Language Exposure Questionnaire PDF
Language Exposure Questionnaire PDF
Language Exposure Questionnaire PDF
Viorica Marian
Henrike K. Blumenfeld
Purpose: To develop a reliable and valid questionnaire of bilingual language status
Margarita Kaushanskaya
with predictable relationships between self-reported and behavioral measures.
Northwestern University, Evanston, IL
Method: In Study 1, the internal validity of the Language Experience and Proficiency
Questionnaire (LEAP-Q) was established on the basis of self-reported data from
52 multilingual adult participants. In Study 2, criterion-based validity was established
on the basis of standardized language tests and self-reported measures from 50 adult
Spanish–English bilinguals. Reliability and validity of the questionnaire were
established on healthy adults whose literacy levels were equivalent to that of someone
with a high school education or higher.
Results: Factor analyses revealed consistent factors across both studies and suggested
that the LEAP-Q was internally valid. Multiple regression and correlation analyses
established criterion-based validity and suggested that self-reports were reliable
indicators of language performance. Self-reported reading proficiency was a more
accurate predictor of first-language performance, and self-reported speaking proficiency
was a more accurate predictor of second-language performance. Although global
measures of self-reported proficiency were generally predictive of language ability,
deriving a precise estimate of performance on a particular task required that specific
aspects of language history be taken into account.
Conclusion: The LEAP-Q is a valid, reliable, and efficient tool for assessing the
language profiles of multilingual, neurologically intact adult populations in research
settings.
KEY WORDS: bilingualism, self-assessment, second language,
language proficiency, questionnaire development
B
ilingualism and multilingualism are the norm rather than the excep-
tion in today’s world (Harris & McGhee-Nelson, 1992), and the pro-
portion of linguistically diverse populations is increasing in the United
States (U.S. Bureau of the Census, 2003). These demographic changes are
reflected in the growing representation of multilingual and multicultural
populations in research and applied settings. However, research with bilin-
guals often yields inconsistent findings (e.g., Grosjean, 2004; Marian, in
press; Romaine, 1995). For example, bilingual cortical organization (e.g.,
Kim, Relkin, Lee, & Hirsch, 1997; Marian, Spivey, & Hirsch, 2003; Perani
et al., 1998; Vaid & Hull, 2002), lexical processing (e.g., Chapnik-Smith,
1997; Chen, 1992; Kroll & de Groot, 1997), and phonological and ortho-
graphic processing (e.g., Doctor & Klein, 1992; Grainger, 1993; Macnamara
& Kushnir, 1971; Marian & Spivey, 2003) have all been found to differ
940 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007 • D American Speech-Language-Hearing Association
1092-4388/07/5004-0940
depending on bilinguals’ ages of language acquisition, questionnaire that assessed the following four areas:
mode(s) of acquisition, history of use, and degree of pro- (a) age/time variables associated with L2 acquisition;
ficiency and dominance. These inconsistencies are further (b) environmental variables (e.g., number of L2 speakers
exacerbated by the absence of uniform assessment instru- at home; frequency with which L2 is spoken at home and
ments in bilingualism research. Those who work with bi- in the workplace; and father’s, mother’s, and siblings’ pro-
linguals and multilinguals often face the challenge of ficiency in speaking, reading, and writing L2); (c) affective
testing individuals whose language they do not speak variables (e.g., self-consiousness, cultural preference and
(Roseberry-McKibbin, Brice, & O’Hanlon, 2005) and thus identity, and motivation); and (d) self-evaluated L1 and L2
have to rely exclusively on self-assessed information, usu- proficiency in speaking, reading, and writing. Jia at al.
ally collected with improvised questionnaires. The need found that self-reported ratings of language proficiency
for a language self-assessment tool that is comprehen- were positively correlated with behavioral performance.
sive, valid, and reliable across bilingual populations and Similarly, in a series of studies on how language domi-
settings prompted a systematic approach to developing nance affects the degree of foreign accent and grammatical
the present Language Experience and Proficiency Ques- ability, Flege, MacKay, and Piske (2002) and Flege
tionnaire (LEAP-Q; see Appendix). et al. (1999) used a language history questionnaire that
targeted participants’ self-reported age of arrival in the
L2-speaking country / initial L2 learning; age of attained
Previous Self-Assessment Studies L2 proficiency; duration of L2 immersion; number of years
of L2 schooling; percentage use of L1/L2; frequency of ex-
In general, previous research suggests that self- posure to L2 TV, movies/videos, and radio; frequency of
reported language measures are indicative of linguistic use of L1/L2 in a working environment; and ability to
ability (e.g., Bachman & Palmar, 1985; MacIntyre, Noels, imitate foreign accents. Flege et al. (1999) found signif-
& Clement, 1997; Ross, 1998; Shameem, 1998; Stefani, icant correlations between language history and degree of
1994). Existing self-assessment tools for studying bilin- foreign accent in L2 and between language history and
guals span both domain-general (e.g., Bahrick, Hall, performance on a grammaticality judgment task (Flege
Goggin, Bahrick, & Berger, 1994; Delgado, Guerrero, et al., 2002). In a study of compuational language use,
Goggin, & Ellis, 1999) and domain-specific proficiency Vaid and Menon (2000) used a questionnaire that focused
(e.g., Flege, Yeni-Komishian, & Liu, 1999; Jia, Aaronson, on language preference for mental arithmetic (e.g., count-
& Wu, 2002; Vaid & Menon, 2000). For instance, in a ing, noting the time, remembering a telephone number).
study of the relationship between self-reported profi- The questionnaire also yielded self-reported ratings of
ciency and language performance, Delgado et al. (1999) language proficiency speaking, comprehending, reading,
tested Spanish–English bilinguals and correlated self- and writing, as well as participants’ age of arrival in the
assessed proficiency in English and Spanish with per- L2-speaking country/initial L2 learning; the setting of
formance on the Woodcock–Muñoz Language Survey language acquisition (home, school, other); duration of L2
( Woodcock & Muñoz-Sandoval, 1993). Delgado et al. immersion; the language of instruction in elementary and
found that participants assessed first-language ( L1) secondary school; and frequency of use of L1/L2 at work,
skills more accurately than they did second-language (L2) with parents, and with siblings, while thinking to self,
skills. Woodcock–Muñoz scores correlated with all self- and while dreaming. Language preference for mental
reported measures of L1 proficiency but with only self- arithmetic was found to correlate with variables in the
reported measures of L2 reading and writing (and not bilinguals’ language history, with the strongest predictor
with L2 speaking and understanding). Similarly, Bahrick being the language of early formal instruction followed by
et al. (1994) found that language dominance ratings corre- length of residence in the L2 country, onset of bilingual-
lated highly with performance on some tasks (e.g., category ism, and relative language dominance.
generation and vocabulary recognition) but correlated less A consistent aspect of these studies was their focus
with performance on other tasks (e.g., oral comprehension). on proficiency and history-related variables. However,
Together, studies of domain-general self-assesement in the studies diverged in three notable ways: (a) The dis-
bilinguals suggest that the relationship between self- tinction among language proficiency, dominance, and pre-
reported and behavioral measures of language perfor- ference remained largely unexplored, (b) behavioral tasks
mance varies across languages and tasks (e.g., Bahrick used to validate the questionnaires were limited, and
et al., 1994; Delgado et al., 1999). (c) questions and scales were not consistent across stud-
Studies of domain-specific proficiency assessment in ies. There is currently no uniform procedure for deter-
bilinguals have focused on grammatical ability (e.g., Jia mining bilingual language dominance and proficiency.
et al., 2002), the degree of foreign accent (e.g., Flege et al., Researchers frequently use distinct aspects of language
1999), and computational language use (e.g., Vaid status and performance to delineate the two, or they use
& Menon, 2000). For instance, Jia et al. used a 32-item the same measure (e.g., L1:L2 proficiency ratio) to define
942 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
environmental factors (e.g., Bialystok & Hakuta, 1994, and preference) was not within the scope of this study, their
1999; Snow, 1983; Snow & Hoefnagel-Höhle, 1978). There- availability in the LEAP-Q enables questionnaire users
fore, in this study, both language proficiency and language to weigh each of these measures against their variable of
history variables were considered necessary for specifying interest.
bilingual language status. Language acquisition. Age of acquisition has been
Given a theoretical framework that incorporates both shown to be tightly connected to language learning, to
language proficiency and language history, the LEAP-Q influence bilinguals’ ratings of language dominance, and
aims to capture factors that previously have been iden- to predict their performance on behavioral tasks (e.g.,
tified as important contributors to bilingual status: lan- Hyltenstam & Abrahamsson, 2003; Johnson & Newport,
guage competence (including proficiency, dominance, and 1989). For example, Flege et al. (2002) found that age
preference ratings); age of language acquisition; modes of acquisition influenced bilinguals’ dominance classifi-
of language acquisition; prior language exposure; and cur- cation and correlated with bilinguals’ sentence duration
rent language use. The LEAP-Q is based on question types ratios in both languages. Consistent with studies demon-
previously used in questionnaires assessing bilinguals strating maturation effects in L2 acquisition, the LEAP-Q
(e.g., Flege et al., 1999, 2002; Jia et al., 2002; Marian & elicited four age-of-acquisition measures for each language
Spivey, 2003; Vaid & Menon, 2000). spoken: (a) age of initial language learning, (b) age of at-
Language competence. Traditionally, the self- tained fluency, (c) age of initial reading (i.e., age at which
assessment literature has used three distinct measures participants started to read in each language), and (d) age
to index bilingual language competence: (a) language pro- of attained reading fluency.
