb and Susser [1989] did
not report the number of words or books their participants read) and achieved reading rate gains
ranging from 7.24 wpm to 65 wpm. The larger reading rate gains, particularly in the short-term
studies, might be overstated due to participants’ lack of familiarity with the pretests, resulting in
artificially low initial reading rate estimates.

Two important conclusions can be drawn from Table 1. The first is that extensive reading has a
consistently beneficial effect on reading rate development and that this effect can occur in less
than one academic year. We would note that studies conducted for one year or longer show more
moderate reading rate gains and, because of their longitudinal nature, they might provide a more
accurate indication of the rate of reading improvement resulting from extensive reading. The
second general conclusion is that extensive reading is effective for students at low reading
proficiency levels, as most of the participants were initially reading below 100 wpm, a rate that is
far below the 200 wpm reading goal suggested by some second language reading authorities (e.g.,
Anderson, 2008, p. 67).

Reading in a Foreign Language 26(1)

 Beglar & Hunt: Pleasure reading and reading rate gains 31

Table 1. Past studies of extensive reading reporting reading rate gains

Length of the Mean pre-reading and post-
Researcher Participants Amount read treatment reading rates (reading rate
gain in words per minute)
Robb & Susser Unspecified number of first-year M = 641 pages of books designed for One academic 79.31 / 86.55 (7.24)
(1989) Japanese university students American teenagers year

Lai (1993) 207 grade 7-9 Hong Kong M = 16.0, 18.5, and 14.2 books read Four-week 7th grade 165 / 226 (61)
secondary school students (who by the 7th, 8th, and 9th grade summer reading 8th grade 85 / 181 (96)
completed the speed reading students, respectively program 9th grade 106 / 121 (15)
tests)

Lao & Krashen 91 first-year university students Six books; approximately 388,000 One 14-week 235 / 327 (92)
(2000) in Hong Kong who graduated words (329,800 standard words) semester
from a high school in which
English was the medium of
instruction.

Bell (2001) 14 elementary level learners in Unspecified; The reading program Two academic 68.10 / 127.53 (59.43)
Yemen was 36 hours. semesters

Sheu (2003) 31 second-year junior high GR Group: Nine graded readers Two academic GR Group 59.7 / 95.8 (36.1)
school students in Taiwan read BNESC Group: Nine books written semesters BNESC Group 98.6 / 136.0
graded readers (GR Group) for English speaking children (37.4)
34 students read books written
for native English speaking
children (BNESC Group)

Nishino (2007) Fumi and Mako, two Japanese Fumi 36 books (402,000 standard 2.5 years Fumi 72 / 137 (65)
junior high school students words) Mako 58 / 111 (53)
Mako 42 books (333,000 standard
words)
Iwahori (2008) 33 Japanese high school students 28 graded readers Seven weeks 84.18 / 112.82 (28.64)

Imamura 38 Japanese high school students Group that read more (n = 19): M = Seven months Group that read more 77.60
(2012) 45,447 words (38,630 standard /100.55 (22.95)
words) Group that read less 86.74 /
Group that read less (n = 19): M = 96.90 (10.16)
14,279 words (12,137 standard
words)

Beglar, Hunt, First-year Japanese university PR Group 1: M = 136,029.07 standard One academic PR Group 1 89.71 / 97.73
& Kite (2012) students: Pleasure reading (PR) words (M = 9.13 books; 439.43 year (8.02)
Group 1 (n = 23), PR Group 2 (n pages) PR Group 2 94.50 / 107.34
= 22); PR Group 3 (n= 35) PR Group 2: M = 158,993.56 standard (12.84)
words (M = 14.82 books; 840.36 PR Group 3 103.09 / 119.93
pages) (16.84)
PR Group 3: M = 200,170.00 standard
words (M = 24.34 books; 1,095.23
pages)

Reading Targets and Actual Amounts of Extensive Reading Completed

Second language reading authorities have repeatedly stated that reading fluency development is
built on a foundation of large amounts of reading. For instance, Grabe and Stoller (2011) stated
that “Most L2 readers are simply not exposed to enough L2 print (through reading) to build
fluent L2 processing” (p. 50). Day and Bamford (1998) suggested that reading one book per
week is a reasonable goal, provided that the books are short and easily comprehensible. Nation
(2009a, p. 50) proposed a target of 500,000 running words (~425,000 standard words) per year
and suggested that this rate be continued for several years (I. S. P. Nation [personal
communication, February 19, 2014] stated that this goal is not meant to be restricted to English
as a Foreign Language [EFL] contexts). Finally, for the purpose of vocabulary learning, Nation

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 32

and Wang (1999) concluded that, to obtain sufficient lexical repetition, L2 learners should read
one book per week at Levels 2 and 3, 1.5 books per week at level 4, and two books per week at
levels 5 and 6 as they progress through incrementally higher levels of extensive readers. Based
on research into fluency development conducted by cognitive psychologists (Anderson, 1987;
Logan, 1997), we believe that learners must read greater amounts to acquire and automatize less
frequent vocabulary.

Nation and Wang’s (1999) reading goals should provide for lexical and fluency development
over the long term; however, longitudinal empirical studies are needed to detail how much
learners read in various educational contexts and the rate at which their reading fluency develops.
Of the eight studies listed in Table 1, the three short-term studies in which students read the
greatest amount are Lao and Krashen (2000; 388,000 words in one semester; 329,800 standard
words), Iwahori (2008; 28 graded readers in seven weeks), and Lai (1993; 14.2-18.5 books in
four weeks). If such rates had been sustained over one academic year, these three groups of
students would probably have read considerably more than Nation’s 500,000-word annual goal.
However, these studies beg the question of whether EFL readers in non-intensive language
programs can sustain such amounts of reading over longer periods. This is an important issue,
given that “The ability to read extended texts for long periods of time is a hallmark of fluent
reading,” and that this ability “develops incrementally over a long period of time” (Grabe, 2009,
p. 311, italics added).

