Varying levels of difficulty in L2 reading materials in the EFL classroom: Impact on comprehension and motivation

In this project, 54 Iranian pre-intermediate EFL learners (16 to 21 years) participated in 15 sessions in which they were asked to read and study books from a structured series of L2 reading materials. Half the students studied with relatively easy materials (‘i − 1’), that is, materials rated as being below their competency level. The others studied with reading materials rated at slightly above their competency level at the outset (‘i + 1’). Using a before and after design, students were retested after 15 weeks. On a test of L2 reading comprehension, means testing revealed both groups showed marked increases, but the ‘i + 1’ group was higher than the other group (effect size 1.1). On an assessment of reading motivation, the ‘i + 1’ group increased significantly, whereas the other group showed no change (effect size 1.58). We suggest that, under supporting conditions, there can be clear benefits for EFL learners to spend time mastering L2 materials above their level of competency. Subjects: Education; Language & Literature; Literature

ABOUT THE AUTHOR Ehsan Namaziandost was born in Shiraz Province of Iran in 1985. He holds an MA degree in TEFL from Islamic Azad University of Ahvaz. Now, he is a Ph.D. student in TEFL at Islamic Azad University of Shahr-e Kord. His main interests of research are second language acquisition, EFL teaching, teaching and testing, teaching speaking, communication and language learning strategies. His research papers and articles have been published by different international journals.

PUBLIC INTEREST STATEMENT
Language input is considered as a highly essential factor in the language learning process. Reading is the most readily available form of comprehensible input, especially in places where there is hardly any contact with the target language. The only reliable way to learn a language is through massive and repeated exposure to it in context: precisely what Extensive Reading (ER) provides. With regard to the vital role of input in language teaching and learning, this study aimed to examine the effects of different text difficulty levels on learners' foreign language reading motivation (FLRM) and reading comprehension. After implementing the treatment sessions, it was revealed that input higher than the students' current level, i.e., i + 1 significantly improved the participants' reading comprehension. The findings also indicated that, at the end of the study, the 'i + 1' group's FLRM increased, while that of the 'i − 1' group indicated no change.