ficiency, (b) language dominance, and (c) language pref- Moreover, the environment in which a language is
erence. Because conflating the three measures can render learned also influences proficiency attainment. For in-
interpretation of results difficult, each of them was stance, Flege et al. (1999) found that the number of years
probed separately in the LEAP-Q. Previous studies have of education received in an L2 country, years of residence
construed proficiency as an index of general abilities in an L2 country, average self-estimated use of L1 and
across language processing domains (e.g., Bachman & L2, and chronological age all influenced age-of-acquisition
Palmar, 1985; Stefani, 1994), including literacy-oriented effects on bilingual language dominance. The importance
proficiency, grammatical proficiency, vocabulary knowl- of environmental and contextual variables in language
edge, and discourse abilities (Bachman, 1990; Harley, acquisition was demonstrated by Carroll (1967), who found
Cummins, Swain, & Allen, 1990). Consistent with other a significant relationship between language performance
studies of bilingual self-assessment (e.g., Bahrick et al., and the extent to which the target language was used in
1994; Flege et al., 1999, 2002; Grosjean, 2004; Jia et al., the home. Therefore, the LEAP-Q elicits descriptions of
2002; Vaid & Menon, 2000), the LEAP-Q elicited profi- acquisition modes in terms of the learning environments
ciency ratings in speaking, listening, reading, and writing. and in terms of the extent to which these learning envi-
However, instead of collapsing proficiency ratings along ronments contributed to language acquisition.
the different performance domains into a cumulative Prior and current language exposure. The degree of
score (e.g., Flege et al., 2002), proficiency ratings ob- prior exposure to a language has been shown to influence
tained in this study were analyzed separately and were research findings (e.g., Birdsong, 2005; Genesee, 1985;
expected to yield different predictive information for dif- Kohnert, Bates, & Hernandez, 1999; MacKay & Flege,
ferent linguistic skills. For language dominance, partici- 2004; McDonald, 2000; Weber-Fox & Neville, 1999). For
pants in this study indicated dominance order for each of example, Flege et al. (1999) found that length of residence
the languages spoken. The debate around the utility of a in the United States influenced bilinguals’ sentence-level
single global measure, such as language dominance (e.g., performance, with various language abilities differen-
Oyama, 1978; Spolsky, Sigurd, Sako, Walker, & Arterburn, tially susceptible to language exposure (e.g., groups with
1968), versus multiple task-specific measures, such as more U.S. education had significantly higher rule-based
linguistic proficiency across domains (e.g., Bahrick et al., morphosyntax scores than groups with less U.S. educa-
1994; Fishman & Cooper, 1969), prompted the inclusion of tion but showed no differences in foreign accent ratings
both global (dominance) and specific (proficiency) mea- or in lexically based morphosyntax scores). Given the evi-
sures of language competence, making it possible for the dence that prior language exposure influences bilingual
LEAP-Q to examine their effectiveness in indexing actual performance, the LEAP-Q assessed exposure to a language
linguistic skills. Finally, LEAP-Q questions targeting pref- in four different environments: (a) in a country, (b) at school,
erence were posed in specific terms (e.g., preference regard- (c) at work, and (d) at home.
ing reading a text available in all languages) rather than as In addition to prior exposure, ongoing language use
general questions about the overall preferred language can influence research findings. For example, Jia at al.
to maximize reliability and interpretation. Although a (2002) found that mothers’ L2 proficiency and frequency of
comparison of the three measures (proficiency, dominance, speaking L2 at home were predictive of bilingual children’s
944 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
answers to LEAP-Q questions designed to measure a sin- sciences and disorders–related fields were reviewed (e.g.,
gle construct were predicted to cluster together during Demorest & Erdman, 1987; Kohnert et al., 2003; Kuk,
factor analysis, to show distinct patterns for each language, Tyler, Russel, & Jordan, 1990; Roseberry-McKibbin et al.,
and to yield factors indicative of L1 and L2 proficiency. It 2005). Second, questions were chosen and prepared on
was also expected that bilinguals’ language history would the basis of assessment materials previously used in ex-
predict self-reported proficiency levels in L1 and L2. perimental studies conducted in our laboratory, other lab-
oratories, and as described in the literature (Chincotta &
Underwood, 1998; Flege et al., 1999, 2002; Jia et al., 2002;
Method Vaid & Menon, 2000). Metacognitive judgment questions,
such as evaluations of the relative contribution of differ-
Participants. The questionnaire was administered to ent exposure variables to learning a language (e.g., in-
52 multilingual individuals (M = 27.29 years, SD = 5.92; teracting with family/friends, reading, etc.), also were
29 women, 23 men). Participants were recruited from the included. Third, the draft version of the LEAP-Q was
Northwestern University campus and the greater Chi- piloted with 8 bilingual and multilingual participants
cago metropolitan area communities. Participants varied (Marian, Blumenfeld, & Kaushanskaya, 2003). All pilot
in their education level from 2 years of college to a doc- participants spoke two or more languages and were rep-
toral degree (M = 18.04 years of education, SE = 2.62; resentative of the LEAP-Q target population. Fourth, on
range = 15–27 years). None reported a hearing, language, the basis of participants’ responses and feedback, the
or learning disability. Of the 52 participants, 11 spoke two LEAP-Q was revised for clarity and succinctness to ac-
languages, 19 spoke three languages, 12 spoke four lan- commodate efficient self-reporting across different lan-
guages, and 10 spoke five languages. Across participants, guages. Questions were considered on an item-by-item
34 languages were represented: American Sign Language, basis to ensure that none yielded missing values, outliers,
Belorussian, Bengali, Cantonese, Croatian, Czech, Dutch, or insufficient variability. Fifth, the resulting items were
English, Filipino, French, German, Hebrew, Hindi, piloted with members of the Bilingualism and Psycho-
Hungarian, Italian, Japanese, Korean, Mandarin, linguistic Research Laboratory and with members of the
Malayalam, Marathi, Norwegian, Polish, Portuguese, Communication Sciences and Disorders department at
Punjabi, Romanian, Russian, Spanish, Swahili, Tamil, Northwestern University. On the basis of the feedback
Taiwanese, Telugu, Thai, Ukrainian, and Welsh. received, the LEAP-Q was revised and distributed to the
Participants’ self-reported language history and pro- participants in Study 1.
ficiency measures can be found in Table 1. L2 acquisition Domains assessed by the LEAP-Q included acquisi-
ages ranged from 0 to 15 years, representing both simul- tion history, contexts of acquisition, present language use,
taneous and sequential bilinguals. Participants reported language preference and proficiency ratings (across the
being exposed to L1 most in the context of family, followed four domains of language use: speaking, understanding,
by friends, reading, radio, TV, and independent language reading, and writing), and accent ratings. Some questions,
study. Participants reported being exposed to L2 most in such as those inquiring about ages of L2 acquisition, were
the context of reading, followed by friends, TV, radio, applicable to all bilinguals tested; other questions, such
family, and independent language study. When asked as those inquiring about L1 learning from tapes, applied
to report how different factors contributed to language to a subgroup of bilinguals only (e.g., individuals in im-
learning, participants reported that learning L1 relied migrant communities who learned their first language
most on family, followed by friends, reading, TV, radio, incompletely from their families and attempted to
and self-instruction, whereas learning L2 relied most on maintain and expand L1 proficiency by means of self-
reading, followed by friends, TV, radio, family, and self- instruction). Questions pertaining to each language were
instruction (see Table 1). When asked to report proficiency designed to be identical, to accommodate variability in
in each language, participants reported highest proficiency histories of L1 learning, and to maintain maximal flex-
for understanding, followed by speaking, reading, and ibility of the questionnaire. Participants completed the
writing for both L1 and L2. questionnaire independently in approximately 25 min.
Materials and procedure. The development of the Analyses. Factor analysis was conducted to compare
questionnaire followed the steps outlined in Question- statistical clustering of questions with accepted dimen-
naires in Second Language Research (Dörnyei, 2003, sions of bilingual status. Seventy-seven attributes were
pp. 66–69) and included compilation of an initial item entered into the principal components analysis, which
pool, discussion of questions, omission of jargon, and clar- served as the extraction method, and a varimax rotation
ification and simplification of instructions and questions. method was applied. The statistical software was given a
First, an expert with experience in questionnaire develop- maximum of 100 iterations to converge on a factor solu-
ment was consulted (the expert’s primary area of exper- tion, and the rotation converged in 51 iterations. Patterns
tise was audiology), and questionnaires in communication of variables within a single construct were examined,
L1 history L2 history
commonalities underlying clusterings of variables within constructs were expected to appear a priori, others were
a single factor were identified, and the factor name was determined only after they emerged from analyses.
