Running Head: UNSTROOPING THE STROOP EFFEC T 1 Unstrooping t he Stroop Effect Can Bilinguals Reduce Stroop Interference Through Other Language Mediation? Ana C. Oliveira University of Florida Submitted under the supervision of Dr. Jorge Valdes Kroff and Dr. Lise Abrams to the University Honors Program at the University of Florida in partial fulfillment of the requirements for the degree of Bachelor of Sciences, summa cum laude in Psychology.
UNSTROOPING THE STROOP EFFECT 2 Abstract Variations of the Stroop task continue to be employed as a robust, automatic measure of interference (MacLeod, 1991) 54 Spanish English bilingual speakers and 54 English monolingual speakers participated in a button press Stroop task administered in English (stimuli and response) The task presented participants with f our conditions ; baseline (BL), English labels (EL), Spanish labels (SL), and alternative labels (AL). Participants saw words in the computer screen in either blue green or yellow ink and had to press the c o r r esponding option in a button response box The SL condition served as a suggestion to bilingual participants to use their second language to help them complete the task more efficiently The Stroop interference effect was defined by comparing reaction time (RT) between incongruent and neutral trials for eligible stimuli. Results showed that bilinguals were overall faster and had a smaller Stroop interference effect compared to monolinguals but there was no significant difference between groups for the SL condition Bilingual participants presenting faster overall RT compared to monolinguals could be evidence to a n advantage in executive functioning. Keywords : Stroop effect interference bilingual conflict adaptation
UNSTROOPING THE STROOP EFFECT 3 Unstrooping t he Stroop Effect Can Bilinguals Reduce Stroop Interference Through Other Language Mediation? A growing body of evidence supports the claim that bilingual speakers present executive functioning advantages due to their experience with more than one language (Morales, Gmez Ariza & Bajo 2013) Among cognitive tasks that investigate executive functioning t he Stroop t ask measures inhibitory control by measuring interference. P articipants must name the ink color despite the color word presented ( blue presented in yellow) while their reaction time is recorded and compared with control stimuli such as control lists (color words in black ink), neutral words in different color inks or congruent trials ( green presented in green. Tzelgov, Henik. & Leiser, 1990 ). Despite different manipulations in the task, it generally serves as a robust measure of interference between word reading, a task irrelevant stimulus, and naming colors ( Milham et al. 2001 ) Several studies have used the task for different purposes including the study of selective attention, conflict resolution, executive control in normal and special populations ( Naylor, Stanley, & Wicha, 2012 ). Most studies using the Stroop task to investig at e bilingual populations have tested different groups of speakers within each of their languages (stimuli and response in the same language for separate conditions ; Milham et al., 2001; Rosselli et al., 2002; Wang, Fan, Liu, & Cai, 2016 ) as well as between them (stimuli and response in different languages in the same condition ; Tzelgov et al., 1990; Costa, Albareda & Santesteban 2008 ; Preston and Lambert, 1969 ) Therefore researchers have created and adapted the Stroop task in various ways to better understand how bilinguals deal with automaticity and to what extent the possible activation of their languages could correlate with the magnitude of the Stroop effect (see Francis, 1999 for a review) Bilinguals ex perience a within language Stroop superiority effect where the magnitude
UNSTROOPING THE STROOP EFFECT 4 of the Stroop effect for within language tasks is generally twice as much as between languages if there is no orthographic overlap for those languages (Naylor et al., 2012). Although there is no consensus about how bilinguals deal with the conflict of having more than one language in the brain, a mechanism of inhibition proposed by Green (1998) and Kroll et al. (2012) would allow bilinguals to use only one language despite a l ikely simultaneous activation of their languages which would explain the reduced Stroop interference between languages. To further explore this idea, w e propose a new design for the task where bilingual speakers will perform the task in English only (stimu li and response) and receive the suggestion to use their second language to reduce the interference from word reading. One of the most challenging aspects about working with bilingual groups is defining the level of proficiency and dominance for the language s they speak. Researchers usually use standardized grammar/vocabulary tests self report questionnaires and age of acquisition of a second language to determine proficiency and dominanc e ( Portocarrero, Burright, & Donovick 2007) Using those measures, analyses usually include an y interaction between the magnitude of the interference effect across languages and bilingual proficiency in each language (Naylor et al., 2012). However, using grammar tests alone could be deceiving if you have a highly proficient bilingual that did not have formal education on their second language Self report could skew the level of proficiency and age of acquisition could mask the real stage of acquisition in cases where speakers acquire a second la nguage at a certain age but do not use that language in daily life (e.g. English Spanish bilinguals that mostly speak Spanish with distant relatives ) We used several measures to determine level of proficiency for the current task (see Methods section) to assess how language dominance would correlate with performance in the task.
