Foreign language learning and its impact on wider academic outcomes: A rapid evidence assessment (2020)

I was interested to come across this 2020 report for the EEF (Education Endowment Foundation) written by the University of Oxford Education Department. For some reason, it passed me last year. I don't recall it being widely publicised, which is a shame since it has interesting things to say - and not just about what the title refers to. Of most interest to me were its findings about MFL pedagogy.

First. here's the link:

Foreign_language_learning_and_its_impact_on_wider_academic_outcomes_-_A_rapid_evidence_assessment.pdf (d2tic4wvo1iusb.cloudfront.net)

The authors were:

Victoria Murphy, Henriette Arndt, Jessica Briggs Baffoe-Djan, Hamish Chalmers, Ernesto Macaro, Heath Rose, Robert Vanderplank and Robert Woore.

The report attempted to answer three research questions, the first of which interested me most. This is it:

i) the research identifying what approaches to teaching FLs are being used and what variables impact on the effectiveness of these approaches (p. 2)

All I'm going to do in this post is pick out some useful quotations from the report, adding a few comments here and there. Their approach was to do a trawl for studies and meta-analyses in two phases. In Phase 1 the authors identified six 'seed reviews' listed below.

Fitzpatrick Morris, Clark, Needs, Tanguay and Tovey (2019); Fox, Corretjer, Webb and Tian (2019); Fox, Corretjer and Webb (2019); Goris, Denessen and Verhoeven (2019); Graham, Choi, Davoodi, Razmeh and Dixon (2018); Harris and Ó’Duibhir (2011); and Lo and Lo (2014). 

Phase 2 consisted in researching any updates to the original reviews, including a sort of triage system to ensure the most reliable studies were included.

So what sort of things did the report come up with, as far as teaching is concerned?

The Executive Summary states on p.4

The general findings from this research indicate that more important than the specific method used is the way in which it is delivered and by whom. In other words, the programme characteristics and practitioner skills are key in impacting on successful FL learning and teaching. In general, approaches that are largely meaning-oriented, providing rich, authentic, and stimulating FL input for students, which increases the involvement load (how engaged the learner is with the task/language) tend to be more successful. 

I find that quotation slightly self-contradictory, actually, since on the one hand they say that teacher skill trumps method, but that both 'programme characteristics' and practitioner skills are key factors in outcomes. But I get their point and it's something I have always felt myself - whether you are an EPI teacher, a TPRS teacher, or communicative teacher, it's the skill of delivery which counts most. I'd even say that an excellent grammar-translation teacher is likely to produce better outcomes than an unskilled communicative teacher. Heresy?

On p.4 again

At the same time, however, there are numerous studies indicating that within this meaning-oriented approach there is scope for careful attention to specific linguistics features - often referred to as Focus on Form (FonF). These approaches need to be strategically employed and their effectiveness very much depends on characteristics of both the teacher (in terms of their skills as a practitioner and their proficiency in the FL) as well as the proficiency of the learner.

This is line with the common belief that some focus on grammar, phonics and other formal aspects of the language is useful, perhaps even necessary. This is suitably vague, since it doesn't get into the issue of how grammar might be taught, e.g. in a PPP fashion or a 'pop-up' fashion, for example.

The authors advocate (p. 5)

...the careful use of both implicit and explicit instructional approaches, appropriate use of technology, and strategy instruction. 

As a reminder, by explicit they mean directly teaching vocabulary, grammar and phonics and by implicit they mean let students pick up the language by repeated exposure and use. Researchers generally agree that the implicit route is the most powerful by far.

Pages 32-78 of the main review contain more details to build on the above general points. A good number of studies are briefly summarised and I should make clear that the results of studies on many issues (such as the use of images and gesture for vocabulary acquisition) tend to be mixed. But let me pick out some quotations and recommendations which are pretty much self-explanatory and which may help teachers put their own practice in context.