logically deduced. For factors that included variables It should be noted that the factor analysis in this
with both positive and negative loadings, positive load- study was not used for combining items on the question-
ings provided inclusionary criteria and described the naire into a single representative score. Although the
underlying construct reflected by the factor, whereas neg- LEAP-Q was constructed so that multiple questions mea-
ative loadings provided exclusionary criteria and indi- sured a single domain of language proficiency (e.g., four
cated an inverse relationship to the construct reflected questions targeted age of acquisition, four questions tar-
by the factor. The names were intended to capture the geted proficiency), the questions were not intended to
nature of the variables that clustered together and to yield multi-item scales. For example, although the ques-
suggest underlying commonalities among them. This tions that clustered together in the L2 Competence fac-
procedure is the standard way for performing factor anal- tor were representative of a single underlying construct,
ysis on large data sets (e.g., Tapia, 2001; Tran, 1994; Wu combining these questions into a subscale of language
et al., 2006; Zea et al., 2003). In this study, although some competence may not be effective at measuring L2
946 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
proficiency in a different sample of bilingual participants. their components and are listed in order of variance ac-
Theoretically, the construct of L2 Competence can be counted for. Cronbach’s alphas were calculated for each
expected to involve at least three different subdomains: factor to assess consistency of components within each
(a) age of acquisition, (b) length of immersion, and (c) self- and to assess the extent to which questions captured
reported degree of language proficiency. Differences similar information. (Cronbach’s alphas should be .7 or
across any of these domains would yield different bilin- higher for items in a set to be considered internally con-
gual profiles. Moreover, differences in population statis- sistent with each other, although cutoffs ranging from .60
tics would likely result in different factor structures to .80 have been used; see Miller, 1995.)
across studies. Given the complex and varied nature of L2 The first factor, accounting for the most variance,
acquisition across populations, questions in the LEAP-Q included self-reported proficiency and comfort with speak-
were intended to be considered separately and not to be ing, understanding, reading, and writing in L1; identifi-
reduced to a limited number of subscales. Factor analysis cation with L1-associated culture; and preference for
served as a tool for the determination of whether LEAP-Q reading in L1 (all positive loadings), as well as L1 accent
questions contributed to underlying constructs shaping (negative loading; Cronbach’s a = .85). The positive load-
bilingual status and was, therefore, valid; it was not used ings of L1 proficiency variables together with the negative
to define these constructs as universal scales. loadings of L1 accent ratings suggested that this factor
In addition, multiple regression analyses were con- was an index of cross-modal L1 Competence.
ducted to examine the relationship between language his- The second factor (in order of variance accounted for)
tory and language proficiency. Specifically, 16 attributes included age of initial L2 acquisition and age of attained
of language history (i.e., acquisition and fluency ages, L2 fluency (positive loadings), and comfort and proficien-
learning environments, exposure variables) were entered cy in understanding L2 (negative loadings). Cronbach’s
as independent variables into stepwise multiple regression alpha could not be calculated for this factor because of
analyses, with self-reported proficiency in understanding, negative average covariance among items, which violated
speaking, reading, and writing as dependent variables. reliability model assumptions. The negative covariance
Pearson R values, F and p values, and regression co- resulted from the inverse relationship between age of
efficients for the best predictor models are reported. (Note acquisition and language competence—that is, later
that magnitudes of beta values in regression analyses acquisition was associated with lower competence. Pos-
are, in part, dependent on the rating scale of the self- itive loadings for ages of acquisition suggested late
reported or behavioral measure in question.) Regression language development, whereas negative loadings for
analyses in this study were used as exploratory tools proficiency variables suggested incomplete acquisition of
rather than for theory testing. Although some researchers the second language. Together, the clustering of these
have found stepwise multiple regression analyses to be variables suggested that oral comprehension presented
controversial (e.g., Menard, 1995; Tabachnick & Fidell, a special challenge for late learners, and this factor was
2007), these analyses were the most appropriate way of interpreted as a measure of Late L2 Learning.
examining the current data set because of the high num- The third factor included total time exposed to L2;
ber of independent variables (i.e., 16 language history exposure to L2 via TV, friends, radio, family, reading, and
variables). Stepwise multiple regression analyses made it classroom; proficiency and comfort with writing in L2;
possible to zero in on the variables that could account for and a preference for speaking L2 (all positive loadings)
the most variability in self-ratings, which is an especially as well as learning L2 from reading (negative loading;
useful approach when the number of independent vari- Cronbach’s a = .92). Positive loadings for L2 immersion
ables is large (e.g., Mendenhall & Sincich, 1996). The pre- and proficiency variables indicated a common underlying
dictive power of independent variables is marked by R competence factor, whereas a negative loading for learn-
values, where R values of .5 or higher are considered ing L2 from reading suggested that, in these bilinguals,
large, R values between .3 and .5 are considered mod- competence in L2 was related to social L2 immersion.
erate, and R values between .1 and .3 are considered small Together, these patterns were interpreted as an index
(Cohen, 1988). of L2 Competence.
The fourth factor included total time exposed to L1;
exposure to L1 via classroom, TV, radio, reading, and
Results friends (positive loadings); and learning L2 from reading
(negative loading; Cronbach’s a = .80). The presence of an
Sixteen factors with eigenvalues greater than 1 were L2 learning variable localized the underlying phenome-
extracted from the data set by means of factor analysis. non to an L2 context, whereas the positive loadings of L1
Of these factors, the first 8 had eigenvalues greater than 3 exposure variables suggested continued immersion in L1.
and accounted for 76% of all variance (see Table 2). These Therefore, this clustering was interpreted as a measure
8 factors were assigned construct names indicative of of continued L1 exposure—that is, L1 Maintenance.
Proficiency reading .947 Age became fluent .864 Exposure (% time) .923 L1 exposure to classes .916
Comfort understanding .910 Age began acquiring .859 Exposure to TV .908 L1 exposure to TV .914
Proficiency understanding .910 Age became fluent reader .855 Exposure to friends .861 L1 exposure to radio .831
Comfort writing .903 Comfort understanding –.803 Exposure to radio .772 L1 exposure to reading .776
Proficiency writing .896 Age began reading .751 Writing proficiency .660 L2 learning from reading –.727
Comfort reading .884 Proficiency understanding –.697 Exposure to family .621 L1 exposure (% time) .627
Identified accent –.788 Years in a country –.681 Comfort writing .592 L1 exposure to friends .530
Comfort speaking .748 Learning from tapes .601 Preference to speak .590
Proficiency speaking .704 Proficiency speaking –.580 Exposure to reading .564
Cultural identification .526 Exposure to classes .543
Perceived accent –.517 Learning from reading –.519
Preference to read .457
% variance 23.480 13.383 9.625 7.534
Cumulative variance 23.480 36.862 46.488 54.021
Factor 5: Loading Factor 6: Loading Factor 7: Loading Factor 8: Loading
Late L2 Immersion values Media-Based Learning values Non-Native Status values Balanced Immersion values
L1 years of class learning .728 L1 learning from TV .866 L2 perceived accent .839 L1 learning from friends –.813
L2 years in workplace .725 L2 learning from TV .838 L2 identified accent .615 L2 years of schooling .627
L1 years in workplace .714 L1 learning from the radio .741 L2 cultural identification –.602 L1 years in a family .622
Proficiency reading L2 –.687 L2 learning from the radio .652 L1 age became fluent .590 L2 years in a classroom .541
L2 learning from friends –.683 L2 comfort reading –.476 L2 learning from family –.519 L1 years in a country .499
L2 learning in a classroom –.556
L1 years in school .476
% variance 6.424 6.226 5.049 4.232
Cumulative variance 60.445 66.671 71.720 75.952
The fifth factor included years exposed to L1 in a schooled mostly in an L2 country, and it was interpreted
language classroom, workplace, and general school set- to reflect Balanced Immersion.
ting, and years exposed to L2 in the workplace (all posi- The language history measures that predicted pro-
tive loadings), as well as proficiency in reading L2 and ficiency in understanding, speaking, reading, and writing
learning L2 from friends and in a foreign language in L1 and L2 are reported in Table 3. The results of the
classroom (negative loadings; Cronbach’s a = .30; how- stepwise multiple regression analyses include regression
ever, note that Cronbach’s alphas are likely to be under- coefficients (marking the relative importance of each in-
estimated in factors where both positive and negative dependent variable that entered the model) and statistics
loadings are present). The positive loading of L2 work- describing the fit of the model. In addition to standard-
place immersion suggested adult L2 acquisition, whereas ized regression beta, variation inflation factors (VIFs) are
positive loadings of a range of L1 immersion variables included to measure the correlation among independent
(including workplace and school settings) suggested late variables (lack of such correlation is a basic assumption in
immigration from the L1-speaking country. This cluster- regression analysis). It is generally accepted that VIF val-
ing of variables is likely to describe a subset of bilinguals ues greater than 10 signal multicollinearity and singu-
consisting of adult immigrants, who spent their formative larity problems (Mendenhall & Sincich, 1996). In this
years in an L1-speaking environment and who were im- study, VIF values of independent variables ranged from
mersed in L2 later in life. To account for the presence of 1.0 to 1.4, suggesting that no multicollinearity/singularity
both L1 and L2 variables within this cluster, this factor problems were present.
was interpreted to index Late L2 Immersion.
The sixth factor included learning L1 and L2 from
radio and TV (positive loadings), as well as comfort with
reading in L2 (negative loading; Cronbach’s a = .75).