UNSTROOPING THE STROOP EFFECT 5 In addition to the classical Stroop interference that measures cognitive inhibition in the response level (eligible trials), we were interested in comparing that to a non response level block (ineligible trials). Milham et al. (2001) showed that the anteri or cingulate cortex (ACC) and the dorsolateral prefrontal cortex (PFC) which are involved with attentional control were activated differently depending on whether the level of conflict was at the response level or not. Furthermore, a response set effect su ggests that response eligible words cause a larger Stroop effect due to greater conflict occurring at the response level rather than the stimulus evaluation phase (Naylor et al., 2012). A response set effect for monolinguals would be equivalent to a betwee n within language Stroop difference for bilinguals (Roelofs, 2010). Having a non response level block will allow us to analyze if the proposed strategy is relevant at the response level. The goal for this study is to further explore differences within the bilingual population and between bilingual and monolingual participants. B oth groups were presented the same conditions where the manipulation is the use of labels as a proposed strategy where b ilinguals could use their non target language to map their responses and monolinguals could use words that prototypically represent colors to reduce the Stroop interference (e.g. peas for green ) In a previous study with the same type of stimuli and similar conditions, researchers tested 24 Spanish En glish bilinguals in a button press response Stroop task. Preliminary accidental findings suggested that bilinguals could use the non test language as an alternative response mapping strategy, i.e. bilinguals may recruit Spanish to map their button response s (i.e., associate responses to azul amarillo verde effect. Monolinguals showed stable Stroop interference for the later two sessions while bilinguals showed greater reduction of RT for incongruent and smaller difference between
UNSTROOPING THE STROOP EFFECT 6 incongruent and neutral for later sessions. As a follow up, we extended the set of stimuli and tested more participants in the same conditions ( see methods section). Ultimately, this research will be another stepping stone for a better understanding about the interaction between bilingualism, cognitive control, and the purported bilingual advantage. Research questions and hypothes e s Studies that used the Stroop task as a measure of conflict adaptation testing bilingual samples have explored many facets of how a second language in the brain relates to reaction time and interference. However, the study of bilingualism still raises many question as to which extent a second language changes the way speakers deal with conflict We fo cuse d on th r ee main questions : 1. Could bilinguals adopt an other language strategy to reduce Stroop interference? 2. Is there a same language analog for monolinguals? 3. What aspects of bilingualism correlate with the Stroop effect and the application of this strategy? In agreement with previous findings, results should reveal overall slower RT for bilinguals compared to monolinguals (Rosselli et al. 2002). However, bilinguals should also have smaller Stroop interference effect compared to monolinguals and a gr eater reduction of that interference in later sessions compared to BL especially for the SL condition agreeing with preliminary findings. Monolinguals could also present a reduction of interference for the AL condition but smaller in magnitude than biling uals for the last two conditions. This prediction is based on previous findings that bilinguals have better conflict adaptation skills from the constant juggling of two
