powerpoint 프레젠테이션€¦ · 3 support for unrelated sets interference effect (higa, 1963)...
TRANSCRIPT
1
3
1 Support for semantic sets
Hierarchical Network Model (Collins & Quilian, 1969)
• Lexical knowledge is organized hierarchically
• Advanced organizer
• Each concept is only linked to the closest concept
4
Spreading Activation Model (Collins & Loftus, 1975)
• Concepts are connected via nodes w/ different lengths of line
• Based on the degree of their association
• The stronger the association,
the closer the connection between the two concepts.
Support for semantic sets (cont.)
Hashemi & Gowdasiaei (2005)
• Group : semantically related (SR) vs. semantically unrelated (SUR)
• Proficiency: lower level vs. upper level
• 60 English learners in Iran (30 for each group)
• Both lower and upper level students in SR treatment reached
a significantly higher score than ones in SUR treatment
• Upper level students made greater gains than lower level ones.
5
2
3 Support for unrelated sets
Interference effect (Higa, 1963)
• When several similar words were introduced at the same time, these words will interfere with each other, impairing retention of them.
6
4 Support for unrelated sets (cont.)
Is the role of verbal ability explains exceptions to a strong age effect?
7
• Tinkham(1993) • Waring (1997)*
N=20 / L1 English, L1 Japanese* psuedowords with L1 translation
• Finkbeiner and Nicol (2003) N=47/ L1 English pseudowords with pictures
• Erten and Tekin (2008) N= 60/ L1 Turkish English vocabulary with pictures
• Bolger and Zapata (2011) N=66/ L1 English pseudowords using contextualized texts Eye-tracking study
8
1 “Real” Korean words
limited to English or Roman alphabet
the ecological validity
2
9
3 an immediate and delayed posttest
to see the effect of grouping words persists in the long term
2 a repeated-measures design
the error variance associated with individual differences
statistical power
1 In terms of accuracy on immediate and delayed posttests,
do SR and SUR groupings of Korean vocabulary differ in their effectiveness for
L2 learning? 2
10
3 Do pictorial and textual cues of testing have a differential effect on participants’ performance in terms of accuracy and reaction times?
2 In terms of reaction times on immediate and delayed posttests,
do SR and SUR groupings of Korean vocabulary differ in their effectiveness for
L2 learning?
11
12
Gender 9 males / 19 females
L1 English
Age 20. 54 years (SD: 2.50)
Formal instruction 2.36 semesters (SD: 1.31)
First exposure 18.50 years (SD: 4.07)
Weekly use 9.7 hours (SD: 7.78)
31 -> 28 students from the beginner and intermediate Korean classes
heritage speakers were excluded
48 unknown words
• selected from 120 basic words (word familiarity scale)
• elementary school level
• from K-6 vocabulary list
• concrete nouns
• # of letters, syllables, and frequency
the SuperLab software
A male/ female native speakers of Korean
Duration of recording = 1500 ms/word x 2
a voice recorder for interviews
13
14
SR condition Natural disaster
• 6 words
Kitchen utensils
• 6words
Shape
• 6words
Diseases
• 6words = 24 words
SR (Semantically Related) condition
SR condition A
• 6 words
B
• 6words
C
• 6words
D
• 6words = 24 words
SUR (Semantically Unrelated) condition
15
16
Picture
L1 English word
L2 Korean word
Sound clip of L2 Korean
• Repeat the word after listening
• Each block (6 words) repeated 3 times
17
Picture
L1 English word
L2 Korean word
Sound clip of L2 Korean
• Repeated twice
• 3 pictures in the same block
• Immediate feedback is provided Good job!
