Levels of Processing Effects in Bilinguals Recognition Memory

1
Levels of Processing Effects in Bilinguals
Recognition Memory

• Marisela Gutierrez
• Thesis Director Dr. Francis, Ph.D.
• University of Texas at El Paso

• Supported by a Teachers for a New Era Mini-Grant

2
Purpose

• Study the effects of levels of processing in
  bilinguals recognition memory based on their
  dominant and non-dominant languages.

3
Introduction

• Bilinguals recognition memory has not been well
  studied. We can think of working in the
  non-dominant language as working with less
  attention or with less familiar words.
• Bilinguals recognition memory can be understood
  by the effects of divided attention and word
  frequency on memory recognition.

4
Levels of Processing

• Memory storage varies in how deeply the items are
  processed.
• Deep processing- thinking about the meaning
• Shallow processing- visual features (e.g. vowels)
• The deeper the level of processing, the higher
  the probability that the information will be
  later retrieved.

5
Previous Studies

• Divided attention effect on recognition
• Memory performance is reduced.
• LOP effect is reduced.
• Word frequency effect on recognition
• Better memory performance for low frequency
  words.
• LOP effect stronger for low frequency words.

6
Predictions

• Based on effects of divided attention
• Memory performance was expected to be lower in
  the less fluent language.
• LOP effect was expected to be weaker in the less
  fluent language.
• Based on effects of word frequency
• Memory performance was expected to be better in
  the less fluent language.
• LOP effect was expected to be stronger in the
  less fluent language.

7
Method

• Participants
• Middle school students from TexPrep summer
  Program.
• N 46 bilinguals.
• Average age 14
• UTEP students
• N 64 bilinguals.
• Average age 20

8
Design

• 2 (language) x 3 (encoding condition)
• Languages English and Spanish
• Encoding conditions deep, shallow and not
  studied.
• Dependent variable recognition performance and
  response time.

9
Materials

• TexPrep students
• Instruction sheet
• Study sheet
• Recognition sheet
• Language Background questionnaire
• UTEP students
• Consent form
• PsyScope program
• Language background questionnaire

10
Procedure

• Study phase
• 54 words
• NP natural or made by people
• V count the number of vowels
• Recognition phase
• 108 words
• 2 studied sets 2 unstudied sets
• Different procedures

11
Hit and false alarm rates
Proportion YES responses
TexPrep students
UTEP students
12
Signal detection analysis
TexPrep students
13
Signal detection analysis
UTEP students
14
Response times (UTEP students)
15
Summary of results

• The TexPrep students performed better in the
  dominant language, but UTEP students did not show
  a language effect.
• The level of processing effects were significant
  for both groups in both languages.
• Levels of processing did not interact with
  languages.

16
Discussion

• Memory performance in bilinguals less fluent
  language cannot be adequately explained by either
  divided attention or by low word frequency.
• For future research, factors like type of
  materials and environment may be addressed.

17
Summary Table-TexPrep
Table 1. Recognition performance of TexPrep
students as function of language condition
Language Condition yes responses Corrected recognition d
L1 Deep 82.3 75.9 2.8
Shallow 54.6 48.2 1.8
False alarm 6.4
L2 Deep 75.7 67.1 2.3
Shallow 49.8 41.2 1.6
False alarm 8.6
18
Summary Table-UTEP
Table 2. Recognition performance of UTEP students
as function of language condition
Language Condition yes responses Corrected recognition d RT
L1 Deep 85.9 75.7 2.6 967
Shallow 63.4 53.1 1.8 1155
False alarm 10.3
L2 Deep 87.8 77.6 2.7 1006
Shallow 64.4 54.2 1.8 1161
False alarm 10.2