UNSTROOPING THE STROOP EFFECT 7 languages in the brain (Kroll, Dussias, Bogulski & Kroff, 2012) and therefore present b etter suppression of task irrelevant stimulus (Milham et al., 2001). Suarez et. al (2014) found that English fluency predicted better performance of English Spanish bilinguals for incongruent trials in a Stroop task in Spanish. Similarly, we should find a correlation between performance in the Stroop task in English and fluency in Spanish. Methods Participants This study was approved by the Institutional Review Board (IRB) of the University of Florida. Each participant was given a description of the experiment and then signed an informed iment number, tasks completed, and any other notes necessary. Participants were recruited in two ways; IRB approved flyers were posted on boards around campus inviting Spanish English bilingual s peakers to participate in the study for a small monetary comp ensation whereas English monolingual speakers were recruited through the Undergraduate Research Pool, SONA. Researchers were able to recruit 5 4 bilinguals ( 39 females ) with ages rang ing from 18 to 30 years ( M = 21. 3 years SD = 2.91 ) and 5 4 monolinguals ( 34 females ) with ages ranging from 18 to 29 years ( M = 19.93 years SD = 2.05) Due to missing data, five participants were excluded from the bilingual group and four from the monolingual definitions that are not used consistently in the literature. She defines bilingualism according to Grosjean (1992, p. 51): "Bilingualism is the regular use of two (or more) langu ages, and bilinguals are those people who need and use two (or more) languages in their everyday lives."
UNSTROOPING THE STROOP EFFECT 8 Aligned with this definition, we recruited Spanish English bilinguals speakers who learned Spanish before or at the same time as English and speak both languages in their daily lives. English monolinguals are speakers who can only carry a conversation in English with limited knowledge of another language (up to school/college requirement level). To determine proficiency level, bilingual participants comp leted different language tests in both languages which will be further discussed in the materials section. Materials The different tasks in this project are either computer based or filled out on paper by research assistants The Stroop task is performed on a computer where participants see on e word at a time on the screen. Words appear in either blue, yellow, or green ink and p articipants must select which ink color is being presented in a push button response box The task design and stimuli were adapted from Milham et al. (2001). There are three types of stimuli ; f or eligible stimuli the color word is also a possible response option (e.g. green in blue ink ) while ineligible stimuli consist of color words that are not a response option (e.g. orange in yellow ink). The last type of stimuli are neutral words ( e.g. farmer ) used for comparison to establish the amplitude of the Stroop effect These three types of stimuli ( S ee Table 1) are s hown pseudo r andomly and organized into four sessions with 192 trials each ; first, participants must memorize the order for Table 1: Types of stimuli for Stroop task
UNSTROOPING THE STROOP EFFECT 9 what colors should be pressed in the button response box which will serve the baseline (BL) for the other sessions In the s econd session English labels (EL) for each color are placed right above the buttons. For the third and fourth sessions labels for the colors in Spanish (SL) and alternat ive words (AL) tha t are prototypically associated with the colors (e.g ocean for blue) are used and the sessions are counterbalanced between participants (s ee Figure 1) After the task is done participants answer a post experiment questionnaire (Appendix A) that assess es of experiment goals and their use of strategies including bilinguals use of their second language. The second computer based task, the AX Continuous Performance T ask (AX CPT), measure reactive (inhibitory) and proactive (monitoring) cognitive control. L etters appear on the screen one at a time where participants must press either yes or no in a push button response box after each letter according to a few conditions After the four first letters participants must regardless of what letter appears in which order. For the fifth let (cue) and (probe) In any other case, participants will after the fifth Figure 1: the four conditions for the Stroop task
