|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 THE EFFECTS OF INTENSIVE STUDY ABROAD AND AT HOME LANGUAGE PROGRAMS ON SECOND LANGUAGE ACQUISITION OF SPANISH By A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FUL FILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010
4 ACKNOWLEDGMENTS I wish to thank first an d foremost my committee chair, Dr. Gillian Lord. Her exceptional guidance and continuous assistance were what made this project feasible. I would also like to thank the members of my committee (both current and former members) Dr. Theresa Antes, Dr. David Pharies, Dr. Paula Golombek Dr. Diana Boxer, and Dr. Virginia LoCastro for their support and direction. Additionally, I thank the students who chose to participate in this study without whom this investigation would not have been possible. Finally, I wish to thank my family and my friends, your steadfast encouragement kept me inspired and your love helped me achieve this great goal.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 ABSTRACT ................................ ................................ ................................ ................... 11 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 13 Interest in Study Abroad ................................ ................................ ......................... 13 Previous Second Language Acquisition Research on Study Abroad ...................... 16 Acquisition of Linguistic Features ................................ ................................ ..... 17 Acquisition of Pragmatic Features ................................ ................................ .... 19 Impact of Study Abroad on Extra Linguistic Features ................................ ...... 21 Theoretical Background ................................ ................................ .......................... 24 The Input and Interaction Hypothesis ................................ ............................... 24 The Output Hypothesis ................................ ................................ ..................... 27 Conclusion ................................ ................................ ................................ .............. 28 2 PRIOR RESEARCH IN FLUENCY ................................ ................................ ......... 31 Introduction ................................ ................................ ................................ ............. 31 Investigating Fluency ................................ ................................ .............................. 31 Defining Fluency ................................ ................................ ............................... 31 Measuring Fluency ................................ ................................ ........................... 37 Learning Context ................................ ................................ .............................. 41 Study Abroad and Oral Pro ficiency Level ................................ ............................... 55 Conclusion ................................ ................................ ................................ .............. 60 3 PRIOR INVESTIGATION OF SECOND LANGUAGE WILLINGNESS TO COMMUNICATE ................................ ................................ ................................ ..... 62 Introduction ................................ ................................ ................................ ............. 62 Development of a Second Language Willingness to Communicate Model ...... 63 Will ingness to Communicate and the Canadian Studies ................................ .. 67 Willingness to Communicate Research Outside of Canada ............................. 71 Conclusion ................................ ................................ ................................ .............. 80
6 4 METHODOLOGY ................................ ................................ ................................ ... 82 Research Questions and Hypotheses ................................ ................................ ..... 82 Participants ................................ ................................ ................................ ............. 84 Program Information ................................ ................................ ............................... 86 Study Abroad Program ................................ ................................ ..................... 86 At Home Program ................................ ................................ ............................. 86 Data Collection ................................ ................................ ................................ ....... 87 Language Background Questionnaire ................................ .............................. 87 Oral Interviews ................................ ................................ ................................ 88 Willingness to Communicate Questionnaires ................................ ................... 89 Language Use Questionnaire ................................ ................................ ........... 91 Reflection Blogs ................................ ................................ ............................... 92 Summary ................................ ................................ ................................ .......... 94 Data Analysis ................................ ................................ ................................ .......... 95 Oral Interviews ................................ ................................ ................................ 95 Willingness to Communicate Questionnaires ................................ ................... 98 Language Use Questionnaires ................................ ................................ ......... 99 Reflection Blogs ................................ ................................ ............................... 99 Conclusion ................................ ................................ ................................ ............ 101 5 RESULTS AND DISCUSSION ................................ ................................ ............. 103 Quantitative Results and Discussion ................................ ................................ .... 103 Fluency Measures for Study Abroad Learners ................................ ............... 103 F luency Measures for At Home Learners ................................ ....................... 107 Comparing Study Abroad Learners to At Home Learners .............................. 110 Willingness to Communicate for Study Abroad Learners ............................... 113 Willingness to Communicate for At Home Learners ................................ ....... 115 Comparing Study Abroad Learners to At Home Learner s .............................. 116 Correlations between Data Sets ................................ ................................ ..... 118 Fluency and amount of Spanish used ................................ ...................... 118 Willingness to Communicate and the amount of Spanish used ............... 119 Correlations between fluency and Willingnes s to Communicate .............. 120 Additional Information Results and Discussion ................................ ..................... 120 Common Tendencies ................................ ................................ ..................... 121 Willingness to Communicate specific tenden cies ................................ .... 122 Non specific common tendencies ................................ ............................ 129 Differences between Study Abroad Learners and At Home Learners ............ 13 6 Specific Individual Differences ................................ ................................ ........ 144 Conclusion ................................ ................................ ................................ ............ 148 6 CONCLUSION ................................ ................................ ................................ ...... 150 Answers to Research Questions ................................ ................................ ........... 150 Question 1 Fluency Research ................................ ................................ ...... 150
7 Question 2 Willingness to Communicate Research ................................ ..... 154 Question 3 Relationship between Fluency and Willingness to Communicate ................................ ................................ .............................. 157 Limitat ions ................................ ................................ ................................ ............. 158 Future Studies ................................ ................................ ................................ ...... 161 Conclusion ................................ ................................ ................................ ............ 162 APPENDIX A LANGUAGE B ACKGROUND QUESTIONNAIRE ................................ ................. 164 B ORAL INTERVIEW QUESTIONS ................................ ................................ ......... 166 C WILLINGNESS TO COMMUNICATE QUESTIONNAIRES ................................ .. 169 D LANGUAGE USE QUESTIONNAIRE ................................ ................................ ... 173 E INTERVIEW TRANSCRIPTIONS ................................ ................................ ......... 175 F CODING SCHEME FOR B LOGS ................................ ................................ ......... 191 LIST OF REFERENCES ................................ ................................ ............................. 193 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 200
8 LIST OF TABLES Table page 4 1 Example of speech analysis ................................ ................................ ............... 97 5 1 Simple means of fluency measures for study abroad learners ......................... 104 5 2 Paired T tests of fluency measures for study abroad learners .......................... 105 5 3 Simple means of fluency measures for at home learners. ................................ 108 5 4 Paired T tests of fluency measures for at home learners ................................ 108 5 5 Simple means of fluency measures for at home learners with two Spanish courses ................................ ................................ ................................ ............. 109 5 6 Paired T tests of fluency measures for at home learners with two Spanish courses ................................ ................................ ................................ ............. 110 5 7 Comparison of study abroad learners t o at home learners for changes in fluency over time ................................ ................................ .............................. 112 5 8 Comparison of study abroad learners to at home learners with two Spanish courses for changes in fluency over time ................................ ......................... 113 5 9 Simple means of Willingness to Communicate in Spanish for study abroad learners ................................ ................................ ................................ ............ 114 5 10 Paired T Test of Willingness to Communicate in Sp anish for study abroad learners ................................ ................................ ................................ ............ 114 5 11 Simple means of Willingness to Communicate in Spanish for at home learners ................................ ................................ ................................ ............ 115 5 12 Pai red T Test of Willingness to Communicate in Spanish for at home learners ................................ ................................ ................................ ............ 115 5 1 3 Comparison of study abroad learners to at home learners for change in Willingness to Communicate over time ................................ ............................. 116
9 LIST OF FIGURES Figure page 3 1 Model of Second Language Willingness to Communicate. MacInytre and Charos (1996) ................................ ................................ ................................ ..... 64 3 2 Heuristic Model of Variables Influencing Willingness to Communicate (MacIntyre, Clment, Drnyei, and Noels, 1998) ................................ ................ 66
10 LIST OF ABBREVIATIONS ACTFL/ILR American Council on the Teaching o f Foreign Languages/Interagency Language Roundtable AH At h ome study language study that takes place at the home institution L1 First l anguage the language first acquired from birth L2 Second l anguage an additional language acquired after the first NS(s) Native speaker(s) speaker of a particular language that was first acquired from birth NNS(s) Non n ative speaker(s) speaker who has learned a particular language after their first language OPI Oral Proficiency Interview An oral interview d evelo ped by ACTFL/IRL to be used for assessing L2 learner proficiency based on their guidelines SA Study abroad language study that takes place in a country where the second language is spoken as a native language WTC Willingness to Communicate the willingn ess a person has to interact with an interlocutor
11 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE EFFECTS OF I NTENSIVE STUDY ABROAD AND AT HOME LANGUAGE PROGRAMS ON SECOND LANGUAGE ACQUISITION OF SPANISH By December 2010 Chair: Gillian Lord Major: Romance Languages Spanish This study explores the effects of short term Spanish study in s tudy abroad ( SA ) and at home (AH) contexts on oral fluency and Willingness to Communicate (WTC). Following the Input and Interaction Hypothesis and the Output Hypothesis, it is hypothesized that learners from short term SA programs will demonstrate larger gains in fluency and WTC due to their greater interaction with native speakers and second language (L2) artifacts. Previous fluency research (Freed, 1995; Towell, 2002; Freed, So and Lazar, 2003; Isabelli Garca, 2003; Segalowitz and Freed, 2004; Segalowit z, Freed, Collentine, Lafford, Lazar, and Daz Campos, 2004; Freed, Segalowitz and Dewey, 2004; Juan Garau and Prez Vidal, 2007; and Collentine, 2007) has been positive for long er term SA programs ; it is the aim of this researc h to add to the s e findings by exploring a short term program. At the time of the current study only one investigation of WTC (Yashima and Zenuk Nishide, 2008) ha s sought a possible interaction with L2 acquisition ; however, a belief persists that learners
12 who are more willing to use the L2 show more improvement linguistically. The findings of this study provide the field with empirical research about the connection between the two. Participants are 23 intermediate level students (9 in SA, 14 in AH). All a re L1 English speakers and are no t bilingual. Data were collected using pre and post program interviews, pre and post program questionnaire s examining WTC in Spanish, a questionnaire examining WTC in English, weekly Language Contact questionnaires, and personal reflection b logs discussing WTC in Spanish. Results somewhat positively favor the SA context showing that the SA learners improved their fluency significantly more than the AH learners. Yet when compared to AH learners who were enrolled in the s ame amount of credit hours fewer differences were seen. SA learners were able to show a faster rate of speech than AH learners, but no other significant differences were seen. Similarly, WTC significantly increased over time for the SA learners but not for the AH learners. Quantitative results do not reveal a correlation between WTC and higher fluency scores although f urther research is needed to determine the nature of the relationship between WTC and fluency.
13 CHAPTER 1 INTRODUCTION The present study e participating in a six week study abroad (SA) program while comparing them to learners in a six week at home (AH) program. Furthermore, it investigates the extra linguistic feature of Willingness to Communicate (WTC) to discover if and how SA may influence possible connection that may exist between WTC and fluency. In this first chapter, an introduction is presente d of the interest in SA today, previous second language acquisition (SLA) research of SA, and previous research concerned with defining and measuring fluency. In the following two chapter s a discussion of existing research is presented; in the second chap ter the factors that impact L2 fluency, SA and oral abilities are discussed and in the third chapter the concept of L2 WTC is presented along with research into elements that may impact WTC The later chapters are set up as follows: Chapter Four explains the methodology used for the study, Chapter F ive gives the results of the study followed by a discussion of those results, and Chapter Six concludes by discussing the implications of this work as well as suggesting future avenues of research. In all, the i nvestigation presented herein seeks to add to the body of SA research to help support previous findings while also providing new information about this distinctive language learning experience. Interest in Study Abroad As United States ( US ) college student s prepare themselves to enter the workforce, many of them recognize the need to be globally aware in both the public and private sectors. Furthermore, these students appreciate the value of being linguistically
14 competent in a L2 and see this knowledge as a key asset to finding employment (Open Doors, 2009). In order to achieve these goals, students are looking toward SA to experience the unique opportunity of immersing themselves in another culture and language. Over the past decade, US college students hav e increased their participation in SA programs by 150% (Open Doors, 2009). Encouragement to study in a foreign country has also increased; for example, President Obama and other US government officials have encouraged more students to take advantage of thi s unique learning our language skills (AIFS, 2009). The US Under Secretary of State for Public Diplomacy and Public Affairs, Judith A. McHale, recently stated that, more than ever before, SA can help our students to understand our interconnected world and to participate Current SA programs are also experiencing greater diversity than those of 10 years ago for example students may participat e in shorter programs, six weeks instead of the traditional semester long program. In the 2007 2008 educational year, 56% of SA students participated in a short term program ranging from two to eight weeks long (Open Doors, 2009). This change is most likely due to the need for more economical opportunities for SA, a trend that is likely to continue in the coming years given the current financial situation in the US. Moreover, short term SA is becoming increasingly popular wit h students who feel unable to dedicate an entire semester away from their home institution (Open Doors, 2009). As interest in SA increases and as SA programs
15 experienc e impacts learning and to attempt to better understand what students are gaining from their stays abroad. SA has always been an important facet of L2 study as many educators believe that total immersion in the L2 community is a necessary step to reach high er, more native like L2 abilities ( e.g., Freed, 1995). In this type of immersion, learners have the opportunity for learning in both a formal and a natural setting, providing them with a chance to interact with native speakers (NSs) on a daily basis, while still having the opportunity to focus on the L2 in a classroom setting. In many cases, this SA experience could be the first time that learners have access to this amount and quality of contact with their L2. Why is it though, that so many instructors and language learners believe that a L2 cannot be truly learned without this type of immersion? How does immersion impact language learning and acquisition? What factors will help students receive the most benefit from a SA program? These questions have been raised by several researchers of SLA in recent years in hopes of determining the ways in which immersion may shape the acquisition process. Research has been conducted on the effects of immersion and SA on linguistic factors, both grammatical and pragmatic as well as on extra linguistic factors, for example, anxiety. As will be discussed in future sections positive results have been found in some studies; however, there are also several instances of mixed results or inconclusive results, making it difficu lt to pinpoint the advantages of SA or immersion over a more traditional formal instruction in the home country. For this reason, there is a need for additional empirical research to allow researchers better understand the effects of SA on SLA.