Introduction
The ability to read in a foreign language presents a daunting challenge to the language learner. In the present project, Iranian students undergoing a course in mastering English (L2) were asked to spend one lesson period each week, over a 15-week period, reading materials from a structured series of English teaching reading materials. This methodology relates to the practice of extensive reading (ER). It is well recognized that EFL students need to devote substantial time interacting with written materials in L2, to so achieve advanced levels of fluency and comprehension within the foreign language. However, in the context of the present project, the students' L2 reading materials were deliberately constrained during the duration of the course. One half of the students were asked to read materials below their natural competence level. The other half were asked to devote the same time to materials deemed to be above their competence level at the outset of the project.
In recent years, extensive reading (ER) has gained consideration as an effective way of expanding foreign language skills (Yamashita, 2013). It is recognized that ER aims "to progress good reading habits to form knowledge of vocabulary and grammar and to encourage a liking for reading" (Richards & Schmidt, 2010, p. 194). The major purpose in ER is to allow the EFL student to achieve a deeper understanding of the text (Richards & Schmidt, 2010). The general recommendations are that ER is for general comprehending in which "the minimum 95% comprehension figure" (Meng, 2009, p. 134) is admissible and the reading velocity is between 100 and 150 words per minute (Mikeladze, 2014). A variety of studies (e.g., Bell, 2001;Chiang, 2015;Hitosugi & Day, 2004;Iwahori, 2008;Leung, 2002;Mason & Krashen, 1997;Tanaka & Stapleton, 2007) have noted that ER significantly enhanced foreign language reading comprehension and general proficiency.
Whilst it is widely recognized that ER opportunities represent key aspects in language development programmes, one interesting issue remains both topical and unresolved. Do EFL students benefit more so from reading relatively easy or relatively difficult L2 materials? Indeed, does this variable matter at all? Within the existing literature, it appears possible to identify two coherent positions on such issues. Whilst not necessarily in contradiction, arguments have been made for the value of easy L2 reading materials. However, clear arguments also exist for the value of using more difficult or challenging L2 reading materials.
One position, associated with Krashen (1982), is the reading input to which learners are exposed ought to be a little above their current level of competence. This is referred to as "i + 1," in which "i" alludes to the present language capacity of learner, and "+ 1' alludes to the input that is somewhat above the learners" present language ability. On the other hand, Day and Bamford (1998) suggest, "ER is efficacious if it furnishes students with input which is marginally beneath their current level of competence (i.e., 'i − 1')" (Day & Bamford, 1998, p. 36). This way language learners can swiftly develop their reading certainty, reading fluency and construct sight words and focus upon highfrequency words.
The position that L2 readings can be set to a level slightly lower than a learner's competency is one that makes clear sense and is readily seen as consistent with the commonly expressed goals of ER. One explicit purpose of ER is for reading to become easy and enjoyable. In essence, a fundamental objective of any ER program is to provide a situation for learners to enjoy reading in L2, and so be able to read unfamiliar authentic texts silently at their own pace, with sufficient understanding (Day & Bamford, 1998).
Based on this scheme, "ER is advantageous if it furnishes the students with input which is somewhat beneath their current level of competence (i.e., 'i − 1')" (Bahmani & Farvardin, 2017, p. 4). Moreover, "'i − 1' creates a condition for automaticity in conveying and extending sight vocabulary rather than learning new target structures" (Mikeladze, 2014, p. 5). From this perspective, "i − 1' is considered as the learners" comfort zone in that they may develop in their fluency and confidence which may be associated with positive attitudes (Chiang, 2015).
However, a different view is possible. According to Krashen's (1982) input hypothesis, adequate exposure to comprehensible input is essential for language learners to learn the language. According to this hypothesis, the input to which learners are exposed should be a little beyond their current level of language competence, i.e., "i + 1." Based on this notion, when learners repeatedly are exposed to, and concentrate upon a large number of messages (input) which is a little beyond their level of competence, they gradually acquire the forms necessary for understanding to advance.
Furthermore, based on Krashen's affective filter hypothesis (1982), language acquisition occurs in low-anxiety situations. Foreign language learners with a low affective filter (e.g., anxiety) will attain the language acquisition or comprehension more easily (Huang, 2001). In the same vein, Krashen (1994) proposed the pleasure hypothesis, arguing that the pleasurable activities are both effective and facilitating for language and literacy development. Based on this hypothesis, ER provides a low-anxiety situation for learners to learn a foreign language. Krashen's hypotheses have encouraged different universities and institutions to do research in ER and utilize ER programs in foreign language teaching (Chiang, 2015).
One assumption underpinning Krashen's (1982) input hypothesis is that adequate exposure to understandable input is essential for language students to further develop their existing language. Considering this speculation, the input to which students are exposed to ought to be a little past their current level of language ability, i.e., 'i + 1'. When learners constantly and repeatedly confront and concentrate on an expansive quantity of input which is a little higher than their level of capability, they inchmeal obtain the necessary structures.
So, should ER materials in L2 be relatively easy, or relatively difficult? Two prior studies are directly relevant to this matter. Bahmani and Farvardin (2017) were able to compare EFL classes who studied under either 'i + 1' or 'i − 1' conditions. The EFL students were young adult Iranians striving to develop their English language skills. It was found that comparable classes showed increases in L2 reading comprehension and did not differ in test scores after a 4-month instructional period. However, the 'i + 1' students did report higher levels of anxiety about reading with foreign language materials.
In a similar study design with first-year university students in Taiwan, Chiang (2015) also compared 'i + 1' to 'i − 1' classes and found no significant differences in terms of reading comprehension levels in English. However, relative to the 'i + 1' students, those who participated within the 'i − 1' class did report significant gains in their attitudes to reading.
In the present project, it was possible to programme this variable ('i + 1' vs 'i -1') into an existing language development course with adolescents and young adults attempting to master English within an EFL context. It was arranged that half the students participated using reading material below their level as measured at the outset. The other half worked with materials slightly above their initial measured competency level. Since all the students participated over a 15-week period, it was assumed that reading comprehension levels would increase over time in both groups of participants.
In addition to being asked to complete reading comprehension assessments both before and after the ER sessions, the students were asked to complete questionnaires enabling them to self-report how motivated they felt as readers. Although a singularly clear definition of reading motivation is not viable, Guthrie and Wigfield (2000, p.405) suggest that "reading motivation is the individual's personal objectives, values, and beliefs regarding the topics, processes, and outcomes of reading". Furthermore, "Not only does reading motivation relate to reading comprehension, but it also relates to both the amount of reading and students' reading achievement" (Guthrie & Wigfield, 2005, p. 76). Within the context of mastering a foreign language, it is often suggested that motivation for reading is a significant attribute associated with students' emotional responses and their willingness to learn from further reading activities in L2 (Ismail, Ahmadi, & Gilakjani, 2012).
The key research question driving the present project was would one group display a relatively higher level of reading comprehension on L2 materials by the end of the project? Would the relative difficulty of the programmed reading material make a difference to either the student achievement or reading motivation levels? Although it can be anticipated that students who participated in the project should increase their comprehension levels over a 15-week period, we did not presume to predict which treatment would be relatively more effective.
In relation to the motivation variable, we expected that being asked to read and spend time studying relatively easier materials would result in increased motivation. This notion follows from the expectation that since the reading material was facile and more readily accessible, then levels of skill and automaticity should increase naturally. But whether motivation would increase over time, when students are asked to read more difficult material in a foreign language, we felt remained an open question.