Longitudinal research on extensive reading presents a different picture than the three short-term
studies described above. It shows that, although students can read in a sustained fashion for one
or more years, the total amount read is less than what might be expected from the results of these
short-term studies. Robb and Susser (1989), Beglar, Hunt, and Kite (2012), and Burrows (2012)
each conducted studies of extensive reading over one academic year. As shown in Table 1, Rob
and Susser’s participants read an average of 641 pages, while the group that read the most in
Beglar, Hunt, and Kite’s (2012) study read approximately 200,000 standard words, an amount of
extensive reading also reported by Burrows. These amounts are well below those suggested in
the literature, and less than that reported by some of the shorter-term studies. The only study in
Table 1 that exceeded one year was Nishino’s (2007), in which two participants read
approximately 402,000 and 333,000 standard words in 2.5 years. In addition, though they did not
measure reading rate gains, Nishizawa, Yoshioka, and Fukuda (2010) conducted a 4-year study
to examine the long-term effects of extensive reading on improving TOEIC scores. They found
that 75% of the 37 Japanese technical college students with low starting reading proficiency read
more than 300,000 words (~255,000 standard words) over three years, and 50% had read more
than 690,000 words (~586,500 standard words) after four years. Despite these impressive results,
these amounts are still well below the yearly targets suggested by many reading authorities.

The Appropriate Level of Reading Materials for Reading Rate Development

A widespread assumption in the second language reading literature—but one that has yet to be
empirically demonstrated—is that texts designed to enhance reading fluency should be
dominated by known lexis and morpho-syntax and be easily comprehensible. For instance, Day
and Bamford (1998) stated that reading materials should be “well within the linguistic
competence of the students” (p. 8, italics in the original). Additionally, when using speed reading

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 33

to develop reading rate, Nation (2009a) emphasized that “there should be little or no unknown
vocabulary or grammatical features” (p. 2).

Comprehensibility is frequently defined from a lexical perspective; numerous researchers and

educators have stated that learners need to know between 95 to 100% of the lexis in a text for
successful extensive reading (Hu & Nation, 2000; Nation, 2009a). Furthermore, knowing at least
98% of the vocabulary in a text is necessary for unassisted comprehension (Hu & Nation, 2000)
and for providing learners a reasonable chance of inferring the meaning of unknown vocabulary
(Hirsh & Nation, 1992). For low- and intermediate-proficiency L2 readers, the lexical and
morpho-syntactic characteristics of simplified texts, in which large amounts of known lexis and
morpho-syntax are embedded repeatedly in meaningful, potentially engaging contexts, should
provide a more supportive environment for reading rate development than unsimplified texts.

Gaps in the Literature and Purposes of the Study

Given the widely varying empirical results shown in Table 1, the amount that must be read to
achieve adequate reading rate gains has not been sufficiently researched. One reason for this is
the imprecise way in which amount of reading is commonly measured in the extensive reading
literature. Six of the nine studies shown in Table 1 do not report the amount read or report it in
terms of pages and books read rather than running words or standard words. This is problematic
because the amount of text on a single page or in a book varies greatly. Thus, the first purpose of
this study is to determine the total number of standard words read by groups of learners who
made greater or lesser reading rate gains through pleasure reading over one academic year. These
data will allow us to arrive at a tentative criterion regarding the minimum annual amount of
extensive reading that lower-intermediate EFL learners need to read to achieve substantial
reading rate gains.

A second gap in the literature concerns the lack of empirical studies distinguishing the relative
contributions of the amount read and type of texts read (i.e., simplified versus unsimplified texts)
on reading rate gains. Thus, our second purpose is to determine whether the amount and type of
pleasure reading make independent contributions to reading rate gains.

A third gap in the literature concerns the lack of empirical support for the widespread belief that
easy simplified texts are more beneficial than more difficult simplified texts for reading fluency
development. While many second language reading authorities assume that easy texts are most
effective for fluency development, they cite neither theory nor empirical evidence to support this
position. Thus, the third purpose of this study is to investigate how various levels of simplified
texts affect reading rate gains and to provide empirical evidence in support of the use of
simplified texts that are easy relative to the learners’ receptive vocabulary knowledge.

Research Hypotheses

The following hypotheses were investigated in this study.

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 34

Hypothesis 1: Greater amounts of reading are associated with greater L2 reading rate gains. This
hypothesis is based on research indicating that the amount of processing is one key to fluency
development (e.g., See Logan, 1997, p. 139 for a summary). To date, only two researchers,
Iwahori (2008) and Beglar, Hunt, and Kite (2012) have provided empirical data that shed direct
light on this hypothesis.

Hypothesis 2: Participants who display greater reading rate gains read more simplified texts and
few or no unsimplified texts. This hypothesis is based on the idea that fluency is largely
developed by recycling language at multiple levels (e.g., orthography, lexis, morpho-syntax,
semantics, and genre) and that a greater amount of lexical recycling takes place in simplified
texts written within the first 2,000 high frequency words of English (Cobb, 2007; Nation &
Wang, 1999).

Hypothesis 3: Participants who make greater reading rate gains read a greater number of lower
level simplified books. Reading lower level books can provide two advantages. First, they can be
read more quickly, allowing participants to read a greater number of standard words over the
academic year. Second, these books provide greater repetition of high frequency lexis and syntax;
hence, learners encounter more opportunities for developing sight vocabulary and processing
larger linguistic units, such as collocations and lexical phrases, more rapidly and with less
cognitive effort.

Methods

Participants

The participants were 76 first-year Japanese students aged 19–20 (57 female and 19 male
students) attending a large, prestigious, private university in western Japan.5 All the participants
had studied English formally for six years in Japanese secondary schools, and they were enrolled
in one 90-minute listening and speaking course and one 90-minute reading course per week at
the time this study was conducted. The 90-minute elective reading courses met once a week 28
times over two semesters (i.e., one academic year). None of the participants reported having any
experience with either extensive or pleasure reading before attending this university. The
participants’ mean starting reading rate was approximately 97 wpm.

The participants were in three intact classes. One class engaged in a combination of intensive and
pleasure reading. These students translated two pages per week of the intensive reading text, The
History of European Fairy Tales (Brown, 1992), as homework and then presented their
translations in class. The instructor primarily explained the content of the stories and sometimes
commented on grammar and vocabulary. In addition to the intensive reading text, the
participants in this class also read self-selected books outside of class. The participants in the
other two classes engaged only in pleasure reading both inside and outside of class and were
instructed to read at least one book every two weeks.

