Many language teachers observe that listening often falls behind other L2 competences, and bemoan a lack of truly effective methods for training listening skills. Jiang, Kalyuga and Sweller (2018) thus investigated what constitutes effective training in listening skills for L2 learners: reading, listening, or both. Their results show that certain approaches to instruction may benefit learners at different levels.

Background

The research takes cognitive load theory as a basis. The theory derives from work into human memory and the ‘architecture’ of cognitive structures. The term ‘cognitive load’ refers to the demands of particular tasks on working memory. It is important here to distinguish between ‘intrinsic’ and ‘extraneous’ load: Whereby intrinsic cognitive load stems from the inherent complexity of information, extraneous cognitive load stems from the general limits of working memory to manage large amounts of information simultaneously.

Various factors can affect cognitive load:

1. The Transient Information Effect –  When long and complex information cannot be retained for long enough to be processed by working memory. This makes listening, i.e. processing auditory information, difficult.
2. The Redundancy Effect – When pieces of information can be understood in isolation, without needing to be integrated mentally, multiple sources of information (e.g. a graph plus a written summary of the same data) become redundant. Processing such redundant information can create unnecessary cognitive load and use up working memory resources.
3.  The Expertise Reversal Effect –  Instructional designs which are effective for lower-level learners may become ineffective or even detrimental for more advanced learners.  This is linked to redundancy “in that information that is essential for novice learners may possibly become redundant for more knowledgeable learners” (p. 1143).

Research

The overall aim of the study was to investigate whether training involving just reading, just listening, or reading and listening differed in its effectiveness at improving listening  comprehension skills of L2 learners at different levels. The study’s four hypotheses reflect the researchers’ expectation to find evidence of the Expertise Retrieval Effect, showing that different instructional designs in language skills training may be beneficial for different groups of learners. This expectation is based on the results of other studies that looked at the combination of audio and visual materials for training listening skills, for example videos with sub-titles. Such research has shown that this combination can help lower-level learners in learning to ‘map’ orthographic onto phonological representations of words and thus in acquiring a sound-form-meaning knowledge of L2 forms. Because higher-level learners, on the other hand, are likely to have mastered basic form-sound-meaning ‘mapping’ in the L2, the visual support may be less necessary for them and may even lead to redundancy, the researchers surmise. However, the nature of auditory input can still lead to the Transient Information Effect, whereas written texts are more permanent and may offer more opportunity to maximise acquisition of new language. Thus the researchers set up the following hypotheses:

• Hypothesis 1. Foreign language learners with lower listening proficiency provided with
  read-and-listen instructions will outperform lower proficiency learners provided with
  listen-only or read-only instructions in the listening performance test.
• Hypothesis 2. Foreign language learners with lower listening proficiency provided with
  read-and-listen instructions will experience a lower level of extraneous cognitive load
  than lower proficiency learners provided with read-only and listen-only instructions.
• Hypothesis 3. Foreign language learners with higher listening proficiency provided with
  read-only instructions will outperform higher proficiency learners provided with listen-only
  and read-and-listen instructions in the listening performance test.
• Hypothesis 4. Foreign language learners with higher listening proficiency provided with
  read-only instructions will experience a lower level of extraneous cognitive load than
  higher proficiency learners provided with listen-only and read-and-listen instructions.

The study was divided into four experiments. These followed the same procedure: 1) Training on new language (using reading only, listening only, or both). 2) Cognitive load questionnaire. 3) Listening comprehension test – summarise information in your L1.

Experiment 1: compared lower-level (IELTS < band 5) and higher-level (IELTS > band 6) learners. Participants were 160 undergraduate students at a Chinese university.

Experiment 2: repeated experiment 1 with lower-level learners and less complex/abridged materials. Participants were 96 undergraduate students at a Chinese university.

Experiment 3: repeated experiment 2 with even lower level learners. Participants were 105 secondary-school pupils (aged 15-16).

Experiment 4: repeated experiment 3 with adult learners. Participants were 96 low-level adult learners of French (undergraduate students at a Chinese university).

Results

Experiment 1

Statistical analyses of comprehension task scores showed that the higher-level students who had only read texts during the learning phase significantly outperformed those who had only listened to audios while learning the new language. Among the lower-level learners, though, no significant effect was found of the type of training on students’ scores.

Regarding cognitive load, a statistical analysis of students’ questionnaire answers found no significant connection between the students’ language proficiency and the type of training on measures of intrinsic cognitive load. However, the read-only groups reported a significantly lower level of extraneous cognitive load than the read-and-listen groups, though no significant differences were found between the read-only groups and the listen-only groups. There were no significant differences in reported levels of cognitive load between the less and more proficient students.