16 Furthermore through research into the effects of SA it will be possible to allow learners to better understand what achievements can be expected from participating in such a program. Generally, all learners have an expectation of improvement in their L2 abilities af ter SA; in particular, there is a belief that the greatest improvement will be in internet search for information about SA reveals that there is a tendency for SA websites to state that SA is the best way to learn a L2 and/or indicate that participants will increase in fluency (see, for example, www.vistawide.com/ studyabroad www.globaled.us/now/why studyabroad intro.html www.goabroad.com and www.transitionsabroad.com ). Due to these beliefs, many learners may develop unre alistic goals for their SA experience; for example, one participant in the current study stated that he would be a perfect speaker after his stay abroad. Discovering how much of an impact SA can have on communication skills and fluency is an important as pect of SA research and will allow both researchers and participants to have a better understanding of what is possible during this special learning experience. Previous Second Language Acquisition Research o n Study Abroad Previous SLA research into the e ffects of SA can be classified as pertaining to four primary categories: acquisition of specific linguistic features, acquisition of pragmatic features, the impact of SA on extra linguistic features, and the acquisition of oral abilities. The acquisition o f oral abilities, in particular fluency, is the main subject of the current study, and will be explored in greater detail in the following chapter. In this chapter, a brief summary of findings is presented for each of the other three categories.
17 Acquisitio n of L inguistic F eatures In this category, investigators may focus on the acquisition of either specific linguistic features of the L2 (such as those in the areas of pronunciation or syntax) or take a more wide progress in all linguistic features in general. For example, one study may investigate the acquisition of the preterit aspect of the past tense only; while another may use a general grammar ariety of structures. The majority of these investigations research the differences between the SA learning context and the more traditional at home (AH) foreign language learning context in an attempt to explain what learners will gain from SA that they a re unable to gain from AH study. Findings for these studies have produced mixed results with some investigators finding an advantage for SA and others finding no advantage or even, in some cases, a disadvantage to SA Some investigations, for example, hav e demonstrated that learners who have participated in SA produce more complex discourse providing more contexts to use more advanced structures (Freed, 1995; Marriott, 1995; Regan, 1995; Allen and Herron, 2003; Freed, So, and Lazar, 2003; Collentine, 2004; Freed, Segalowitz and Dewey, 2004; Isabelli and Nishida, 2005; Howard, 2005; Barron, 2007; Magnan and Back, 2007). However, in many cases learners still fail to use the correct L2 structure in the context they have created (Marriott, 1995; Regan, 1995; C ollentine, 2004) Likewise, looking at pronunciation and the acquisition of a native like accent, s tudents are able to also achieve similar levels of native like pronunciation Likewise,
18 Lord showed that explicit instruction can produce a comparable benefit. I n an investigation of Spanish pronunciation Daz Campos (2004) found that both SA and AH stude nts showed a similar pattern of improvement or lack of improvement in acquiring target like pronunciation of a variety of consonants. This result demonstrates that for certain elements SA may not be able to provide any special benefit that may not also be achieved in traditional classroom study. Segalowitz, Freed, Collentine, Lafford, Lazar and Daz and, as was seen with Daz Campos (2004), no advantage was found for SA since both le arner groups were able to make significant improvements. Turning to possible changes in morphosyntactic structures i n some instances, as was found for Isabelli and Nishida (2005) s tudy of the Spanish subjunctive and ast tense, significant improvement was made by the SA learners but not the AH learners demonstrating an advantage for the SA context. Nonetheless, in Collentine (2004) stud y of Spanish morphosyntactic features and Geeslin and Guijarro Fuentes (2005) s tudy of the Spanish copula verbs ( ser and estar ) no overall advantage is sh own for the SA context. Collentine suggests that because linguistic (and specifically morphosyntactic ) features are widely emphasized in AH classes and assessment learners are cap able of acquiring these structures and making significant progress in a traditional setting, and therefore, SA may be more effective in other respects Despite an overall lack of positive results, Collentine reported that in general SA learners were able to produce more semantically dense discourse than the AH group, although they continued to remain less accurate morphosyntactically than the AH group. Cubillos, Chieffo and Fan (2008) found that on
19 average AH learners were able to make equally significant improvement in listening comprehension as the SA learners; however, when comparing the advanced learners only, SA learners were found to have made greater improvement. In all investigations, nevertheless, it is important to note that even though significa nt improvement may have been made, the overall level of morphosyntactic accuracy still remain ed below NS levels, indicating a need for additional L2 study beyond the SA experience. It is important to keep in mind that participation in SA is not a fast trac k to native like unrealistic. The results seen in these existing studies can help linguists, instructors and learners become more aware of what achievements are realistically p ossible and should promote practical goals. Acquisition of P ragmatic F eatures It has been suggested by several researchers that the greatest benefits that can be seen for SA, and other immersion environments that provide access to NSs, are in the form of pragmatic gains. With the pressures of covering large amounts of grammar and vocabulary, many traditional foreign language classrooms may not devote much time to pragmatic instruction (Kasper and Schmidt, 1996). Kasper and Schmidt (1996) proposed that: A s econd language environment is more likely to provide learners with the diverse and frequent input they need for pragmatic development than a foreign language learning context (p. 159 160) Because SA learners are provided with the L2 in its full social co ntext, they have opportunities to see real language in use on a recurrent basis which may lead to developments of their own pragmatic competence (Barron and Warga, 2007). It is
20 important to note though, that most research has shown that while the developm ent that does occur approaches the native norm, it rarely reaches full native like pragmatic competence (Kasper and Rose, 2002). As with SA and the morphosyntactic domain results have been mixed for studies researching SA and pragmatic features. Positive results have shown that learners are able to produce more native like speech acts after participating in SA; likewise, they are capable of producing more complex and content rich discourse (Lafford, 1995; Hassall, 2006; Barron, 2007) For example, Lafford discovered that SA learners were better able to initiate conversations, showed more diversity of suitable structures within a conversational context, and were more proficient at completing conversations; overall, their conversations were more natural and realistic. On the other hand, AH learners were less adept at managing conversations and preferred to take on a passive role; moreover, their conversations lacked realism and adhered to memorized speech chunks whenever possible. Likewise, researching his ow n L2 improvement during SA, Hassall (2006) found that he began with formulaic expressions of leave taking, and A s he continued to observe NS responses and his own successes, Hassall broke away from his formula to embrace a spontaneous approach allowing his interlocutors to guide him. This change in behavior resulted in more highly developed discourse and use of more native like structures. Similarly, i n a study of upgrading refus als, Barron (2007) found that as learners spent more time in the NS community, they produced more native like speech in their refusals, particularly when speaking with strangers. Overall, these
21 investigations find that as learners spend more time in the L2 community, they were able to change their speech to reflect target language norms. P ositive results were also found in helping learners improve on specific speech acts such as refusing offers (Barron, 2007), negation (Regan 1995), requests and apologies (Cohen and Shively, 2007) leave taking (Hassall, 2006) and offering advice (Matsumura, 2007) In these studies, SA learners were able to make significant improvements over time with these speech acts, abandoning their foreign sounding structures for more native like speech. Nonetheless, the learners still remained below NS standards, as was seen with morphosyntactic structures. I n some instances it was hindered a more native like acquisition of certain features such as politeness (Marriott, 1995; Matsumura, 2007) Based on the comments provided by the participants in t appeared that although learners expressed a desire to become competent L2 speak ers they we re not necessarily prepared to give up their L1 belie f s as to what is acceptable speech behavior. Impact of Study Abroad on E xtra L inguistic F eatures R ecent SA and immersion environment research has focused on extra linguistic gains that impact acquisition of an L2. The most widely studied extra linguistic factors are self confidence in the L2, self perception of L2 abilities, motivation, and anxiety. By investigating these features, researcher s have sought to better understand how they might re late to progress in linguistic features ; for example if a learner experiences high anxiety while in the SA environment, he or she may not be able to appreciate the interaction possibilities with NSs and therefore may be less likely to make linguistic devel opments (Churchill and DuFon, 2006).
22 As with morphosyntactic and pragmatic features, results can vary for extra linguistic features and SA although, in general, the results appear to indicate that the SA context has a favorable influence on learners Res ults show that self perception of abilities and self confidence both increase d for learners after participating in a SA program (Tanaka and Ellis, 2003; Allen and Herron, 2003; Dewey, 2004; Magnan and Back, 2007; and Cubillos, Chieffo and Fan, 2008) I n so me cases, such as Magnan and Back, self confidence and self perception can follow a u shaped pattern with a period of decrease early on in the program that is later followed by a rise. Often in the second or third week of a program, after the excitement of being in the new country has waned, learners may become discouraged because they are suddenly aware of gaps in their knowledge and are coping with linguistic culture shock. However, as they continue to interact with NSs and are successful with communicati on in the L2, their self confidence rises and often surpasses their initial confidence level. Cubillos, Chieffo and Fan (2008) conducted one of the few studies to compare changes in self confidence and self perception between a SA group and an AH group. Th eir findings show that SA students were more confident and secure in their abilities to speak and comprehend Spanish than AH students; however, although many of the SA students claimed to have improved their listening skills, they were not able to out perf orm the AH students on the proficiency measure. In a similar trend to increased self confidence and self perception, a nxiety was found to decrease overall in SA learners (Allen and Herron, 2003; Dewey, 2004; and Cubillos, Chieffo and Fan, 2008 ). After par ticipating in SA, in particular, learners were less worried about communicating in their L2 and were not as concerned with making
23 errors as they were prior to their SA experience Additionally, learners were found to be less anxious in both classroom and n atural settings. This result is not surprising since learners have more practice with the L2 in a variety of settings and with a variety of interlocutors we can expect that they will become more comfortable overall using the L2. On the other hand, when c onsidering the extra linguistic feature of motivation, some studies have found n o change after SA (Allen and Herron, 2003; Dewey, 2004) I t may be that these learners were already very highly motivated to learn their L2 hence their desire to study in the NS community in the first place. In addition, the competency level of the learners may have also been a factor in these results. In both studies the learners were at the intermediate level or higher, indicating that they had already invested time and effor t in learning the L2 and therefore were fairly highly motivated to learn the L2. The positive result that can be gleaned from these two studies is that no decrease in motivation is seen, indicating that after the SA program, learners continue to be equally motivated to continue studying the L2. Overall, further research into these extra linguistic features and other similar features such as Willingness to Communicate (WTC) in the L2, can allow investigators more insight into the SA experience and take in to account the many facets that impact language acquisition. Future studies should take into account the importance of a comparison AH group to allow researchers to better understand the different effects that can be seen as a result of a SA environment. F inally, it is important to look for connections between extra linguistic features and linguistic features to further explore how self confidence, self perception, anxiety, and motivation relate to acquisition.
