Transcript Sample Final Presentation - Minnesota State University Moorhead

Effects of Shared
Sound and Spelling on
False Word
Recognition
Christine Malone, Kristen Bouwman,
Susan Johnson, Julie Przekwas
Minnesota State University, Moorhead
Background Information


False memories can be induced in a list learning paradigm.
The most commonly used paradigm is the DRM Paradigm

Studied words with common semantic (bed, rest, wake) or sound
(paradise) information can induce false recognitions of their
corresponding base words on the test list (sleep and paradox,
respectively) (e.g., Roediger & McDermott, 1995; Wallace, Malone, & Spoo,
2000).

Both sound and spelling influence spoken word
recognition in a priming paradigm
(e.g., Slowiaczek, Soltano, Wieting, & Bishop, 2003).

example
Rationale for Study



Previous work has shown both sound and
spelling to influence spoken word recognition in
a priming paradigm.
Previous false memory research has
manipulated only spelling or sound and left the
other factor free to vary.
We aimed to systematically study sound and
spelling information to understand their
individual and combined contributions in
creating false recognitions of unpresented
words.
Sample Stimuli







Initial Sound: centigrade – sentiment
Ending Sound: peppermint – sentiment
Initial Spelling: gentle – genius
Ending Spelling: radius – genius
Initial Both: tickle – ticket
Ending Both: pocket – ticket
Unrelated: welfare – ticket
Hypothesis

If sound and spelling are automatically
activated as the study word unfolds, study
words with both shared sound and spelling
(parasite) should activate their
corresponding target (paragraph) to a
greater extent during study and seem
more familiar at test. This should lead to a
greater proportion of false recognition
errors to their target (e.g., report having
heard paragraph on the study list, when, in
fact, it was parasite).
Method

Participants


86 MSUM psychology undergraduates
Design

2 (position: initial & end) x 3 (shared
information: sound, spelling, & both)
within-subjects design
Procedure

Recognition Memory Paradigm



Participants were asked to listen carefully to a study list
consisting of 186 words.
The study list contained 90 words which were each
randomly presented twice. Also, the list included 3 buffer
words at the beginning and end of presentation.
Participants were then presented a second, shorter list and
were asked to respond “old” or “new” after each word.
 “Old” indicated that the word appeared on the first list
 “New” indicated that the word did not appear on the first
list
Results


There was a significant difference
between the three shared information
variables, F (2, 85)= 23.15, p < .0001.
An interaction was found between
position and shared information,
F (2, 85)= 3.40, p < .036.
Error rates across conditions
0.35
Error Rate
0.3
0.25
0.285
0.241
0.197
0.2
0.191
0.15
0.1
0.165
0.165
0.05
0
Initial
Ending
Position of Shared Information
Sound
Spelling
Both
Conclusions


Sound and spelling
connections do
influence recognition
memory
Both together
produce the highest
error rates

Semantic
connections are
most often studied
in long-term
memory, but sound
and spelling need to
be investigated as
well
Questions?