Discussion
Positive loadings for both L1 and L2 suggested that the Participants in Study 1 reported high levels of pro-
construct underlying this factor was not specific to just ficiency and extensive immersion in both languages. Across
one language; instead, it was more likely to capture a modalities, L1 proficiency was higher than L2 proficiency,
general trend of language learning from the media. This with the largest difference in understanding and the
pattern of variable loadings indicated a measure of lan- smallest difference in reading. It is possible that the em-
guage learning within a popular culture framework and phasis on reading in this study is reflective of the specific
was interpreted as a measure of Media-Based Learning. sample of bilinguals in Study 1, who were recruited pri-
The seventh factor included L2 accent as perceived marily from academic communities. In order to succeed in
by the participant and as identified by others, age of at- an L2 academic environment, reading in L2 was likely to
tained L1 fluency (positive loadings), and identification be particularly important. Therefore, bilinguals in this
with L2 culture and learning L2 from family (negative study may have been more likely to judge their overall L2
loadings; Cronbach’s a = .24). Positive loadings for acqui- competence on the basis of L2 reading skills versus other
sition ages and for accent judgments indicated late L2 skills. The emphasis on L2 reading evident in proficiency
acquisition, and negative loadings for identification with ratings was echoed in participants’ reports that reading
L2 culture and learning L2 from family indicated lack of experiences (e.g., learning via reading and exposure to
assimilation into the L2 environment. Age of attained L1 reading) contributed most to L2 competence, whereas
fluency loaded positively, suggesting that this cluster was family-based experiences (e.g., learning language from
specific to a subgroup of bilinguals who experienced a family and exposure to family) contributed most to L1
discontinuity in use of their L1. Together, these patterns competence. These patterns of reading-oriented L2 acqui-
were interpreted to index Non-Native Status. sition accurately reflect typical language learning pat-
The eighth factor included years spent in an L1 fam- terns for sequential bilinguals, where a native language
ily and country, years exposed to L2 in a language class- is acquired within a family environment and a second
room and in a general school setting (all positive loadings), language is often acquired on entrance into schooling and
and learning L1 from friends (negative loading; Cronbach’s takes place in a classroom environment involving explicit
a = .27). Positive loadings from both L1 and L2 variables reading instructions.
suggested that the underlying phenomenon was common Factor analysis yielded component groupings that
to both languages, whereas the negative loading for learn- accounted for most of the variance in bilinguals’ self-
ing L1 from friends suggested relatively early immi- reported data, suggesting that questions on the LEAP-Q
gration from the L1-speaking country. Together, these were broad enough to capture variability in the bilingual
patterns suggested a balanced bilingual profile, where population that was sampled in this study. The clusters
both L1 immersion and L2 immersion were important. that emerged reflected underlying dimensions of bilingual-
Therefore, this factor was probably descriptive of a subset ism, with questions that clustered together measuring the
of bilinguals who were born in an L1 country but were same construct. The first four factors (L1 Competence,
Late L2 Learning, L2 Competence, and L1 Maintenance) The structure of the L1 Competence and L2 Compe-
were general in nature, provided information shared by tence factors suggests that the constructs of proficiency
all bilinguals, and accounted for more than half of the in the first and second language share subcomponents.
variance in the data. The remaining factors (Late L2 However, some differences in factorial structures were
Immersion, Media-Based Learning, Non-Native Status, also observed. For instance, unlike the L1 Competence
and Balanced Immersion) grew increasingly specific and factor, the L2 Competence factor included a large number
appeared to be driven by bilingual subgroups. In addition, of age-of-acquisition variables. The significant contribu-
the measures of variance in the data set accounted for tion of maturation variables to competence in L2 is con-
by each factor (indicated by eigenvalues) were augmented sistent with previous studies that have identified L2
by measures of consistency within each observed factor acquisition age as an important determiner of L2 com-
(indicated by Cronbach’s alphas). The four factors with petence (e.g., Flege et al., 1999, 2002; Jia et al., 2002;
highest Cronbach’s alphas were general in nature: L1 Johnson & Newport, 1989; Vaid & Menon, 2000). More-
Competence, L2 Competence, L1 Maintenance, and over, the structure of the L2 Competence factor con-
Media-Based Learning. Conversely, the three factors firms Hyltenstam and Abrahamsson’s (2003) model of
with lower Cronbach’s alphas—Late L2 Immersion, Non- L2 acquisition, which postulates interplay between age
Native Status, and Balanced Immersion—might have of acquisition and environmental variables in shaping
emerged from characteristics of specific subgroups in the L2 attainment.
sample tested. Together, the eigenvalues and Cronbach’s Although some factors may include variables that
alphas reveal that L1 Competence, L2 Competence, and appear similar, they represent constructs that differ in
L1 Maintenance accounted for much of variance in the nuanced ways. For example, L2 learning in a classroom
data and were highly consistent internally. ( Late L2 loads negatively onto Factor 5 (Late L2 Immersion), and
Learning accounted for approximately 13% of the vari- years in an L2 classroom loads positively onto Factor 8
ance but did not lend itself to statistical evaluation of in- (Balanced Immersion). The two questions may contribute
ternal consistency because of negative covariance across to overall language proficiency in distinct ways. Classroom
age and ability components. Finally, Media-Based Learn- experience as a contributor to learning L2 is a subjective,
ing had high internal consistency but accounted for rela- metacognitive self-assessment measure, whereas dura-
tively little variance in the data, about 6%.) tion of L2 classroom exposure is an objective temporal
950 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
measure. Contribution to learning may reflect learning proficiency was excluded because of its close relationship
style, whereas years of exposure measures duration of to reading proficiency—specifically, proficiency in reading
experience (amount of time spent in a classroom is not L1 predicted proficiency in writing L1, R = .9, F(1, 50) =
always indicative of language proficiency). This distinc- 365, p < .001, and proficiency in reading L2 predicted
tion is reflected in differential loadings of these two vari- proficiency in writing L2, R = .8, F(1, 50) = 108, p < .001.
ables onto Factors 5 and 8. The high predictability of writing proficiency from
Multiple regression analyses were used to generate measures of reading proficiency suggested that a separate
predictive equations for self-reported proficiency levels in question assessing writing proficiency would not pro-
L1 and L2. The results suggested that different experien- vide additional new information. Third, current classroom
tial variables predicted proficiency in the two languages. exposure was excluded because of its close relationship to
For instance, L1 proficiency levels across modalities were current reading exposure, with reading exposure in L1
consistently predicted by years spent in an L1 country, predicting ratings of classroom exposure in L1, R = .7,
whereas L2 proficiency levels were predicted by L2 F(1, 50) = 34, p < .001, and reading exposure in L2 predict-
acquisition ages. In contrast to L1 regression models, ing ratings of classroom exposure in L2, R = .6, F(1, 50) =
years of immersion (in this case, length of time spent in an 33, p < .001. Because classroom exposure was generally
L2 workplace) predicted only proficiency in writing L2, not predictive of self-reported proficiency ratings, and
but not proficiency in speaking, understanding, or read- because the extent of classroom exposure could be reli-
ing L2. Moreover, although similar experiential variables ably deduced on the basis of reading exposure, it was
predicted proficiency levels in L1 across understanding, omitted from the final version of the LEAP-Q. Fourth,
reading, speaking, and writing, predictors of proficiency percentage of bilingual contacts was excluded because it
levels in L2 were more varied. For instance, L1 pro- did not correlate with any other measures and did not load
ficiency levels were consistently predicted by time spent onto any factor in the factor analysis. Because the aim of
in an L1 country. Conversely, predictors of L2 proficiency this study was to construct an internally consistent self-
differed: For example, proficiency in understanding was assessment measure, a question that was not related to
predicted primarily by age when participants began read- any other questions in the questionnaire was deemed
ing L2, and proficiency in speaking was predicted unreliable and uninformative. As a result of these changes,
primarily by exposure to L2 in family and classroom seven questions per language were omitted, shortening
environments. The greater variability and larger number questionnaire completion time without losing predictive
of language history predictors for L2 is likely due to value. In addition, because of limited variability (1–5) in
different acquisition patterns for the two languages, with questions that required responses on a scale, the ranges of
L2 acquisition more varied across settings and modalities values on all scales were increased (1–10). Finally, the
relative to L1. Greater variability in L2 acquisition pat- questionnaire was transferred into digital format. Pull-
terns is probably due to highly diverse experiences asso- down menus were added to questions that required
ciated with learning a second language (relative to a responses on scales, thus facilitating questionnaire com-
native language, the acquisition of which is less variable). pletion and subsequent data extraction for analyses. In
For example, whereas some participants learned L2 in an the digitized version, once the participant indicated a spe-
immersion-type setting ( because of immigration, or cific language, the language was filled in automatically
studying and working abroad), others learned L2 in a throughout the questionnaire. The revised version of the
classroom environment. questionnaire was used in Study 2 and required approx-
imately 15 min for bilinguals to complete.
