Page 39 - ELG2009 Sep Issue 472

THOUGHT LEADER . FEATURES & COMMENT

Trusting the eviidence community to do better than it has tended to

do so far.

Is there any unbiased research to
improves outcomes?
Ron Ragsdale talks to Hamish Chalmers, BRIAN A JACKSON / SHUTTERSTOCK suggest that using L1 in the classroom
In my systematic review of research on using
language education researcher and former the L1, I found very little research (11 studies
in total) that met my inclusion criteria. Of
international school teacher, about the these, only four studies found that using the
L1 helped improve L2 vocabulary acquisition.
power of randomised trials However, in all of these studies the L1 being
used was the same for all pupils, a language
that was shared with their teachers. In the
An Oxford Professor once told the been routinely adopted by researchers. It is UK it is far more common to find many
Gazette that you “came from an common to find research that fails to take into different languages represented.
evidence-based family”. Can you account the potential for biases. We routinely Technology could allow us to use ebooks
enlighten us? see research that compares groups that are set up so that when you click on a word, an
t's true, funny as that sounds. My dad is a systematically different to one another. This Randomly assigning participants to L1 gloss of that word appears. Alternatively,
medical researcher and uses randomised makes it very difficult to determine whether groups removes bias and increases an approach that does not rely on technology
trials to assess alternative approaches to the differences in outcomes are a result of the the validity of results might be encouraging students who share the
Ihealthcare, so I was brought up on the systematic difference between groups or a same L1 to discuss concepts in the classroom
idea of fairness in decision making. For result of the different teaching approaches The risk, therefore, to L1 maintenance and However, parents have chosen which schools Is it practically possible to do this? Is using that language before carrying out work
example, when deciding whether it was mine being compared. cultural identity is lower than it is in schools to send their children to. It is plausible that there a good reason that this research around those concepts in English.
or my brother’s turn to do the washing up, we We also often see research that does not in the anglosphere with high proportions of bilingual education is responsible for superior doesn’t exist? In both cases, these could be easily
would both voice competing reasons why it even have a comparison group. Researchers students who speak non-English languages at performance, but it is just as plausible that It may be ignorance of the importance of bias compared in a randomised trial, where
shouldn’t be ours. To resolve this fairly, Dad assess students, teach them in a particular home, such as government schools in there is something about families who elect reduction in educational research. It might classrooms are allocated to either use L1 in
would flip a coin. way, then assess them again. Any differences Southern California, state schools in to elect to send their children to bilingual be that researchers in our field are less well this way or not, then compare outcomes at
Naturally, I would complain that it still between before and after is attributed to the suburban London, or independent schools or schools that predisposes them to perform versed in randomised trial methodology. the end.
wasn’t fair, especially if it was the second or approach, when in fact the difference could private language schools in the UK with large better generally. There is also a myth that randomised trials
third time in a row that it was my turn. So, be explained by any one of a number of proportions of students from overseas. Another issue is the use of proxy are difficult and expensive. The only thing You have been involved in a number of
Dad flipped the coin 100 times and things. I don’t think this is good enough. From an educational sense, international outcomes. For example, we often see self- that is unique about randomised trials is how systematic reviews in language
demonstrated that over time we would end schools must consider whether there might reported enjoyment, feelings of participants are allocated to comparison teaching. Can you tell us about your
up doing the washing up the same number of be an academic advantage in incorporating empowerment, notions of respect for groups. What can be trickier is getting the most recent review?
times. As I grew up, I came to appreciate the Do you want students’ L1s into their teaching. Firstly, multilingualism and so on as outcomes. scale needed to produce trustworthy results. I was part of a team at Oxford that prepared
power of randomisation. trustworthy research there is a self-evident advantage if both the These are all really important, but they are However, this is not unique to randomised a rapid evidence assessment focussed on
L1 and English are used in academic not the same as better test scores, or higher trials. foreign language instruction. We took a series
How does this relate to your own that gives you an contexts, as graduates will be better equipped graduation rates, or whatever metric you The important element here is that we do of published reviews, assessed the quality of
research? accurate answer (even than students who have developed these believe is reflective of an effective education as much as we can to reduce bias when we their methods, chose the most rigorous and
Since my days studying to become a teacher, I skills in English only. system. We must be careful about what we conduct research so that we can be more updated them.
have wanted to understand the effects of when that is There is also the idea that the L1 can can legitimately claim on the basis of existing confident in our findings. That’s the power of This will be published shortly by the