The participants initially selected graded readers that were generally well below the 2,000
headword level (see the Appendix in Beglar, Hunt, and Kite [2012] for a list of these simplified

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 35

readers, their levels, and the number of standard words per page). Although the participants were
advised to read graded readers at around the 600-headword level, they were free to read books at
higher or lower levels. As a result, although some participants chose higher-level graded readers
early in the first semester, most chose some higher-level graded readers or unsimplified books
late in the second semester. By the end of the academic year, 47 (62%) of the 76 participants had
read 57 unsimplified texts, including Harry Potter and the Philosopher’s Stone (Rowling, 1997)
and Bridget Jones’s Diary (Fielding, 1996). They also read four instructor-selected graded
readers in the first semester and two in the second semester and completed a variety of
comprehension tasks. The participants’ out-of-class pleasure reading was regularly monitored by
having them submit written reports for each book they read. None of the participants engaged in
any reading activities designed to increase reading rate (e.g., timed or paced readings).

Instrumentation

Vocabulary Levels Test. A 24-item version of the Vocabulary Levels Test (Nation, 1990)
covering the second and third 1,000 word frequency levels was administered at the start of the
first semester in April to confirm that the participants had sufficient knowledge of the high-
frequency English vocabulary needed to read the graded readers and the reading rate test
passages with minimal difficulty. A longer version of the test was analyzed statistically and the
best-performing 24 items were selected for use in this study. The test was analyzed using
WINSTEPS 3.60.1 (Linacre, 2006) and was found to have a Rasch item reliability estimate
of .94.

Reading Rate Test. The participants completed a reading rate test as a pretest in May and again
as a posttest in December. This test provided estimates of the participants’ reading rates and
passage comprehension. The test consisted of four approximately 400-word passages selected
from Reading Power (Mikulecky & Jeffries, 1998). Each passage was recalculated using
standard word units (Carver, 1982, 1990) to increase measurement precision; as a result, the total
length of the four passages was determined to be 1,400 standard words. Students completed a
practice reading rate test prior to taking both the pretests and posttests to familiarize them with
the test format in order to obtain accurate measurements of their reading rates.

The reading rate test consisted of a double-sided page with the reading passage on the front and
eight objectively scored multiple-choice comprehension questions on the back of each passage
(four passages x eight comprehension questions per passage = 32 total questions); thus, the
instructors would have easily noticed any participant attempting to look back at the reading
passage while answering the comprehension questions.

The multiple-choice comprehension questions had four answer options (a-d). The first question
was about the topic using the stem “This passage is about” and the remaining questions asked
about specific details in the passage. Example stems are “Susan and Sam liked,” “At the pub,
there were some,” and “Susan and Sam thought the food was….” A Range (Nation & Heatley,
2002) analysis using the BNC word lists showed that 96.93% of the words used in the questions
were within the 2,000-word level. Eight native speakers of Japanese who were highly proficient
in English (TOEFL paper-based test score > 575) answered the comprehension questions without
viewing the reading rate passages to determine how many questions they could answer correctly.

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 36

A criterion of four or more persons answering the same question correctly was considered
evidence that the question could be answered without reading the accompanying passage. Six of
the 32 questions (18.75%) could be answered correctly without reading the passage. This figure
indicated that the participants had to read and comprehend the reading passages to surpass the
75% criterion for comprehension set in this study. The Rasch item reliability estimate for the 32
multiple-choice questions was .86.

The reading rate passages were selected for three reasons. First, their difficulty level was
considered to be well within the participants’ reading level. The Flesch Reading Ease estimate
was 85.5, the Flesch-Kincaid grade level was 3.3, and the first 2,000 words of the BNC plus
proper nouns provided 97.29% coverage of the reading rate test passages. Thus, the passages
were similar to a Level 4 Oxford Bookworms graded reader in terms of Flesch-Kincaid grade
level and lexical composition. Second, narrative passages were selected because the graded
readers the participants read were primarily narratives. Third, the lexical composition and Flesch
reading difficulty estimate of the reading rate passages were similar to those of most of the
graded readers the participants read (See Table 1 in Beglar, Hunt, and Kite [2012] for a detailed
breakdown).

Procedures

The 24-item Vocabulary Levels Test was administered during the second week of class in April,
and an initial practice reading rate passage was administered in the third week to familiarize the
participants with the procedure. The 32-item reading rate pretest was then administered over a
two-week period; two passages were administered during weeks four and five of the first
semester. In December of the same year, following the two-semester treatment, the instructors
re-administered the reading rate practice test and posttests using the same procedure as at the
start of the academic year. Reading comprehension measures were also obtained from the
reading rate pre- and posttests.

Preliminary Analyses

The Vocabulary Levels Test results indicated that the participants knew approximately 89% or
more of the items at the 2,000-word level and an average of approximately 75% of the items at
the combined 2,000 and 3,000 word levels. Given that 97.29% of the running words in the
reading rate test passages consisted of the 2,000 high-frequency words of English plus proper
nouns, the participants probably had sufficient lexical knowledge to read them easily; they
correctly answered an average of 83.7% (M = 26.78, SD = 2.72) of the reading rate passage
questions at the beginning of the study and 86.7% (M = 27.75, SD = 2.87) at the end. In addition,
the vast majority of graded readers available to them were written using between 500 to 1,900
headwords; thus, the participants likely met few unknown lexical items when reading at those
levels.

Addressing hypothesis 1 required classifying students into groups based on their reading rate
gains. To create the groups, the participants’ raw reading rate gain scores were converted into z-
scores, which were transformed to percentile ranks. The percentile ranks were then used to create
five groups based on their reading rate gains over the academic year: Groups 1-5 had reading rate

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 37

gains above the 78th centile, between the 56th and 77th centile, between the 33rd and 55th
centile, between the 20th and 32nd centile, and below the 20th centile, respectively.

In order to determine whether the reading rate gains of these five groups were statistically
distinct from one another, a one-way analysis of variance was conducted. The independent
variable, group, had five levels (the five percentile groups), and the dependent variable was
reading rate gain over the academic year. The assumptions for the analysis were met except that
the variances among the groups were unequal (Levene statistic (4,72) = 9.50, p < .001); therefore,
the Welch and Brown Forsythe tests were utilized. As both tests were statistically significant,
only the results of the Welch test are reported. The ANOVA was significant, F(4,34.36) =
178.61, p < .001, partial eta-squared = .91, so follow-up tests were conducted with Dunnett’s T3
test. All pairwise comparisons were significant at p < .001; thus, the Group 1 participants, who
are the focus of this study, made significantly greater reading rate gains than the other
participants.