First, they sum up findings on vocabulary learning and teaching (p.49):

• Focusing attention and intentional learning activity on form and meaning of individual vocabulary items enhances vocabulary uptake, but this must be strategically applied. 
•  While input-only instruction is effective in terms of (limited) vocabulary uptake, learning gains are greatly enhanced when input is supplemented with some pre-teaching and further interaction.
•  Vocabulary learning is facilitated by tasks with high involvement load. 
•  The use of well-selected and graded L2 video clips and TV programmes, especially with L2 captions and some guidance from teachers, is an effective means of learning new vocabulary. 
•  Integrating creative imagery can boost vocabulary uptake as may songs; integrating gesture and movement yields more modest gains. 
•  Timing and variety of mini-interventions in learning has a significant impact on effective learning, regardless of approach/method. 
• Teacher experience and competence is a key variable in successful vocabulary teaching and learning.
•  Teachers should carefully consider learners’ proficiency level when adopting extensive or intensive reading approaches to vocabulary development.

Next, when it comes to grammar:

On the whole, the evidence base reported by Fitzpatrick et al. (2018) supports explicit instruction of grammatical structures over learners becoming aware of such structures implicitly (p.50).

Their summary of effective approaches to grammar teaching is as follows (p. 55):

• While explicit instruction in grammatical features is effective, it is rarely more effective than other types of instruction in developing grammatical competence and should be tailored according to age and level of proficiency. 
•  Both inductive and deductive types of explicit instruction are effective, but inductive may be slightly more effective under certain conditions. 
•  Attention to prosodic features (e.g. rhyme, rhythm) in oral input can aid the development of grammatical competence. 
•  Both oracy and literacy approaches in primary school can be effective in developing grammatical competence. 
•  Individual developmental differences in ability to handle linguistic concepts should be allowed for in primary school level teaching. 
•  Teacher language competence, experience and number of hours’ instruction are more influential factors than instruction type. 
•  Tutoring systems which provide immediate feedback on grammatical exercises can assist learners in achieving competence in grammatical structures. 
• Processing instruction may be worth considering as a structured approach to presenting difficult grammatical concepts. 

By the way, Processing Instruction is an approach advocated by NCELP and used to some extent in their lessons (freely available at ncelp.org). It involves designing input language and exercises in such a way that students are forced to make form-meaning links in the input. The classic example is omitting time phrases from drills or exercises, forcing learners to work out time frames by the verb form only. For example, if a student hears or reads:

"Last Saturday I went shopping."

They can understand the time frame with reference to the phrase 'last Saturday' which is clearly discernable at the start of the sentence and need not pay attention to the verb form. The theory is that by forcing students to notice the verb form, they are more likely to acquire it (internalise it). So 'structured input' is used to to reinforce form-meaning relationships - PI was first proposed by Bill VanPatten, but has probably featured in teaching before that.

Regarding reading and reading aloud:

Three studies reported by Fitzpatrick et al. (2018) confirm the value of phonics instruction for reading words, correct pronunciation as well as fluent and accurate reading aloud, if not improvement in comprehension (p. 55).

On higher level processing skills for reading:

Six studies of reading strategy instruction are reported by Fitzpatrick et al. (2018), all finding that it can assist in developing comprehension skills (Harris, 2007; Macaro & Erler, 2008; Macaro & Mutton, 2009; Manoli, Papadopoulou, & Metallidou, 2016; Martínez & de Zarobe, 2017; Mistar, Zuhairi, & Yanti, 2016) (p.57).

The summary on the effective teaching of reading is summarised on p. 59

• Effective approaches include explicit attention to both lower-level and higher-level processes since it cannot be assumed that either will be transferred from a student’s L1 without instruction. 
•  Phonological training can help beginner learners to process word forms, but not necessarily word meanings. 
•  Technology-supported learning can contribute to the development of oral reading fluency by facilitating student-centred learning. 
•  Watching video material supported by captions can assist learners in aural word recognition and “phonological recoding” of foreign language aural input. 
•  Computer-assisted learning of reading may enhance focus on task, reduce teacher dependency, and benefit lower ability learners. 
•  Explicit instruction in reading strategies and skills is effective in developing reading comprehension skills. A range of strategies can be taught together, but not in excessive quantities. 
•  Reading materials which are adapted to reduce cognitive load can help to support comprehension.