UNSTROOPING THE STROOP EFFECT 10 letter as well The AX trials will be the target response and compose 70% of trials. AY trials compose 10% of trials and require participants to use proactive control due to the cue that prepares them BX trials also appear 10% of the time and demand participants to engage in reactive control where they need after the probe (see Figure 2 ) The Backward Digit Span Task (Working Memory Tes t from Wechsler Memory Scale Revised) is administered by research assistants where they read a string of numbers out loud and instruct participants to repeat it back to them in the reverse order. This task has been used by researchers as a measure of working memory capabilities ( Ramsay & Reynolds 1995). Figure 2 : Procedure for the AX CPT. (a) The series of events in a typical target trial. (b) The 4 different types of trial, the correct response for each of them, and the proportion of a given trial during the task. From Morales et. al (2013)
UNSTROOPING THE STROOP EFFECT 11 To determine language profici ency, two multiple choice standardized grammar tests are used; an adapted version of the Diploma de Espaol como Lengua Extranjera (DELE) is the Spanish grammar assessment and the Michigan English Language Institute College Entrance Test (MELICET) is the English grammar version. In addition to these grammar tests, research assistants administer ed a modified bilingual versio n of the Boston Naming Task ( BNT; Kaplan et al., 1983 ) where participants s ee images on the screen and ha ve to name them in English and Spanish in separate block s with distinct images This task attempts to quantify vocabulary size and the ratio found by dividing Spanish by English scores indicated if a participant was Spanish dominant (ratio 0.8) or English dominant (ratio < 0.8) The Language History Questionnaire (LHQ Appendix B ) present s participants with questions about their language background including the number of languages they speak, the frequency in which they speak those languages, age of acquisition, language preference, and whether they code switch These questions will help researchers determine in addition to grammar tests how profic ient/ dominant participants are in each of their languages and it will serve to double Procedure Testing occurred at the Bilingual Sentence Processing Lab at Dauer Hall and last ed approximately two hours for bilinguals and one hour and a half for monolinguals R esearch assistants explain ed each section of the informed consent to participants and ha d them sign it. Participants went through each task on the lab computer except for the post experiment questionnaire and digit span task. The software E prime ran both t he Stroop task and the AX CPT where it recorded RT and accuracy on to the computer We designed t he LHQ, DELE, and MELICET through the Qualtrics website that re cords and compiles the answers Finally, the
UNSTROOPING THE STROOP EFFECT 12 BNT (images presented on the computer screen) the digit span task, and the post experiment questionnaire were collected by research assistants on paper. Results We used r epeated measures analys i s of variance (ANOVA) as the main statistical test Post hoc t tests helped determine which variables were driving the interactions. B ivariate correlation tests were used to find associations between Stroop variables language proficiency and cognitive measures. We averaged RT for each participant for every level and excluded any trials over or under 2.5 standard deviations from each individual mean. The error analysis revealed an approximate 2% error rate and did not show any significant differences between groups. Analysis 1 had 16 levels 4 (Session) x 2 (Congruency) x 2 (Response Type) x 2 (Group) Analysis 2 consisted of 8 levels 4 (Session) x 1(Incongruent neutral) x 2 (Response Type) x 2 (Group) We ran c orrelations within the levels of analysis 2 with the digit span task and all proficiency tests (MELICET, DELE, BNT, and BNT ratio) for bilinguals to determine if proficiency had any correlation with RT for those levels. Analysis 3 had 6 levels 3 (Session: AL BL; SL BL; EL BL) x 2 (Congruency) x 1 (Eligible ). We ran the same correlations from analysis 2 but with the levels of analysis 3. Analysis 1 After running repeated mea s ures ANOV A results showed a m ain effect for c ongruency F ( 1,106) = 90.48, MSE = 7019.84, p < .01 where incongruent trials had higher mean RT than neutral mean RT indicating that both groups presented the Stroop interference effect This difference was present for all session s (p < .01) in both eligible and ineligible blocks as pairwise comparisons showed in post hoc t tests. Th e ANOVA also revealed an interaction between session,
