Revisiting the Relationship Between Receptive Vocabulary Knowledge and Receptive Language Proficiency – a Multi-Group SEM Approach - presented by Hung Tan Ha

Revisiting the Relationship Between Receptive Vocabulary Knowledge and Receptive Language Proficiency – a Multi-Group SEM Approach

Hung Tan Ha

Abstract

Speaker

AI Outline

Discussion

REVISITING THE RELATIONSHIP
BETWEEN RECEPTIVE VOCABULARY
KNOWLEDGE AND RECEPTIVE
LANGUAGE PROFICIENCY - A MULTI-
GROUP SEM APPROACH
HUNG TAN HA (& DUYEN THI BICH NGUYEN)

02:35

WHAT'S ON THE MENU?
I. Current literature & Research gaps
2. Our hypotheses
3. What we did & what we found

03:00 - 03:10

1

1

1

1

1

RECEPTIVE VOCABULARY
KNOWLEDGE?
Vocabulary knowledge
Productive:Writing & Speaking
Receptive: Reading & Listening
Breadth: Quantity
Depth: Quality
(Nation, 2022)

1

03:15 - 04:20

THE FOCUS OF OUR STUDY: A RELATIONSHIP
Receptive
Receptive
Vocabulary
Language
Knowledge
Proficiency

04:30 - 04:36

A COMPLICATED RELATIONSHIP...

04:43 - 04:45

CURRENT LITERATURE &
RESEARCH GAPS
The relationship between L2 listening and reading
comprehension and phonological and orthographic
vocabulary knowledge is very complicated because:
I. Issues of modality
2. Questions concerning Skills vs. Language

04:59 - 05:05

1

2

1

2

1

2

3

4

CURRENT LITERATURE & RESEARCH GAPS
MODALITIES
Theoretically, the orthographic or auditory modalities of vocabulary measures are sensitive to the respective modalities of L2 comprehension tests (Milton, 2009; Read, 2000).
Researchers are encouraged to use the corresponding modalities of vocabulary knowledge
when investigating the relationship between vocabulary knowledge and L2 comprehension.
Orthographic vocabulary tests should be used in reading research.
Aural vocabulary tests should be used in listening research.
Some researchers followed this (Cheng, Matthews, Lange, & McLean, 2022; Ha, 2021a)
And some just didn't (Reynolds, Xie & Pham, 2022; Staehr, 2009)
Tests of phonological vocabulary didn't exist, accent, other reasons

1

2

3

4

5

6

05:15 - 06:57

CURRENT LITERATURE & RESEARCH GAPS
MODALITIES
Receptive Vocabulary Knowledge
Phonological Vocabulary Knowledge
Orthographic Vocabulary Knowledge
0.60 (Meta-analytic data)
0.52 (Meta-analytic data)
(Zhang & Zhang, 2022)
(Zhang & Zhang, 2022)
0.765 (Meaning-recall)
Listening
(Stewart, McLean & Batty, 2021)
Comprehension
0.671 (Meaning-recognition)
Receptive
(Stewart, McLean & Batty, 2021)
Language
0.652 (Meaning-recognition)
0.648 (Meaning-recognition)
Proficiency
(Ha, 202la)
(Ha, 202la)
0.49 (Meta-analytic data)
0.60 (Meta-analytic data)
(Zhang & Zhang, 2022)
(Zhang & Zhang, 2022)
0.778 (Meaning g-recall)
Reading
(Stewart, McLean & Batty, 2021)
Comprehension
0.699 (Meaning-recognition)
(Stewart, McLean & Batty, 2021)
0.609 (Meaning-recognition)
624(Meaning-recognition)
(Ha,2021a)
(Ha,2021a)

1

2

3

08:16

1

1

2

CURRENT LITERATURE & RESEARCH GAPS
MODALITIES
Ha (2021a) carried out a Z-test and found no statistically significant differences between
the correlations.
Zhang & Zhang (2022) conducted a Q-test and only found significant differences
between the .49 and .60 correlations.
Stewart, McLean & Batty (2021) didn't conduct a test of significance for these correlations, but rather wrote: "The findings show that correlations of forms of written
vocabulary knowledge to listening skills are very similar to their correlations to reading proficiency, albeit slightly lower." (p. 59)

1

2

3

09:58 - 10:25

1

2

3

4

5

1

2

3

4

5

1

2

3

4

5

6

1

2

3

4

5

6

CURRENT LITERATURE & RESEARCH GAPS
MODALITIES
In fact, the correlations between phonological and orthographic word knowledge have been found to be from .70 (Cheng & Matthews, 2018; Hamada & Yanagawa, 2023) to as
large as 0.89 (Ha, 2021; Mizumoto & Shimamoto, 2008; Shin, Lee & Choi, 2023)
.90 => the degree at which multicollinearity occurs
Unidimensionality? (González-Fernández, 2022)
=> The two modalities of vocabulary knowledge could be unidimensional and contribute to a latent variable of "Receptive Vocabulary Knowledge".
Question: What is the contributory role of receptive vocabulary knowledge in
listening and reading comprehension?