Participants
The participants were 54 EFL learners (25 male and 27 female) ranging in age from 16 to 21 years. These students were enrolled in a private language institute in Ahvaz, Iran, and were using the textbook series American Headway (Soars & Soars, 2010) as part of their L2 studies. The course of study involved completing an initial placement test. On this assessment, it was established that all the students had achieved a level of competency equivalent to B1 (between 66% and 74%) on precourse materials. This represents a moderate level of competency in English, described as the preintermediate level on the Common European Framework of Reference (CEFR) framework (Bahmani & Farvardin, 2017).

Graded reader series
As one aspect of the course syllabus, students were to devote time, each week, to expanding their reading agenda through exposure to the Oxford Bookworm Series of graded readers. This series of texts, published by Oxford University Press, consists of seven structured levels with increasing difficulty, as defined in terms of word count, complexity of grammar, and vocabulary. Using the CEFR framework, the pre-intermediate level students in this project were regarded, at the outset of the project, as being at level 3 on the Oxford series, in terms of their natural L2 reading capabilities.

Reading comprehension test
The reading comprehension component of the Cambridge First Certificate in English (FCE, 2008) was used to measure the participants' reading comprehension ability. This comprised of three reading passages which include both macro and micro questions, such as the expression of opinion, attitude, purpose, main idea, detail, tone, and gist. The reading section of the FCE includes 30 items to be completed in 30 minutes. Two forms of this test are available, as equivalent forms. Hence, one form was used as the pretest, the other as posttest. A Pearson correlation of .94 between the two forms was reported in the published manual test material.

Motivation for reading questionnaire (MRQ)
The students completed the Motivated Reading Questionnaire at both the beginning and completion of the project. This instrument asks students to rate themselves as readers but is specifically designed to tap students' attitudes towards themselves as good readers with ability and confidence to use reading to learn new information. Wigfield (2000, 2005) developed this scale as an index of how motivated students are as readers. The original scale had 53 items, but this was shortened to 30 for purposes of this project, and translated into Persian. Three sample items are "I like to read about new things", "In comparison to other activities I do, it is important to be a good reader", and "If a book is interesting, I do not care how hard it is". For each of the 30 items, students rated level of agreement using the following scale: 1 "I strongly agree"; 2 "I agree"; 3 "I do not know"; 4 "I disagree; and 5 "I strongly disagree".

