A Small Scale Study of Reading Aloud Enabled by Technology Use
Christopher Walker has worked as an EFL teacher since 2006. He completed his CELTA at International House Krakow and spent the following year there, before moving to IH Bielsko-Biala, where he has remained since 2008. He has spoken at numerous conferences since 2019, and has had several articles published in some of the top EFL journals. Email: firstname.lastname@example.org
The argument over the place of reading aloud (RA) in TESOL continues. While supported by research in L1 education, it is an under-researched area in L2 studies. In this study, five A2-level students in at a language school in Poland underwent a one-month programme involving RA and peer training. To avoid the complaints associated with RA, such as it being a ‘filler’ activity, Zoom Breakout Rooms and the voice recording website Vocaroo were used, along with elements of peer training to encourage learner autonomy. At the end of the experiment, four of the five students showed an improvement in their pronunciation at a phonemic level, with two of the students showing a large improvement. One student showed no improvement compared to the start of the experiment, but had made progress in ‘noticing’ their errors. This study suggests that, if approached appropriately, RA can be a useful technique for improving pronunciation.
On the technology side of TESOL, this study also demonstrates that the appropriate use of technology (in the present case, Zoom Break-Out Rooms and Vocaroo) can ameliorate many issues connected with conducting experiments in the classroom, as well as problems of synchronicity with RA practices. By splitting students into Break-Out Rooms and having students record their spoken production using Vocaroo, much classroom time can be saved.
I remember my own school experiences of reading aloud. It was Dickens, and we took turns reading a few paragraphs to the class. None us seemed much interested in the story, or in what was said by our peers, instead focusing on where they had reached so we knew where to begin when it was our turn. It is likely that this experience is shared by many TESOL teachers, and this might account for the general reticence to engage in reading aloud activities in their classroom. But there are those for whom this practice is justified – their own memories are far more positive, perhaps - and when the two sides meet a loud disagreement often results. This was certainly the case recently on Twitter, as described by Anderson (2020), and it is his blog post that inspired the present study.
Reading aloud (RA) here is understood as the student reading a text out loud, either in the presence of others or on their own. Word-length and sentence-length pronunciation practice cannot be considered under this umbrella term; instead, short texts are chosen by the teacher and read in their entirety by the students. How this is done in the classroom is a management issue for the teacher, and many approaches exist.
Much research about RA is available, though there is less within the context of L2 learning. In mainstream L1 education, RA is an accepted part of the curriculum and much support exists for teachers to make the most of the practice (Strickland, Ganske, and Monroe 2002 is one of many examples). In terms of more specific support, Koskinen and Blum (1986) look at how poor readers can improve their reading ability through pair-reading activities; but again, this is support designed for students in L1, where the reading skills being developed might not be the same as for weak readers in L2. Their ideas, however, can be adapted for use in the TESOL arena. Within ELT, Martinez Adrian (2014) found that students at the University of the Basque Country improved their English pronunciation through an RA task. However, the subjects of this action research were well-informed, more mature students than the majority of those commonly encountered in TESOL contexts and therefore more likely to be receptive to the overall aims of RA.
The most general study of RA is that presented by Gibson (2008). Though it contains no direct research, instead focussing on attitudes to RA and the literature that examines the topic, it does highlight the supposed advantages and disadvantages of the practice. If done for no other reason than to fill empty time in the lesson, RA surely will not be of benefit to students. However, as Gibson points out, there are definite advantages to RA, such as making explicit the connection between English orthography and pronunciation. Silent reading does not necessarily make that connection for students. But beyond summarising the positions taken by many in TESOL as pertains to RA, there is little to note in terms of the practical application of the various ideas and activities suggested.
Technology was used to ameliorate two of the central issues surrounding RA: that it takes up too much classroom time, and that it focuses on one student at a time, with the other students left with little to do as they await their turn. To solve this first problem, the cloud-based audio recording service Vocaroo (https://www.vocaroo.com) was chosen. Vocaroo is a simple service with a minimal interface that students with low-level English skills would be perfectly capable of using after only a few minutes of training. It was used successfully in a study by Serafini and Blair (2017), where it enabled teachers to supply verbal feedback quickly and efficiently. As Gerich (2013) found, sharing links to recordings via Vocaroo is a simple process, and this assisted the researcher in collecting the data needed from each of the participants; it also made it possible for the students to share their recordings with one another, and allowed the students to track their progress for the duration of the experiment. The second issue was overcome through the use of the Zoom platform’s Break-Out Rooms. As Kohnke and Moorhouse (2020) write, these allow the teacher of online classes to break their larger groups into smaller units much as they would have done in the offline, face-to-face classroom. One advantage of Break-Out Rooms is that students in one Room cannot hear or be disrupted by students in any of the other Rooms – which is an improvement over the classroom experience, where proximity can lead to one group being disrupted by another that is physically nearby. The biggest disadvantage, however, is that the teacher cannot monitor what is happening in any of the Break-Out Rooms without being present in that Room; so when students are sent to their Rooms to perform a task, the teacher must trust that they actually will.