Results

Hypothesis 1 was addressed by inspecting the descriptive statistics for the five groups, which are
displayed in Table 2. The total amount read (standard words total) indicates that more reading
generally resulted in greater reading rate gains. The participants in Group 1, who made the
greatest gains, also read the most (Mean number of total standard words read = 208,607; SD =
47,669), while the participants in Groups 2 and 3 read somewhat less, and the participants in the
bottom two groups (Groups 4 and 5), who displayed only slight increases or slight losses in their
mean reading rate gains, read the least.

Table 2. The average amount read and mean gain scores for the five groups
Group Standard Standard Ratio of Standard Simplified Unsimplified Total Total (Mean posttest
words words simplified to words books books books pages reading rate –
simplified unsimplified unsimplified total (mean) (mean) (mean) (mean) mean pretest
standard reading rate)
words Reading rate gain
in WPM
1 181,032 27,575 6.57:1 208,607 22.93 .40 23.33 1,116.67 (130.26-97.27)
(n=14) 32.99
2 144,315 32,175 4.49:1 176,490 18.27 .53 18.73 943.73 (115.61-96.90)
(n=15) 18.71
3 131,605 52,880 2.49:1 184,485 17.44 .75 18.19 905.75 (104.56-93.31)
(n=16) 11.24
4 101,923 60,215 1.69:1 162,138 13.75 1.13 14.88 749.00 (94.76-90.26)
(n=16) 4.51
5 115,162 47,387 2.43:1 162,549 14.47 .87 15.33 734.20 (103.99-107.90)
(n=15) -3.91
Note. N = 76

There are two exceptions to the trend that reading more resulted in greater reading rate gains.
First, despite the statistically significant differences in mean gain scores in favor of Group 2, the
participants in Group 3 read approximately 8,000 more standard words (total) on average than
the participants in Group 2. Second, although the participants in Groups 4 and 5 read almost

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 38

identical amounts, the Group 4 participants made significantly greater reading rate gains. These
exceptions to the general trend suggest that the amount of reading, while important, was not the
sole determinant of reading rate gains.

An additional indication that the total amount read is not the only factor influencing reading rate
gains is that the differences in the total amount read among the groups appear to be too small to
adequately account for such large differences in reading rate gains. For instance, Group 1 read
24,122 more standard words on average than Group 3, but the two groups’ mean reading rate
gains differed substantially, at 32.99 wpm and 11.24 wpm, respectively. A second example is the
small difference in total amount read by Groups 2 and 5 (13,941 standard words) and the 22.62
wpm difference in mean reading rate gains made by the two groups. These exceptions suggest
that the primary reason for these differences lies elsewhere, possibly in the type of reading (i.e.,
simplified versus unsimplified), the level of reading (i.e., lower versus higher-level simplified
books) the participants engaged in, or both.

Hypothesis 2, which stated that participants displaying greater reading rate gains read more
simplified texts and few or no unsimplified texts, was first addressed by inspecting the amount of
simplified and unsimplified reading each group engaged in. The amount of simplified and
unsimplified texts read by each group and the resulting ratios are shown in columns 2, 3, and 4
respectively in Table 2. Reading rate gains generally paralleled the number of simplified
standard words read, with Group 1 reading the most standard words (181,032) from simplified
texts and Group 4 the least (101,923). This trend stands in contrast to the amount of unsimplified
reading the participants in the five groups engaged in; the participants in Groups 3, 4, and 5 read
approximately twice the number of standard words in unsimplified books as the participants in
Group 1. The differences in the reading patterns of the five groups are also apparent in the ratio
of simplified to unsimplified standard words read, with Group 1 displaying a ratio of 6.57:1 and
Groups 2 through 5 displaying ratios between 4.49:1 to 1.69:1. With the exception of Groups 4
and 5, the ratio decreases steadily, indicating yet again that greater reading rate gains were
generally associated with reading more simplified and fewer unsimplified texts.

Hypothesis 2 was also investigated by calculating the differences in reading rate gains between
14 pairs of participants matched on total standard words read, but who differed in terms of
whether they read only simplified texts or a combination of simplified and unsimplified texts.
This analysis clarified the effect of the type of text read on reading rate gains by holding the total
amount read steady. The 14 participants who read unsimplified texts were chosen based on the
criterion that at least 25% of the total number of standard words read was from unsimplified texts.
An average of 53.47% (SD = 20.81; range = 28.47–93.19%) of the reading completed by these
participants was from unsimplified texts; thus, they differed distinctively in this respect from the
participants who chose not to read any unsimplified texts.

The results for the 14 matched pairs are shown in Table 3, which is organized based on the
reading rate gains of the participants who read only simplified texts, with the top gainer (ID = 3)
at the top of the table and the remaining 13 participants arranged in descending order. The top
part of the table lists participants who read no unsimplified texts, as indicated by the zeros in the
third column (Standard words unsimplified), while the bottom part of the table shows the
matching participant in terms of total amount read.

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 39

Table 3. Comparison of pairs of student who read approximately equivalent amounts

Participants who read only simplified texts
ID Standard words Standard words Standard words Reading rate gain
simplified unsimplified total
3 251,490 0 251,490 40.51
7 230,965 0 230,965 32.97
8 193,093 0 193,093 32.87
10 187,208 0 187,208 30.70
13 270,349 0 270,349 23.39
18 199,880 0 199,880 20.35
19 177,866 0 177,866 20.14
22 221,031 0 221,031 19.73
23 162,174 0 162,174 19.04
24 209,535 0 209,535 18.18
25 136,569 0 136,569 18.09
27 140,961 0 140,961 17.00
34 202,880 0 202,880 12.51
42 172,782 0 172,782 9.62
M 196,913 0 196,913 22.51
SD 38,580 0 38,579 8.65
Participants who read both simplified and unsimplified texts
ID Standard words Standard words Standard words Reading rate Reading rate
simplified unsimplified total gain gain
difference
31 123,514 129,598 (51.20%) 253,112 14.36 + 26.15
36 164,953 65,647 (28.47%) 230,600 11.91 + 21.06
66 101,758 96,407 (48.65%) 198,165 0.03 + 32.84
46 107,627 77,356 (41.82%) 184,982 7.15 + 23.55
52 165,349 103,194 (38.43%) 268,543 4.89 + 18.50
40 118,255 71,461 (37.67%) 189,715 11.33 + 9.02
35 109,614 69,353 (32.37%) 178,966 12.05 + 8.09
61 67,316 144,205 (68.18%) 211,521 2.10 + 17.63
54 18,759 144,447 (88.51%) 163,206 4.28 +14.76
5 144,899 69,353 (32.37%) 214,252 36.16 - 17.98
56 53,600 77,356 (59.07%) 130,956 3.98 + 14.11
49 64,235 77,356 (54.63% 141,590 6.59 + 10.41
16 50,688 144,205 (73.99%) 194,893 22.06 - 9.55
58 11,418 156,245 (93.19%) 167,663 3.71 + 5.91
M 92,999 101,870 194,869 10.04 12.46
SD 49,798 34,316 39,031 9.51 13.47
Note. For the participants who read both simplified and unsimplified texts, the percentage in the
Standard words unsimplified column indicates what percentage of the total amount read were
from unsimplified texts.