The final bullet point appears to support Nation's idea that input should be highly comprehensible (95-98%). It's in line with what Gianfranco and I suggest too.

The summary of effective teaching of writing is as follows (p. 64):

• Combining literacy and oracy teaching can be beneficial, especially in mixed attainment classes.
•  Technology has a potentially important role to play in writing development if carefully implemented but it cannot be assumed to be advantageous merely by its presence; its use must be informed and planned. 
•  Learners require training in using IT-based information for writing. 
•  Use of IT requires specific classroom management techniques. Pupils with limited technological skills must be appropriately supported. 
•  Interactive writing tasks promote proficiency and engagement. 
•  Individual differences between learners indicate the need for a wide variety of activities and strategies to develop writing skills. 
•  Strategy training (e.g. planning, morphological awareness, cognitive and metacognitive strategies) can improve writing performance.

I'd pick out the first bullet point above. Lessons benefit from practising the same language through listening, speaking + reading/writing modes. Good lesson plans often do this.

Concerning speaking and listening (p. 72):

• Approaches that create meaningful interaction more positively affect speaking skills than activities where the focus is on form. 
• Game-based activities (in contrast to e.g. drill-based) have a positive effect on pronunciation and communicative achievement. 
•  Online interaction and interaction with robots can be equally as effective as face-to-face interaction in creating an effective learning environment for spoken interaction. 
•  Opportunities to interact and to make language errors are important to the development of speaking skills. 
•  Use of automatic speech recognition (ASR) technology to provide feedback aids pronunciation and is more efficient but not necessarily more effective than a teacher. 
•  Task repetition aids pronunciation. Supporting authentic listening material (even with L2 input in a different medium) and moderating input both aid comprehension. 
• Viewing appropriately selected and graded L2 video material with or without L2 captions, either with teacher guidance or under learner control, can help to improve listening comprehension, listening skills and phonological recoding. 
•  Pre- listening support aids comprehension. 
• Strategy Instruction or providing opportunities for students to develop strategies for speaking and listening improves both competences.

Note in the above the references to technology - this is a recurrent theme in the whole report. The general message is: use technology where it helps learning, not for the sake of it.

In terms of general methodological approaches:

According to Fitzpatrick et al. (2018), the main conclusions of these reviews (e.g. Bauckham’s (2016) modern foreign languages pedagogy review concentrating on teaching practices in 73 Years 7, 8 and 9 in maintained schools in England and Wales) and Edelenbos, Johnstone & Kubanek’s (2007) report on early language learning) do not point to any particular methods or approaches or the most effective way of teaching; so much depends on the skill and competence of the teacher delivering the teaching irrespective of method. Our view is that this position certainly reflects the reality of language teaching and learning not only in the UK but all over the world (p.72-3).

I must say that this sensible conclusion seems to fly in the face of what we have been seeing from NCELP in England. While they acknowledge that there's 'no best method in all circumstances', the resources they produce do seem to promote a methodology based on very explicit teaching of phonics, vocabulary and grammar, with communication apparently relegated to a lower status - or at least thought to develop later. That's the way I see it, anyway.

Finally, here are a few further points with regard to timetabling and other issues. I have highlighted the last bullet point (p.77):

• Short-term intensive language programmes may be more effective than programmes where time is distributed over a longer period, but there are likely to be trade-offs and the evidence is inconclusive.
• Form-focused instruction should be strategically deployed, and may be more beneficial to older children. 
• Data driven learning such as training learners to use corpora for contextual word searches can develop language sensitivity, noticing, and inductive skills, and encourages autonomous learning and engagement with authentic language. 
• In general, considering the evidence from all six sections, approaches and methods can be less influential on learning than factors such as teachers’ language confidence, opportunities for language exposure, societal and educational context for learning, effective teacher training and ongoing skills development, motivating learners and making the language relevant to them in terms of their interests, community and identity.