The results of Study 1 prompted exclusion of four
measures from the questionnaire. First, measures of com-
fort across modalities (e.g., “How comfortable are you
speaking, understanding, reading, and writing in a lan-
guage?”) were excluded. These questions yielded val-
Study 2: Establishing
ues that were similar to those yielded by proficiency Criterion-Based Validity
measures—that is, the two patterned identically, corre-
lated significantly, and were highly predictive of each The objectives of Study 2 were to confirm the inter-
other. Specifically, proficiency measures were predictive nal validity of the LEAP-Q in a more homogeneous sam-
of comfort measures for speaking (L1: R = .9, F[1, 50] = ple of bilinguals through factor analysis as well as to
167, p < .001; L2: R = .9, F[1, 50] = 264, p < .001), establish criterion-referenced validity by comparing self-
understanding (L1: R = .95, F[1, 50] = 468, p < .001; L2: reported and standardized proficiency measures using
R = .9, F[1, 50] = 332, p < .001), reading (L1: R = .97, correlation and regression analyses. We made the
F[1, 50] = 881, p < .001; L2: R = .9, F[1, 50] = 181, p < .001), following five predictions:
and writing (L1: R = .99, F[1, 50] = 1760, p < .001; 1. Bilinguals’ answers to questions that referred to the
L2: R = .96, F[1, 50] = 561, p < .001). Second, writing same underlying aspects of bilingual profiles would
L1 history L2 history
Self-reported proficiencya
Understanding 9.58 0.93 6.00–10.00 7.92 2.33 2.00 –10.00
Speaking 9.32 1.15 6.00–10.00 7.74 2.05 2.00 –10.00
Reading 9.26 1.26 5.00–10.00 8.02 1.97 2.00 –10.00
Age milestones (years)
Started learning 1.08 1.75 0.00–11.00 8.25 5.95 0.00–23.00
Attained fluency 4.00 2.81 0.00–13.00 15.10 7.44 2.00–39.00
Started reading 5.5 2.24 3.00–14.00 11.52 5.44 3.00–25.00
Became fluent reading 7.92 3.27 4.00–41.00 15.09 7.44 2.00–39.00
a
Range: 0 (none) to 10 ( perfect). b Range: 0 (not a contributor ) to 10 (most important contributor ). c Range: 0 (never) to 10 (always).
d
Range: 0 (none) to 10 ( pervasive).
952 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
Table 5. Standardized proficiency measures for Study 2.
Performance Performance
Measure M SD range M SD range t tests and effect sizes ( partial h 2 )
Note. PPVT = Peabody Picture Vocabulary Test; TVIP = Test de Vocabulario en Imágenes Peabody.
representing both simultaneous and sequential bilinguals. Johnson & Newport, 1989). All measures were adminis-
Participants varied in their education levels from high tered in language blocks, with half the participants re-
school to graduate school (M = 16 years of education, ceiving the Spanish measures first and half receiving the
SE = 2.5, range = 11–22 years; note that in some countries, English measures first. Test administrators were highly
a high school education is equivalent to fewer than proficient in both languages. Specifically, the following
12 years). Participants reported being exposed to L1 most seven behavioral measures were administered:
in the context of family, followed by friends, radio, read-
1. A reading fluency test (Subtest 2 of the Woodcock–
ing, TV, and self-instruction. They reported being exposed
Johnson Test of Achievement in English and its equiv-
to L2 most in the context of reading, followed by friends,
alent Woodcock–Muñoz version in Spanish). This test
radio, TV, family, and self-instruction. When asked to re-
required participants to read as many sentences as
port how different factors contributed to language learn-
possible within a 3-min interval and to decide whether
ing, participants reported that learning L1 relied most on
each sentence was true or false.
family, followed by reading, friends, TV, radio, and self-
instruction, and they reported that learning L2 relied 2. A passage comprehension test (Subtest 9 of the
most on reading, followed by friends, TV, radio, family, Woodcock–Johnson Test of Achievement in English
and self-instruction. Participants performed better in L1 and its equivalent Woodcock–Muñoz version in
than in L2 across a range of behavioral measures (see Spanish). This test required participants to read
Table 5 for means and statistical comparisons). passages and supply missing words.
Materials and procedure. The revised version of the 3. A productive picture vocabulary test (Subtest 14 of the
LEAP-Q was administered to all participants at the start Woodcock–Johnson Test of Achievement in English
of the experimental session on a computer. Participants and its equivalent Woodcock–Muñoz version in Span-
independently completed the LEAP-Q in English. On ish). This test required participants to name pictures.
completion, participants were administered a battery of
4. An oral comprehension test (Subtest 15 of the
standardized behavioral measures of language ability.
Woodcock–Johnson Test of Achievement in English
These included subtests from the Woodcock–Johnson
and its equivalent Woodcock–Muñoz version in
Tests of Achievement (Woodcock, McGrew, & Mather,
Spanish). This test required participants to listen to
2001), the Woodcock–Muñoz Tests of Achievement
passages and supply missing words.
(Muñoz-Sandoval, Woodcock, McGrew, & Mather, 2005),
and the Peabody Picture Vocabulary Test in English 5. A sound awareness test (Subtest 21 of the Woodcock–
(PPVT, Dunn & Dunn, 1997) and Spanish (Test de Johnson Test of Achievement in English and its
Vocabulario en Imágenes Peabody [TVIP]; Dunn, Padilla, equivalent Woodcock–Muñoz version in Spanish).
Lugo, & Dunn, 1986). In addition, sentence grammati- This test required participants to complete a rhym-
cality judgment tasks were constructed on the basis of ing task, a sound deletion task, a sound substitution
previous materials (Bedoya et al., 2005; DeKeyser, 2000; task, and a sound reversal task.
954 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
Table 6. Factors yielded in Study 2.
L1 exposure to reading –.872 Speaking proficiency –.929 L2 learning from the radio .755 L1 age began acquiring .871
L2 exposure to TV .866 Age began reading .824 L2 age when fluent reader .724 L1 learned from family –.823
L2 exposure to reading .844 Reading proficiency –.747 L2 age when fluent .654 L1 age when fluent .624
L1 exposure to TV –.811 Age when fluent reader .745 L2 age began reading .605 L2 preference to speak .577
L2 exposure to friends .804 Comprehension proficiency –.709 L2 exposure to self-instruction .561 L1 identified accent .526
L2 exposure to radio .777 Learning from tapes .629 L2 learning from tapes .557 L1 exposure to family –.468
L1 exposure to radio –.773 L1 perceived accent –.548
L1 exposure (% time) –.771 L2 perceived accent .452
L1 exposure to friends –.762
L2 exposure (% time) .759
L2 reading proficiency .754
L1 preference to read –.713
L2 preference to read .696
L2 speaking proficiency .580
L1 learning from reading –.522
L2 identified accent –.496
L2 years in the country .474
% variance 25.296 12.425 9.615 7.471
Cumulative variance 25.296 37.722 47.337 54.807
Factor 5: Loading Factor 6: Loading Factor 7: Loading Factor 8: Loading
Marian et al.: Bilingual LEAP Questionnaire
Exposure to family .898 L1 years of schooling .885 L2 age when began acquiring 0.854 L1 Learning from radio .845
Years in family .894 L1 years in family .818 L2 cultural identification –0.713 L1 Learning from TV .816
Learning from family .747 Chronological age .766
Years of schooling .537 L1 years in country .740
L2 learning from TV –.503
% variance 5.933 4.860 4.095 3.835
Cumulative variance 60.740 65.600 69.695 73.530
955
learned L1 at home but lived and received schooling from country. Therefore, this factor was interpreted as index-
an early age in an L2 environment and made an active ing L2 Nonacculturation.
effort to maintain L1. These patterns were interpreted as The eighth factor included learning L1 from radio
reflecting L1 Learning. and TV (positive loadings, Cronbach’s a = .77) and was
The third factor included learning L2 from radio interpreted to reflect Media-Based L1 Learning.
and language tapes, exposure to L2 through independent Establishing predictive relationships using multiple
study, ages of becoming a fluent L2 speaker and reader, regressions. Regression analyses are reported in Table 7,
and self-perceived accent in L2 (all positive loadings), as together with regression coefficients (marking the rela-
well as self-perceived accent in L1 (negative loading; tive importance of each independent variable that
Cronbach’s a = .77). Positive loadings of acquisition age entered the model) and statistics describing the fit of
and self-instruction variables suggest late and incom- the model (B and b coefficients, VIF values, R and R2
plete acquisition of the second language, whereas positive values, and F tests). VIFs associated with each indepen-
loadings of L2 accent and negative loadings of L1 accent dent variable ranged from 1.00 to 2.46, suggesting that no
suggest higher L1 fluency than L2 fluency. Together, these multicollinearity or singularity problems were present.
patterns were interpreted as indexing Late L2 Learning. The language history measures that predicted
The fourth factor included ages of L1 acquisition, ages self-reported proficiency in understanding, speaking,
of attained L1 fluency, L1 accent as identified by others, and reading are reported in the top panel of Table 7, self-
and preference to speak L2 (all positive loadings), as well reported proficiency measures that predicted behavioral
as exposure to and learning from an L1-speaking fam- performance are reported in the second panel of Table 7,
ily (negative loadings), suggesting limited L1 exposure. language history measures that predicted behavioral
Cronbach’s alpha could not be calculated for this factor proficiency are reported in the third panel of Table 7,
because of the negative average covariance among items, and behavioral measures that predicted self-reported
a violation of reliability model assumptions. Positive load- proficiency are reported in the bottom panel of Table 7.
ings of acquisition age and accent variables indicate lack Correlations between behavioral and self-reported
of fluency in L1, whereas negative loadings of family ex- measures. To establish criterion-based validity of self-
posure in L1 suggest lack of immersion in an L1 environ- reported proficiency measures, Pearson’s R correlation
ment. These variables are likely to describe a subset of analyses between self-reported and behavioral proficien-
bilinguals for whom L1 is no longer a dominant language cy measures were conducted within each language and
and who prefer to use L2 in daily life. Therefore, this fac- processing modality (see Table 8). The results yielded
tor was interpreted as indexing L1 Nondominant Status. strong positive correlations between standardized behav-
The fifth factor included L2 family-based components ioral measures (i.e., reading fluency, passage comprehen-
(e.g., current exposure to, years spent in, and learning sion, productive vocabulary, oral comprehension, and
from an L2-speaking family) as well as years of L2 grammaticality judgments) and self-reported measures
schooling (all positive loadings; Cronbach’s a = .86). This of understanding, speaking, and reading L1 and L2. Al-
factor was interpreted to suggest an overall measure of though performance on standardized measures of sound
interactive L2 Immersion. awareness and receptive vocabulary did not relate to self-
The sixth factor included years spent in an L1- reported L1 proficiency, it was significantly related to
speaking school, family, and country; participant’s chro- self-reported L2 proficiency. The majority of standardized
nological age (all positive loadings); and learning L2 from measures correlated more strongly with self-reported L2
TV (negative loading: Cronbach’s a = .50). The negative proficiency than with self-reported L1 proficiency (with
loading of L2 learning variables may be indicative of a the exception of grammaticality judgment latencies,
subset of bilinguals who maintained their first language which correlated stronger with self-reported L1 proficiency).
and had minimal exposure to their second language. This For L1, self-reported proficiency measures correlated most
factor was interpreted as an overall measure of interac- strongly with standardized behavioral measures of oral
tive L1 Immersion. comprehension. For L2, the highest correlation val-
ues were obtained for passage comprehension and oral
The seventh factor included age of L2 acquisition
comprehension.