The mean difference for the total amount read by the 14 pairs was 2,044 standard words, with the
greatest difference being 10,165 standard words, or a 6% difference, for participants 18 and 40.
Although no objective criterion exists for deciding when such a difference would result in
differential reading rate gains, we do not believe that a difference of 10,000 standard words or
less would exert a measureable effect on changes in reading rate over one academic year for two

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 40

reasons. First, if reading rates were so easily increased, the gains found in this and other studies
would have been far higher. Second, both first and second language reading authorities agree that
the development of reading fluency requires large amounts of reading over significant periods of
time (e.g., Grabe, 2010; Nation, 2009a).

The reading rate gain differences between the pairs can be seen in the rightmost column in the
bottom half of Table 3. The positive numbers indicate that in twelve of the fourteen pairs, those
who read simplified books exclusively made greater reading rate gains than those who did not.
The two negative numbers indicate that two participants (participants 5 and 16) who read
unsimplified texts outgained their counterparts who read only simplified texts (M = 13.77),
though we would note that participant 5 did complete a large amount of simplified reading (i.e.,
144,899 standard words). The mean difference in reading rate gains for these 12 pairs was 16.84
wpm, and the mean difference for the 14 pairs was 12.46 wpm. A one-way ANOVA was run to
investigate whether the reading rate gains between the 14 matched pairs were significantly
different. The independent variable was group (the participants who engaged in no unsimplified
reading versus those who did) and the dependent variable was reading rate gains across the
academic year. The participants who read no unsimplified texts significantly outgained their
matched counterparts, F(df = 1, 26) = 13.16, p = .001, η2 = .68. Thus, after controlling for the
total amount read, simplified texts were significantly more effective for increasing reading rate
gains than a mixture of simplified and unsimplified texts.

Hypothesis 3, which stated that participants making greater reading rate gains read a greater
number of lower level simplified books, was investigated by placing the books read into five
levels based roughly on descriptions provided by commercial publishers: 300–800 headwords
(Elementary), 1,000-1,600 headwords (Lower intermediate), 1,700–2,500 headwords
(Intermediate), 2,800-3,800 headwords (Upper Intermediate and Advanced), and unsimplified
readers, which we estimated to contain approximately 5,000 headwords.6 The number of books
read by each group in each of the five levels is displayed in Table 4, and the average number of
books read in each level by each participant in the group is reported in parentheses. Of the books
read by the participants in group 1, 94.9% (336 out of 354 books) were at the 300–800 and
1,000–1,600 headword levels, and they read from 75 to 129 more books at these two levels than
the participants in the other groups. Given that the vocabulary test administered at the outset of
the study confirmed that the participants had adequate knowledge of the 2,000 high frequency
words of English, we can be reasonably certain that they read books composed almost entirely of
words whose primary meanings were known to them.

Table 4. Number of books read at five difficulty levels (Average number of books read per person)
Group 300–800 1,000–1,600 1,700–2,500 2,800–3,800 5,000
1 188 (12.53) 148 (9.87) 12 (0.80) 0 (0.00) 6 (0.40)
2 138 (9.20) 108 (7.20) 23 (1.53) 0 (0.00) 8 (0.53)
3 150 (9.38) 111 (6.94) 14 (0.88) 2 (0.13) 12 (0.75)
4 129 (8.06) 72 (4.50) 16 (1.00) 3 (0.19) 18 (1.13)
5 125 (8.33) 72 (4.50) 13 (0.87) 5 (0.33) 13 (0.87)
Note. 5,000 = Estimated number of headwords in the unsimplified texts.

In contrast to Group 1, the individual members of the other groups read an average of 3.15 to
4.47 fewer books at the 300–800 headword levels and 2.67 to 5.37 fewer books at the 1,000–

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 41

1,600 headword levels. Thus, the most successful participants read an average of 5.82 to 9.84
additional books containing more comprehensible vocabulary and syntax than found in many of
the texts read by the participants in the other groups. This pattern reverses at the 1,700–2,500
headword level texts and beyond. The participants in Group 1 read from 10 to 21 fewer books at
each of the three most difficult levels than the participants in the other groups. We would also
note that the participants in Group 1 chose to read low-level simplified texts at the beginning of
the academic year. An analysis of the first four texts read indicated that Groups 1 to 5 read books
composed of an average of 652, 700, 755, 1,123, and 778 headwords, respectively. Thus, the
participants with the greatest reading gains were the most conservative in terms of the levels of
simplified books they read initially and, with few exceptions, they maintained this conservative
approach throughout the academic year.

Discussion

The discussion of the results is divided into two sections concerning the major findings of the
study: the quantity of reading and the optimal type of texts needed to make significant reading
rate gains.

Quantity of Pleasure Reading and Reading Rate Gains

Hypothesis 1 stated that greater amounts of reading would be associated with greater reading rate
gains. This hypothesis was generally supported; however, Group 2, which read 4.33% less than
Group 3, made greater reading rate gains (18.71 wpm versus 11.24 wpm), and Groups 4 and 5,
which read nearly the same total amount, displayed significantly different gains. Group 1, the top
gainers, read an average of 208,607 standard words over the academic year—a figure that is
roughly equivalent to reading one book per week during two 14-week academic semesters (i.e.,
between 25–30 books, mostly graded readers). Although this amount of reading is less than
Nation’s 500,000-word (~425,000 standard words) annual goal, it compares favorably with the
amounts reported in Table 1, and these participants completed this amount of reading in twenty-
eight 90-minute classes. Learners in programs with longer semesters or a greater emphasis on
extensive reading could read more than the participants in this study.

