Comparison of Psychometric Properties of Foreign Language Learner Strategy Inventories:
SILL, LSUS, and LASSI in their Czech Adaptation
Katerina Vlckova, Iva Hudeckova, and Katerina Svejdikova (Institute for Research in School Education, Faculty of Education, Masaryk University, Brno, Czech Republic)
Research question
• For measuring declared FLLS use, inventories based on previous inductive research from early stages of FLLS research were developed.
• It is important to know how precisely these instrument measure strategies, if and how the instruments are inter-correlated and which is better in predicting achievement.
• This study compares psychometric properties of 3 mainly used FLLS inventories:
1) SILL - Strategy Inventory for Language Learning (Oxford, 1990).
2) LSUS - Language Strategy Use Survey (Cohen, Oxford & Chi, 2002).
3) LASSI - Learning and Study Strategies Inventory (Weinstein, Schulte & Palmer, 2002).
Research sample
Reliability coefficient Cronbach’s alpha
Discussion
• The FLLS use was highest at LSUS, than SILL and lowest by LASSI.
• Regarding concurrent validity, the scores of strategy use of all 3inventories were
inter-correlated. The strongest correlation were between SILL and LSUS.
• Reliability coefficient Cronbach´s alpha reached an acceptable level for all the
inventories. In all inventories, the reliability for men was slightly higher.
Acknowledgement: This study was funded by Czech Science Foundation - GAP407/12/0432 Foreign Language Learning Strategies and Achievement: Analysis of Strategy Clusters and Sequences.
Foreign language learning strategies (FLLS)
• Important concept in theory, research of SLA, language learning and teaching since 1960s:
• Capture a wide range of linguistic behaviours.
• Operations to acquire, retain, retrieve information or perform (Rigney, 1978).
• Ways in which the learner selects, acquires, or integrates new knowledge
(Weinstein, Mayer, 1986).
• Sets of conscious thoughts and actions that a learner takes to achieve a learning goal
(Chamot, 2004).
• Connected concepts :
• Self-regulation, metacognition, learning styles, cognitive style.
• Strategy classification:
• Most often classified according to psychological functions - cognitive, metacognitive,
and socio-affective (O'Malley, Chamot 1990),
• or 4 language skills (Cohen, Weaver 2006).
• Strategy choice and use is influenced by different variables:
• e.g. gender, experience, motivation, language proficiency.
Examples of invertory items
„To understand unfamiliar words, I make guesses.“
„I first skim an English passage (read over the passage quickly) then go back and read
carefully.“
„I find the meaning of an English word by dividing it into parts that I understand.“
Method
• Translation and adaptation of the 3 inventories for the Czech conditions (SILL: Vlčková, 2007; LASSI: Hudečková, 2012)
• Partial standardization (LSUS: Vlčková & Přikrylová , 2011).
Data collection
• All 3 inventories were completed in a random order one week after each other by the same 126 students .
• Non-random sampling.
• Students reported their strategies of their preferred FL .
Year of sampling 2012
N 126
Schools level upper secondary
Schools type comprehensive
Students´age 17-18
Female 69 %
Preffered FL English 73 %
Years of preffered FL learning mostly 8-10
Results
vlckova@ped.muni.cz
Average SD Scale
(points)
SILL 3.02 .41 5
LSUS 2.66 .31 4
LASSI 2.77 .28 5
Declared FLLS use
α α
male
α
female
No.
items
Scale
(points)
SILL .91 .92 .89 72 5
LSUS .91 .92 .91 89 4
LASSI .82 .87 .79 80 5
R
SILL/LSUS .66
LASSI/LSUS .43
LASSI/SILL .41
Scores of strategy
use of all 3
inventories intercorrelated
at p < .05.
The order of administration of inventories affected the reliability:
• LSUS – the lowest reliability when administered as first inventory.
• LASSI – the lowest reliability when administered as the last one.
• SILL – reliability was the highest when administered as the last one.
Effect of administration order on reliability
R
School mark
R
Self-assessment of FL competence
SILL - .20 .12
LSUS - .20 - .01
LASSI .22 (negative) .34
Concurrent instruments´ validity Correlation of strategy use scores with achievement indicators
Significant at p < .05.
• The order of administration of inventories affected the reliability of each instrument.
• Predictive power of the inventories for students´ achievement was very low, though
statistically significant in all cases. Self-assessment correlated best with LASSI scores
and school mark correlated best with LASSI, but negatively – students with worse
marks used strategies measured by LASSI more.
Combination of administration
order of the inventories
Number of
students
ABC SILL-LASSI-LSUS 20
ACB SILL-LSUS-LASSI 23
BAC LASSI-SILL-LSUS 15
BCA LASSI-LSUS-SILL 18
CAB LSUS-SILL-LASSI 19
CBA LSUS-LASSI-SILL 25