(positive loading) as well as identification with L2 culture
(negative loading). Cronbach’s alpha could not be calcu-
lated for this factor because of the negative average
covariance among items, a violation of reliability model
Discussion
assumptions. The positive loading of L2 acquisition age The results of Study 2 confirmed questionnaire-
onto this factor indicated a late L2 learning profile, based predictors of self-reported proficiency, identified
whereas the negative loading of L2 cultural identification questionnaire-based predictors of behavioral language
suggested lack of acculturation within the L2-speaking performance, and revealed behavioral predictors of
956 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
Table 7. Multiple regression analyses for Study 2: Language history predictors of self -reported proficiency, self-reported proficiency predictors of
behavioral performance, language history predictors of behavioral performance, and behavioral predictors of self-reported proficiency.
F (1, 50) = 10.7, p < .001 Years in L1 schools 1.91 0.51 .58 2.07 .62 .38
Years in L1 family –1.26 0.52 –.36 1.94 .66 .43
Learning L1 from tapes –3.84 1.39 –.31 1.10 .71 .50
L1 Oral Comprehension ( WJ/M) Years in an L1 country 0.67 0.28 .34 1.20 .46 .21
F (1, 50) = 8.9, p < .001 Exposure to L1 reading 1.90 0.87 .30 1.20 .53 .28
L1 Grammaticality Accuracy Exposure to L1 reading 0.02 0.01 .34 2.46 .75 .56
F (1, 50) = 30.5, p < .001 Learning L1 from reading 0.02 0.01 .28 1.42 .79 .63
L1 general exposure 0.001 0.001 .35 2.23 .83 .68
L1 Grammaticality Latency Exposure to L1 friends –134.3 37.04 –.46 1.25 .60 .63
F (1, 50) = 17.8, p < .001 Learning L1 from reading –120.4 46.29 –.33 1.25 .67 .45
L1 Receptive Vocabulary Learning L1 from TV –2.02 0.76 –.35 1.00 .36 .13
F (1, 50) = 5.8, p < .01 Learning L1 from tapes –1.92 0.94 –.27 1.00 .45 .21
L1 Sound Awareness Years in L1 schools 0.75 0.35 .33 1.00 .33 .11
F (1, 50) = 4.6, p < .05
L2 Reading Fluency (WJ/M) General L2 exposure 0.57 0.08 .76 1.04 .71 .51
F (2, 50) = 23.7, p < .001 L2 learning from radio –2.06 0.90 –.26 1.04 .75 .57
L2 Passage Comprehension ( WJ/M) Exposure to L2 friends 4.75 1.04 .56 1.17 .68 .46
F (1, 50) = 20.8, p < .001 Years in an L2 family 0.83 0.34 .30 1.17 .73 .54
L2 Productive Vocabulary ( WJ/M) Exposure to L2 friends 3.39 0.70 .57 1.14 .68 .47
F (2, 50) = 22.9, p < .001 Age attained L2 fluency –1.03 0.37 –.33 1.14 .75 .56
L2 Oral Comprehension ( WJ/M) General L2 exposure 0.38 0.12 .45 1.29 .63 .40
F (1, 50) = 18.1, p < .001 Years in an L2 family 0.10 0.35 .39 1.29 .72 .52
L2 Grammaticality Accuracy Exposure to L2 TV 0.02 0.004 .51 1.77 .82 .67
General L2 exposure 0.002 0.000 .37 1.77 .88 .77
F (1, 50) = 50.4, p < .001 Age attained L2 fluency 0.01 0.002 –.23 1.13 .90 .82
L2 Grammaticality Latency Age attained L2 fluency 81.47 30.93 .38 1.23 .54 .29
F (2, 50) = 11.49, p < .001 L2 exposure (% time) –15.23 6.17 –.36 1.23 .63 .40
L2 Receptive Vocabulary Age started reading L2 –1.79 0.69 –.37 1.00 .39 .15
F (2, 50) = 6.7, p < .01 Learning L2 via friends 2.63 1.08 .35 1.00 .52 .27
L2 Sound Awareness Years in an L2 country 1.08 0.22 .64 1.04 .58 .34
F (3, 50) = 13.15, p < .001 Learning L2 from radio –3.31 0.89 –.47 1.03 .74 .54
L2 self -instruction –2.32 1.10 –.27 1.01 .78 .61
Behavioral predictors of
self -reported proficiency
Comprehending L1 Grammaticality latency 0.00 0.00 –.48 1.00 .48 .23
F (1, 50) = 11.5, p < .01
Proficiency Speaking L1 Oral comprehension 0.04 0.01 .68 1.25 .54 .30
F (2, 50) = 11.10, p < .001 Sound awareness –0.02 0.01 –.31 1.25 .61 0.37
Reading L1 Oral comprehension 0.03 0.01 .44 1.59 .64 .40
F (2, 50) = 16.18, p < .001 Grammaticality latency 0.00 0.00 –.32 1.59 .68 .47
Comprehending L2 Passage comprehension 0.04 0.01 .60 1.24 .72 .52
F (2, 50) = 24.2, p < .001 Grammaticality latency 0.00 0.00 –.28 1.24 .76 .58
Speaking L2 Passage comprehension 0.03 0.01 .42 2.58 .78 .61
F (3, 50) = 26.22, p < .001 Grammaticality accuracy 3.74 1.47 .31 1.66 .81 .66
Receptive vocabulary 0.02 0.01 .26 1.78 .84 .70
Reading L2 Passage comprehension 0.03 0.01 .45 1.64 .67 .45
F (1, 50) = 29.69, p < .001 Grammaticality accuracy 3.73 1.57 .35 1.64 .73 .53
958 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
Table 8. Self -reported and behavioral proficiency correlations in Study 2.
Self -reported Reading Passage Productive Oral Sound Receptive Grammaticality Grammaticality
measures fluency comprehension vocabulary comprehension awareness vocabulary judgments (Acc) judgment ( RT )
L1 Proficiency
Speaking .409** .441** .448** .541** 0.008 0.236 .377** –.434**
Comprehension .346* .428** .413** .481** – 0.025 .307* .337* –.432**
Reading .417** .520** .495** .661** 0.190 0.179 .517** –.555**
L2 Proficiency
Speaking .642** .741** .640** .739** .499** .640** .667** –.339*
Comprehension .542** .562** .535** .621** .466** .494** .502** –.377**
Reading .532** .634** .586** .575** .529** .574** .603** –.286*
self-reported language proficiency. Similar to partici- more than half of the variance in the data. The remaining
pants in Study 1, participants in Study 2 reported high factors (L2 Immersion, L2 Nonacculturation, and Media-
levels of proficiency and extensive immersion in both Based L1 Learning) accounted for less variance and may
languages. Participants in Study 2 ranked exposure to an be characteristic of bilingual subgroups. Unlike L1 and
L1-speaking family as the most significant contributor to L2 Competence in Study 1, L1 and L2 Competence in
learning L1 and ranked reading in L2 as the most sig- Study 2 did not cluster as separate factors. However,
nificant contributor to learning L2. Consistent with these within the same factor, L1 and L2 Competence variables
rankings, participants reported that their highest L1 generated distinct patterns, with positive loadings for L2
proficiency was in understanding, whereas their highest variables and negative loadings for L1 variables. These
L2 proficiency was in reading. Participants’ reports for L2 findings replicate the central nature of L1 and L2 com-
proficiency were lower than their reports for L1 profi- petencies in defining bilingual profiles and suggest that
ciency, with the largest differences in understanding and the two competencies may be interconnected depending on
the smallest differences in reading. Differences between sample characteristics. Moreover, high Cronbach’s alphas
self-reported proficiency levels in L1 and L2 were re- suggested that the majority of factors had high internal
flected in participants’ performance on behavioral measures consistency. The four factors with highest Cronbach’s
of linguistic ability, with significantly lower performance alphas were Relative L2–L1 Competence, Late L2
on standardized proficiency measures in L2 than in L1. Learning, L2 Immersion, and Media-Based L1 Learning.