Procedure
The study design is that of a quasi-experiment in that the treatments were administered to two intact groups. Students were assigned into their groups, of roughly equal size and gender proportion, by course administrators blind to the nature of the project at the outset. During the first two weeks of the course, the two groups were treated similarly in that they completed the placement test, the comprehension test, and the MRQ, and consent forms in which the participants agreed to allow their data to be used for research purposes. Beginning in the third week, the groups were treated differently. The 25 students in one group, 'i − 1', were asked to work through reading materials drawn from levels 1 and 2 in the Oxford Bookworm Series. Within each class period session, time was devoted to reading the set material and then several minutes of group discussion. The students were asked to openly discuss what they had read. The 27 students in the other group, 'i + 1', were treated similarly, except they were requested to use the more difficult materials drawn from levels 4 and 5 in the Oxford Series. In the final week of the course, the 15 th class period, all students completed the MRQ in their L1, and the second test of reading comprehension in L2.

Results
Prior to analyses, it was established, using Kolmogorov-Smirnov tests, that all dependent variables, at both pretest and posttest levels, could be considered to possess normal distributional properties. Kurtosis and skewness coefficients in all instances were below 1 and so all variables were deemed acceptable for parametric analyses. For the reading comprehension tests, the means and deviations are shown in Table 1.

Reading comprehension test
A 2 × 2 repeated measures ANOVA was applied to the reading comprehension scores with groups and time as independent variables. A significant effect was found for groups, F (1,52) = 8.9, p < .01. Time was also significant, F (I,52) = 107, p < .01. The time by group interaction was significant, F (1,52) = 6.3, p = .015. This significant interaction, effect, depicted in terms of means in Figure 1, indicates the impact of the treatment differed markedly between the two groups.  Paired t-tests clearly established that both groups improved on the test over time (p< .01 in both groups). Mean testing indicated that the two groups did not differ significantly at the pretest level. However, it was evident that, at the posttest level, the 'i + 1' group scored higher than the 'i − 1' group, t (52) = 3.6, p< .01. This difference, expressed in terms of effect size using Cohen's d procedure (i.e. per standard deviation units), was found to be 1.1.

Motivation for reading questionnaire
A 2 × 2 repeated measures ANOVA was applied to the reading motivation (MRQ) data with groups and time as independent variables. A significant effect was found for group, F (1,52) = 33, p < .01. The effect for time did not achieve significant effect, F (1.52) = 3.3, p = .07. Most importantly, however, the interaction term was significant, F (1,52) = 5.8, p = .02. This interaction effect is clearly shown in Figure 2.
Independent t-tests revealed that although the two group means did not differ on the MRQ at the pretest level (p = .25), the respective means were significantly different at the posttest level, t (52) = 5.8, p< .01. It was apparent that, in the case of the 'i − 1' group, the motivation scores did not alter significantly over time (p = .6). On the other hand, the 'i + 1' group means showed a significant increase from pretest to posttest, paired t (26) = 3, p < .01. At the posttest level, this manifest difference between the two groups, when expressed in terms of effect size using Cohen's d procedure, was found to be 1.58.