This study aims to answer two Research Questions:
- Is RA an effective way to improve the pronunciation skills of A2-level students?
- Can technology be used to ameliorate certain of the conditions of RA in the classroom – namely, that RA takes too long and that it only involves one student at a time?
There were six participants in this study, but one was absent from five of the eight lessons in the period concerned, and thus was removed from the final analysis. Another student joined the class after the initial test, and thus will not be considered in this study; however, the student did participate in the majority of the reading practice, and contributed to the peer correction described below. The five remaining participants (labelled S1 through S5) were all members of the same class, an A2 level teenage class (ages 13 to 15) at a private language school in Poland. The class met twice a week (on Tuesdays and Thursdays) for lessons of ninety minutes’ duration. For the period studied, all of the lessons took place online using Zoom.
In terms of the character or personality of the students involved in this study, there are some important similarities to consider. All of the students had lessons online at their state school, and joined my lesson at 15.30 after six or more hours already spent in front of a computer or other device. Whether because of this or for some other reason, the students were all much quieter than at the start of the academic year, when the lessons were held in person at the language school. None of the students turned on their cameras during the lessons, and remained muted unless called upon by the teacher. The majority of in-class communication was between the teacher and the students individually, and took place through the Zoom chat. It was this reticence to speak most of all that led to the choice of this class for the study – I felt that the pronunciation practice offered by RA would at least mean some time spent speaking English, and that of all of the classes I have taught under the conditions imposed by COVD-19, this one would suffer the least disruption to their natural desire for communication as a result of taking part in this study.
There is another aspect to consider, and that has to do with experimental validity. Performing experiments on live subjects is always fraught. If the researcher believes in the efficacy of what is being tested, surely it is unethical to exclude students from the experience, and how then can the researcher establish a control group? And how can the variable under test be isolated, when the students themselves are engaging in further learning in every lesson they attend? In this study I believe that the two questions can be answered satisfactorily. In the first case, arranging the students into groups according to how likely they were to speak to one another in these online lessons meant that nobody was at risk of being excluded by the researcher – if they were excluded from the experiment at all, it was because of their own attitude and approach to the twin tasks of RA practice and peer training. In the second, given how reluctant the students have been to speak either to the teacher or to the class in the time since COVID-19 forced all of our lessons online, the variable of phonological production has to an extent been isolated: during the month that was spent reading aloud, the only time the students actually spent producing language orally was during each practice segment. Otherwise they remained resolutely silent.
The study began on November 12th, 2020, and concluded on December 17th, 2020. First, the students were recorded reading a sample text from the current unit of the class coursebook, and graded along the lines described below. Over the course of the next several weeks, the students were given six opportunities to practice their reading aloud, at the end of which practice they would record themselves using Vocaroo. The links to these recordings were then sent to the teacher, who maintained a Google Sheets spreadsheet with all the relevant information. At the end of the period, the students read aloud without practice; this was their final test, as it were, and stood to compare against the first test in November.
The texts chosen for the students to read aloud were all taken from the coursebook, English in Mind 2 (Puchta and Stranks 2010), and the RA practice was always subsequent to the teaching of material found in the text – the students were never expected to pronounce new items of vocabulary before they had been studied in the lesson. The procedure was as follows:
- Before the lesson, the teacher recorded himself reading the text at a slow and steady rate, using Vocaroo.
- At the appropriate point in the lesson (after the grammar or vocabulary introduced or practiced in the text had already been covered), the teacher shared the Vocaroo link with the students via the Zoom chat.
- The teacher then instructed the students to work in pairs in the Breakout Rooms. First the students were to listen to the recording, and then they were asked to practice reading the same text aloud to their peers in the Breakout Room. Their peers were then supposed to correct the students as they read, stopping them when they heard a mistake, and by this means of peer correction to improve the pronunciation of the individual items in the reading text. This part of the procedure follows the advice in Koskinen and Blum (1986), though it goes beyond pairs reading a text to one another, and introduces an element of peer correction.
- The groups were given approximately ten minutes in which to practice – this time included an allowance for listening to the source recording (which usually ran to about one minute in duration) and the time required for each member of the group to read the text through twice with corrections.
- At the end of the practice period, the students re-assembled in the main room of the Zoom meeting, and were then invited to record themselves reading the text in Vocaroo.