24 Theoretical Background Several theoretical con cepts of second language acquisition (SLA) value interaction with other L2 speakers as an important process of acquiring an L2. Two such concepts are the Input and Interaction Hypothesis (Gass, 1997) and the Output Hypothesis (Swain, 1993, 2000). For both, an important part of the language acquisition process involves interacting with native speakers (NSs) and/or other non native speakers (NNSs) of the L2, acquiring input and producing output, and learning from this interaction. Given that the SA context av ails L2 learners of a wealth of interaction occasions, for students who take advantage of these opportunities there is a great potential for much language acquisition. What follows is a brief summary of these constructs, with special attention to how the S A context may particularly aid acquisition in light of these proposals. The Input and Interaction Hypothesis The Input and Interaction Hypothesis developed by Gass (1997) takes into account the importance of interaction in the L2 as one of two key compone nts of the acquisition process. With the wealth of opportunity for interaction provided by the SA environment, this hypothesis is a logical choice to examine the potential for language acquisition during SA. As well, the SA context provides a great amount of input, this developed as part of the Monitor model, which states that a learner must be exposed to comprehensible input in order to acquire a L2. He stipulates that in order for a learner to make progress, the L2 input must be at the appropriate level of comprehensibility which is described as being one or l evel i +1. While
25 S he believes that all input is valuable to the acquisition process and that it is through i nteraction that learners are able to understand input that may be much more advanced than their current level of knowledge. Gass further explains that while input provides the opportunity for learning, it is not enough for acquisition to occur. Rather, it is through the act of communicating in the L2, receiving input and working with it, that learners are able to develop their language abilities. As part of the notion of interaction, Gass introduces the concept of negotiation, or problem as opposed to communication in which there is a free flowing exchange of to arise due to the learn problems, the two interlocutors must negotiate between themselves to discover the intended message. Two types of negotiation are discussed under this hypothesis, negotiation of form and negotiat ion of meaning; though often, these two are connected. Negotiation of form refers to a miscommunication with linguistic form. For example, many novice learners have a tendency to rely on temporal adverbs when speaking about the past tense while using the p resent tense of the verb as in (1). (1) *Ayer yo voy a la playa. *Yesterday I go to the beach. This use of the present tense may prompt a more knowledgeable interlocutor to question this verb form and negotiate for the correct form, for example, by provid ing the correct form fui went
26 significance of the intended message. In the same example above, the more knowledgeable speaker could ask for clarification regarding when the action took place, placing attention on the overall message rather than the specific linguistic form. Gass believes that by redirecting attention to the problem, learners are able to become aware of additional L2 information that may be lacking from their interlanguage system, or n otice the gap in their knowledge This negotiation may be explicit, if one of the interlocutors makes it clear that there is a misunderstanding, or may be implicit if there is only an implied problem. Additionally, the negotiation may be initiated by th e learner or the other interlocutor (a NS, a more knowledgeable speaker, another learner). In any case, it is important to note that the miscommunication may not always be resolved through the negotiation. It is possible that the negotiation merely highlig hts a particular aspect of the L2 for the learner, and that additional input and interaction are necessary before learners are able to fully comprehend that element. Certain learning contexts can provide learners with a richer opportunity for interaction and negotiation. In a traditional classroom learners will spend the majority of their time interacting with the instructor and other learners. The negotiation possibilities are therefore limited, especially when the interaction is learner to learner, since there is a possibility that learners will not notice communication problems. Conversely, in a SA environment, there is more potential for learners to interact with NSs and other more knowledgeable speakers that can provide them with more valuable negotiat ion opportunities. It is important to note that learners may not take full advantage of this potential. In all instances, learners must be able to see the miscommunication and negotiation process as part of the learning experience, and understand that it s erves to
27 help them improve. This concept is not addressed in depth by Gass, yet it stands to reason that if learners are not aware of this process, they may become discouraged when interacting in the L2 and thus become less inclined to communicate. The Out put Hypothesis During communication in the L2, learners are not only provided with input and an opportunity to learn from their interlocutors, they are also active in producing output. In her Output Hypothesis, Swain (1993) argues that output pushes learne rs to analyze language on a more in depth level; and furthermore, provides more linguistic development than input, due to the extra mental effort required to produce a statement t o produce [languag e], learners need to do something. They need to create linguistic form and meaning, and in the input they receive, learners must be able to manipulate that input and work w ithin the language system to be able to produce comprehensible output. observed by other researchers (Schmidt and Frota, 1986; Ellis, 1994), the ability to notice what is mis acquisition; moreover, linguistic elements that go unnoticed will not be acquired. It is language an d discover that they are not able to convey their message in precisely the way in which they had hoped. From their production, learners can progress in their abilities based on the type of feedback they receive from other speakers. When met with positive f eedback, or a lack of misunderstandings, learners will know that what they have produced is correct and will continue to produce such language. After receiving
28 negative feedback, learner s can become aware that something is incorrect in their system and not ice a need for additional information. Output also provides learners with an opportunity for hypothesis testing B eliefs about how to use linguistic elements change over time as learners receive more information and develop their interlanguage Swain (200 0) argues that learners may test out various linguistic elements when producing output to discover how these elements work and when they may be used. Moreover, as learners produce output they discover the ways in which forms may be manipulated and changed to suit various situations. The more learners work with different structures and produce output that is received by NSs the more opportunity they will have for feedback about those structures. From that feedback then, learners can better understand which changes/manipulations are possible. This hypothesis testing is not limited to grammatical forms, but may include all facets of the L2 from pronunciation to pragmatics. Although it may be possible for learners to notice how these elements are used in input, Swain postulates that once again learners will be more aware of these elements in their own output and how that output is received by more knowledgeable interlocutors. Conclusion As it becomes more common for US university students and other language lear ners to study in the target language community, it is necessary for the field of SLA to continue to explore the gains that can be realistically be made during that study. From the research reviewed here, it can be seen that the impact of SA on L2 developme nt is intricate and not easy to define. While in some investigations SA is shown to be clearly beneficial (i.e. Isabelli and Nishida 2005), for others there is a tendency to see a mixture of elements where learners may show improvement in one area yet lag in
29 others (i.e. Cubillos, Chieffo and Fan, 2008). These results demonstrate that the simple unrealistic, and the research seen here can assist in creating more attainable go als for students who choose to participate in SA. At the same time, this research has also served to demonstrate that there are benefits to traditional AH classroom instruction and that for certain linguistic elements AH learners can be equal to or even mo re successful than SA learners. Furthermore, as investigation continues to explore SA, researchers may be able to further ascertain the unique advantages provided by the immersion context. As SA research continues, we should seek to continue comparing SA t o the more traditional AH setting to allow linguistics to better understand the differences between the two contexts and to ensure that positive results that are seen in SA investigations can truly be attributed to that context. When considering extra ling uistic features, researchers should seek out possible links between these features and linguistic abilities in hopes of explaining the role they play in the acquisition process. In addition, investigators should be aware of current trends in SA programs, s uch as short term programs, and consider how these changes may influence the acquisition process. In the two hypotheses discussed above, the Input and Interaction Hypothesis, and the Output Hypothesis, communication and interaction with NSs and more knowle dgeable NNSs is a vital part of the acquisition process. It is through this communication that learners are able to receive valuable input and linguistic information but also they are able to have the chance to produce the necessary output to convey their thoughts and needs. Moreover, in producing that output learners work to become more efficient and eloquent in responding to more knowledgeable speakers. This
30 improved output production can be seen as part of improving learner fluency. In a traditional clas sroom environment communication is limited due to the fixed number of interlocutors and the availability of only one highly knowledgeable speaker. In the SA context there is greater potential for learners to interact with a variety of highly knowledgeable speakers both native and non native. Following the main tenets of these two hypotheses, it stands to reason that learners in a SA context would benefit from the additional input, interaction, and opportunity to produce output provided by their learning en vironment. The following chapter continues to explore the SA context and the possible impact it may have on improving learner fluency in the L2. First, a discussion of the definition of fluency is presented; and second, the manner in which fluency can be m easured is provided. In addition, the next chapter considers prior research on oral proficiency and the possible influence of the SA environment
31 CHAPTER 2 PRIOR RESEARCH IN FLUENCY Introduction In this section, the definition of fluency and the ways in which fluency can be measured are discussed. Next, the findings of prior fluency research on learning contexts ( study abroad ( SA ) and other types of immersion) on fluency are examined. This is followed by a discussion of SA research that has investigated o ral proficiency, in addition to fluency, and the possible impact that contact with the second language ( L2 ) has on these gains. While oral proficiency is not considered to be the same as fluency in the current study previous research has often intertwined the two concepts and therefore is considered here. Investigating Fluency U cy a fluent style of speech have been a major foc us of study in the field of language acquisition. As has been previously mentione d, there is also a general belief that SA will help learners become fluent in the L2. Following this idea, fluency research has investigated the impacts of learning contexts on oral fluency to discern how different experiences with the target language can In this section, first, the definition of fluency is explored this followed by a discussion of previous research into how fluency may be measured and then the findings of research on t he impact of learning context on fluency are presented Defining F luency Perhaps the most difficult aspect of studying fluency is the complexity in defining
32 definitio ns can be found; yet, there does not appear to be a consensus that one definition is preferable over another. Rather, it becomes apparent that the concept of fluency involve s many features, for instance temporal and phonetic features, language proficiency, or automaticity. This plethora of features results in definitions of fluency that are conceptually distinct from one another. Koponen and Riggenbach (2000) argue that d ifference in ideas Lennon (1990, 2000) continues with the notion that there is not one is based on the perception of the listener, and may be perceived differently by different listeners. As a result of this plasticity of meaning rather than focus on the search for one single definition of fluency, it i s necessary to consider the idea that there may be several types of fluency; furthermore, when researching fluency it is necessary to clearly state the idea that is being explored. One possible definition that takes a more general aspect states that fluency movement and being defined with t (Koponen and Riggenbach, 2000). T here is frequently a sense of ease and automaticity that comes with fluent speech, indicating that there does not appear to be a great pressure or struggle to produce langu age (Crystal and Varley, 1993; Koponen and Riggenbach, 2000; Taguchi, 2008). This definition also alludes to the belief that fluid From this general description, researchers have taken a more technical approach and define fluency by temporal, phonetic, and acoustic features, such as rate of speech,
33 number of silent pauses, number of filled pauses ( hesitations that consist of a non word sound or spe ech drawls (lengthening of a syllable) This definition of fluency is frequently applied to L2 language speakers, and many aspects of this meaning are often employed in judging L2 proficiency. For example, the American Council on the Teaching of Foreign La (OPI) utilizes similar defining guidelines for assessing fluency as part of overall oral proficiency (ACTFL, 1999; Koponen and Riggenbach, 2000). Looking at fluency from this point of view also allows researchers to collect data in order to quantify fluency based on a pre determined set of features, a method that is used frequently in fluency studies Furthermore, this definition suggests that there are abilities that can be learned and practiced to improve fluenc y (Koponen and Riggenbach, 2000), a notion that is appealing both to language learners and instructors. However, several researchers argue that this temporal definition is an oversimplification of the concept of fluency because it lacks a sense of the mor e sophisticated elements of speech that demonstrate language proficiency (Lennon 1990, 2000; Koponen and Riggenbach, 2000). Lennon re names the temporal definition is another more encompassing definition of fluency, which delves further into a well a speaker employs his/her linguistic and pragmatic knowledge of the language, in conjunct ion with temporal features, constitutes fluency (Koponen and Riggenbach, 2000). Nonetheless, this definition remains ambiguous in that it is unclear if proficien cy
34 is only a part of fluency or if fluency can be used as a synonym for proficiency. Moreover, the features of this type of fluency appear endless and highly variable, for example Wenerstrom (2000) considers intonation to be a key feature of fluency, while Bavelas (2000) considers non verbal aspects that convey meaning (such as a hand gesture to indicate direction) an important feature. A third definition of fluency considers the processing necessary to produce language and equates fluency with automaticit y. Under this meaning, highly fluent speakers possess the ability to automatically produce syntactical and lexical forms whereas lower level speakers must resort to the slower process of applying rules to produce language (Koponen and Riggenbach, 2000; Len non, 1990, 2000; Pawley and Syder, 2000; Segalowitz, 2000; Taguchi, 2008). This definition is more in line with what what constrain fluency rather than a lack of linguistic kn owledge. In this case, being able to access that knowledge quickly allows a speaker to more rapidly connect ideas together and produce fluent speech. Segalowitz (2000) further explores this definition by stating that there is cognitive fluency and performa act that this definition probes deeper into what may cause the differences seen in temporal features, from a research standpoint, it is difficult to collect data that would allow for knowing whether or not dysfluencies are a result of processing/cognitive mechanisms or of insufficient linguistic knowledge.