Both first- and second-language reading authorities assume that processing large amounts of
written input is necessary for developing high degrees of fluency. Logan (1988, 1990), in his
formulation of instance theory, stated that fluency development follows a power function
(Newell & Rosenbloom, 1981), in which the repeated processing of the same information results
in mathematically predictable decreases in retrieval times. The power function, which has also
been identified in the acquisition of foreign language listening and speaking skills (DeKeyser,
1997), suggests that, while there are no major shortcuts to fluency development, certain
principles, such as consisent practice and the repetition of embedded linguistic forms in
communicative contexts, should be implemented over long periods of time to ensure progress.

Despite this widespread agreement that processing large amounts of written input is necessary
for the development of reading fluency, there have been few proposals by second language
reading authorities concerning minimum yearly reading targets. This is possibly due to the lack

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 42

of empirical data available and the potentially large differences in reading fluency development
among individual second language learners caused by such factors as aptitude, first language (L1)
reading proficiency, L1 orthography, L2 reading motivation, and learning context. Given the
results of this and several previous studies, we tentatively propose a minimum yearly reading
target of 200,000 standard words of highly comprehensible texts for EFL students at a similar
proficiency level as those in this study. Achieving this minimum target would plausibly result in
substantial reading rate gains for such learners.

The justification for this minimum reading goal is twofold. First, the top quintile, who gained an
average of 32.99 wpm, read an average of 208,607 standard words over the academic year.
Second, a goal of reading 200,000 standard words in one year has been shown to be feasible in a
variety of educational contexts. For instance, the Japanese participants in Burrows’ (2012) study
read an average of 195,620 standard words over one academic year, the Hong Kong Chinese
university participants in Lao and Krashen’s (2000) study read an average of 388,000 words (i.e.,
approximately 329,800 standard words) over one academic semester, and the Vietnamese
government officials in Renandya, Rajan, and Jacob’s (1999) study read an average of 728.68
pages (i.e., approximately 145,000 standard words) over two months.

One way to illustrate what reading 200,000 standard words per year entails is to examine how
the goal breaks down on a daily, weekly, and academic semester basis for individuals with
different reading rates (see Table 5).

Table 5. Amounts of yearly reading for eight reading rates

Number of standard Number of standard Number of standard
Reading rate (wpm) words read in 20 words read per five-day words read in one
minutes per day week academic year
60 1,200 6,000 168,000
80 1,600 8,000 224,000
100 2,000 10,000 280,000
120 2,400 12,000 336,000
140 2,800 14,000 392,000
160 3,200 16,000 448,000
Note. One academic year is defined as two fourteen-week semesters (i.e., 28 weeks)

Although Table 5 simplifies a complex situation, it highlights two important aspects of a reading
curriculum: the approximate amount of time needed to achieve a specific reading goal and the
fact that less fluent readers need to read more than 20 minutes per day to achieve the minimum
target of 200,000 standard words per year. In situations where learners, for reasons of
proficiency or curricular time constrains, are unable to read daily, it might take several years to
attain the goal of reading 200,000 standard words or more per year. For instance, Nishizawa et al.
(2010) reported that after three years 75% of the students had read more than 300,000 words
(255,000 standard words), which they concluded was the necessary “threshold for the subjects to
feel at ease while reading English texts” (p. 632). Feeling at ease with L2 texts likely entails a
degree of fluency development, given that a sense of effortlessness is commonly listed as an
aspect of automatic processing (see Logan [1997] for a discussion of the property-list approach
to defining automaticity).

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 43

Extensive reading should be emphasized with students at both lower and intermediate reading
proficiency levels because simplified texts provide them with more opportunities to meet high
and mid-frequency vocabulary than do unsimplified texts. Cobb’s (2007) results indicate that (a)
graded readers are well designed for second language readers who have yet to automatize the
basic syntactic patterns, the high frequency affixes of English, and the high frequency 2,000
words of English; and (b) intermediate learners need to read considerably greater amounts of
more difficult text to automatize less frequent syntactic patterns and lexis. We believe it is best to
include a substantial fluency development strand in foreign language reading programs until
learners are able to read unsimplified texts with a degree of ease.

Optimal Text Type for Reading Fluency Development

The group analysis indicated that, for EFL learners at this proficiency level, reading simplified
texts was more effective for fluency development than reading unsimplified texts because
learners who read the highest ratio of simplified to unsimplified standard words also made the
greatest reading rate gains (see the fourth column in Table 2). Particularly for Group 1, greater
reading rate gains were associated with reading lower rather than higher levels of simplified
readers (i.e., 1,600 headwords and below) and reading fewer upper-level simplified texts (2,800
headwords and above) and fewer unsimplified texts (see Table 4). The importance of reading
simplified texts is further supported by the analysis of the 14 matched pairs who read a similar
number of standard words. In 12 out of the 14 pairs, those who read only simplified texts
achieved greater reading rate gains than those who read both simplified and unsimplified texts.

There are several possible reasons why lower-level texts are advantageous for reading fluency
development. The first reason concerns the opportunity cost associated with reading higher-level
texts. Given the added lexical and syntactic complexity of more difficult texts, individuals’
reading rates are likely slower when reading those texts. Thus, they read smaller amounts of text
in a given amount of time and have fewer opportunities to repeatedly process linguistic features
(e.g., letter combinations, lexis, collocations, and morpho-syntax).

The second reason is that automaticity develops with practice in consistent environments (i.e.,
those in which repetition occurs more frequently). Consistency is important when we consider
that the probability of repetition is lower as the level of processing becomes higher; that is, letter
combinations repeat more often than single words, which repeat more often than phrasal units,
which repeat more often than complex grammatical constructions. Thus, automatization occurs
more slowly at higher levels of processing. This is important if, as noted above, higher level
processes partly determine reading fluency (See Logan [1997] for a discussion of this possibility).
The large amount of recycling found in lower-level graded readers might be particularly
beneficial for lower proficiency learners, such as those in this study, because it might allow them
to gain control over linguistic forms and coordinate multiple types of processing more quickly
than if they were reading more difficult texts. Evidence for this possibility was provided by Horst
(2009), who found that even modest amounts of extensive reading (three graded readers)
produced lower lexical response times for high frequency vocabulary, particularly for words with
15 or more occurrences in the texts.