UNSTROOPING THE STROOP EFFECT 13 congruency, and response type F (3,318) = 2.63, MSE = 1104.447, p < .05 in which incongruent eligible trials differed significantly from BL to all other conditions ( p < .04) and E L differed Graph 1: Error bars represent + 1 SD from the mean 600 625 650 675 700 725 750 775 800 Baseline English Spanish Alternative Monolingual Mean RT for Eligible trials Inc-Eli Neu-Eli Inc-Ine Neu-Ine 600 625 650 675 700 725 750 775 800 Baseline English Spanish Alternative Bilinguals Mean RT Inc-Eli Neu-Eli Inc-Ine Neu-Ine Graph 2 : Error bars represent + 1 SD from the mean
UNSTROOPING THE STROOP EFFECT 14 from AL significantly ( p < .05). No other difference between sessions was significant ps>. 05 This analysis also showed an interaction for congruency, response type, and group F (1,106) = 7.81 M SE = 2880.09 p < 01 where a post hoc t test revealed that the mean difference between incongruent and neutral trials for bilinguals in the eligible condition was 46.915 ( p < .01) while for monolinguals was 69.916 ( p < .01) showing a bigger Stroop effect for monolinguals overall. For an overview of mean RT for each group see Graph 1 and Graph 2 No other effects were signific ant, ps > .05. Analysis 2 After subtracting mean neutral trial RT from incongruent trial RT for each session and response type, the repeated measures ANOVA indicated a main effect for session F ( 3,318) = Graph 3 : Error bars represent + 1 SD from the mean -5 5 15 25 35 45 55 65 75 85 95 Eli Ine Eli Ine Eli Ine Eli Ine Baseline English Spanish Alternative Incongruent Neutral trials Bilinguals Monolingual
UNSTROOPING THE STROOP EFFECT 15 10.076 MSE = 2159 736 p < .001 and response type F (1,106) = 60.422 MSE = 5760.177 p < .00 1 where eligible trials have higher RT than ineligible Analysis also showed an interaction between response type and group F ( 1 106 ) = 7.808 MSE = 5760.177 p < .0 06 where bilinguals presented smaller difference than monolinguals for eligible trials (See Graph 2). The ANOVA also displayed an interaction between session and response type F (3, 318) = 2.63, MSE = 2208.893, p < .05 where BL differed from SL ( p <.001) and AL ( p <.001), EL differed from SL ( p <.002) and AL ( p <.001), but SL showed no significant difference from AL ( p < .745) for eligible stimuli No other effects were significant, ps > 05 All correlations between Stroop variables and the digit span task and proficiency tasks were not significant ps > .05. Analysis 3 The difference between each session and baseline was calculated for incongruent and neutral trials for the eligible condition only R epeated measures ANOVA revealed a main effect for congruency F (1,106) = 3389.317, MSE = 3389.317, p < .001 where incongruent trials had higher RT than neutral The analysis showed an interaction between session and congruency F (2,212) = 8.186, MSE = 1107.749, p < .001 where incongruent trials were only significantly different from neutral for SL minus BL ( p < .001 ) and AL minus BL ( p < .001 ) with mean differences of 28.03 and 30.09 respectively. No other effects were significant, ps > .05. Graph 3 shows the overall mean for incongruent trials for both groups. Bivariate tests found a negative correlation between both SL BL r (108) = .189, p<.05 and AL BL r (108) = .213, p<.03 and t he digit span test showing that participants who performed better in the digit span task presented smaller difference between these two conditions and BL No other correlations w ere significant ps > 05.