35 Returning to the idea of different types of fluency, Fillmore (1979, 2000) introduces (2000, p. 51). Fillmore offers the example of a radio disc jockey or sports announcer who must continue to fill air time with his/her voice. This type of speech lends itself more towards exposition and should be accessible to a variety of listeners. This i s suggestive of/similar to the first definition of fluency mentioned above which considers what a speaker is able to produce in a given time talk in coherent, reas oned, and semantically dense skill takes speaking a step further in that it becomes not only important to be able to produce language but that language should be logical, using a variety of vocabulary. In essence there is a focus on the quality of the speech rather than the quantity. Fillmore gives the example of William Buckley or Noam Chomsky for this type of speaker. A modern example would be Secretary of State Hillary Clinton who must convey her message in a rational and articulate manner while being aware of a limited speaking time. The third (2000, p. 51) and gives the example of Barbara Walters who has had the experience of interviewing an array of public figures from politicians such as Fidel ability s ome people have to be creative and imaginative in their language use, to express ideas in novel ways, to pun, to make up jokes, to attend to the sound
36 (2000, p. 51). T his type of speaker represents an ultimate level of fluency, a level that might not be achieved by every speaker. Fillmore gives two different examples : the comedian Mort Sahl, and the linguist Joshua Whatmough. A modern example would be the comedian Steph en Colbert, who is frequently creative with language, inventing new words and employing different styles of register. According to Fillmore, a speaker who (2000, p. 52). native language, his four styles of fluency can also be applied to L2 learners. As was mentioned previously, these four types of fluency can represent a continuum, similar to learning stages; as a learner progresses in his/her L2, s/he can reach a new type of fluency. A beginning or low intermediate level learner may only be able to master the first type of fluency, the ability to fill time with their L2 speech. With continu ed study of the L2, a learner can acquire more sophisticated syntactic structures and vocabulary and thus s/he may be able to progress to the higher quality second type of fluency. Asking w hether or not a L2 learner may be able to reach the highest type of fluency is akin to wondering if a L2 learner may ever reach native like language use. As language learners, teachers, and linguists we would like to believe that it is possible with a sufficient amount of study and linguistic knowledge but many would ackn owledge that it is entirely possible to be a competent L2 speaker without ever reaching this level of fluency. Overall, the lack of a single definition of fluency is not necessarily problematic when investigating fluency. It allows the researcher the free dom of investigating a
37 myriad of facets of fluency and to scrutinize this linguistic feature from numerous points of view. Nonetheless, a careful researcher should elucidate which definition of fluency s/he is exploring, and explain why that definition was chosen, in order for others in the field to fully understand and appreciate the results of the study. This being stated, for the current study, fluency is evaluated here similar to lowest level and is defined as the ability to produce a constan t rate of speech that is free from dysfluencies such as silent pauses (longer than 0.4 ms), filled pauses ( um uh ), repetitions, first language ( L1 ) words, self corrections, and clusters of dysfluencies. Measuring F luency In order to determine what may or may not lead to progress in L2 fluency, researchers need a way in which to measure fluency. As with other second language acquisition ( SLA ) research methods, it is important that these measures c an be used again in other studies and with a variety of part icipant populations. It is not surprising that the majority of the measures used to evaluate fluency are based on temporal, phonetic, and acoustic features; these features are easy to obtain in recorded speech samples and can be quantified, allowing for st atistical analysis. Of the many possible features that can be measured, five have been validated repeatedly in previous empirical research. These five features are: speech rate, mean length of fluent runs, phonation time ratio, average length of paus es, an d amount of filled pauses; each is discussed in turn below. Speech rate refers to the quantity of speech produced in a given amount of time. This is most often calculated as non repeated words per minute or non repeated syllables per second. Since repetiti ons are seen as dysfluencies it is important here to distinguish between non repeated and repeated words/syllables. In their studies on
38 fluency measures, Raupach (1987), Lennon (1990), Riggenbach (1991), Freed (1995, 2000), Towell et al. (1996), Ejezenberg (2000), and Kormos and Dnes (2004) all find that speech rate is an indicator of fluency both when comparing highly fluent speakers results. When calculating the mean leng th of fluent runs, researchers are looking for groups of words that are uninterrupted by dysfluencies. For example in the following pause] uh mi familia y yo [2 secon fluent runs have been identified the mean length of words (or syllables) is calculated. In a study of L2 learners and the development of fluency, Raupach (1987) discovered that more fluent speakers produce d longer fluent runs and that as learners developed in fluency there were capable of lengthening their fluent runs. Kormos and Dnes (2004) found that the advanced learners in their study had a significantly higher mean length of fluent runs than the low i ntermediate learners. Additionally, mean length of fluent runs strongly correlated with native speaker ( NS ) and non native speaker ( NNS ) scores for fluency. Towell et al. (1996) discovered that in a study on the impact of SA on fluency changes over time (from pre to post program) were seen for speech rate and were primarily due to the increase in the mean length of fluent runs, making it a core aspect of measuring fluency. W hen compared to the mean lengt h of fluent runs in the L1, the L2 mean is still significantly lower, indicating that additional L2 study/experience is needed to reach target like fluency. Lennon (1990) found that for three of his four participants mean length of fluent
39 run was a signifi cant factor showing improvement in fluency. Although Ejzenberg (2000) did not consider the mean length of all fluent runs, she made a similar finding with speech continuity and automaticity. Phonation time ratio is defined as the percentage of time spent speaking as a percentage proportion of the time taken to produce the speech sample (Raupach, 1987; Towell et al., 1996; Kormos and Dnes, 2004). Consequently in h ighly fluent speakers the phonation time ratio will be higher because these speakers are able to fill more time with speech rather than dysfluencies. Raupach found that as fluency increased (as measured by speech rate and length of fluent runs) phonation t ime ratio also increased. As was seen with speech rate and mean length of fluent runs, Kormos and Dnes discovered that phonation time ratio was significantly greater for advanced speakers over low intermediate speakers, as well as strongly correlating wit for fluency. Similarly for Towell et al., phonation time ratio is found to increase over time but once again when compared to the L1 ratio, the L2 results do not reach the same level of fluency. Calculating the average length of pause s is of interest in that speakers who pause for longer periods of time are perceived as less fluent (Kormos and Dnes, 2004) In the L1 a natural pause while speaking has been defined as a micropause that can only be measured by acoustic analysis. As a mea surement, micropauses have been defined as lasti ng less than 0.2 seconds (Kormos and Dnes, 2004), or less than 0.4 seconds (Freed, 1995, 2000). Pauses which are longer than these measurements can be considered dysfluent. Kormos and Dnes (2004) and Towell et al. (1996) found that the
40 average length of pauses is a significant factor in fluency, indicating that more fluent results showed that average pause length decreased sig nificantly over time for three of his four participants. However, average pause length may not be as strong an indicator of fluency as other features. Kormos and Dnes were not able to show correlations th as they were with speech rate, mean length of fluent runs, and phonation time ratio. Likewise other studies, such as Freed (1995, 2000), have not been able to show the same significant results suggesting that average pause length may not be a salient f eature for fluency judges. The number of filled pauses in a speech sample has also been shown to be an indication of oral fluency. Filled pauses are most often defined as any sound that does not constitute a word or a morpheme in the language, for example considered to be filled pauses. Similar to unfilled pauses, filled pauses allow the speaker to hesitate while speaking. It stands to reason that a more fluent speaker will produce fewer filled pauses. Lennon (1990) and Riggenbach (1991) both found that the number of filled pauses produced was a viable measure of fluency, in that less fluent speakers produced more filled pauses than those speakers who were found to be more fluent. However, as with phonation time ratio as well, this result has not always been duplicated (for example, Freed, 1995; Towel et al., 1996). These five measures have become an important set of measurements for fluency studies along with other similar temporal measures. In the current study, three of the five measures were used to evaluate fluency: speech rate, average length of fluent
41 runs, and amount of filled pauses. (Other fluency measures were also used and are discussed i n Chapter 3.) Speech rate and average length of fluent runs were chosen due to their successful validations in the existing research, and because it is believed by the researcher that these measures are most easily noticed by listeners. Furthermore, the am ount of filled pauses was also thought to be a salient feature of learner speech and is a factor that marks the speech as foreign sounding or non native. Learning C ontext In exploring outside element s that may impact learner L2 fluency, learning context h as been investigated as an important factor S tudies have primarily considered immersion contexts (SA or second language learning) and compared them to the traditional foreign language classroom. Anecdotal evidence from learners, instructors, SA program di rectors, and others involved with language learning point to immersion as a key step in achieving fluency. It is the goal of this research to determine if and why that may be true. Two earlier studies that investigated the influence of learning context on L2 fluency Lennon (19 90) and Freed (1995) utilized both NS judges and quantitative analysis of temporal factors. Lennon investigated the oral fluency of four German students learning English as a L2. Oral data were collected by means of a picture story ret elling task and were collected using a pre and post program design The nine judges for this study were instructed to attempt to distinguish between pre and post program recording and to determine if the post program recordings were more fluent. In addit ion to the judging, Lennon also performed a quantitative analysis of the following fluency features: unpruned words (all words) per minute, pruned words (excluding self corrected words) per minute, repetitions per T unit self corrections per T unit fille d pauses per T unit
42 percentage of repetitions and self corrections, total unfilled pause time, total filled pause time, mean length of speech runs, percentage of T unit s followed by pause (filled and unfilled), percentage of total pause time (filled and u nfilled) at T unit boundaries, and mean pause time (filled and unfilled) at T unit boundaries. Results showed that judges correctly identified the post program recordings and judged them to be more fluent. T he quantitative analysis revealed that the four p articipants significantly increased in speech rate and had significantly fewer filled pauses in their second recordings. Additionally, three of the four participants were able to significantly increase the percentage of T unit s following a pause; the fourt h maintained the same percentage. O n five other factors (unpruned words per minute, repetitions per T unit total unfilled pause time, mean length of speech runs, and mean pause time at T unit boundaries) three of the participants showed improvement toward s target like fluency ranging from intermediate to advanced level studying French for one semester; they were divided into an at home ( AH ) group and a SA group. Oral data were collected from learners us ing the ACTFL OPI, and s ix NS judges used a Likert scale to rate each learner Although a definition of fluency was not provided l inguistic analyses consider ed the amount of speech, the rate of speech, unfilled pauses, filled pa uses, the length of fluent speech runs, repairs made, and clusters of dysfluencies. Overall, there was no significant difference in the fluency measurements of the SA and AH group s over time ; nor was there a difference between the two groups. Upon this fin ding, Freed decided to reassess the results by eliminating the advanced learners from her study based on the
43 is, it was shown that SA students received higher scores of fluency at the end of the semester than the AH students Learners from the SA group had longer and faster speech that contained fewer dysfluent silent and non lexical filled pauses These results were similar to the Some researchers may criticize Freed for eliminating the advanced learners from her comparison, but, by doing so, she allows for the possibility that advanced students may not benefit from SA in terms of fluency at least as it is measured here It is possible that these learners had already achieved close to native like fluency in the elements measured in this study through their previous extensive study. A different definition of fluency and/or other me asures may have shown the areas in which the advanced learners still needed improvement. For both studies the SA environment appeared to have a positive impact on the learners since they were able to show some type of significant improvement. However, in comparable control group of traditional learners is a limitation in that it can not be known to what extent the SA environment was a factor in these results. Although unlikely, it is possible that the additional study was suffic ient for the learners to make these improvements. advantage for the SA group over the AH group for the lower level learners, however as learners were more advanced no significant differences were seen. While it is unkno wn for these studies the amount or type of contact the learners had with native speakers during their time abroad the SA environment appears to have provided additional opportunities for the learners to have contact with NSs and the L2 resulting in their improved fluency. Nonetheless, details into the extent and type of contact would further
44 reinforce these beliefs. The results of the NS judges appears to have followed the analysis of the temporal features, but as the instructions to the judges was somewha t vague with a lack of a fluency definition or a set of criteria with which to evaluate fluency, it is difficult to know if the judges were considering the same or similar features of speech. This limitation reflects the possibility of several definitions of fluency and may result in inconsistencies Following her earlier study, Freed continued to explore the effects of SA with several colleagues using the method of analysis of temporal fluency features to measure and identify changes in oral fluency but w ithout the use of NS judges as a comparison point Freed, So and Lazar (2003), Freed, Segalowitz and Dewey (2004), Segalowitz and Freed (2004), Segalowitz, Freed, Collentine, Lafford, Lazar and Daz and Collent ine (2007 ) investigated university students in semester long (14 16 weeks) SA programs studying either French or Spanish (M any of these studies us the same participant pool ) All students were considered to be intermediate to advanced level. As with Freed (1995), data for these studies were collected from pre and post program OPIs. These investigations have shown similar results to Freed (1995) and Lennon (1990) in that after participating in a SA program for one semester learners generally demonstrated lo nger fluent speech runs, had a faster rate of speech, and had fewer pauses (both filled and unfilled). As was mentioned previously, for Freed, these results were significant for lower level learners only, although more advanced learners were also able to s how improvement (albeit non significant); nonetheless in the se more recent studies no such distinction was seen between the learner levels.