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 44

An additional reason for using easier texts is based on Perfetti’s (1985) verbal efficiency theory,
which states that the multitude of cognitive activities that occur in reading take place within a
limited-capacity working memory system. Lower-level texts place relatively few demands on
orthographic, lexical, and syntactic processing, leaving more cognitive capacity available for
automatizing top-down processes, such as creating an interpretative situation model of the text
(Grabe, 2009). This is important if we conceptualize bottom-up and top-down processing as
being complementary and reciprocal aspects of fluency development. In this view texts that are
more comprehensible facilitate fluent reading (Hudson, Pullen, Lance, & Torgesen, 2009), in
part because familiarity with genre organization and the concepts embedded in the text are
important aspects of reading fluency.

This paper has attempted to set a benchmark for future research by empirically measuring the
amounts, types, and levels of texts read and their effects on reading rate gains. These data are a
necessary starting point for tracking the effect of extensive reading on L2 learners reading rate
development over multi-year periods. It will also give educators, researchers, and administrators
a more precise understanding of the task faced by L2 learners and how best to design curricula to
address the on-going failure of many educational institutions to provide learners with
opportunities for reading rate and fluency development.

Conclusion

This study has produced three main findings: (a) the most successful participants read 200,000
standard words or more over the academic year; (b) simplified graded readers provided
significantly better results than unsimplified texts; and (c) lower-level simplified texts were more
effective than higher-level simplified texts. The first finding provides a goal for the annual
minimum amount to be read by EFL learners at this proficiency level. The last two findings
provide empirical support for the widely held belief that reading easy texts is optimal for reading
fluency development.

The results raise a number of issues in need of further investigation.

1. Longitudinal studies spanning three or more years are needed to gain an understanding
of reading rate growth curves.
2. The assumption that L2 learners read higher-level books more slowly is in need of
empirical verification.
3. Studies in which extensive reading is compared to or combined with other approaches
to fluency development, such as speed reading (Macalister, 2008) or repeated reading
(Nation, 2009b) need to be conducted to determine the relative effectiveness of extensive
reading and these approaches.
4. The specific sources of second language reading fluency are still unknown.
Researchers need to investigate how the components and processes of reading develop
and interact as learners become increasingly fluent.

We believe that answers to these questions will help illuminate the ways in which second
language reading fluency develops.

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 45

Acknowledgments

The authors would like to acknowledge Professor Nevitt Reagan of Kansai Gaidai University
and Professor Catherine LeBlanc of Kwansei University for their constructive suggestions on
earlier versions of this paper.

Notes

1. The term pleasure reading, instead of extensive reading, is used to describe the type of reading
in this study because many of the participants chose to read both simplified and unsimplified
texts (e.g., Harry Potter), which were likely too difficult to have been read fluently with full
comprehension. In this paper, extensive reading refers to the reading of simplified materials.

2. Grabe (2009) defined automaticity as “...processing operations that are rapid, relatively
resource-free, not subject to interference, unconscious, and hard to suppress” (p. 291). We agree
with Kuhn and Stahl (2003) that automaticity, along with accuracy and reading rate, is a sub-
component of reading fluency. Automaticity is generally applied to sub-lexical and lexical
processing, whereas fluency also includes efficient processing beyond the lexical level (e.g.,
multi-word units, morpho-syntax, and discourse organization).

3. Rauding (Carver, 1977), which occurs when a person comprehends while reading, is another
option for operationalizing fluency; however, it was not used in this study because it requires
data for both reading and aural processing. Aural processing data were not collected in this
study.

4. Carver (1982, 1990) defined one standard word as six character spaces (i.e., letters,
punctuation, and spaces). The use of standard words as a unit of measurement is more accurate
than counts of words, pages, or books because the number of words on one page and the number
of pages in one book can vary significantly. The use of standard words also allows for precise
comparisons across studies. The conversion ratio of running words from graded readers to a
standard word unit in this study was determined by randomly selecting one 500-token passage
from five books at each of the six levels in the Oxford Bookworms series. We then determined
the number of standard words in the 30 passages. We concluded that the conversion of running
words to standard words in these graded readers could be estimated by reducing the number of
running words by 15%.

5. Of the 97 participants in the Beglar, Hunt, and Kite (2012) study 17 participants in the
intensive reading (control) group and four additional participants who did little or no
independent reading outside of teacher-selected texts were omitted from this study.

6. The 5,000 word figure was arrived at by calculating the lexical composition of the three
unsimplified books read most frequently by the participants in this study and by using Table 6 in
Nation (2006), which indicated that 4,000 headwords and proper nouns accounted for an average
of 94–95% of the words in five unsimplified novels. Knowledge of 5,000 headwords represents

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 46

95-98% coverage of the lexis in the unsimplified books used in this study.

References

Anderson, J. R. (1987). Skill acquisition: Compilation of weak-method problem solutions.