UNSTROOPING THE STROOP EFFECT 16 Graph 5 : Error bars represent + 1 SD from the mean 600 625 650 675 700 725 750 775 800 Bilingual Monolingual Eligible X Ineligible by Group Inc-Eli Neu-Eli Inc-Ine Neu-Ine 0 10 20 30 40 50 60 70 Incongruent Incongruent Incongruent BL-EL BL-SL BL-AL Difference from Baseline Bilinguals Monolingual Graph 4 : Error bars represent + 1 SD from the mean
UNSTROOPING THE STROOP EFFECT 17 Discussi on The current study aimed to investigate whether bilinguals would reduce Stroop interference in later sessions by utilizing their non target language as a strategy, how they compare to monolinguals, and whether monolinguals could find a same lang uage strategy to reduce interference. We found that the Stroop interference effect was present for both groups as participants had slower RT for incongruent eligible stimuli compared to neutral eligible stimuli across all sessions. Bilingual participants w ere overall faster than monolingual participants for both incongruent and neutral trials for eligible stimuli only although results lacked significance for any difference across sessions between the two groups. Monolinguals showed an overall larger Stroop interference effect than bilinguals. For both groups combined, BL had higher RT than the other three conditions while EL had higher RT than AL for incongruent eligible stimuli. The difference between incongruent and neutral RT differed significantly among sessions except between SL and AL. The difference between each session and BL for incongruent and neutral trials was significantly different from each other for SL and AL conditions only. According to previous findings, bilinguals should be overall slower than monolinguals (Rosselli et al, 2002; Prior, 2012); however, they showed faster RT compared to monolinguals and sustained that pattern throughout the study which agrees with findings for different attentional tasks ( Bialystok & Craik, 2010; Bialystok, Craik, & Luk, 2008). Bilinguals have long been the object of research for executive functioning and different studies have demonstrated that having two languages in the brain can result in better monitoring and inhibitory control (Prior, 2012, Morales et. al, 2013). Additionally, in the present task bilinguals could be using Spanish as a strategy to reduce RT for both incongruent and neutral trials without the need for suggestion. Language proficiency measures did not predict reduced Stroop interference but more
UNSTROOPING THE STROOP EFFECT 18 analyses between proficiency and RT for incongruent trials could reveal the correlation between Spanish dominance and faster RT based on results from Suarez et. al (2014) and Tse & Altarriba (2012). Moreover, the difference in motivation for completi ng the study could also explain this unexpected result as bilinguals received a small monetary compensation for their participation and have limited opportunities to participate in studies compared to monolinguals that were completing the study for class c redit. Another important limitation for any language study is the method of assessing proficiency. Having different proficiency measures could serve as a better way to determining whether a participant is balanced or dominant in one of their languages. How ever, current tests present their own limitations; for instance, the BNT task used here is outdated and grammar tests might not be the most effective way to detect language dominance, although results for the BNT in Spanish showed correlation with scores f or DELE and the BNT in English was correlated with scores for the MELICET. Self rating through the LHQ also presented challenges and might not be a clear representation of proficiency. Moreover, bilinguals could be overtly introduced to the strategy of usi ng their non target language instead of using labels as a suggestion. The Stroop effect is an automatic robust measure of interference that is not easily reduced; therefore, even if the strategy was introduced directly it would only be effective if bilingu als can use the non target language to map their responses. The lack of interaction between group and session for the Stroop effect was also unexpected as the main hypothesis predicted that the SL condition would show reduced Stroop interference for biling uals. A possible explanation is the lack of power of 54 bilingual participants for the desirable effect. Opposing the idea that bilinguals used the non target language strategy early on, proficiency measures did not predict their performance on the Stroop
UNSTROOPING THE STROOP EFFECT 19 task. The BNT ratio revealed that only 26% of bilingual participants were Spanish dominant and neither that nor their language preference correlated with a reduced Stroop effect. Age of acquisition and time in the US have not been analyzed with Stroop data Other analyses between proficiency measures and Stroop data are needed and will be explored in the future. Another possible confounding variable is participants using squinting or looking away in an effort to avoid the conflict with word reading. Preston and Lambert (1969) asked their participants to avoid squinting or looking from the corner of their eyes which was reported in the post experiment questionnaire as a strategy some participants used in the present study. This strategy could help participant s perform better but it could also hinder their performance by causing more fatigue; therefore, all participants should be instructed to avoid squinting and looking from the corner of the eye to achieve uniformity across participants in future studies. Fut ure Directions For now, results have shown that bilinguals are faster and had less Stroop interference than monolinguals, but further analyses are still needed to draw stronger conclusions. The AX CPT data still needs to be organized and analyzed. We hypot hesize that better inhibitory control in the AX CPT task will predict better performance on the Stroop task. Another way to look at the progression of cognitive control would be to analyze the change within each session. The first third of trials could be used as baseline for the latter third of trials. This analysis could reveal how participants progressed within a same session and what that progression looked like as a group. From the many proficiency measures collected, there might be better and more effective ways to rate bilingual proficiency and these differences could reveal important interactions with
UNSTROOPING THE STROOP EFFECT 20 Stroop data. We need to look at age of acquisition and find how the dif ferent measures agree with each other. We will run more correlation tests between proficiency measures and decide on how to use them to further analyze differences among bilinguals and contrast that to monolinguals. The data still contain some participants that could be acting as outliers for the group and raising the mean RT. Preliminary analyses showed that r emov ing participants that had mean RT over 2.5 SD from group mean for more than one session would reduced the overall mean by 10 30ms. Careful analys is of outliers could change later interactions and correlations for the group. Acknowledgements My sincere gratitude to Dr. Jorge Valdes Kroff who put me in charge of this project from its initial stages and guided me every step of the way allowing me to have autonomy and gain valuable research experience. And Dr. Lise Abrams who worked closely with us since the beginning of this project and shared her lab resources with us guidance was decisive for the long hours of statistical analyses an d making sense of the data. I am also grateful to every research assistant that helped us test participants throughout the 3 semester testing period.