45 In addition to the above mentioned features, Freed et al. (2003) also found that overall the SA learners spoke more in their post program interviews, indicating that the SA experience helped the learners to be able to be more descriptive and offer more information when speaking. It is also possible that the participants were more interested in the topics of the second OPI which would have led to their additional speech. A s was seen with Lennon (1990) and Freed (1995) it is difficult to determine how strongly the SA environment impacted learner fluency for Freed et al. (2003) without knowing how frequently learners inte racted with NSs. In studies where this was considered (Segalowitz and Freed, 2004; Segalowitz et al., 2004; and Freed et al., 2004), the increased L2 contact provided by the SA context showed a positive effect for fluency. Segalowitz and Freed only showed a slight effect for the impact of NS and learner interaction, whereas Segalowitz et al. and Freed et al. were able to find a stronger connection between the two in that learners with more contact made greater improvements. As discussed by Swain (2000) the pressure to produce understandable output in a timely manner while interacting with NSs or more advanced NNSs most likely assisted these learners in improving their fluency. In the case of Freed, Segalowitz and Dewey, this study also considered learners i n a n intense immersion context (IM) at the home institution where learners participated in both courses and out of classroom amount of L2 contact during their programs, t he IM learners reported having a higher level of contact than those in the SA context who tended to spend time using their L1 with other students. Their results showed that although the SA learners made improvement s in length of turns, speech rate, filled pauses, repetitions, repairs, and
46 mean length of fluent runs the IM learners had greater significant gains in all of these elements along with a larger amount of total words post program Since both groups had a similar amount of classroom based instructi on, it appears that the increased L2 contact enabled the IM learners to make that additional progress in fluency. In their reported L2 contact activities, the IM learners indicated that they spent more time producing output both in speaking and writing. Th is additional output appears to have had a positive influence o n oral fluency. Perhaps by having the additional time to practice and manipulate various structures in an environment of lower time pressure, this assisted the IM learners in being able to more efficiently produce these structures during their post program interviews. Also showing a positive effect for L2 contact on fluency, Isabelli Garca (2003) found that participants who reported having more frequent and more meaningful contact with NSs were able to significantly improve in temporal fluency measures over those with less reported contact. The participants in this study were three students starting out in the intermediate level. In order to measure fluency, Isabelli Garca assessed number of wo rds per response, number of pauses per utterance, and number of times the learner struggled with the language (struggling with the language was defined as having difficulty deciding between forms and as asking the interviewer for a lexical item). Results s howed that learners increased in their average number of words per response, as well as decreased their pauses and the number of times they struggled with the language. Further analysis showed that the two learners who had more interaction with NSs were fo und to make more progress on fluency than the learner with limited NS interaction. The learner who showed limited change in fluency reported spending more
47 time with other L1 speakers and did not develop personal relationships with NSs. Once again following Gass (1997), as with the learners from Segalowitz and Freed (2004), Segalowitz et al. (2004), and Freed et al. (2004), the learners in this study benefited from the frequent and in depth interaction with NSs providing them with practice, input, and feedba ck that enabled them to reach a higher level of fluency. Nonetheless, as was mentioned for Lennon (1990), the small number of participants in this study weakens the results somewhat in that it is difficult to know whether or not the same results would be seen with a larger participant group. This may be especially true if all participants are part of the same study group at the same location abroad. As can be seen with larger participant populations in the SA context (for example as in Magnan and Back, 200 7), learners have a tendency to cluster together and avoid forming deep personal relationships with NSs. In some ways it may be beneficial for learners to be part of a smaller learner group because they may be more willing to branch out from their safety z one. In a two year study of beginning level English as a second language ( ESL ) learners in Canada, Derwing, Munro, and Thomson (2007) show ed similar results to the previously mentioned works Participants were sixteen Mandarin language speakers (MAS) and s ixteen Russian and Ukrainian (Slavic) language speakers (SLS). T hree data sets were analyzed, Time 2 : after 3 to 5 months of study ; Time 6 : 8 months after Time 2 ; and Time 7 : a year after Time 6. Data were collected by means of a story telling speaking tas k, an exposure to English questionnaire, and an interview (at Time 7 only) about their experience learning and using English in Canada. The speaking task was evaluated by 33 native English speaker raters using 7 point Likert scales. The raters
48 were instruc ted to judge fluency based on speech rate, filled pauses, self corrections, silent pauses, and the overall flow of speech. They were also instructed to disregard grammar and vocabulary. While these judges were provided with more direction than in previous investigations (Lennon, 1990; Freed, 1995), there is still the question of the judges being truly able to ignore grammar and vocabulary as these types of errors are often salient. Results showed that while initially there was no difference between the two groups, only the SLS learners improved significantly over time. It was also revealed that SLS learners spent more time interacting in English with NSs and NNSs as well as spending more time listening to English talk radio ; t his additional L2 contact appea rs to have positively impacted their fluency. T he MAS learners reported difficulties in interacting with NSs based on a lack of common interests. It appears that here perhaps more investigation into the differences of the L1s and the L1 cultures would have been constructive in helping to explain why the SLS learners were able to make connections with the NSs while the MAS learners were not. In addition to these results, the MAS learners also gave the impression of being less confident in their language abil ities and less motivated to seek out interaction with NSs (again, likely due to their inability or unwillingness to connect to the NS culture). These findings relate to the Input and Interaction and the Output H ypothes e s in that through interaction and th e opportunity to learn from NSs, the SLS learners appear to have benefited from their additional L2 contact allowing for an increase in fluency Collecting data over a three year period, Towell, Hawkins and Bazergui (1996) and Towell (2002) investigated t he changes in oral fluency over time for twelve intermediate level university students of French and found similar results to those of Freed and her
49 colleagues. Data were collected as speech recordings by means of a story completion task made during all th ree years. Between the Year 2 and Year 3 recordings, participants spent six months studying in France. As well, in Year 3 participants recorded a personal adventure monologue. Data were analyzed quantitatively by measuring the following temporal variables: speech rate, mean length of fluent run, phonation/time ratio (a global measure of pausing), and average length of pauses. In samples were analyzed qualitatively to explore the differences in fluency. Resul ts showed that speaking rate and mean length of fluent run were significantly higher at Year 3 than at Year s 1 and 2. The average length of pauses showed no significant change over time. This indicate d that after participating in SA learners were able to m ake significant progress in fluency. When looking at the individuals themselves, a pattern was found showing learners whose scores were initially lower experienced the greatest improvement in fluency ; this finding is similar to results seen by Freed (1995) While these findings suggest that SA had an overall significant impact on oral fluency, the lack of an AH control group makes it difficult to generalize the results. T hese studies show that the greatest gains in fluency were seen after SA as no such chan ge was seen between the AH years. However, as was seen with Lennon (1990), it may be that the additional study not the SA setting, helped the learners to become more fluent. Several other studies consider temporal measures of fluency, but differ slightly in these measures. Llanes and Muoz (2009) included filled pauses per minute, silent pauses per minute, longest fluent run, syllables per minute, other language word ratio (code switching), and articulation rate. Juan Garau and Prez Vidal (2007) chose to
50 assess only speech rate, and considered words per clause and words per sentence as their fluency measures. Similarly, Garca Amaya (2008) decided to evaluate only syllables per second as a measure of fluency, and did not consider any dysfluencies. Taking a slightly different route, Trenchs Parera (2009) selected the dysfluencies and considered only unfilled pauses, self repetitions, lexicalized phrases, single words, drawls, vocalizations, and laughter. The majority of these studies show similar positive results to the above mentioned studies finding that SA appears to positively affect speech rate, fluent runs, and that it can reduce dysfluencies. Llanes and Muoz (2009) researched 24 L1 Spanish /Catalan learners of English participating in a short term ( three to four weeks) SA program. The participants had studied English for at least five years prior to this SA program, and of the 24, eleven of them had previously spent time abroad from as little as two weeks to as much as five months. Data were collecte d through an informal interview and a picture elicited story task. Llanes and Muoz discovered that learners demonstrated statistically significant differences in the measures of syllables per minute, other language word ratio (code switching), articulatio n rate, and longest fluent run. Including the measure of other language use is a newer addition to the study of fluency that can be important. Although some speakers and communities may be accustomed to code switching, for many NSs it can be confusing and dysfluent when they do not speak the other language. Furthermore, code switching may be more noticeable when learners do it because it can be seen as an error or a lack of vocabulary, ra ther than a natural phenomenon. While bilingual and NSs may code switc h due to cultural and heritage reasons, learners likely switch to their L1 in attempt to resolve a deficiency in their
51 speech. What is surprising in this study is that Llanes and Muoz did not choose to look at two groups of learners separately throughout their study : those with prior extensive time abroad and those without such experience. There is no indication that the total group was equal in their fluency abilities at the beginning of the programs and so it is possible that the participants with prior SA experience were more fluent than the others. Additionally, without considering two separate groups it is difficult to know the extent to which prior SA experience impacted the reported results. These researchers did run tests to determine predictors of fluency features and found that prior SA experience was a strong predictor of gains in length of fluent runs. Like Llanes and Muoz, Juan Garau and Prez Vidal (2007) were also able to demonstrate a positive influence of SA on speech rate for twelve L1 Sp anish/Catalan learners of English The participants from this study come from the larger Stay Abroad the impact of periods abroad on foreign language development, through c riterion 122). Data were collected by means of a pre and post program role play task. The results showed significantly more words per clause and words per sentence for lea rners at the end of their three month long SA program. In addition, the researchers found that the length and number of main and coordinate clauses increased in post SA speech. These reported findings are similar to the previously discussed results from ot her studies that also found that learners spoke fasters and had longer strings of uninterrupted speech, as well as overall longer speech.
52 Garca Amaya (2008) also concluded that learners who have studied abroad have a faster rate of speech. Participants in this study were 20 L1 English learners of Spanish divided into four groups: AH Group 1 (6 years of study), AH Group 2 (8 years of study), IM Group (6 years of study), and SA Group (8 years of study). The two AH groups had never participated in SA, the IM group participated in a seven week program, and the SA group in a ten month program. Data were collected by means of a sociolinguistic interview. I t is difficult to truly compare the study of Garca Amaya with other research in that the methodologies do no t follow the same design. Rather than using a pre and post test protocol, Garca moment in time. Without knowing their rate of speech prior to their SA program, it is impossible to know if the SA students have faster speech as a result of their experience abroad. Furthermore, simply based on the variety of years of previous study it is possible that there was a variety of fluency levels. For two of these studies there is a lack of a comparison AH group as was seen with Lennon (1990), Towell et al. (1996), and Towell (2002), making it difficult to attribute the fluency gains entirely to the SA context. For the Garca Amaya study although there were AH groups used, the lack of a time frame for the st udy makes it somewhat unclear how the groups were being compared. In all three of the studies there are no measures of L2 contact during the SA period which does not allow for additional insight into why these improvements were seen. The findings of Tren chs Parera (2009 ) are somewhat conflicting to previous studies Results showed that in the pre significantly lower than at the end of their three month SA program. Nevertheless when
53 looking at the data coll ection overall, an interesting pattern emerges. Three d ata sets were collected : at Time 1 (T1) ( pre university formal instruction ) Time 2 (T2) ( post university formal instruction ) and Time 3 (T3) ( post SA program ) Trenchs Parera finds the lowest frequen cy of dysfluencies at T1 followed by a dramatic increase at T2 and then a decrease by T3, re approaching more target like fluency. The researcher postulates that as l earners progressed they may have become more conscious of producing grammatical speech I n order to achieve this, they make use of dysfluencies to allow for sufficient time to formulate their output. It appears that the time these learners spent abroad helped them make a shift back towards more target like behavior. The increased interaction wi th NSs may have allowed them be more focused on producing fluent and timely output than having perfectly formed sentences. It would be prudent here to analyze learner speech from a grammatical standpoint to see if that is in fact the case, and/or to ask le arners themselves if they were more or less conscious of their speech in certain settings. Trenchs Parera further speculates that the level of the learners in the study may also be a reason for the difference seen between this and other research (Trenchs P studies). While this is certainly a possibility, it is important to note that the learners from Trenchs Garau and Prez learners and are cons idered to be at the same starting advanced level. (Advanced is not clearly defined here by any particular standards or guidelines. All participants had passed a university L2 language test and thus were determined to be advanced.) Therefore a similar patte rn could be expected for both studies, yet this is not the case. What can be said here with comparison to the other investigations is that, overall, SA
54 appears to have had a positive influence re gain aspects of more fluent speech after their semester abroad. In a rare study that does not use temporal analysis of fluency, Allen and Herron (2003) investigated the effects of short term SA on oral fluency To evaluate fluency of L2 French university students in t heir study, Allen and These judges were two non native speakers with native like fluency who rated fluency as part of pre and post oral skills tests. There is no explanation offered as to the guidelines used for rating. These j udges found significant improvements in fluency for all students at the end of their six week SA program. Likewise they found significant communication, features which are reminiscent of the fluency features outlined above. Given that these students participated in a shorter SA program, the comparable results to those of other studies are somewhat promising, indicating that even in shorter time periods learners can perhaps make signifi cant improvement in fluency. Nonetheless, a clear definition of what is meant by fluency in this study would assist others in better understanding the extent of the improvement seen by these learners and allow for future researchers to compare their result s to those found by Allen and Herron The majority of the conclusions discussed above for the effects of learning context show that after participating in an immersion program, learners are found to be more fluent in their L2 as judged by NS and as measur ed by temporal fluency features. In sum, learners are able to show faster rates of speech, fewer hesitations and dysfluencies, and in many cases produce longer fluent strings of speech. However, there is some evidence (Trenchs Parera, 2009) that this may n ot always be the result;
