Can a word-recognition quirk predict foreign-language reading success?
Beginners who leaned more on look-alike words shared with Spanish went on to read English better, but for intermediate learners the reverse was true.
Source
Foreign language comprehension achievement: insights from the cognate facilitation effect
Study at a glance
- Design
- Cohort — Prospective predictive study: an English lexical decision task (cognates vs non-cognates), digit span and non-verbal IQ were measured mid-year, then used in backward regressions to predict official reading exam scores months later, separately for A2 and B1 learners.
- N
- Two separate groups: 62 A2 (low-proficiency) learners in Experiment 1 and one hundred and five B1 (intermediate) learners in Experiment 2.
- Population
- Adult native Spanish speakers (aged roughly 19 to 69) taking English classes at a public language school in the Basque Country, Spain.
- Outcome
- Official end-of-year English reading comprehension exam score, predicted from the size of each learner's cognate effect (reaction-time difference between non-cognate and cognate words) plus IQ, working memory, age, age of acquisition and self-rated proficiency.
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Spanish-speaking adults in English classes at two levels (beginner A2 and intermediate B1) did a computer task deciding whether letter strings were real English words; half of the real words were cognates that look like their Spanish translations (such as 'minute'/'minuto') and half were not. They also did a digit-span working-memory test and a short non-verbal IQ test. About 4 months later, their official reading comprehension exam scores were collected and predicted from these measures.
What they found
Both groups recognised cognates faster and more accurately than non-cognates. Among beginners, a larger cognate advantage predicted better exam scores, alongside higher IQ and higher self-rated proficiency, with the model explaining about 25% of the variance. Among intermediate learners, a smaller cognate advantage predicted better scores, along with earlier age of first English exposure, better working memory and higher IQ, explaining about 20% of the variance; chronological age did not predict success in either group.
The limits
What it doesn't show
The design is correlational, so it does not show that changing how much learners rely on their native language would change their reading scores. The two proficiency levels were different people, not the same learners followed over time, and they also differed in working memory and IQ, so the 'shift' from positive to negative is inferred rather than observed. Backward stepwise regression in modest samples can capitalise on chance, working memory was measured with a simple forward digit span, and all participants lived in a bilingual Spanish-Basque region with one language pairing tested.
Key terms
- Cognate
- A word that looks and/or sounds similar and means the same in two languages, such as 'guitar' and 'guitarra'.
- Cognate facilitation effect
- The finding that bilinguals recognise cognates faster and more accurately than non-cognate words; here measured as the reaction-time difference between the two.
- Lexical decision task
- A task in which people decide as quickly as possible whether a letter string is a real word.
- Revised Hierarchical Model (RHM)
- A model proposing that beginning second-language learners access meaning through native-language translations, with direct links to meaning strengthening as proficiency grows.
- CEFR levels
- The Common European Framework's proficiency bands; A2 is elementary and B1 is intermediate.
Flashcards
0 of 9 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
Which pair of words is an example of Spanish-English cognates?
Common questions
Why would relying on cognates help beginners but not intermediate learners?
Early on, linking new English words to similar Spanish words is a useful scaffold for getting to meaning; at higher levels, good readers seem to access meaning directly from English words, so heavy reliance on the native language signals less developed direct links.
Is this a true longitudinal study of the same people?
Each learner was measured once and their exam score was collected months later, but the beginner and intermediate groups were different people, so the change across proficiency levels is a comparison between groups.
Does older age make learning a language harder according to this study?
Chronological age did not predict exam scores in either group, although earlier age of first exposure to English did predict better reading among intermediate learners.
More on Bilingualism