The Breakout Rooms were assigned according to what I believed to be the students’ likely communicative potential – in other words, the participants were each placed with somebody else likely to engage to a similar degree with the task in hand. Three groups were thus formed in this way: the first group, containing S1 and S2, showed the greatest reluctance to turn on their microphones, and only participated in the peer correction component of this study to a minimal degree. S3 and S4 were far more involved in peer correction, with S4 in particular making a very great effort to both improve pronunciation and to assist their partner. S5 worked with the late addition to the class and with the student whose absences caused the withdrawal from this study; S5’s participation in the peer correction aspect of the study varied from practice to practice, as will become clear in the Results described later.
Assessment of pronunciation
At A2 level, it cannot be expected that pronunciation will be perfectly natural, or that it will be a perfect reproduction either of the teacher’s model or of a neutral Received Pronunciation (RP) standard – and indeed RP is not the goal of these lessons. Jenkins (2002) introduced the Lingua Franca Core to ELT, and it is the constituent parts of that core that I selected as a reasonable set of criteria against which to judge the students’ pronunciation.
The following were counted as mistakes:
- Missing words (for example, dropping the indefinite article before a noun)
- Saying the wrong word (for example, ‘population’ instead of ‘popular’)
- Inappropriate choice of phoneme (for example, pronouncing ‘scheme’ as /ʃiːm/ instead of /skiːm/)
- L1 intrusion (for example, pronouncing ‘your’ as /jɔːr/ instead of /jɔː/)
The following were not penalised as mistakes:
- Mild deviations from RP, especially with names (for example, ‘Toronto’ pronounced as /tuːrɒntəʊ/ instead of /tərɒntəʊ/)
- Inappropriate word stress patterns where the meaning of the word was not affected and the word could still be understood (for example, ‘bicycle’ pronounced as /baɪˈsɪkl/ instead of /ˈbaɪsɪkl/)
- Substitution of /f/ or /v/ for /θ/ or /ð/.
Following each practice, I listened to the Vocaroo recordings, counting for each student how many mistakes they made and tabulating the results against the total number of words in each text (these ranging between 90 and 140 words).
Results and discussion
In each of the charts presented in this section, the percentage reflects the number of mistakes in production made by each student; therefore in the first test, where S5 scored 26%, that equates to making in pronunciation mistake in roughly one quarter of the words in the text. Gaps in each chart show that the student was absent for the lesson.
Group S1-S2 recorded the following performance:
The two participants in this group generally suffered the least from pronunciation problems at the start of the study. Their performance was generally very good during all of the practice sessions, and their final test performance represented a small improvement over their initial test. The two participants did not engage much in peer correction; whenever I visited them in the Breakout Room, their microphones were muted, and even when I reminded them to practice reading the text together they seemed reluctant and did not correct each other when a mistake was made.
It is hard to say whether the RA practice made any difference to the students in this group. The practice of peer correction was almost certainly ineffective.
Group S3-S4 recorded the following performance:
The participants in this group responded most positively to the instruction to peer correct. On my visits to their Breakout Room, they were always engaged in reading the text aloud to one another, though I failed to observe much actual peer correction. It is likely that there was none – perhaps since neither student felt qualified to correct the other’s pronunciation. However, both of the students in this group made progress with their pronunciation, and although this study was limited only to the correct production of individual phonemes, I did notice that prosodic features of pronunciation were more in evidence towards the end of the study than they were at the beginning.
It is highly likely that these students benefitted from the practice afforded them in this study, but the impact of peer correction itself was negligible.
Finally, the following chart shows the progress made by student S5, who worked with the two students whose performances have not been included in this study.
At the start of the study, student S5 had the worst pronunciation of all the students in the class. As well as producing L2 that was severely impacted by L1 (the majority of S5’s mistakes were connected with over-production of /r/), S5 also produced language that suggested incomplete learning of grammatical forms (such as the overly literal reading of past simple -ed endings, with ‘walked’ produced as / wɔːlkɛd/ instead of /wɔːkt/).
S5’s performance was the most variable of any of the students, though by the end of the study the number of mistakes appeared to match that recorded at the start. Although this might at first glance seem to suggest that practicing reading aloud has no effect in the present case, attention must be drawn to Practice 4. Here the number of mistakes dropped to a level that would position S5 among the best performers in the class, and listening to the recording immediately reveals why. Here S5 read the text more carefully than in any other case, concentrating especially on controlling the /r/ phoneme and eliminating it as necessary.
Unfortunately, whatever progress was recorded in Practice 4 had disappeared by Practice 5; excepting Practice 4, the average performance of S5 shows almost no change whatsoever across the whole period of study.
I spoke to S5 after the final test, and asked what might have been learned from the experience. It appears that this month spent practicing reading aloud has had an effect, though it has yet to bear fruit: S5 is now at least aware of the problems with their pronunciation, but lacks the discipline when speaking to control phonemic production.