55 possibly as learners increase in their proficiency level and are more aware of grammatical features they may sacrifice fluency for grammatical accuracy. Further investigation into a variety of learners and programs is needed to hel p researchers better understand which situations favor increased fluency and which may hinder fluent speech. These results can be better understood when L2 contact during the SA experience is considered. As several studies have been able to demonstrate, le arners who were able to have more frequent and meaningful contact in the L2 were able to show more improvement in fluency. Following the Input and Interaction and Output Hypotheses, the increased interaction and communication with NSs and/or advanced NNSs dysfluencies, and in greater quantities thus overall leading toward more native sounding L2 speech. Study Abroad and Oral Proficiency Level In addition to research into SA and oral f luency, several studies have investigated oral proficiency level. The differences between proficiency level and fluency may be rather vague at times, especially when some definitions of fluency include proficiency and conversely, some definitions of profic iency include fluency. As a way to describe guidelines to speaking proficiency (1999). These guidelines divide learners into the ten proficiency levels based on their ability to perform certain speech acts, use L2 grammar, use correct L2 pronunciation, speak about a variety of topics, and in general make themselves comprehensible to native speakers (NSs). Freed (1995), Isabelli Garca (2003), Segalowitz, Freed, Collenti ne, Lafford, Lazar, and Daz Campos (2004), Segalowitz and Freed (2004), Magnan and Back (2007), and Llanes and Muoz (2009)
56 SA program. These SA programs took p lace over one typical university semester ranging from 13 to 15 weeks long, with the exception of Llanes and Muoz who researched a four week summer program. These researchers find that post program, many, but not all, students were able to increase their proficiency by at least one level; likewise, some learners were able to increase more than one level. The non increasing learners were able to maintain their levels and in no cases did learners decrease in their proficiency level. The reasons why learners were or were not able to increase in proficiency level are not clearly seen across all of these studies. Freed (1995) and Llanes and Muoz (2009) discovered that lower level learners were able to increase significantly, whereas more advanced learners, des pite showing improvements, could not. This finding is in that more advanced learners may already be at a very high level of oral competency and so there was less opportuni ty for significant improvements to be made. They also suggest that SA may not be enough to propel these learners into higher levels and that another type of study may be necessary. Lower level learners however had a much greater opportunity for significant improvements to be made and it appears that the SA experience was beneficial in bringing about that increase in proficiency level. In her study, Freed also considered AH learners and when evaluating the lower level learners, a significant difference was s een between the two groups in favor of the SA learners. This finding supports the idea that SA can more greatly impact oral proficiency than traditional AH study. Nonetheless these two studies lack information concerning how
57 the SA environment impacted the changes that were seen, in that there is no measure of the amount of L2 contact while abroad. It would be useful to know the amount and type of interaction these learners had with NSs to help determine how much the context may have affected these learners Particularly in the case of Llanes and Muoz, it is important to know how much L2 contact learners were able to have during their shorter four week program and how that enabled them to achieve similar results to learners who spent a greater amount of tim e abroad. In general, for these two studies it appears that context was beneficial, but to what extent is unknown. Magnan and Back (2007), on the other hand, f ou nd that more advanced learners were able to show significant increases in proficiency while low er level learners tended to maintain their level. With further examination the investigators determined that more advanced learners had more frequent and longer communication with NSs. This result s were able to show significant improvement in their oral abilities after extensive interaction in the L2. Magnan and Back further speculated that these learners were able to forge more and stronger social connections due to their ability to have more comp lex and meaningful conversations. spent more time speaking their L2 (French in this study) with other students from their same L1 (English in this study) showed the least amount of improvement in proficiency. It seems that even though these learners were producing output and receiving input, the quality of this input was insufficien t to effect proficiency change; in order for learning to take place it appears that a learne r needs to have interaction with a more knowledgeable interlocutor who will help the learner reach higher ability levels. In the
58 instances of two L1 English students conversing in French, these data showed that even if one of the students were to be at a s lightly higher proficiency level, this would not be enough to allow the learners to reach a higher level of performance as they might with the help of a more advanced or expert speaker. In fact, Magnan and Back addressed this when comparing the American Am erican relationships to those between the American learners and more fluent immigrants to France. Unlike students who remained in mainly American social circles, students who had frequent social interaction with these immigrants made linguistic improvement during SA. Magnan and Black American learners] apprentice into the new community of practice they are trying to 56). Similar to Magnan and Back and parallel to her own fluency results, Isabelli Garca (2003) also found that participants who interacted more with NSs increased their proficiency levels. The participants in this study were three students starting out in the intermediate level. Of the two with extens ive NS interaction, one was able to increase two levels from intermediate low to intermediate high and the other increase one level from intermediate mid to intermediate high. As with fluency, the learner who showed no change in proficiency level was the l earner who spent more time with other L1 speakers and did not develop personal relationships with NSs. Contrasting from Magnan and Back and Isabelli Garca, Segalowitz and Freed (2004) were not able to fully show that learners who had more contact with th e L2 were class and out of
59 (p.192). Segalowitz and Freed do not completely rule out the importance of L 2 contact while abroad however, stating that the contact learners had during this SA program may have been insufficient to fully impact proficiency improvements. Their results show that the majority of learner to NS interaction was between the learners and their host families. They hypothesize that this type of interaction tended to be somewhat formulaic not challenged to go beyond what they already know and seek to test out their abilities then little progress can be made. Rather than L2 contact, these researchers discovered access processing speed and efficiency. In other words, lea rners who were initially faster and more efficient at recalling lexical items in the L2 were able to make significant improvements in proficiency. This finding further proposes that the learners who made significant improvement were more cognitively ready to gain from being in the SA context because they were better able to handle the demands of that context. Segalowitz et al. (2004) explain their findings in comparison to the explanations offered by Magnan and Back, Isabelli Garca, and Segalowitz and Free d, indicating that both cognitive abilities and L2 contact lead to increases in proficiency level. As with Segalowitz and Freed, they find that the L2 contact had less impact on proficiency level for these learners, however they suggest that L2 contact ca n be an important factor in advancing to a higher level of attainment. They state that: The more the adult learner is able to communicate in the target language the more he or she will do so. As a result the very act of communicating will further enhance learning, leading to more communication, which should promote furthe r learning (p. 14).
60 This statement is suggestive of the willingness of the learners to participate in the contact and communication in the L2 and how it may affect the linguistic gains mad e during a SA program. Overall, when reviewing previous research, SA can be shown to positively affect oral proficiency. Learners who actively participate in the SA experience and fully explore the opportunities to interact with NSs or more advanced L2 spe akers are more likely to increase in their proficiency, often increasing more than one level. These findings tie into the Input and Interaction and the Output Hypothes e s that cite the importance of frequent and meaningful interaction with NSs that allows l earners to learn through communication and challenges them to produce better and more proficient language. Conclusion As can be seen from the literature discussed above, the concept and definitions of fluency are endless due to the complexity of this lingu istic feature. While the general concept of fluency considers more temporal features and evokes a sense of smoothness and liquidity to the speech, other definitions delve deeper into a concept that encompasses grammaticality, as well as sociolinguistic and pragmatic knowledge. It has been suggested here that rather than view fluency as one definition, it can be seen as a continuum with different fluency levels not unlike proficiency levels. Although a single agreed upon definition of fluency may never be de termined, researchers should clearly state their own concept of fluency in order to provide others with a comprehensible starting point when reviewing the investigation. Due to the many elements that may influence L2 fluency, this construct can be studied from a myriad of viewpoints. In quantifying fluency, the five temporal measures of speech rate, mean length of fluent runs, phonation time ratio, average length of
61 pauses, and amount of filled pauses have been shown to effectively assess fluency. Other mea sures may also be indicative of fluent speech and it is the job of future studies to attempt to recreate positive results that will allow us to validate these additional assessments. In the case of speaking abilities, both oral proficiency and oral fluency the SA SA experience is only a few weeks. These findings support the ideas put forth by the Input and Interaction Hypothesis and the Output Hypothesis that indicate tha t language acquisition is driven by interaction with NSs. Again, due to the wealth of interaction possible in a SA environment it is likely that the learners were able to show such gains because of their increased contact with NSs. However there is a conti nued need to explore the reasons for this improvement and examine the differences between students who elect to go abroad and those who remain at home. The more researchers can discover about the type and frequency of learner NS interaction the better our understanding will be of the impact of SA on the acquisition process. Furthermore, it is important to remember that the majority of SA research has been done with longer term semester length programs, and it is therefore compulsory to continue to research shorter programs, especially as their popularity rises with students.
62 CHAPTER 3 PRIOR INVESTIGATION OF SECOND LANGUAGE WILLINGNESS TO COMMUNICATE Introduction As alluded to by Segalowitz et al. (2004), as learners have more success in their oral abilitie s in the second language ( L2 ) they will likely seek out more communication. This desire to seek out communication can be defined as Willingness to Communicate (WTC) in the L2. Taking into account both the influences of motivation and language anxiety, the concept of WTC was first developed from the field of communication psychology and was originally used to describe how willing people are to interact with one another in their first language (L1) (McCroskey and Richmond, 1985; 1990). Considered to be part of the personality construct, WTC referred to a trait that indicated how talkative a person would be when given the opportunity. Early studies, such as Sallinen Kuparinen, McCroskey and Richmond, (1991), focused on WTC differences between cultures and look ed for tendencies between other personality traits and WTC. In general, these L1 studies found that people who were more extraverted, exhibited less language anxiety, and were less shy overall, were shown to have higher WTC. Following this research, MacInt yre (1994) developed a model of WTC that hypothesized that WTC is primarily affected by perceived communicative competence and communicate to the extent that they are not apprehen sive about it and perceive 138). The model included the global personality traits of introversion and extraversion, indicating that introversion would lead to higher language anxiety and therefore less WTC, whereas extraversion would have the opposite result. Furthermore, introverts were
63 found to have lower levels of perceived competence leading to less WTC, while extraverts had higher levels of perceived competence leading to higher W TC. In addressing the limitations of this model, MacIntyre states that there is a lack of situation considered and notes that variables that can influence WTC may often come from external sources. As will be seen in L2 WTC research, these external sources and situations will be more greatly considered than personality traits. Development of a Second Language Willingness to Communicate M odel Following these L1 studies and considering the Socio Educational model of Gardner (1985) and the WTC model of MacInty re (1994), MacIntyre and Charos (1996) applied WTC to L2 contexts. Their main reason for expanding WTC to L2 research was use the language to develop proficiency, that is reasoning, learners must be willing to communicate in the L2 in order to improve their language abilities and acquire more native like speech. This concept resonates with other popular theoretical c onstructs of sec ond language acquisition ( SLA ) including the Input and Interaction Hypothesis (Gass, 1997) and the Output Hypothesis (Swain, 2000), as well as current classroom pedagogical principles (i.e. communicative language teaching), all of which take into account t he importance of communicating in the L2. In adapting the model to L2 contexts, seen in Figure 2 1 below, some changes were made to include context (the opportunities for L2 communication), motivation, and frequency of communication. According to the model context is predicted to influence both WTC and motivation; while WTC, together with integrativeness and attitudes towards the learning situation, influences motivation. Finally, WTC and motivation, together, lead to L2 communication frequency.
64 Figure 3 1 Model of Second Language Willingness to Communicate. MacInytre and Charos (1996) In order to test this L2 WTC model, MacIntyre and Charos collected data from Canadian L1 English adult learners of French. These participants were volunteers from adult ev ening courses. Data were collected by means of survey using a 7 point Likert scale which measured the following model elements: the Big Five personality traits (De Raad, 1992 including extraversion, agreeableness, conscientiousness, emotional stability, a nd intellect) ; frequency of L2 communication ; L2 WTC ; perceived competence in the L2 ; attitudes ; motivation ; and the amount of French present in the work and home
65 context. Results demonstrated that L2 frequency was indeed impacted by WTC and motivation alo ng with perceived communicative competence and context; specifically, learners who were more highly motivated and were more willing to communicate exhibited higher frequencies of L2 communication. Not surprisingly, having additional opportunities to commun icate in French also led to more L2 communication. Finally, having a high level of perceived competence was found to be the largest single effect for predicting the frequency of L2 communication. This suggests that if a learner believes that s/he has the n ecessary skills to communicate, s/he will be more likely to participate in L2 communication regardless of proficiency level. Looking at WTC alone, it was found that language anxiety, perceived competence, and context had direct effects. When learners had l ower language anxiety, higher perceived competence, and/or more opportunities for L2 use they displayed higher levels of WTC. In further developing a L2 WTC model, MacIntyre, Clment, Drnyei, and Noels (1998) developed a situational model. Breaking away f rom the more trait like models of L1 WTC, this newer model considered the effects of context on WTC and the likelihood of communication. Shaped like a pyramid, this model is divided into six layers and is composed of twelve variables as seen below in Figur e 3 2.