what we do in the classroom. unexciting) or support academic learning in English- research. randomised trials. Education Endowment Foundation. It
If you are trying to assess whether a medium international schools. If top IGCSE This is even more problematic where lots Another criticism of randomised trials is brings together hundreds of studies and we
particular way of teaching is better than untrustworthy results are a goal for the school, for example, of different L1s are spoken. There is no that they tend to report slim or no differences believe that it fairly assesses the state of our
another, you have to use the two approaches, research that gives does using the L1 make that more or less research that robustly examines the in effectiveness of alternative approaches. knowledge about what we can claim about
observe what happens, then compare likely? From the research that we have, it educative effects of using L1 where more This is blaming the tool for the shortcomings the effects of language teaching.
observations. However, you must create you a clear answer seems likely that using the L1 to explain than one L1 is represented. Taking the of the worker. Biased studies can exaggerate
groups of students that are as alike as possible. that is probably abstract concepts helps students understand findings of research from bilingual schools differences and mislead us.
Groups of people are messy. Some possess content and might even improve English where everyone speaks the same L1 and Remember, in this case we are comparing Hamish Chalmers
characteristics that influence how they learn wrong? language skills. extrapolating it to monolingual schools with using the L1 with whatever teaching is course director of
in one way, and others possess different The Achilles heel is the purchasing power lots of different L1s is problematic, regardless approaches are already being used, which the MSc in Applied
characteristics that will influence how they of parents. Because ‘English only’ tends to be of the quality of the research. may be very effective. Dramatic effects are Linguistics and
learn in a different way. The trouble is, when You are particularly interested in the assumed by parents (and many teachers) to If we want to claim that using L1s is more rare, so it is likely that using the L1 will have Second Language
we create comparison groups we can see and use of L1 in the school classroom, yet be the most effective way of teaching, schools effective, we need research conducted in a slim effect. Acquisition in the
measure some of these characteristics, but we many schools are known for insisting who don’t offer this risk losing out. those contexts. Despite an enormous number Do you want trustworthy research that Department of
can’t see or measure others. So, how do we on English only. Does research give of international schools and local schools gives you an accurate answer (even when Education at the
account for differences that we can’t see or any insights on how this should You have posited that there is a lack of with significant numbers of language minority that is unexciting) or untrustworthy research University of
measure, so that we know our comparison change? evidence-based findings to students, this either hasn’t been done or that gives you a clear answer that is probably Oxford. He is also
groups are similar? From an ideological perspective, substantiate what “use of L1” should suffers from proxy outcomes and biased wrong? Vice-Chair of
We randomise. Random allocation to international schools in particular must look like in the L2 classroom. Do we comparisons. Randomised trials are not difficult. Fair NALDIC (The National Association for
groups spreads out the mess fairly. We then consider whether it is a priority for them to need to have the results of more This is not some esoteric point about tests are how we get closest to ‘the truth’. In Language Development in the Curriculum).
have confidence that any differences in honour the linguistic and therefore cultural rigorous research in order to justify methodology, it has real world consequences: education, far too many studies are unfair He was a primary school teacher prior to
average characteristics of those groups are backgrounds of their students or whether any particular stance at all? if we tell teachers they should move to a tests; wasting time, energy and resources and moving into academia, most recently working
chance differences and not systematic they want to inculcate an ‘anglo-cultural’ Much of the research on the effects of using multilingual model, there will be costs, in failing to adequately inform the very people as Director of English as an Additional
differences (or biases). milieu in the school. the L1 has either not adequately controlled terms of time, energy, money and that researchers are ultimately trying to help: Language (EAL) at an international School
Actually, in many international schools for potential biases or has measured proxy opportunity. Schools should be confident that teachers and students. in Thailand. His research interests are use of
Isn’t this standard research there is a legal requirement to teach the outcomes for actual attainment. a switch to multilingual pedagogies will be an I want teachers to have confidence in the L1 as a pedagogic tool in the mainstream
methodology? home languages to home students. Moreover, For example, children who go to bilingual improvement before making these using the results of randomised trials to education of EAL learners, fair tests for
It should be, but I have been astounded to opportunities to use the L1 outside schools schools do as well as or better than children investments. This can only come via good inform their practice. I see nothing difficult causal inferences, and evidence synthesis.
find that this most basic consideration has not are plentiful. who go to monolingual English schools. quality research. or controversial about asking the research

38 September 2020 editorial@elgazette.com 39