The scale of this study is too small to offer a generalisation, but that is not to say that lessons cannot be learned from the results of the research.
To answer RQ1, it appears that the worst thing that can happen by practicing reading aloud in this manner is nothing at all – in other words, there is no downside other than that the students have spent some time in the lesson doing something that probably did not improve their pronunciation. None of the students were worse at producing spoken English (albeit not spontaneous spoken English) by the end of the study compared to at the start, and in many cases they showed improvement of their production of individual phonemes, or of prosodic features of production (though these were not tested), or they at least reduced the incursion of L1 features into their L2 production (for example by producing ‘Wi-Fi’ as /waɪfaɪ/ instead of the inappropriate /wiːfiː/).
The answer to RQ2 is far more positive. The procedure adopted in this study can ameliorate the concerns that Gibson (2008) broached but did not address directly. One danger of reading aloud in open class is that it involves one student at a time, and that when the students are not engaged in reading aloud, they might be engaged in nothing at all. If each student in a class of twelve were required to read aloud from a text for a minute at a time, that would involve the use of at least twelve minutes of classroom time – and more still would be required for correction and further practice. By making use of Vocaroo and Zoom Breakout Rooms, it is clearly possible to run RA practice with groups working in parallel, meaning that in those same twelve minutes, all of the students might practice RA. The teacher can still check their production, though this will happen later – either during a quiet period of the lesson, or afterwards, and thus adds to the burden of teaching – and can prepare appropriate feedback or further discussion.
It is not clear from this study whether peer correction has any effect. I believe it depends largely on the students’ attitudes to the practice; if they don’t wish to engage, they cannot be forced to, and will not benefit from any notional advantages bestowed by the approach. The idea of peer training encompasses both research questions: RA should be more effective if more proficient students help less proficient students by providing feedback (RQ1); but if a student chooses to take advantage of the limitations of Zoom’s Break-Out Rooms they will not make the best use of the time allotted to them (RQ2) and that time will indeed have been wasted – and, worse still, the teacher might not be aware of this, as they can only observe the Zoom Break-Out Room they are currently in.
As a small scale study, the present research can only answer either of the research questions in a limited manner; however, it has at least highlighted the potential that technology can offer, especially in such situations as that faced by the students in this study, where lessons have been moved online due to the COVID-19 pandemic. If, in the future, some lessons remain online, the use of such technological support will likely continue, hopefully with improvements to those areas of difficulty discovered in this research.
Anderson, J. (2020), “Reading aloud”: What it’s really called and why it’s essential to formal language learning, Retrieved from https://speakinggames.wordpress.com/2020/11/11/reading-aloud-what-its-really-called-and-why-its-essential-to-formal-language-learning/ (Accessed 20/12/2020)
Gerich, D. (2013). Beyond the Class Blog: Creative and Practical Uses of Blogger for the ESL Classroom. TESOL Journal 4.1: 175-181
Gibson, S. (2008). Reading aloud: a useful learning tool? ELT Journal 62/1: 29-36
Jenkins, J. (2002). A Sociolinguistically Based, Empirically Researched Pronunciation Syllabus for English as an International Language. Applied Linguistics, 23/1: 83-103.
Kohnke, and Moorhouse (2020). Facilitating Synchronous Online Language Learning through Zoom. RELC Journal 1-6.
Koskinen, P. & Blum, I. (1986). Paired repeated reading: A classroom strategy for developing fluent reading. The Reading Teacher, 40/1: 70-75.
Martinez Adrian, M. (2014). The Efficacy of a Reading Aloud Task in the Teaching of Pronunciation. Journal of English Studies, 12: 95-112.
Puchta, H. and Stranks, J. (2010). English in Mind 2, 2nd Edition. Cambridge University Press.
Serafini, T. M. and Blair, R. (2017). Can You Hear Me Now? An Innovative Approach to Assess and Build Connections with Online Learners. Systemics, Cybernetics and Informatics, 15/6: 7-11.
Strickland, D. S., Ganske, K., & Monroe, J. K. (2002). Supporting struggling readers and writers: Strategies for classroom intervention 3-6. Portland, ME: Stenhouse.
Please check the Pilgrims f2f courses at Pilgrims website.
Please check the Pilgrims online courses at Pilgrims website.
A Small Scale Study of Reading Aloud Enabled by Technology Use
Christopher Walker, Poland
Extensive Reading Using Online Resources: Case Study
Rasha Abdel Haliem Osman, Egypt
Poster Presentation in the Language Classroom
Ian Michael Robinson, Italy
Use of Self-regulated Learning Strategies in Paragraph Writing at Van Lang University
Mai Tran Thi Thanh, Vietnam