66 Figure 3 2. Heuristic Model of Variables Influencing Willingness to Communicate (MacIntyre, Clment, Drnyei, and Noels, 1998) Layer I, the top layer, represents Communication Behavior and considers L2 use of all kinds. Layer II represents Behavi oral Intention and is comprised of L2 WTC which the (p. 547). Layer III is Situa ted Antecedents and contains desire to communicate with a specific person and state communicative self confidence both of which demonstrat e the situation specific reasons that can lead to WTC. Layer IV is Motivational Propensities including interpersonal motivation, intergroup motivation, and self confidence; these variables include both internal and external motivation as in his/her L2 abilities. Layer V corresponds to Affective and Cognitive Context and contains intergroup attitudes, social situations, and communicative competence; the
67 variables in this level are both individual and broader based attitudes, and while they may be further from the specific situations that lead to L2 use, they are still important in affecting t hose specific situations. Finally, Layer VI is Societal and Individual Context and includes intergroup climate and personality; this layer represents communication through the basic interaction of society and the individual and it is from this interaction that the other variables are seen. In all, this model attempts to take both theory and practical considerations into account and offers a more complete depiction of the factors that lead to WTC and actual language use. W illingness to Communicate and the Ca nadian S tudies Approximately half of the existing L2 WTC studies thus far have been conducted at Canadian institutions by MacIntyre and his colleagues. The majority of these studies investigate L2 WTC and immersion programs in Canada; due to the bilingual English/French environment found in Canada there has been ample opportunity to study immersion programs and bilingual schools. These investigations have all focused on WTC and frequency of L2 use, and the variables that can affect them. All data were colle cted by surveying large groups of learners by means of questionnaires about WTC, language anxiety, motivation, frequency of language use, perceived competence, integrativeness, attitudes towards the learning situation, and other similar topics. MacIntyre, Baker, Clment, and Conrod (2001) and MacIntyre, Baker, Clment, and Donovan (2003b) both investigated L2 WTC of learners in a junior high school immersion setting, in both cases L1 English students learning French as an L2. MacIntyre et al. (2001) specifi cally researched types of orientations that may affect WTC both inside and outside of the classroom. Overall, WTC was higher inside the classroom than outside the classroom. Results showed that strong social L2
68 connections were necessary for a higher level of WTC. Their results further demonstrated that job related and travel orientations highly correlated with WTC inside the classroom, indicating that learners saw a need to practice communication skills in classroom activities in order to be successful wit h regards to careers and/or travel. One may ask how this differs from a simple motivation to learn. The key element here is that learners are recognizing that it is through the communication that they are able to make progress that will assist them in the future. It is not just that the learners value language learning and see it as a valuable tool, but that by practicing communication skills and being willing to use the L2 they will be able to flourish. It should be noted here, that this is a learner belie f and that the researchers did not test proficiency levels. Thus, the results here do not state that being more willing to use the L2 is correlated with making improvements in L2 abilities, but rather that the learners believe this to be true. Outside of t he classroom, school achievement orientation was found to highly correlate with WTC. This finding illuminates the possibility that learners may see authentic communication as a means to improve academically. MacIntyre et al. (2003b) looked at WTC, languag e anxiety, perceived competence, and motivation with additional effects of sex and age group (grades 7, 8, and 9). Their findings showed that, overall, the strongest correlate of L2 WTC was perceived competence. When considering effects of sex, a significa nt difference was seen only for anxiety in that males showed a constant level of anxiety whereas females showed a decrease in anxiety from grades 8 to 9. When considering age group, increases were seen in WTC, perceived competence, and frequency of communi cation from grades 7 to 8. Additionally a decrease in motivation was seen from grade 7 to 8. No significant
69 differences were seen between grades 8 and 9. It may be that there is less difference in learner proficiency level between these grades which may ha ve some effect on their WTC since perceived competence was such a strong predictor of WTC. Also, since data were collected cross sectionally it may be that the learners in grades 8 and 9 possessed more WTC overall. To gain a better understanding of WTC ov er time though, a longitudinal study that follows the same population of learners is necessary. In three other studies that have been completed on WTC, Baker and MacIntyre (2000), MacIntyre, Baker, Clment, and Donovan (2003a), and Clment, Baker, and MacI ntyre (2003) investigated the effects of learning environment on WTC. Baker and MacIntyre compared English L1 learners of French in an immersion program to learners in a non immersion program. Results identified a clear distinction between the two groups: immersion learners had higher levels of WTC, lower levels of anxiety, higher perceived competence, and more positive attitudes/motivation towards French and the French speaking community than did the non immersion learners. In looking at what factor influe nced WTC the most, results showed that anxiety impacted WTC more for immersion students, but that perceived competence impacted WTC more for non immersion students. When further exploring these results it was found that immersion students felt more pressur e to communicate well and felt anxious that they would not live up to expectations. MacIntyre et al. (2003a) looked at three groups of English L1 learners of French, immersion students, non immersion students, and students from an intensive summer immersi on program. Results demonstrated that learners from the immersion programs had higher levels of WTC, higher perceived competence, and higher frequency of L2
70 communication. T here was no significant difference in anxiety between the three groups. Further inv estigation into WTC by itself revealed that motivation was strongly correlated with WTC in the immersion groups but not for the non immersion students. In the intensive group there was a negative correlation between anxiety and WTC, indicating that learner s who are anxious will be less willing to communicate. It was also found that for both of the immersion groups anxiety was the leading predictor of WTC. It appears that these learners felt more pressured to be better communicators in the L2 due to the expe ctations of others. The driving factor for WTC for non immersion learners, on the other hand, was perceived competence, indicating that learners were more willing to participate in communication when they felt competent that they could be successful. In a study comparing learners at a bilingual university, Clment et al. (2003), researched both L1 location in Ottawa, a predominantly English speaking province, the Francophones were learning Englis h in a more second language environment whereas the Anglophones were learning French in a more foreign language environment. Results showed clear differences between the two groups. Francophones had higher levels of WTC, more frequent contact with their L2 higher L2 confidence, higher L2 subjective norms, and identified more with the L2 culture than did Anglophones. Further analysis revealed that for both groups of learners, frequency and quality of contact with the L2 group was a predictor of L2 confidenc e. In turn, L2 confidence was a predictor of WTC and identification with the L2 community. It was also seen that WTC and identification with the L2 community both led to more frequent and higher levels of L2 use. It appears that
71 due to the vast opportunity for communication in a second language environment, WTC may be higher than in a foreign language environment. Overall, the Canadian studies have shown that perceived competence and anxiety appear to be strongly correlated with WTC. It also seems that in s ome instances motivation is also strongly correlated with WTC; however, due to some conflicting results additional research is warranted. Finally, it appears that immersion environments tend to promote WTC as well as promoting perceived competence and lowe ring anxiety. It is important to note though, that in all of these studies the data for WTC, anxiety, motivation, frequency of communication, and other factors have been collected through questionnaires only As will be seen below in other WTC investigatio ns, other methods of data collection will offer a different and much needed perspective. Willingness to Communicate R esearch O utside of Canada Recently, other researchers have become interested in WTC and have expanded investigation to other language learn ing groups. A few researchers are continuing in the same vein as the Canadian studies and utilize a similar methodology of data collection by questionnaire from large groups of learners (Yashima, 2002; Yashima, Zenuk Nishide, and Shimizu, 2004; Yashima and Zenuk Nishide, 2008). However, others (Kang, 2005; Cao and Philp, 2006; Cao, 2006; Peng, 2007) are beginning to assess WTC with different methodology that intends to capture real communication data rather than self reported data only. The three studies th at follow the Canadian style research, Yashima (2002), Yashima et al. (2004), and Yashima and Zenuk Nishide (2008) investigated Japanese learners of English. Yashima collected data from university students and employed similar questionnaires measuring WTC, communication anxiety, motivation, and
72 perceived communicative competence. Additional information was obtained and included measures of intercultural friendship orientation, interest in foreign affairs, interest in international vocation/activities, and a pproach avoidance tendency (to were similar to Canadian findings, in that learners with higher WTC experienced lower levels of anxiety and perceived themselves to be mo re competent in English. Additionally, results showed that learners with more positive attitudes toward intercultural communication or who claimed to have more international interests were also more willing to communicate. Further analysis revealed that pe rceived competence was the strongest predictor of WTC. This outcome reflects the findings for non immersion students from the Canadian studies suggesting that similar formal language learning environments will find comparable results. Yashima et al. (2004 ) continued the research of Yashima (2002) by expanding the measures to include frequency of communication. In addition, Yashima et al. chose to investigate high school learners in a formal at home ( AH ) learning environment, as well as in a study abroad ( S A ) environment. Results for the AH learners showed that frequency of communication was relatively low and that learners rarely used English outside of class. Nonetheless, for learners who had a higher level of WTC, frequency of communication was greatest, indicating that they communicated more in the classroom and outside of class than learners with lower WTC. Once again, it was found that learners with higher perceived competence were more willing to communicate. Results for SA learners were similar to AH learners and confirmed that learners who had higher WTC more frequently participated in communication in the SA environment. It was
73 found that higher WTC led to a development of strong interpersonal relationships during SA which then led to more frequent c ommunication. Interestingly, Yashima et al. collected data both at the beginning and end of the SA program, yet they did not analyze their data for changes over time. Continuing with their research into the effects of learning context on WTC and proficien cy, Yashima and Zenuk Nishide (2008) once again investigated high school learners. These learners were 165 students enrolled in two classes : Course A which was communication focused and Course B which was grammar focused and was designed to help student s with university entrance exams. Out of this large group of participants, 16 participated in ten month SA programs in Australia, Canada, New Zealand, the United Kingdom, or the United States. Thus in this study there were three groups of participants: SA from Course A, AH from Course A, and AH from Course B. From a methodological standpoint the researchers continued to use questionnaires to assess WTC adding a frequency of communication component to their questionnaire, and used the Institutional Testing Programme Test of English as a Foreign Language ( ITP TOEFL ) exam to assess language proficiency. For this investigation, a pre and post test design was employed to determine changes over time. As in previous research, results revealed that learners from the SA group demonstrated significantly higher levels of WTC than both AH groups and this desire to communicate was maintained over time. Additionally, results showed that the AH students from Course A (the more communicative course) had significantly high er levels of WTC than the students in the more grammar centered course. When looking at their initial levels of WTC, it was also seen that the students with the highest levels of WTC
74 were the ones who chose to participate in SA; additionally, they were fol lowed in level of WTC by the students in Course A and then by the students in Course B. Therefore, the results suggest that level of WTC can determine the likelihood that students will choose to go abroad and to what extent they will choose to participate in content based courses. Linked to WTC, frequency of communication in English showed a similar trend with SA students having the highest frequency of communication followed by the AH Course A students, and then the AH Course B students. Over time, all gro ups increased in their frequency of communication but the trend between the three groups remained constant. When considering the results of the proficiency test, the SA learners made larger gains than both AH groups. When looking at the AH groups themselve s, an interesting finding showed that the Course A learners had higher proficiency than the Course B learners who were in a class specifically designed to help them succeed in taking standardized tests. Therefore these results are particularly important in that they appear to support the notion that learners with higher levels of WTC in the L2 are able to demonstrate higher levels of proficiency in the L2. At the time of the present study, this research by Yashima and Zenuk Nishide is the only empirical evi dence known to support this belief. In an attempt to research WTC from real communication data, the most recent studies have incorporated qualitative methods such as observations, recorded conversations, interviews, stimulated recalls, and teacher evaluati ons to supplement quantitative learner reported data. The main reasons for this more varied methodology are to allow for an outside view of learner WTC and to explore WTC in the moment, or
75 in other words, to better understand situational WTC. In a study th at considers qualitative data only, Kang (2005) explored WTC of four Korean learners of English at a US university. The participants in this study were participants in a conversation partner program where native speakers ( NSs ) and non native speakers (NNSs ) met to have face to face conversations with the goal of assisting the NNSs with their English. Data were collected by video recording the conversations for eight weeks, by a semi structured interview with each learner, and by stimulated recalls where lea rners watched the recorded conversations and commented on their WTC during the conversations. In analyzing the results, Kang uses her own terminology to define three psychological conditions that impacted WTC: security, excitement, and responsibility. Secu interlocutors well, when the number of interlocutors in a group increased (from two to three to four, etc.), when interlocutors appeared uninterested, and when other NNSs seemed to be more proficient in English. In these cases of weakened security, WTC to be influenced by familiarity and interest of topic, attractiveness of interlocutors (due to their language abilities and/or physical appearance), interlocutor interest, and conversational context. These instances promoted excitement in learners to keep introduced by the learners, when learners were more knowledgeable on a certain topi c,
76 when the number of interlocutors decreased in a conversation (for example, a group of six versus a pair), and when misunderstandings occurred due to a prior statement. On these occasions, learners felt responsible to communicate and to expand their thou ghts by offering additional examples or explanations. Although Kang uses her own terminology, her psychological conditions are similar to those of MacIntyre and his colleagues. Where MacIntyre et al. use anxiety and perceived competence ; Kang uses securit y Likewise, MacIntyre et al. use motivation and attitudes towards the learning environment ; while Kang uses excitement and responsibility pattern of behavior towards communication as t he participants from the various Canadian studies, i.e. learners who were less anxious, had a higher perceived competence, were more motivated, and had better attitudes toward the L2 were more willing to use their L2. While can be compared t o those of previous research, due to a difference in terminology it is impossible to say, for example, that security is entirely equal to anxiety. Two studies that also use qualitative measures are Cao and Philp (2006) and Cao (2006). Both studies collect ed data from the same population of eight learners of English at a New Zealand university. The learners had a variety of L1 backgrounds including Chinese, Korean, Japanese, and Swiss German, and were all enrolled in an intensive General English program. Ca o and Philp chose to compare trait WTC as obtained from self reported data and state WTC as obtained from observational data collected during in class communication. The self reported data were obtained through a questionnaire adapted from the Canadian stu dies; in class communication data were
77 acquired through eight classroom observations, six audio recordings of group and pair work, and an interview with each participant that included stimulated recall. Results found a discrepancy between the self reported WTC and the observed WTC in that learners reported themselves to be more willing to communicate than they actually appeared to be in live conversation. Cao and Philp attributed this finding to the difference between trait WTC and state WTC, indicating tha t learners may feel that they are willing to communicate in a given context but certain factors that arise during the situation may cause them to be less willing to communicate. It is possible too that in these situations learners were unable to communicat e because they did not have the necessary L2 knowledge. Further analysis of the data revealed specific situational communication/group size, familiarity with interlocutor(s), level of i participation in communication, and topic familiarity and interest. When considering context/group size, as participants worked in pairs or in small groups of three, high WTC was observed, whereas when group size increased (to four or more) W TC decreased. For the factor of familiarity, it was seen that when participants had stronger social connections to their interlocutors they were more willing to communicate than if their interlocutors were only acquaintances (in this case there was no inte raction between the participants and strangers since all observations were completed in class). The to high WTC, but if an interlocutor showed disinterest or their own unwillingness to participate in conversation then lower levels of WTC were seen. Finally, when participants were more interested and/or familiar with a topic they were more willing to
78 communicate than when they perceived the topic to be uninteresting or di fficult. These findings are similar to those of Kang (2005), albeit using different terminology, and suggest a strong influence of situation on WTC. Additionally, both groups of learners reported a high level of self confidence in the L2 on the questionnai re as well as reporting themselves to be at a higher level of WTC than was seen in the communication data. could be seen over time from the beginning of an intensive program to the end. Using the same qualitative data that were collected for Cao and Philp (2006), Cao discovered that, over the course of the program, there were changes in WTC during whole class communication, group work communication, and pair work communication. In each context, students had lower WTC at the beginning of the program and tended to have higher WTC as the program continued. Looking at data from the personal interviews, it is possible that this improvement is a result of increased familiarity with int erlocutors, as many learners mentioned feeling more willing to speak to others once they knew them better. It is also possible that as learners acquired more knowledge about English in the program, their self confidence increased, leading to a higher level of WTC. For some students, there were fluctuations seen with levels of WTC from week to week where a student may have shown a sharp decrease or increase in WTC from one week to the next. Cao ascribed this fluctuation to situational factors that strongly i mpacted WTC. The strongest indicator for this fluctuation was found to be task topic, in that if a learner was unfamiliar or uninterested in a topic, s/he was much less willing to communicate.