Experiment 2

Statistical analyses of these lower-level students’ scores showed that those who had only read the abridged texts to learn the new language scored significantly higher than those who both had read and listened to the abridged information. No significant differences were shown between the read-only and listen-only groups, nor between the listen-only and read-and-listen groups.

Experiment 3

Statistical analyses of these seconday-school pupils’ scores showed significant differences in that the read-and-listen group significantly outperformed the two other groups, who had learnt the new langauge either from a text only or from an audio only. No significant difference was found between the read-only and listen-only groups.

The group who only learnt the new language from audio input reported significantly higher levels of intrinsic cognitive load than the other groups. Regarding extraneous cognitive load, the listen-only group reported higher levels than the read-and-listen group. No significant differences were found between the read-only and listen-only groups.

Experiment 4

Statistical analyses of these very low-level students’ listening comprehension scores showed that the read-and-listen group scored significantly higher than the read-only and listen-only groups, whereby there was no significant different between the scores of students in these two groups.

Again, the group who only learnt the new language from audio input reported significantly higher levels of intrinsic cognitive load than the other groups. Regarding extraneous cognitive load, the read-and-listen group reported significantly higher levels of than the read-only and listen-only groups, where by there was no significant difference in reported cognitive load between these two groups.

Conclusions

These experiments demonstrate a link between the effectiveness of different ways of training L2 listening skills and learners’ levels of langauge competence.

Listening Skills

Hypothesis 1: Experiment 1’s results did not support the hypothesis that a combination of visual and auditory input benefits lower-level learners. The researchers surmised that this may due to the overall level of the testing materials being too difficult for these learners.  However, Experiment 2 did not provide support for this hypothesis, either. Researchers believed this was because even the “lower-level” learners in this case already had too much knowledge from their ~7 years of learning English. Experiment 3 therefore replicated Experiment 2, but with teenage learners who had been learning English for under 3.5 years. The results of Experiment 3, did find the combination of visual and auditory input in teaching listening to be more effective for these lower-level learners, but it was unclear whether this result was due to their age or the length of time they had been learning the L2. To discern which of these factors caused the effect, Experiment 4 repeated the procedure, but with adult learners of French who had been learning the language for one year. The results of Experiment 4 support Hypothesis 1, confirming that lower-level langauge learners may benefit more when their listening skills are trained through a combination of reading and listening input.

Hypothesis 3 can be accepted: Experiment 1’s results confirm that teaching new language with written texts benefits higher-level learners’ listening comprehension more than teaching only with audio input.

Cognitive Load

Hypothesis 4: The results of Experiment 1 show that the cognitive load for higher-level learners is reduced when they study from written texts, instead of a combination of written and audio input, thus confirming Hypothesis 2. The researchers explain this result with the fact that a written text alone avoids the Transient Information Effect and Redundancy Effect which both cause increased load on working memory. How can reading improve the comprehension of audio input? The researchers  “assume that learners who read textual information may mentally  rehearse the sounds of the printed text either consciously or unconsciously, thus improving listening skills via this route” (p. 1150).

Hypothesis 2: Experiments 3 and 4 lead the researchers to accept the hypothesis that lower-level learners experience less cognitive load when learning new language from a combination of text and audio input. Since lower-level learners are not yet able to mentally rehearse the sounds of the printed text or successfully map orthographic onto phonological forms of words and phrases, the read-and-listen approach is more beneficial to them as it helps them to bridge this gap: It removes the Transient Information Effect but, for them, the multiple input sources do not cause the Redundancy Effect.

Discussion

Many teachers probably assume that the best way to train listening skills is to make learners listen. The findings here suggest that this is only partly true: This assumption seems only applicable for lower-level students. Indeed, the overall message from the findings of the experiments reported in this article is that “it is important to adjust listening teaching approaches dynamically to learners’ proficiency levels to facilitate meaningful learning” (p. 1139). However, the study is not able to provide a concrete suggestion of the level at which the combination of reading and listening ceases to be the most effective. This seems an interesting area for further research. Until we have answers to this question, though, experienced teachers will have to judge carefully, and encourage students to reflect on their learning.

 

Works Cited

Jian, D., S. Kalyuga & J. Sweller, ‘The Curious Case of Improving Foreign Language Listening Skills by Reading Rather than Listening: an Expertise Reversal Effect’, Educational Psychology Review, 30 (2018), pp. 1139–1165.