Discussion
In this study, we sought to investigate if it would make a difference if EFL learners studied reading materials that were programmed either above or below their initial reading competence level. Two earlier studies (Bahmani & Farvardin, 2017;Chiang, 2015) had reported that, under instructional programmes similar to that of the present study, significant differences between 'i + 1' and 'i − 1' conditions were not found on L2 reading comprehension measures.
In the present study, the findings indicated that reading comprehension levels increased in both groups over a period of almost 3 months. However, it was also found that the students who studied the more advanced books, the 'i + 1' group, scored more highly on a reading comprehension measure that their peers in the 'i − 1' group. That is, the level of change from pretest to posttest was greater for those students exposed to the more advanced L2 reading materials.
In the current dataset, the apparent advantages for the 'i + 1' strategy extended also to a measure of self-reported motivation. The measure we used is a generalized scale allowing students to index their feelings of confidence in being able to use reading skills to learn new material. Students in the 'i − 1' group reported similar levels of such motivation at the outset and conclusion of the 15-week course. However, the 'i + 1' students reported considerably higher motivation levels by the end of the course. Although the difference between the groups on motivation levels was minimal at the outset, the Cohen effect size coefficient was a remarkable 1.58 standard deviation units by the end of the semester programme.
We can interpret the present findings as supporting the input hypothesis, a position associated with Krashen (1982). This position states input which is somewhat above the present level of competence of the language learner can be conducive to learning. If i is the language learner's current level of competence in the foreign language, then i + 1 is the following prompt advance along the improvement continuum. Accordingly, if the objective is to help the language student advance in their task, then it appears that within a supportive context, they can readily benefit from being exposed to L2 reading matter likely to challenge them.
In our study, it could be possible that consistent exposure to written input facilitated the participants' incidental vocabulary learning (Mikeladze, 2014;Waring & Takaki, 2003). The obtained results may be due to the significant role of inputs which the students received. The comprehensible inputs which the students were subjected to, before being retested, greatly helped the students to be able to read English more efficiently.
However, it is important to note the treatments in our project did involve group level discussions each week in addition to silent reading. This is of course in the nature of teaching programme the students were undertaking. The students, as EFL learners, were expected to read and be prepared to briefly discuss what they have read. It is possible that the reading material regarded as 'i + 1' was somehow more engaging or interesting to the students. It is conceivable that exposure to the more difficult materials served to stimulate interest, or otherwise simply the drive to "find out more". It is clearly in the nature of using more difficult materials that students will find the reading task itself relatively more challenging. Indeed, our finding, that the 'i + 1' students reported a clear increase in reading motivation, suggests to us that the use of the more advanced materials may trigger elements of intrinsic motivation. Such motivation, expressed as the urge to "understand", or "find out more" could be driven, in part, by the careful use of challenging tasks and materials within supportive teaching situations.
Why did our findings differ from two earlier studies (Bahmani & Farvardin, 2017;Chiang, 2015) which had found minimal effects associated with the L2 text difficulty variable? We are unable to provide a simple explanation for such a discrepancy. It is apparent that students benefit from both 'i + 1' and 'i − 1' reading materials, when programmed within a structured ER sequence. There is likely to be a natural developmental progression in that achieving mastery and competence on easier materials will encourage individual students to actively seek out more difficult texts, and thereby learn from such input experience. In the present study, we noted that the 'i + 1' class indeed presented a positive and encouraging experience to the participating students, together with several minutes devoted each week to open social discussion about what the group had read. Under such supporting conditions, it is possible that students feel relatively more able to cope with the increasing demands of the more advanced texts, and so advance through their developmental progression with facility.

Conclusion
This study compared the effects of 'i + 1' and "i − 1' materials on Iranian EFL learners" reading comprehension and reading motivation. The 'i + 1' group outperformed the 'i + 1' group on measures of both L2 (English) reading comprehension and L1 reading motivation. Hence, we may suggest that materials of EFL English textbooks could be one level higher than the current level of the students to help to motivate and challenge them. We can suggest that the input hypothesis of Krashen (1982), that learners progress in their knowledge of the language when they comprehend language input slightly more advanced than their current level, represents a viable perspective.
This study provides implications for EFL teachers interested in using ER in their classes. It is apparent that ER is effective in improving EFL learners' reading comprehension and helpful in enhancing vocabulary, grammar, and reading speed regardless of the level of materials learners are using. However, the present study also suggests that students can benefit from tackling materials above their level, provided they are learning within a positively motivating context. We infer that choosing novels based on the participants' own interests can encourage them to raise to the challenge presented by exposure to difficult text. Ideally, teachers may wish to consider the value of self-selected materials as a key to a successful implementation of ER.
Finally, it is important to acknowledge limitations in the study. In particular, this was a quasiexperiment in that the treatments were administered to intact classes, this aspect being both a strength and a weakness. Given this type of project, random placement of individuals into treatments (the "true" experiment) is well-nigh impossible to achieve within educational institutes. Hence, it would be desirable to attempt to replicate the study design in different contexts and with different types of participants. The participants in the present study were relatively advanced and capable learners. The extent to which similar effects would be found with less capable students requires further investigation.
Further, it should be acknowledged that although both 'i + 1', and 'i − 1', L2 experiences can be associated with positive effects, we know little of the conditions under which either strategy could become relatively more advantageous for the individual. For instance, could a period of exposure to 'i − 1' be seen as a clear way to move a student into the 'i + 1' experience? To what extent do students need additional supportive teaching before they can benefit from more autonomously regulated exposure to 'i + 1'? Under what conditions would exposure to more difficult L2 materials become disheartening rather than motivating? Research in the future may help in addressing such issues.