For L2, participants reported high levels of reading pro- Cronbach’s alpha values were somewhat lower for L1
ficiency while at the same time showing low performance Immersion and L1 Learning. Thus, variables describing
on reading fluency and sound awareness tasks (sound Relative L2–L1 Competence and Late L2 Learning had
awareness reflects phonological skills underlying read- high eigenvalues and high alpha scores, suggesting that
ing). The discrepancy between self-reported and behav- they accounted for extensive variance and had high in-
ioral measures of reading suggests that bilinguals may ternal consistency. L2 Immersion and Media-Based L1
overestimate their true L2 reading abilities. This ten- Learning had high internal consistency but accounted for
dency to overestimate L2 reading skill may be due in part relatively little variance in the data set.
to lack of direct feedback on performance during reading, Multiple regression analyses were used to generate
which by its nature is less interactive than speaking or predictive relationships between language history vari-
listening. The finding that bilinguals overestimate their ables and self-reported proficiency for L1 and L2. L1
skill in a weaker language is also consistent with learning proficiency was predicted by age when participants began
research reporting that students overestimate their aca- reading in L1 and by exposure to L1 reading. L2
demic abilities when their performance is poor (Orsmond, proficiency was predicted by exposure to L2-speaking
Merry, & Reiling, 1997). friends and learning L2 from family. The variables with
In Study 2, questions that were expected to measure the highest predictive values for L1 proficiency were the
the same underlying construct (i.e., language proficiency, ones that participants reported as most relevant to
etc.) clustered together in the factor analysis. Four fac- learning L2, and the variables with highest predictive
tors (Relative L2–L1 Competence, L1 Learning, Late L2 values for L2 proficiency were the ones that participants
Learning, and L1 Nondominant Status) accounted for reported as most relevant to learning L1. In other words,
960 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
linguistic abilities. Although it is yet premature to studies) but also population-specific characteristics of
suggest that questions within these factors might be bilingual subsamples (i.e., factors that are idiosyncratic
combined into subscales for assessing L1 and L2 compe- to each study).
tence in diverse bilingual populations, the patterns are In examining the external validity of the LEAP-Q,
promising. For example, studies that target bilinguals’ strong correlations between self-reported proficiency and
L1 proficiency may include questions that contributed to standardized tests of linguistic performance were found,
L1 status factors (e.g., L1 Maintenance and Balanced and preliminary predictive relationships between self-
Immersion in Study 1, and Relative L1–L2 Competence reported and standardized measures were established.
and L1 Immersion in Study 2). These questions, which The fit of the models, which refers to variance in the
were also most predictive of behavioral L1 measures in dependent variable accounted for by variability in the
regression analyses, address years in an L1 country, L1 independent variable(s), averaged 62% for L2 and 50% for
school, and L1 family; general L1 exposure; and exposure L1 and was comparable to values reported in previous
to L1 reading. Similarly, studies that target L2 immer- research (e.g., MacIntyre et al., 1997). Self-reported pro-
sion may include questions that contributed to L2 im- ficiency in reading L1 was a reliable predictor of behav-
mersion factors, such as Late L2 Immersion in Study 1 ioral performance on standardized tests of language
and L2 Immersion in Study 2. These questions address ability in L1. Self-reported proficiency in speaking L2 was
years in an L2 workplace, L2 school, and L2 family; a reliable predictor of behavioral performance on stan-
exposure to an L2 family; and learning from an L2 dardized tests of language ability in L2. We suggest that
family. if questionnaire data are used to make general inferences
Although subscales might be a useful way to assess about language function, then reading proficiency should
specific subsamples of bilinguals in the future, developing be used to index L1, and speaking proficiency should be
such subscales was not the objective of this study. In- used to index L2. Because both simultaneous and se-
stead, the goal was to create a comprehensive question- quential bilinguals were sampled in Study 2, it is likely
naire that is applicable across a diverse group of bilingual that this observation will apply to various groups of
and multilingual populations. Subscale construction as- bilingual individuals, with diverse acquisition histories.
sumes constant relationships between questionnaire items; However, it is possible that this pattern of results, with
however, Harley et al. (1990) demonstrated that the fac- speaking proficiency more indicative of L2 skills and with
torial structure of L2 proficiency varied across bilingual reading proficiency more indicative of L1 skills, is more
groups with different language experiences. Therefore, applicable to sequential bilinguals (who acquired their
data collected in this study illustrate meaningful rela- first language early in life and their second language later
tionships between questions (hence validating them) but in life). Future research may explicitly test the predictive
did not provide evidence for specific subscales. Future power of self-assessed speaking and reading skills across
studies may focus on specific subsamples of bilinguals various bilingual groups and directly compare sequential
and administer questions that clustered together in L1 and simultaneous bilinguals.
and L2 Competence factors to examine their utility as Language history measures predicted self-reported
scales. For instance, if variables that constituted L2 Com- and behavioral proficiency better for L2 than for L1. For
petence factors in Studies 1 and 2 (e.g., length of exposure instance, correlation values ranged from .5 to .68 for L1
to L2, current exposure to L2 in different environments, and from .73 to .84 for L2 in analyses of self-reported
proficiency speaking L2, etc.) formed a reliable subscale, proficiency, and from .45 to .83 for L1 and from .52 to .90
then answers to these questions should reliably predict for L2 in analyses of behavioral performance. In Study 1,
the level of L2 performance across a range of bilingual self-reported L1 proficiency was predicted by immersion
speakers. in L1 environments, and self-reported L2 proficiency was
A number of factors varied across studies; these predicted by L2 acquisition ages. In Study 2, self-reported
included Non-Native Status, L1 Maintenance, and Bal- L1 proficiency was predicted by age when participants
anced Immersion in Study 1 and L1 Nondominant began reading L1 and by exposure to L1 reading, whereas
Status, L1 Immersion, L1 Learning, and L2 Nonaccul- self-reported L2 proficiency was predicted by exposure to
turation in Study 2. It appears that whereas L1 learning, L2 friends and learning L2 from family. Despite the fact
immersion, and nondominant status clustered into three that the specific predictors of language proficiency varied
separate factors in Study 2, they were part of the across studies, a common pattern emerged; namely, in
combined L1 Maintenance factor in Study 1, which likely both studies, self-reported L1 proficiency was predicted
is due to differences in participants’ characteristics. The by only a few language history attributes (e.g., age when
finding that some factors were specific to either Study 1 or participants began reading was a sole predictor for L1
to Study 2 suggests that the LEAP-Q is versatile enough proficiency understanding), whereas L2 proficiency levels
to reflect not only the general characteristics of bilingual were predicted by a more mixed set of predictors (e.g.,
experience (i.e., factors that converge across the two exposure to L2-speaking friends, learning L2 from family,
962 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
Bachman, L. F. (1990). Constructing measures and measur- DeVellis, R. F. (1991). Scale development: Theory and appli-
ing constructs. In B. Harley, P. Allen, J. Cummins, & M. cations. Applied Social Research Methods Series, 26.
Swain ( Eds.), The development of second language profi- London: Sage.
ciency ( pp. 26–49). Cambridge, England: Cambridge
Doctor, E., & Klein, D. (1992). Phonological processing in
University Press.
bilingual word recognition. In R. Harris ( Ed.), Cognitive
Bachman, L., & Palmar, A. (1985). Basic concerns in processing in bilinguals ( pp. 237–252). Amsterdam: Elsevier.
language test validation. Reading, MA: Addison Wesley.
Dörnyei, Z. (2003). Questionnaires in second language re-
Bahrick, H. P., Hall, L. K., Goggin, J. P., Bahrick, L. E., & search. Mahwah, NJ: Erlbaum.
Berger, S. A. (1994). Fifty years of language maintenance
and language dominance in bilingual Hispanic immigrants. Dunn, L. M., & Dunn, L. M. (1997). Peabody Picture Vocab-
Journal of Experimental Psychology: General, 123, 264–283. ulary Test (3rd ed.). Circle Pines, MN: AGS.
Bedoya, E., Braun, G., Craig, L., Rodo, C., Senerth, D., & Dunn, L. M., Padilla, E. R., Lugo, D. E., & Dunn, L. M.
Hughes, E. R. (2005). The best test preparation for the AP (1986). Test de Vocabulario en Imágenes Peabody (Adapta-
Spanish language exam (5th ed.). Piscataway, NJ: Researchers ción Hispanoamericana) [Peabody Picture Vocabulary Test
& Education Association. Hispanic American version]. Circle Pines, MN: AGS.
Bialystok, E., & Hakuta, K. (1994). In other words: The Favreau, M., & Segalowitz, N. (1982). Second language
science and psychology of second language acquisition. reading in fluent bilinguals. Applied Psycholinguistics, 3,
New York: HarperCollins. 329–341.
Bialystok, E., & Hakuta, K. (1999). Confounded age: Lin- Ferrer, M., Cordoba, J., Garin, O., Olive, G., Flavia, M.,
guistic and cognitive factors in age differences for second Vargas, V., et al. (2006). Validity of the Spanish version
language acquisition. In D. Birdsong ( Ed.), Second language of the Chronic Liver Disease Questionnaire as standard
acquisition and the critical period hypothesis ( pp. 161–181). outcome for quality of life assessment. Liver Transplanta-
Mahwah, NJ: Erlbaum. tion, 12, 95–104.
Birdsong, D. (2005). Interpreting age effects in 2nd language Fishman, J., & Cooper, R. (1969). Alternative measures
acquisition. In J. Kroll & A. de Groot ( Eds.), Handbook of of bilingualism. Journal of Verbal Learning and Verbal
bilingualism ( pp. 109–127). New York: Oxford University Behavior, 8, 276–282.
Press.
Flege, J., Frieda, A., & Nozawa, T. (1997). Amount of native-
Carroll, J. (1967). The foreign language attainments of language (L1) use affects the pronunciation of an L2.
language majors in the senior year: A survey conducted in Journal of Phonetics, 25, 169–186.