UNSTROOPING THE STROOP EFFECT 21 References Bialystok, E., Craik, F., & Luk G. (2008). Cognitive control and lexical access in younger and older bilinguals. Journal of Experimental Psychology: Learning, memory, and cognition 34 (4), 859. Bialystok, E., & Craik, F. I. (2010). Cognitive and linguistic processing in the bilingual m ind. Current directions in psychological science 19 (1), 19 23. Braver, T. S. (2012). The variable nature of cognitive control: a dual mechanisms framework. Trends in cognitive sciences 16(2), 106 113. Costa, A., Albareda, B., & Santesteban, M. (2008). As sessing the presence of lexical competition across languages: Evidence from the Stroop task. Bilingualism: Language and Cognition 11 (1), 121 131. Dalrymple Alford, E. C. (1968). Interlingual interference in a color naming task. Psychonomic Science 10(6), 215 216. Francis, W. S. (1999). Cognitive integration of language and memory in bilinguals: semantic representation. Psychol. Bull 125, 193 222. Green, D. W. (1998). Mental control of the bilingual lexico semantic system. Bilingualism: Language and cogni tion 1(2), 67 81. Grosjean, F. (1992). Another view of bilingualism. In R. Harris (Ed.), Cognitive processing in bilinguals (pp. 51 62). Amsterdam: Elsevier. Heidlmayr, K., Moutier, S., Hemforth, B., Courtin, C., Tanzmeister, R., & Isel, F. (2014). Succe ssive bilingualism and executive functions: The effect of second language use on
UNSTROOPING THE STROOP EFFECT 22 inhibitory control in a behavioural Stroop Colour Word task. Bilingualism: Language and Cognition 17 (3), 630 645. Kaplan, E., Goodglass H., & Weintraub, S. (1983). Boston Naming Test (2nd ed.). Boston: Lea & Febiger. Kiyak, H. A. (1982). Interlingual interference in naming color words. Journal of Cross Cultural Psychology 13(1), 125 135. Kroll, J. F., Dussias, P. E., Bogulski, C. A., & Kroff, J. R. V. (2012). Juggling two languages in one mind: What bilinguals tell us about language processing and its consequences for cognition. In Psychology of learning and motivation (Vol. 56, pp. 229 262). Academic Press. Lesh, T. A., Westphal, A. J., Niendam, T. A., Yoon, J. H., Minzenberg, M. J., Ragland, J. D., ... & Carter, C. S. (2013). Proactive and reactive cognitive control and dorsolateral prefrontal cortex dysfunction in first episode schizophrenia. NeuroImage: Clinical 2, 590 599. Milham M. P., Banich, M. T., Webb, A., Barad, V., Cohen, N. J., Wszalek, T., & Kramer, A. F. (2001). The relative involvement of anterior cingulate and prefrontal cortex in attentional control depends on nature of conflict. Cognitive Brain Research 12 (3), 467 473. Morales, J., Gmez Ariza, C. J., & Bajo, M. T. (2013). Dual mechanisms of cognitive control in bilinguals and monolinguals. Journal of Cognitive Psychology 25(5), 531 546. Naylor, L. J., Stanley, E. M., & Wicha, N. Y. Y. (2012). Cognitive and ele ctrophysiological correlates of the bilingual Stroop effect. Frontiers in Psychology 3, 1 18.
