More Related Content
Similar to Event Parsing and Verb Learning in Children
Similar to Event Parsing and Verb Learning in Children (20)
Event Parsing and Verb Learning in Children
- 1. RESEARCH POSTER PRESENTATION DESIGN © 2015
www.PosterPresentations.com
Verb acquisition requires children to segment dynamic scenes and link different
elements to specific verbs. Most prior verb studies have included only relevant
events. Two studies ask 1) how well children learn verbs while seeing relevant and
irrelevant events, and 2) if they can parse events, linking only relevant subevents to
new verbs. Structural Alignment theory (e.g., Markman & Gentner, 1997) predicts
that events (or subevents) with few alignments will be discarded. In Study 1, 2 ½-
and 3 ½-year-olds saw a Relevant First (RRDDR), Distractor First (DRDRR), or
Alternating (RDRDR) set of events, and learned two verbs; 3 ½ year olds succeeded
across conditions. In Study 2, 2 ½- and 3 ½-year-olds saw events with distractor
and relevant subevents (either Relevant First or Distractor First); 3 ½-year-olds
performed best in the Distractor First condition. This is an important new direction
for verb research, with links to Structural Alignment and other emerging theories.
ABSTRACT
Prior research in event processing shows:
• 10-11 month-olds attend to the structure of intentional events (Baldwin et al.,
2001),
• 6-8 month-olds attend to embedded target actions within events (Hespos,
Saylor, & Grossman, 2009) ,
• And, for 14 month olds, the order of relevant/irrelevant events affects infants’
processing of events (Su-hua Wang, 2013).
• In addition, preschoolers can reproduce actions even when two actions are
interleaved (Loucks & Meltzoff, 2013).
What about in verb learning?
Study 1: When learning verbs, can children focus on interleaved relevant events?
• Hypothesis: Events that are not alignable in a set will not be compared.
Study 2: When learning verbs, can children focus on relevant parts of events?
• Hypothesis: Subevents that are not alignable also will be ignored.
INTRODUCTION
Participants
• 2 ½ year olds (range: 25m-35m), n= 20
M=32m; 3 ½ year olds (range: 36m-47m), n =
25, M=43m; 4 1/2-year-olds (range: 48m-58),
n=12, M= 52m
Design
BS: Condition (Distractor First or Distractor
Last)
BS: Age group (2.5, 3.5, or 4.5 years)
DV: total number correct
Materials
• iPad to present stimuli
• Video camera to record participants
Stimuli
Videos were created in either a kitchen or park
setting.
STUDY 1 STUDY 2 RESULTS
GENERAL DISCUSSION
In these studies, children were asked to learn a verb while seeing both
relevant and distractor events. Two main results emerged. One is that
there is developmental change between 2 1/2 and 3 1/2 years. Two ½-
year-olds in Study 1 had trouble focusing on relevant events which were
interleaved with irrelevant events and in Study 2, they had difficulty
parsing relevant events from attached distractor actions. Three-year-olds
succeeded at both of these tasks, which fits predictions made by
Structural Alignment theory (e.g., Markman & Gentner, 1997). In Study
1, no effect of order of events emerged, while in Study 1, children who
succeeded seemed to do best when the distractor event immediately
preceded the relevant event. A goal in future studies is to further
examine whether there are conditions under which 2 1/2-year-olds can
succeed in these types of tasks.
Overall, these findings are important for understanding how children
parse and compare events when learning new verbs. Few studies of verb
learning have asked how children ignore irrelevant events or parse
events, and thus the present findings are important. Verb learning is
essential to children’s language development.
ACKNOWLEDGEMENTS
Thank you!
Parents and their children, undergraduate RAs in the Children’s Research Lab, child
care centers (including: UPCC LHUMC, FPCC, St. Andrews), the Mach Foundation,
and the McNair Foundation
Megan Dolan, Emily Donnan, Laura Flores, Shelly Gordon and Jane B. Childers
Department of Psychology, Trinity University, San Antonio, TX
Event Parsing and Verb Learning:
How Well Do Children Learn Verbs Amidst Distractions?
STUDY 2
Procedure
2 Warm-up Trials
Experimental Phase
Each child learns two novel verbs: gorp and
snarf.
• Learning Phase: depending on condition,
the participant watches 3 sets of videos
showing an attached distractor event and
target event.