Psychological Review, 94, 192–210.
Anderson, J. R., & Lebiere, C. (1998). The atomic components of thought. Mahwah, NJ:
Erlbaum.
Anderson, N. J. (2008). Reading. New York, NY: McGraw-Hill.
Beglar, D., Hunt, A., & Kite, Y. (2012). The effect of pleasure reading on Japanese university
EFL learners‘ reading rates. Language Learning, 62, 665–703. doi:10.1111/j.1467-
9922.2011.00651.x
Bell, T. (2001). Extensive reading: Speed and comprehension. The Reading Matrix, 1.
Breznitz, Z. (2006). Fluency in reading: Synchronization of processes. Mahwah, NJ: Erlbaum.
Brown, S. (1992). The history of European fairy tales. Tokyo: Macmillan Language House.
Burrows, L. (2012). The effects of extensive reading and reading strategies on reading self-
efficacy. Unpublished doctoral dissertation, Temple University, Philadelphia,
Pennsylvania.
Carver, R. P. (1982). Optimal rate of reading prose. Reading Research Quarterly, 18, 56–88.
Carver, R. P. (1990). Reading rate: A review of research and theory. New York, NY: Academic
Press.
Cobb, T. (2007). Computing the vocabulary demands of L2 reading. Language Learning and
Technology, 11, 38–63.
Day, R. R., & Bamford, J. (1998). Extensive reading in the second language classroom.
Cambridge: Cambridge University Press.
DeKeyser, R. M. (1997). Beyond explicit rule learning: Automatizing second language morpho-
syntax. Studies in Second Language Acquisition, 19, 195–221.
DeKeyser, R. M. (Ed.). (2007). Practice in a second language: Perspectives from applied
linguistics and cognitive psychology. Cambridge: Cambridge University Press.
Fielding, H. (1996). Bridget Jones’s diary. New York, NY: Penguin.
Grabe, W. (2009). Reading in a second language. Cambridge: Cambridge University Press.
Grabe, W. (2010). Fluency in reading—Thirty-five years later. Reading in a Foreign Language,
22, 71–83.
Grabe, W., & Stoller, F. L. (2011). Teaching and researching reading (2nd ed.). Harlow, UK:
Pearson Education.
Hirsh, D., & Nation, P. (1992). What vocabulary size is needed to read unsimplified texts for
pleasure? Reading in a Foreign Language, 8, 689–696.
Horst, M. (2009). Developing definitional vocabulary knowledge and lexical access speed
through extensive reading. In Z. Han & N. J. Anderson (Eds.), Second language reading
research and instruction (pp. 40-64). Ann Arbor, MI: The University of Michigan Press.
Hu, M., & Nation, I. S. P. (2000). Unknown vocabulary density and reading comprehension.
Reading in a Foreign Language, 13, 403–430.
Hudson, R. F., Pullen, P. C., Lance, H. B., & Torgesen, J. K. (2009). The complex nature of
reading fluency: A multidimensional view. Reading and Writing Quarterly, 25, 4–32.
doi:10.1080/10573560802491208

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 47

Imamura, K. (2012). How extensive reading, reading span, and reading speed are interrelated.
Extensive Reading World Congress Proceedings, 1, 124–127.
Iwahori, Y. (2008). Developing reading fluency: A study of extensive reading in EFL. Reading
in a Foreign Language, 20, 70–91.
Kuhn, M., & Stahl, S. (2003). Fluency: A review of developmental and remedial practices.
Journal of Educational Psychology, 95, 3–21.
Lai, F-K. (1993). The effect of a summer reading course on reading and writing skills. System,
21, 87–100.
Lao, C. Y., & Krashen, S. (2000). The impact of popular literature study on literacy development
in EFL: More evidence for the power of reading. System, 28, 261–270.
Linacre, J. M. (2006). WINSTEPS: Rasch model computer program [Computer software].
Chicago, IL: Winsteps.com.
Logan, G. D. (1988). Toward an instance theory of automatization. Psychological Review, 95,
492–527.
Logan, G. D. (1990). Repetition priming and automaticity: Common underlying mechanisms?
Cognitive Psychology, 22, 1–35.
Logan, G. D. (1997). Automaticity and reading: Perspectives from the instance theory of
automatization. Reading and Writing Quarterly, 13, 123–146.
Macalister, J. (2008). The effect of a speed reading course in an English as a second language
environment. The TESOLANZ Journal, 16, 23–33.
Mikulecky, B., & Jeffries, L. (1998). Reading power. White Plains, NY: Addison Wesley
Longman.
Nation, I. S. P. (1990). Teaching and learning vocabulary. New York, NY: Newbury House.
Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening? The
Canadian Modern Language Review, 63, 59–82.
Nation, I. S. P. (2009a). Teaching ESL/EFL reading and writing. New York, NY: Routledge.
Nation, I. S. P. (2009b). Reading faster. International Journal of English Studies, 9(2), 131–144.
Nation, I. S. P., & Heatley, A. (2002). Range: A program for the analysis of vocabulary in texts
[Computer software]. Wellington, NZ: LALS, Victoria University of Wellington, New
Zealand. Retrievd from http://www.victoria.ac.nz/lals/staff/paul-nation/nation.aspx
Nation, I. S. P., & Wang, K. (1999). Graded readers and vocabulary. Reading in a Foreign
Language, 12, 355–380.
Newell, A., & Rosenbloom, P. S. (1981). Mechanisms of skill acquisition and the law of practice.
In J. R. Anderson (Ed.), Cognitive skills and their acquisition (pp. 1–55). Hillsdale, NJ:
Erlbaum.
Nishino, T. (2007). Beginning to read extensively: A case study with Mako and Fumi. Reading
in a Foreign Language, 19, 76–105.
Nishizawa, H., Yoshioka, T., & Fukada, M. (2010). The impact of a 4-year extensive reading
program. In A. M. Stoke (Ed.), JALT2009 Conference Proceedings (pp. 632–640). Tokyo,
Japan: JALT.
Perfetti, C. A. (1985). Reading ability. New York, NY: Oxford University Press.
Pikulski, J. J., & Chard, D. (2005). Fluency: Bridge between decoding and reading
comprehension. The Reading Teacher, 58, 510–519. doi:10.1598/RT.58.6.2
Renandya, W. A., Rajan, B. R. S., & Jacobs, G. M. (1999). Extensive reading with adult learners
of English as a second language. RELC Journal, 30, 39–61. doi:
10.1177/003368829903000103

Reading in a Foreign Language 26(1)

Beglar & Hunt: Pleasure reading and reading rate gains 48

Robb, T. N., & Susser, B. (1989). Extensive reading vs. skills building in an EFL context.
Reading in a Foreign Language, 5, 239–251.
Rowling, J. K. (1997). Harry Potter and the philosopher’s stone. London: Bloomsbury.
Segalowitz, N. (2010). Cognitive bases of second language fluency. New York, NY: Routledge.
Sheu, S. P.-H. (2003). Extensive reading with EFL readers at beginning level. TESL Reporter, 36,
8–26.

About the Authors

David Beglar is an Associate Professor and Academic Coordinator of the M.S.Ed. and Ph.D.
programs (TESOL) at Temple University, Japan Campus. His primary research interests are in
the areas of foreign language assessment, vocabulary acquisition, and reading fluency
development. E-mail: [email protected]

Alan Hunt is a professor at Kansai University, Osaka Japan. His research concentrates on testing,
EFL vocabulary acquisition, extensive reading and reading fluency. E-mail: [email protected]

Reading in a Foreign Language 26(1)