U.S. colleges and universities. Cambridge, MA: Harvard
University Press. Flege, J. E., Frieda, E. M., Walley, A. C., & Randazza, L. A.
(1998). Lexical factors and segmental accuracy in second
Chapnik-Smith, M. (1997). How do bilinguals access lexical language speech production. Studies in Second Language
information? In A. de Groot & J. F. Kroll ( Eds.), Tutorials in Acquisition, 20, 155–187.
bilingualism ( pp. 145–168). Mahwah, NJ: Erlbaum.
Flege, J. E., MacKay, I. A., & Piske, T. (2002). Assess-
Chen, H. C. (1992). Lexical processing in bilingual or ing bilingual dominance. Applied Psycholinguistics, 23,
multilingual speakers. In R. J. Harris ( Ed.), Cognitive 567–598.
processing in bilinguals ( pp. 253–264). Amsterdam:
Elsevier. Flege, J. E., Yeni-Komishian, G. H., & Liu, S. (1999). Age
constraints on second-language acquisition. Journal of
Chincotta, D., & Underwood, G. (1998). Non temporal
Memory and Language, 41, 78–104.
determinants of bilingual memory capacity: The role of
long-term representations and fluency. Bilingualism, 1, Genesee, F. (1985). Second language learning through
117–130. immersion: A review of U.S. programs. Review of Educa-
Cohen, J. (1988). Statistical power analysis for the behavioral tional Research, 55, 541–561.
sciences. Mahwah, NJ: Erlbaum. Goggin, J. P., Estrada, P., & Villarreal, R. P. (1994).
Cromdal, J. (1999). Childhood bilingualism and metalin- Picture-naming agreement in monolinguals and bilinguals.
guistic skills: Analysis and control in young Swedish– Applied Psycholinguistics, 15, 177–193.
English bilinguals. Applied Psycholinguistics, 20, 1–20. Grainger, J. (1993). Visual word recognition in bilinguals.
de Houwer, A. (1995). Bilingual language acquisition. In P. In R. Schreuder & B. Weltens ( Eds.), The bilingual lexicon
Fletcher & B. MacWinney ( Eds.), The handbook of child ( pp. 11–26). Amsterdam: John Benjamins.
language ( pp. 219–250). London: Basil Blackwell. Grosjean, F. (1982). Life with two languages. Cambridge, MA:
DeKeyser, R. M. (2000). The robustness of critical period Harvard University Press.
effects in second language acquisition. Studies in Second Grosjean, F. (2004). Studying bilinguals: Methodological and
Language Acquisition, 22, 499–533. conceptual issues. In T. Bhatia & W. Ritchie ( Eds.), The
Delgado, P., Guerrero, G., Goggin, J. P., & Ellis, B. B. handbook of bilingualism ( pp. 32–63). Malden, MA:
(1999). Self-assessment of linguistic skills by bilingual Blackwell.
Hispanics. Hispanic Journal of Behavioral Sciences, 21, Harley, B., Cummins, J., Swain, M., & Allen, P. (1990).
31–46. The nature of language proficiency. In B. Harley, P. Allen,
Demorest, M. E., & Erdman, S. A. (1987). Development J. Cummins, & M. Swain ( Eds.), The development of second
of the Communication Profile for the Hearing Impaired. language proficiency ( pp. 7–25). Cambridge, England:
Journal of Speech and Hearing Disorders, 52, 129–143. Cambridge University Press.
964 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
Shameem, N. (1998). Validating self-reported language Vaid, J., & Menon, R. (2000). Correlates of bilinguals’ pre-
proficiency by testing performance in an immigrant com- ferred language for mental computations. Spanish Applied
munity: The Wellington Indo-Fijians. Language Testing, 15, Linguistics, 4, 325–342.
86–108. Weber-Fox, C., & Neville, H. J. (1999). Functional neural
Snow, C. E. (1983). Age differences in second language acqui- subsystems are differentially affected by delays in 2nd
sition: Research findings and folk psychology. In K. M. Bailey, language immersion. In D. Birdsong ( Ed.), Second language
M. Long, & S. Peck ( Eds.), Second language acquisition acquisition and the critical period hypothesis ( pp. 23–38).
studies ( pp. 141–150). Rowley, MA: Newbury House. Mahwah, NJ: Erlbaum.
Snow, S., & Hoefnagel-Höhle, M. (1978). The critical period Wiebe, J. S., & Penley, J. A. (2005). A psychometric com-
for language acquisition: Evidence from second language parison of the Beck Depression Inventory—II in English and
learning. Child Development, 49, 1114–1128. Spanish. Psychological Assessment, 17, 481–485.
Snow, C. E., & Hoefnagel-Höhle, M. (1979). Individual Woodcock, R. W., McGrew, K. S., & Mather, N. (2001).
differences in second language learning ability: A factor Woodcock– Johnson Tests of Cognitive Abilities (3rd ed.).
analytic study. Language and Speech, 22, 1515–1562. Itasca, IL: Riverside.
Spolsky, B., Sigurd, B., Sako, M., Walker, E., & Arterburn, C. Woodcock, R. W., & Muñoz-Sandoval, F. (1993). Woodcock–
(1968). Preliminary studies in the development of techni- Johnson Language Survey. Itasca, IL: Riverside.
ques for testing overall second language proficiency. Lan- Wu, Y., Larrabee, J. H., & Putman, H. P. (2006). Caring
guage Learning, 18, 79–101. behaviors inventory—A reduction of the 42-item instru-
Stefani, L. (1994). Peer, self and tutor assessment: Relative ment. Nursing Research, 55, 18–25.
reliabilities. Studies in Higher Education, 19, 69–75. Zea, M., Asner-Self, K., Birman, D., & Buki, L. (2003). The
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate Abbreviated Multidimensional Acculturation Scale. Cultural
statistics (5th ed.). Boston: Allyn & Bacon. Diversity and Ethnic Minority Psychology, 9, 107–126.
Talamas, A., Kroll, J., & Dufour, R. (1999). From form to
meaning: Stages in the acquisition of second-language
Received March 13, 2006
vocabulary. Bilingualism: Language and Cognition, 2,
45–58. Revision received July 25, 2006
Tapia, M. (2001). Measuring emotional intelligence. Psycho- Accepted October 2, 2006
logical Reports, 88, 353–364. DOI: 10.1044/1092-4388(2007/067)
Tran, T. V. (1994). Bilingualism and subjective well-being in Contact author: Viorica Marian, Department of Commu-
a sample of elderly Hispanics. Journal of Social Service nication Sciences and Disorders, Northwestern University,
Research, 20(1–2), 1–19.
2240 Campus Drive, Evanston, IL 60208.
U.S. Bureau of the Census. (2003). Language use and E-mail: v-marian@northwestern.edu.
English-speaking ability (Census 2000 Brief C2KBR-29].
Retrieved November 24, 2004, from www.census.gov/
population /www/cen2000/ briefs.html.
Vaid, J., & Hull, R. (2002). Re-envisioning the bilingual brain
using fMRI. In F. Fabbro ( Ed.), Advances in the neurolin-
guistics of bilingualism ( pp. 315–355). Udine, Italy: Forum.
(1) Please list all the languages you know in order of dominance:
(2) Please list all the languages you know in order of acquisition (your native language first):
(3) Please list what percentage of the time you are currently and on average exposed to each language.
(Your percentages should add up to 100% ):
(4) When choosing to read a text available in all your languages, in what percentage of cases would you choose to read it in each of your languages?
Assume that the original was written in another language, which is unknown to you.
(Your percentages should add up to 100% ):
(5) When choosing a language to speak with a person who is equally fluent in all your languages, what percentage of time would you choose to speak each
language? Please report percent of total time.
(Your percentages should add up to 100% ):
(6) Please name the cultures with which you identify. On a scale from zero to ten, please rate the extent to which you identify with each culture. (Examples of
possible cultures include US-American, Chinese, Jewish-Orthodox, etc.):
966 Journal of Speech, Language, and Hearing Research • Vol. 50 • 940–967 • August 2007
Appendix (p. 2 of 2). Language Experience and Proficiency Questionnaire.
Language: Language X
This is my (please select from scroll-down menu: First, Second, Third, etc.) language.
All questions below refer to your knowledge of Language X.
(1) Age when youI:
(2) Please list the number of years and months you spent in each language environment:
Years Months
A country where Language X is spoken
(3) On a scale from zero to ten, please select your level of proficiency in speaking, understanding, and reading Language X from the scroll-down menus:
Speaking (click here for scale) Understand spoken language (click here for scale) Reading (click here for scale)
(4) On a scale from zero to ten, please select how much the following factors contributed to you learning Language X:
Interacting with friends (click here for scale) Language tapes/self instruction (click here for scale)
Interacting with family (click here for scale) Watching TV (click here for scale)
Reading (click here for scale) Listening to the radio (click here for scale)
(5) Please rate to what extent you are currently exposed to Language X in the following contexts:
Interacting with friends (click here for scale) Listening to radio/music (click here for scale)
Interacting with family (click here for scale) Reading (click here for scale)
Watching TV (click here for scale) Language-lab/self-instruction (click here for scale)
(6) In your perception, how much of a foreign accent do you have in Language X?
(click here for scale)
(7) Please rate how frequently others identify you as a non-native speaker based on your accent in Language X:
(click here for scale)