1

2

3

4

5

6

10:47 - 12:23

1

1

2

1

2

CURRENT LITERATURE & RESEARCH GAPS
SKILL OR LANGUAGE?
L2 comprehension is a language problem or skill problem (Anderson, 1984)?
Jeon and Yamashita (2014) showed that reading comprehension had significantly higher
correlations with linguistic resources such as vocabulary (r=.79) and grammar (n = .85) compared to decoding (r = .56), phonological awareness (r = .48), orthographic
knowledge (r = .51), morphological knowledge (r = .61), working memory (r = .42),
metacognition (r = .32), and LI reading comprehension (r = .50).
Hypothesis: A change in linguistic resource should not lead to a change in the degree to
which linguistic resource (vocabulary) could predict L2 comprehension.
Question: To what extent do the contributory role of receptive vocabulary knowledge in listening and reading comprehension differ amongst groups of
students with different vocabulary sizes?

1

2

12:27 - 14:15

1

WHAT WE DID
Establish
Collect Data
Analyze Data
Hypothesis
Ha (202la)
Multi-group

1

14:21 - 14:39

1

1

2

1

2

3

4

1

2

3

4

THE DATA
The present study re-analyzed the dataset used in Ha (2021a) which include the results of 234 second-year university students on:
The Updated Vocabulary Levels Test (Webb, Sasao & Balance, 2017)
The Vietnamese Listening Vocabulary Levels Test (McLean, Kramer & Beglar, 2015; Ha, 2021b)
An IELTS Listening Test
An IELTS Reading Test (Academic Module)
Tests administered in 2 meetings, in two consecutive weeks.
Students did the tests on paper, with pens & pencils
Speakers were used for the listening tests. No technical issues encountered.

1

2

3

4

14:50 - 15:20

DESCRIPTIVE STATISTICS
LVLT
UVLT
IELTS Listening
IELTS Reading
Mean
89.4
86.2
5.16
3.74
4.65
2.47
6.09
5.15
4.63
23.4
26.7
2.65
1.83
1.88
2.08
2.90
3.24
4.12
Skewness
-0.313
-0.104
-0.411
0.604
-0.227
0.857
0.0554
0.414
0.493
Kurtosis
-0.196
-0.615
-0.703
0.410
0.0403
1.13
-0.453
-0.577
-1.07
0.919
0.910
0.805
0.735
0.931
0.921
0.815
0.758

15:27

OUR BASELINE MODEL
—_—
pa
ce
/
ee —

15:38

MODEL FIT
Acceptable Fit
Present study's
González-Fernández (2022)
model
Multidimensional
Unidimensional
17.2
16.117
15.476
x2/df
0.748
1.01
0.97
>.95
0.997
>.95
1.000
1.000
1.000
>.95
1.008
SRMR
<.05
0.018
0.021
0.019
RMSEA
<.05
0.000
0.01
0.00

1

16:21

SINGLE-GROUP SEM
0.80
0.25
UVLT
0.63
0.39
0.87
0,33
0.78
Reading
-0.73
0.47
0.82
Vocabulary
0.83
0.45
0.55
0.83
0.74
Listening
0.64
0.59
0.74
0,86
0.31
0.76
LVLT
0.42
0.26

17:11

SINGLE-GROUP SEM
Phonological and orthographic vocabulary knowledge contributed
equally to the latent variable of Receptive Vocabulary Knowledge
Receptive Vocabulary Knowledge similarly predicted Listening and
Reading comprehension
Hypothesis confirmed

17:38 - 17:42

TWO VOCABULARY SIZE GROUPS
Based on the participants' performance on the UVLT (150 items) High Vocabulary Size (HVS): correctly answered 90 items and above (>3000 word families)
Low Vocabulary Size (LVS): correctly answered 89 items and lower (<3000 word families)
The results of independent t tests for the two vocabulary size groups
Phonological
Orthographic
Listening
Reading
Vocabulary
Vocabulary
Comprehension
Comprehension
Knowledge
Knowledge
N =112
N = 122
N=112
N = 122
N=112
N = 122
N=112
N = 122
Mean
109.0
65.3
107.2
73.1
20.1
12.3
20.6
11.5
13.93
17.68
13.63
16.52
5.10
5.96
6.63
7.36
< 0.001
< 0.001
< 0.001
< 0.001
Cohen's
2.13
2.88
1.39
1.31
Note: HVS = High Vocabulary Size; LVS = Low Vocabulary Size; LC = Listening Comprehension; RC = Reading
comprehension; SD = Standard Deviation

18:09 - 18:32

MODEL FIT (AGAIN
Acceptable Fit
Multi-group SEM
Model I
Model 2
Model 3
Model 4
Model 5
49.3
70.1
71.6
80.8
x2/df
1.07
1.35
1.33
1.39
2.04
>.95
.995
.994
.994
.993
.987
>.95
.995
.971
.972
.964
.887
>.95
.992
.960
.962
.955
.879
SRMR
<.05
.041
.075
.078
.103
.135
RMSEA
<.05
.025
.055
.053
.058
.095
Decision
Accept
Reject
Reject
Reject
Reject
Model I: No constraints
Model 2: measurement weights equal
Model 3: measurement weights and structural weights equal
Model 4: measurement weights, structural weights and structural variance equal
Model
measurement weights, structural weights, structural variance equal and measurement residual equal