• Test Phase: Children point to either a
correct or incorrect action
WARM UP PHASE
See Figure 1
DISTRACTOR LAST LEARNING PHASE DISTRACTOR FIRST LEARNING PHASE
Condition displays relevant action flowing into a
distractor event
“She’s going to gorp it.
She’s gorping it. She
gorped it!”
“Do you see what she’s
doing? Oh…look!”
“She’s going to gorp it.
She’s gorping it. She
gorped it!”
“Now look what she can
do. Wow!”
“She’s going to gorp it.
She’s gorping it. She
gorped it!”
“Oh! Do you see what
she’s doing? Look!”
Figure 5
Condition displays distractor events flowing into
a relevant action
“Do you see what she’s
doing? Oh…look!”
“She’s going to snarf it.
She’s snarfingit. She
snarfed it!”
“Now look what she can
do. Wow!”
“She’s going to snarf it.
She’s snarfingit. She
snarfed it!”
“Oh! Do you see what
she’s doing? Look!”
“She’s going to snarf it.
She’s snarfingit. She
snarfed it!”
TEST PHASES
“Which one is gorping it? Can you show me?
Can you point to it?
“Which one is gorping it? Can you show me?
Can you point to it?
“Which one is snarfing it? Can you show me?
Can you point to it?
“Which one is snarfing it? Can you show me?
Can you point to it?
STUDY 1
METHOD
METHOD
Participants
• 2.5 (range: 25m-35m), n = 23, M= 31m, and
3.5-year-olds (range: 36m-50m), n = 22, M=
42m
Design
BS: Condition (DRRDR, RRDRD, or RDRDR)
BS: Age group (2.5 or 3.5 years)
DV: Total number correct
Materials
• iPad to present stimuli
• Video camera to record participants
Stimuli
Videos were created in either a kitchen or park
setting.
Procedure
2 Warm-up Trials
Experimental Phase: Each child learns two
novel verbs: gorp and snarf.
• Learning Phase: the participant watches
multiple scenes: 3 show examples of the
relevant action, 2 show distractor actions;
the order of scenes depends on condition
• Test Phase: Children point to either a
correct or incorrect action
Figure 1. “Look at
these things. Can you
point to the rolling
fire truck?”
WARM UP PHASE
“Do you see what she’s
doing? Oh…look!”
DRRDR LEARNING PHASE
“Look! She’s going to
gorp it. She’s gorping it.
She gorped it!”
“Hey! She’s going to gorp
it. She’s gorping it. She
gorped it!”
“Now look what she can
do. Wow!”
“Look! She’s going to
gorp it. She’s gorping it.
She gorped it!”
TEST PHASES
Figure 2
Figure 3. Correct
and incorrect
action displayed
in a split-screen
format. Child
asked to point to
the event that
corresponds to
the novel verb
from the learning
phase.
“Which one is gorping it? Can you show me?
Can you point to it?
“Which one is gorping it? Can you show me?
Can you point to it?
STUDY 1 RESULTS
Figure 7. Univariate ANOVA with Age group (3: 2, 3, 4) and Condition (2: D First, D
Last) as between-subjects factors; dv = number of correct points at test. Results
show a main effect of Age group, F(2,45) = 7.12, p<.002, p2 = .26. Post-hoc tests
with Sidak corrections showed 2 ½-year-olds differed significantly from both 3 ½
and 4 ½ year olds (p< .02), which did not differ from each other. One sample t-tests
show responses exceeded chance in both 3 ½-year-old groups, and in the 4 ½-year
old condition, ps< .001.
Figure 6
Figure 4. Univariate ANOVA with Age group (4: younger 2s, older 2s,
younger 3s, older 3s) and Condition (3: DRRDR, RRDRD, RDRDR) as
between-subjects factors; dv = number of correct points at test. Main
effect of age group, F(1,44) = 21.07, p<.001, p2 = .35. One sample t-
tests showed younger 3s, t(12)= 2.22, p<.05, and older 3-year-olds
succeeded, t(17) = 8.04 , p<.001.
0
0.5
1
1.5
2
2.5
3
3.5
4
Younger 2s Older 2s Younger 3s Older 3s
TotalCorrect
*
*
0
0.5
1
1.5
2
2.5
3
3.5
4
2s 3s 4s
TotalCorrect
*
*