Interesting recent research according to Penny Ur - Pt. 2

 Penny Ur shares research she finds interesting at IATEFL 2023

Penny Ur posting for photos with fans
after her IATEFL talk (19.4.23)

This is the second instalment of my two-part post about Penny Ur’s session at IATEFL matter-of-factly entitled Interesting recent research, in which, in a span of 30 minutes, I learned more than during the rest of the conference. 

Click HERE for the first part.

The reason why it’s taken me so long to publish this (IATEFL took place in April) is that it’s not merely a summary of the talk. Upon my return I did some digging into the sources Penny Ur cited; the result is a summary combined with some personal reflections. By her own admission, she whizzed through so many studies in the talk that I only focus on those that are interesting to me.

The previous part covered the age factor, the use of L1 and error correction; in this one we turn to the following three strands of research: inferencing meaning from context, the flipped classroom and the use of pictures.

Inferencing meaning from context

Another Penny Ur’s hobbyhorse. She laid the groundwork by referencing some early studies, such as Laufer & Bensoussan (1984), in which students guessed correctly less than a quarter of target items in a text. Making a case for explicit vocabulary teaching, Laufer and Bensoussan argued that mere exposure to new vocabulary would not lead to vocabulary learning. 

Similarly, in Nassaji's (2003) study, learners could accurately guess less than 50% of the target words, according to Ur (in fact, the number of correct guesses was very similar to that in Laufer & Bensoussan, at 25%). “And he was playing fair with them”, she remarked – 95% of words in the text were familiar to learners (knowing at least 95% of words in a text, or 95% lexical coverage, is commonly considered the threshold for adequate reading comprehension – see, for example, Laufer & Ravenhorst-Kalovski 2010). Nassaji concurs that guessing is not an effective way of enriching vocabulary because of limitations inherent in inferring meaning from contextual clues.

More recent experimental studies compared guessing from context with other vocabulary learning strategies showing that the latter are more effective. In Zou’s experiment (2018), dictionary consultation led to better learning of word meanings while in Ko’s (2012) study providing glosses for unknown words either in L1 (Korean) or L2 (English) resulted in significantly better learning than guessing from context, which the author refers to as “risky and misleading”. Interestingly, the participants expressed a preference for L2 glosses, especially if they are concise and clear.

A recent review (Xiali & Altunel 2018) of empirical research on lexical inferencing concludes that depending solely on contextual clues is not an effective vocabulary learning strategy, particularly for lower L2 proficiency levels. The paper ends, however, with a recommendation to use concordance lines for vocabulary learning. The strategy does rely on contextual clues, but since learners are presented with several examples of the target word in context, they have a greater opportunity to figure out its meaning.

Further thoughts and practical implications

Studying concordance lines not only helps learners with the meaning of the target word, but also, perhaps more importantly, with how the word is used. It provides so called ‘condensed exposure’ or ‘condensed reading’ (Gabrielatos 2005) and is one application of Data-Driven Learning (DDL), which I delve into in my recent chapter in Demystifying Corpus Linguistics for English Language Teachers (Harrington & Ronan 2023). When it comes to resources available online, Nik Peachey has some good ideas HERE.

In general, it is important to distinguish between guessing from context as a vocabulary learning strategy and an exam strategy. Yes, it may not be effective for the former – and personally I have never relied on it as a source of vocabulary development – but it may come in handy when learners grapple with unknown words in a reading passage on an exam. See some useful suggestions on the EAP Foundation website: www.eapfoundation.com/reading/skills/guess/

The flipped classroom

Another 'trendy' topic, which, incidentally, I was discussing with colleagues (other IATEFL attendees) over dinner the night before Penny Ur’s presentation. 

Just to remind you, in the flipped classroom model students engage with instructional materials before class (e.g. a recorded lecture), thereby freeing up class time for discussions, pairwork and other interactive activities as well as practice activities which traditionally might be given for homework.

Although the concept has been around for some time, the trend for flipping was plausibly accelerated by the recent pandemic. According to Ur, good results have been particularly evident at university level, although very often students don’t do the preliminary viewing, which defeats the purpose of flipping.

But what about the L2 classroom? There is some encouraging news too. A new study by Chang (2023) shows the benefits of flipped instruction for language learners at A2 level. The findings indicate a significant improvement in writing and grammar among the group taught using a flipped approach. Flipped instruction also had a positive impact on the learners’ motivation and self-efficacy compared to the group that received traditional instruction.

However, a recent article by Kapur et al (2022) entitled "Fail flip fix and feed" urges practitioners to curb their enthusiasm. The authors, one of whom is the king of meta-analyses John Hattie (famous for his book Visible Learning) do not outright dismiss the value of flipping; quite on the contrary, they propose a new, modified model in which learners engage in problem-solving prior to instruction resulting in “productive failure”, which is then fixed in the instruction stage. Their main criticism is that the perceived success of the flipped model as reported by numerous studies they have analysed can largely be attributed to the additional time on task and additional exposure to the material rather than than the effect of flipping. In other words, students do better in a flipped model not because it’s flipped but because they’ve engaged with the material more.

Further thoughts and practical implications

I find the term “flipped classroom” overused or, rather, misused. Just the other day I had a discussion on Twitter about whether giving students some kind of pre-task, specifically asking them to watch a portion of a film or an episode of a TV series ahead of the lesson – with the aim of later discussing it in class –  constitutes a flipped classroom. As someone who has used a lot of films and TV series with my students (see some of my resources HERE), I fail to see how this is flipped in the sense that how it could possibly work otherwise (i.e. what the unflipped/traditional version would be): watching a whole episode in class and discussing it… at home? For me a flipped lesson implies doing what would normally be assigned as homework (e.g. exercises) in class and vice versa. Perhaps it’s not a relevant concept for language learning in the first place?

Pictures 

Finally, and very briefly, on to the last strand of research chosen by Penny Ur: using images to support vocabulary learning. Here only one study was mentioned – Farley et al (2012), which investigated the effect of visual imagery on the learning of abstract words. Participants – first-year learners of Spanish –  were divided into two groups and taught 12 abstract and 12 concrete words. The group that was shown pictures alongside new words recalled new items better than the no-picture group, suggesting that metaphorical or symbolic images can aid the learning of abstract words. Interestingly,  the same effect wasn’t found for concrete items. The authors attribute this surprising finding to the already existing connections between concrete words and images in our mental lexicons. Consequently, the addition of visual images is of less benefit compared to abstract words, where the association with images is not as readily available.

Further thoughts and practical implications

My attempt at supporting the learning of
the verb refuse with an image

I have occasionally included pictures in my Quizlet sets for concrete nouns (as you may know, I’m an avid user of Quizlet). But it hasn’t occurred to me till now that adding pictures for abstract items could support learning. 

In this set designed for B1/B2 level Business English students quizlet.com/284027830 I have tried adding images in almost every flashcard, including for abstract words such as willing and entitled.

Thank you, Penny, for making me dig into all these research studies and inspiring me to write not one, but two blog posts! 

You can see Penny Ur's Padlet with slides and references HERE

References

Only additional references not mentioned in the talk are listed here:

Laufer, B., & Ravenhorst-Kalovski, G. C. (2010). Lexical threshold revisited: Lexical text coverage, learners’ vocabulary size and reading comprehension. Reading in a Foreign Language, 22(1), 15–30. Available from http://nflrc.hawaii.edu/rfl/

Gabrielatos, C. (2005). Corpora and language teaching: Just a fling or wedding bells? TESL-EJ 8(4): 1-35. Available from http://tesl-ej.org/ej32/a1.htm