19:57 - 19:59

$HIGH VOCABULARY SIZE GROUP / N al / \ ae) 9.8$

20:43

KEY RESULTS OF THE MULTI-GROUP ANALYSIS
Structural Weights and Residuals (in beta values)
High Vocabulary Size Group
Low Vocabulary Size Group
Reading - Vocabulary
Regression weight: 0.798
Regression weight: 0.455
Disturbance: 0.363
Disturbance: 0.802
Listening ~ Vocabulary
Regression weight: 0.608
Regression weight: 0.588
Disturbance: 0.630
Disturbance: 0.654

21:06 - 21:08

1

1

1

1

1

EXPLANATION
Listening:
due to the temporary nature of aural input (Cheng & Matthews, 2018),
learners had to rely on what they can comprehend to answer the questions.
Chances for test-taking strategies or blind guessing, if any, were equal for all
students.
Reading:
Written input is more temporally stable and could be revisited many times
(Cheng & Matthews, 2018).
Students with larger vocabulary size could understand more, and therefore
based their answers on what they could comprehend.
Students with limited vocabulary knowledge could not understand much of
the text, but can somehow give correct answers for the questions.
Reading comprehension is prone to test-taking strategies.

1

24:14 - 25:40

WHAT IT MAY MEAN & WHY WE
SHOULD CARE
On the one hand, for L2 listening comprehension, a change in vocabulary knowledge as a linguistic resource did not lead to a change in the degree to which vocabulary can
predict listening comprehension, which seems to support the view that second language
listening is a language problem.
For L2 reading comprehension, only the other hand, the degree to which vocabulary knowledge can explain the variance in reading comprehension drops as the students'
vocabulary size is low. This signifies a possibility that knowledge or skills on other aspects could make up for the loss in linguistic resource, and therefore partly supports the stance
that second language reading poses a reading problem.
Emphasizes the need to actively control potential covariates that exist naturally in the
reading comprehension construct.

25:50 - 27:09

WHAT IT MAY MEAN & WHY WE
SHOULD CARE
The research provides robust evidence for the unidimensionality of the two key modalities of vocabulary knowledge. However
The unidimensional model does not suggest that assessing only one aspect of vocabulary
knowledge is sufficient when the aim is to fully gauge learners' word knowledge. The findings show that each of the word-knowledge aspects is an essential part of overall vocabulary
knowledge, and thus basing conclusions on the mastery of just one of these aspects provides only partial insights about lexical knowledge and can lead to the underrepresentation of this
construct (González-Fernández, 2022, p.27).

1

27:22 - 28:06

LIMITATIONS
A mismatch between measures of vocabulary knowledge and
reading/ listening passages.
Meaning recognition test formats (multiple-choice & matching)
Measures of grammar knowledge currently missing.
The absence of other factors affecting L2 comprehension.

29:06 - 31:09

REFERENCES
Alderson, J. C. (1984). Reading in a foreign language: A reading problem or a language problem? In J. C. Alderson & A. H. Urquhart (Eds.), Reading in a foreign language (pp. 1-24). London: Longman.
Cheng, J., & Matthews, J. (2018). The relationship between three measures of L2 vocabulary knowledge and L2 listening and reading. Language Testing, 35(1), 3-25 https://doi.org/10.1177/0265532216676851
Cheng, J., Matthews, J., Lange, K., & McLean, S. (2022). Aural single-word and aural phrasal verb knowledge and their relationships to L2 listening comprehension. TESOL Quarterly, Early View.
https://doi.org/10.1002/tesq.3137
González-Fernández, B. (2022). Conceptualizing L2 Vocabulary Knowledge: An Empirical Examination of The Dimensionality Of Word Knowledge. Studies in Second Language Acquisition, 44(4), 1124-1154.
https://doi.org/10.1017/S0272263121000930
González-Fernández, B. (2022). Conceptualizing L2 Vocabulary Knowledge: An Empirical Examination of The Dimensionality Of Word Knowledge. Studies in Second Language Acquisition, 44(4), 1124-1154.
https://doi.org/10.1017/S0272263121000930a

1

2

3

4

31:39

LALS Seminar Series

Linguistics and Applied Language Studies (Victoria University of Wellington)

Cite as

H. Tan Ha (2023, May 26), Revisiting the Relationship Between Receptive Vocabulary Knowledge and Receptive Language Proficiency – a Multi-Group SEM Approach

doi.org/10.52843/cassyni.rszt8r

Share

Timestamped permalink

Details

Listed seminar This seminar is open to all

Recorded Available to all

Video length 42:01

Q&A Now closed

Cassyni impact seminars

Showcase your research, get noticed and get cited.