18
Picture
L2 Korean word
L2 – Pic Test L2 – Word Test
L1 English
L2 Korean word
19
Testing
SR
Word
Picture
UR
Word
Picture
Keep the same order of sets as the one in the training session
The order of modality was counterbalanced
The order of pictures was pseudo-randomized within each set (SR,UR)
20
1 Condition x Time
21
Accuracy Score Reaction Times
Condition* F(1,27)=10.27, p=.003, ηρ²= .28
Time**
F(1,27)=22.37, p<.001, ηρ²= .45
Condition* F(1,27)=11.17, p=.002, ηρ²= .29
Time
F(1,27)=3.06, p=.092, ηρ²= .10
SUR
SR
SUR
SR
2 Mode x Time
22
Accuracy Score Reaction Times
Mode x Time* F(1,27)=9.16, p=.005, ηρ²= .25
Mode* F(1,27)=11.17, p=.002, ηρ²= .29
Mode x Time*
F(1,27)=14.46, p=.001, ηρ²= .35
Picture
Word Picture
Word
*
*
1 “The words in sets were way more difficult. I mean, I knew that ‘hongsu (flood)) was a kind of natural disasters, but didn’t know what exactly it was.” (Participant 4, male)
3
23
Perceived Difficulties of SR sets
2 “The words related to kitchen were the most difficult because I don’t know, well, I guess because I’m not familiar with kitchen supplies. I think they all go.. together, but natural disasters were like…tsunami is a lot different from flood” (Participant 2, male)
Other factors
# of letters origin of the word (e.g. [choong-chee] cavity -> [chee] tooth presence of double consonants (e.g. [bbul], [kki]) -> pronunciation
Preference of testing modes
Picture (20) >>>> Word (8)
1 In terms of accuracy on immediate and delayed posttests,
do SR and SUR groupings of Korean vocabulary differ in their effectiveness for L2 learning?
24
3 Do pictorial and textual cues of testing have a differential effect on participants’ performance in terms of accuracy and reaction times?
2 In terms of reaction times on immediate and delayed posttests,
do SR and SUR groupings of Korean vocabulary differ in their effectiveness for L2 learning?
SUR condition -> a significant higher score
the effect of the SUR condition remained on the delayed posttest
the SUR condition -> shorter reaction times
the effect of the SUR condition remained on the delayed posttest
Presenting semantically grouped L2 words can cause semantic interference effects,
which in turn, can interfere with learning.
Picture testing mode -> higher score and faster response
the effect of the picture was absent on the delayed posttest
25
1 Make a unique contribution to the existing body of literature on the acquisition of Korean as a second language
2 Suggest the need for reevaluation of current pedagogical practice
Grouping related words is not the best way to teach vocabulary.
26
3 Provide some evidence to develop new, creative ways of introducing words to minimize interference effects of semantic relations
27
1 the third/ fourth language?
Asian language share the same cognates
might have facilitated the memorization of the target words
28
2 thematic grouping?
an effective presentation method in L2 vocabulary teaching
the proficiency level of the participants was not high enough
3 young learners?
no effect of clustering words for children (Papathanasiou, 2009)
the presence of concepts in
Ina Choi [email protected]
30
Do
n’t
kn
ow
th
e w
ord
Hav
e se
en o
r h
eard
th
e w
ord
Thin
k I k
no
w t
he
wo
rd
Kn
ow
th
e w
ord
Kn
ow
th
e w
ord
;
Thin
k I k
no
w t
he
mea
nin
g
Kn
ow
th
e w
ord
;
Fair
ly s
ure
of
the
mea
nin
g.
Kn
ow
th
e w
ord
an
d it
s m
ean
ing
1 2 3 4 5 6 7
감자 (potato)
어깨 (shoulder)
달력 (calender)
…
Revised
Hierarchical
Model (RHM)
31
L1
L2
Concepts (picture)
• L1: privileged access to meaning • Direct link :L2 words and L1 words
• L2 to L1 >> L1 to L2 • L1 to concepts >> L2 to concepts
• Backward translation (L2 to L1) Lexically mediated
• Forward translation (L1 to L2) Conceptually mediated
32
SR condition
Natural disaster
Kitchen utensils Shape Diseases
= 24 words
SR (Semantically Related) condition
SR condition = 24 words
SUR (Semantically Unrelated) condition
지진 가뭄 황사 해일 화산 홍수
행주 냄비 수세미 저울 국자 도마
직사각형 마름모 타원 원기둥 각뿔 사다리꼴
여드름 천식 충치 빈혈 다래끼 습진
A Kitchen utensils Shape Diseases
분지 빙하 경적 액체 묘목 이끼
간식 항구 흑연 도끼
송곳 물레방아
과일 망치 확성기
아가미 철봉 필기도구
일출 궁전 용수철 현미
경 합창 장신구