UNSTROOPING THE STROOP EFFECT 23 Portocarrero, J. S., Burright, R. G., & Donovick, P. J. (2007). Vocabulary and verbal fluency of bilingual and monolingual college students. Archives of Clinical Neuropsychology 22 (3), 415 422. Preston, M. S., & Lambert, W. E. (1969). Interlingual interference in a bilingual version of the Stroop color word task Journal of Verbal Learning and Verbal Behavior 8(2), 295 301. Prior, A. (2012). Too much of a good thing: Stronger bilingual inhibition leads to larger lag 2 task repetition costs. Cognition 125 (1), 1 12. Ramsay, M. C., & Reynolds, C. R. (1995). Separate digits tests: A brief history, a literature review, and a reex amination of the factor structure of the Test of Memory and Learning (TOMAL). Neuropsychology Review 5 (3), 151 171. Roelofs, A. (2010). Attention and facilitation: Converging information versus inadvertent reading in Stroop task performance. Journal of Experimental Psychology : Learning, Memory, and Cognition, 36(2), 411. Rosselli, M. Ardila, A. Santisi, M., Arecco, M., Alvatierra, J., Conde, A., Lenis, B (2002). Stroop effect in Spanish English bilinguals. Journal of the International Neuropsychological Society, 8 819 827 Stroop, J. R. (1935). Studies of interference in serial verbal reactions Journal of Experimental Psychology 18, 643 662. Suarez, P. A., Gollan, T. H., Heaton, R., Grant, I., Cherner, M., & Group, H. (2014). Second Language Fluency Predicts Native Language Stroop Effects: Evidence from Spanish
UNSTROOPING THE STROOP EFFECT 24 English Bilinguals. Journal of the International Neuropsychological Society : JINS, 20(3), 342 348. Tse, C. S., & Altarriba, J. (2012). The effects of first and second language proficiency on conflict resolution and goal maintenance in bilinguals: Evidence from reaction time distributional analyses in a Stroop task. Bilingualism: Languag e and Cognition 15(3), 663 676. Tzelgov, J., Henik, A., & Leiser, D. (1990). Controlling Stroop interference: Evidence from a bilingual task. Journal of Experimental Psychology: Learning, Memory, and Cognition 16 (5), 760. Wang, R., Fan, X., Liu, C., & Cai, Z. G. (2016). Cognitive control and word recognition speed influence the Stroop effect in bilinguals. International Journal of Psychology 51 (2), 93 101.
UNSTROOPING THE STROOP EFFECT 25 Appendix A Post Experiment Questionnaire Please try to answer the following questions as honestly as you can. Provide any extra information that you think is relevant and be as specific as possible. If you do not understand a question, please ask. 1. What do you think was the purpose of this experiment? 2. Were all of the questions of the same difficulty, or did you find that some of the questions were easier or harder to answer than other questions? a. If some were easier/harder, what made them easier/harder? b. For those that were harder, did you use any strategies to make the task easier? 3. Did you use any overall strategies to help you complete the task? If so, what strategies did you use? 4. (Spanish speakers) Which language(s) did you use to help you complete the task?
UNSTROOPING THE STROOP EFFECT 26 Appendix B Language History Questionnaire
UNSTROOPING THE STROOP EFFECT 27
UNSTROOPING THE STROOP EFFECT 28
UNSTROOPING THE STROOP EFFECT 29
UNSTROOPING THE STROOP EFFECT 30