79 This finding is consistent with situational factors found in Ca o and (2005) studies. The final study discussed here which combines quantitative data with qualitative observations, is Peng (2007). Peng investigated Chinese learners of English and their WTC in a foreign language classroom setting. Ba sed on number of years of study and course placement guidelines, participants were at four separate proficiency levels: intermediate, high intermediate, low advanced, and advanced. Data were collected using an adaptation of the WTC questionnaire from MacIn tyre et al. (2001) and a group interview comprised of both high and low WTC learners. Results showed that overall, Chinese learners appeared to have low WTC in a classroom setting. Results also showed that females reported higher WTC than males, that as pr oficiency level increased so did WTC, and that learners with higher perceived competence at all levels had higher WTC. All of these findings support previous results seen in the Canadian studies. The group interview revealed that Chinese learners felt that the instructor should be the most influential factor for increasing WTC in the classroom context. Peng determined that for the majority of the participants, these learners desire to be passive and take few risks when communicating in English. Many learner s stated that they would feel odd initiating conversations in class or volunteering their own opinions without being asked first. Others mentioned anxiety, a desire to save face, and concerns about showing off as reasons to refrain from speaking. These rea sons for taking a more passive role in the classroom most likely stem from societal norms in Chinese culture and beliefs of how a proper student should behave. Additional research
80 arner behavior would assist in further understand ing the results seen here. Conclusion Taken as a whole, the studies on WTC have shown that perceived competence and language anxiety are two important factors that predict how willing a learner is to communi cate in the L2. In spite of their actual L2 competence, as long as learners have faith in their L2 abilities they are more willing to communicate. In a similar way, if anxiety is reduced and learners feel at ease, they will have a higher level of WTC. Addi tionally, it appears that learners in an immersion environment will be more affected by anxiety, whereas non immersion learners are more affected by perceived competence. Although not as crucial a factor, attitudes towards the L2 and the learning environme nt also play a studies have been able to show a proportional relationship between WTC and motivation; however, other studies have not been able to replicate these fin dings, demonstrating a need for additional research on these two factors and how they interact. When considering the situation and its effect on WTC it appears that the factors of group size, interlocutor interest, and topic are all key factors in determi ning a A limitation to these findings is that they were mostly determined by learner self reported data and measured how willing to communicate learners think they are. The recent shift to mixed methodologies and the coll ection of real communication behavior data have been able to demonstrate a possible discrepancy between self reports and actual behavior. As findings have shown in mixed methodology studies, even if a learner reports that s/he would be willing to communica te in a certain situation, moment
81 to moment factors may arise to change that WTC. Future WTC studies should bear this in mind and continue to collect real communication data when possible. Although these studies have been able to demonstrate that WTC lea ds to more frequent communication, to our knowledge, as of now, there has been only one empirical study by Yashima and Zenuk Nishide (2008) that has explored a connection between WTC and linguistic abilities. Their findings concur with MacIntyre and his c more willing to communicate will create more opportunities for interaction in the L2 and will therefore become more proficient in linguistic abilities. Nevertheless, there is still a level of WTC and his/her linguistic abilities. It is the intent of the current study to seek out additional evidence toward this potential relationship and explore any connections seen between WTC and oral abilities, in particular, oral fluency. The following chapter presents the specific research questions that motivated this study, and discusses the methodology used in the experiment.
82 CHAPTER 4 METHODO LOGY In light of the findings of previous research and the need for further study in the areas of study abroad (SA), fluency, and Willingness to Communicate (WTC), the present study was designed and carried out. T his chapter explains the methodology used t o conduct this study is explained Initially, the research questions are discussed along with hypotheses. Next, a description of the participants and their programs is provided. Subsequently, the various ways in which data were collected are described alon g with the motivation behind these means. Finally, the methods used to evaluate the data are discussed Research Questions and Hypotheses progress in oral fluency and WTC du ring intensive short term language study. Furthermore, the study seeks to determine if there is a connection between these two factors. There are three research questions that motivate the current investigation: 1. a. How does intensive short term (six weeks) language study relate to oral fluency in Spanish? b. How does oral fluency differ among learners in an at home context versus those in a study abroad context? When considering the first question, it is hypothesized that all learners will be ab le to show some improvement in oral fluency after a period of six weeks. However, it is expected that learners who spend more hours communicating in Spanish outside of the classroom will make greater progress in fluency than learners who report lower level s of communication. This expectation is based on the importance of interaction and communication in the acquisition process as presented by the Input and Interaction Hypothesis (Gass, 1997) and the Output Hypothesis (Swain 1993, 2000). Likewise, it is
83 anti cipated that learners who participated in SA will make greater progress in fluency given that they are expected to have higher levels of communication due to the vast opportunities provided by their learning environment Evidence of these potential gains h as been documented in previous work (Lennon, 1990; Towell, Hawkins and Bazergui, 1996; Freed, 1995; Towell, 2002; Freed, So and Lazar, 2003; Isabelli Garca, 2003; Freed, Segalowitz and Dewey, 2004; Segalowitz and Freed, 2004; Segalowitz, Freed, Collentine Lafford, Lazar and Daz and Collentine, 2007; Derwing, Munro and Thomson, 2007; Juan Garau and Prez Vidal, 2007; Llanes and Muoz, 2009) and discussed in detail in the preceding chapters 2. a. How does intensive s hort term (six weeks) language study relate to willingness to communicate in Spanish? b. How does willingness to communicate differ among learners in an at home context versus those in a study abroad context? In previous WTC research, possible ve been little explored. Therefore it is difficult to determine if these learners will show changes in their levels of WTC in Spanish at the end of their programs (SA and at home ( AH) ) I t is believed that learners wi ll want to continue to study a second language ( L2 ) as they acquire more knowledge and are more successful in that L2 (Gardner, Day, and MacIntyre, 1992). Therefore it is anticipated that, like fluency, all learners will show some increase in WTC in Spani sh at the end of their programs. While there are no existing studies (to my knowledge) that have found this result, it is thought that the additional Spanish study will encourage learners to want to use Spanish more. I t is hypothesized further that learner s who spent more time communicating in Spanish will demonstrate more willingness to communicate at the end of their program. It particular,
84 the SA learners are expected to show greater increases in WTC than AH learners due to the above mentioned opportunit ies for communication warranted by SA. 3. What is the willingness to communicate in Spanish? The third research question is perhaps the most difficult to answer from a hypothetical standpoint a s there is only one previous study (Yashima and Zenuk Nishide, 2008) that has attempted to find such a link between WTC and linguistic ability. Based on their findings that learners with a high level of WTC performed better on a standardized grammar test a nd based on the general belief that learners who are more communicative will be more successful in the L2, it is expected that learners who can demonstrate a strong WTC in Spanish will also show greater fluency in Spanish. Additionally, by looking at a lin guistic skill that is strongly tied to communication we seek to determine if a connection exists between these two factors. Nonetheless, any positive or negative correlation will be useful to this area of research by helping to provide empirical evidence o f a relationship between WTC and linguistic abilities. Participants Participants for this study all come from the same large North American university and were selected based on their decision to participate in either the summer SA program or the summer A H program offered by the Spanish and Portuguese Studies department. At the beginning of the program, the total number of participants was 31 university students, with 9 students in the SA group and 22 students in the AH group. However in the AH group, 3 st udents dropped the course and did not complete the required post tests. Five other students were eliminated from the AH group because of prior SA experience in a Spanish speaking country and/or extensive interaction with
85 family members who are native speak ers (NSs) of Spanish. 1 This brought the final count of participants to 23 total, with 9 students in the SA group and 14 students in the AH group. The 23 participants we re all native speakers ( NSs ) of English and d id not consider themselves proficient in a ny other language. Of these participants, 14 we re female and 9 we re male; in the SA group there we re seven females and two males, in the AH group there we re seven females and seven males. All participants were classified as traditional university students who attend ed classes full time, and we re between the ages of 19 and 22. None of these learners had previously studied abroad in a Spanish speaking country. Likewise, none of these students lived or socialized frequently with Spanish speaking relatives 2 Th e SA participants were recruited at a pre program meeting and were given monetary compensation to participate in the study upon completion of all tasks The AH participants were recruited at the first meeting of their course. These learners were not provid ed with monetary compensation but rather their instructor awarded them participation points for completing the data collection elements of the study. All students were enrolled in the same intermediate high level course, a grammar and composition course de signed to be the starting point for the third year of Spanish study at the university. Both classes, SA and AH, were taught by university instructors and followed a similar lesson plan designed to develop writing skills and review grammatical structures. Of course, given the fact that one program took place on 1 Due to these prior experiences, it was determined that these learners could have an advantage over the other students who did not have these opportunities and that including these participants in the final datase t would be unwise. 2 One student has a Spanish speaking grandmother who resides in Spain; however the student reported meeting her only twice and does not communicate with her
86 the AH campus and the other in the SA setting, there were obvious curricular differences that could not be controlled. Nonetheless, given the common prerequisites and the common syllabus and requirem ents, it was deemed an acceptable comparison between relatively equivalent groups. Program Information Study Abroad P rogram Participants in the SA program spent six weeks in Santander, Spain located in the northern part of Cantabria, Spain. Santander is a small co a stal city with a limited presence of English speakers. The learners took two three credit courses arranged by their home university and were taught by home university instructors. As was mentioned previously, all students were enrolled in the sam e intermediate high grammar and composition course. For the other course, the students had a choice between intermediate Intensive C ommunication S kills and intermediate Spanish for B usiness. Four of the participants were enrolled in the I ntensive C ommunica tions course and the remaining five were enrolled in the B usiness course. All courses were conducted in Spanish. While in Santander, the students lived with Spanish families; all students had one roommate from the same program. T he program included cultura l excursions to other cities in Spain ( Madrid, Toledo, and Salamanca ) Students also had opportunities to participate in special cultural events, such as dance classes, during the program. At Home P rogram The AH program was at the home university and was a lso six weeks long. The participants in this program were, as mentioned before, all enrolled in the same three credit intermediate high grammar and composition course as the students in the SA program. As these students were at their home institution, they were able to take a
87 wide variety of other courses if they so chose. Six students in the AH program were also enrolled in a second Spanish course that was the same intermediate I ntensive C ommunication course that was also offered in the abroad program. Mor eover, students remained in their normal student housing (dormitories, apartments, houses, etc.) for this program. There were no special events or outings planned for the students in this program apart from their regularly scheduled class. Data Collection Data were collected from a variety of sources during the course of the programs. Some data, such as the interviews and WTC in Spanish questionnaires were collected pre and post program, while other data were collected each week, and other data were collec ted one time only. In this section, a description of the materials/tools used is presented in the order in which data were collected and then t he following section explains how data were analyzed. Language B ackground Q uestionnaire All participants complete d a language background questionnaire prior to the programs. This questionnaire (see Appendix A) asked participants to provide information about previous language study, languages spoken, information about any Spanish speaking relatives, previous ly SA expe rience in a Spanish speaking country, and other Spanish speaking experiences. This questionnaire was used to control for a more homogeneous participant population and allowed for the elimination of bilingual speakers, participants who had previous SA exper ience, and/or participants who lived and interacted extensively with Spanish speaking relatives. The questionnaires were carefully read to look for three possible elimination factors. The first factor was bilingualism; a participant would be eliminated if s/he
88 indicated that s/he had grown up speaking more than one language. However, no participant indicated that s/he was bilingual. The second factor was previous SA experience in a Spanish speaking country. Three students were eliminated from the AH group based on a previous SA program. Finally, the third factor was extensive use of Spanish with immediate or extended family members. Two students reported that while they did not consider themselves bilingual in English and Spanish they lived with Spanish spe aking relatives with whom they communicated on a regular basis in Spanish. These two students were eliminated from the AH group for this reason. All questionnaires were also read carefully to see if any participants had any other special learning experien ces that may have given them an advantage over other learners (for example, having studied Spanish since elementary school resulting in a greater amount of years spent learning Spanish) but no such experiences were found. Oral I nterviews In order to colle ct a measure of oral fluency, data were collected in the form of informal oral interviews (see Appendix B) following a pre program and post program design (Isabelli Garca, 2003; Derwing, Munro and Thomson, 2007; Llanes and Muoz, 2009). All interviews we re conducted with the principal investigator in an academic formal setting, the investigator endeavored to make the interviews feel like a casual conversation, as oppos ed to an oral examination interview. This was accomplished through general small talk prior to the recorded interview to help put the participants at ease, use of the informal Spanish second person forms, and maintaining an easy going and approachable deme anor with the participants. Participants were also reminded that their participation in these interviews was not only voluntary but that it was unrelated to
89 their performance in the classes they were taking for a grade. The interviews were approximately 15 minutes in duration and consisted of personal information and opinion questions that covered a variety of topics and grammatical structures. Two sets of interview questions, set A and set B, were devised in order to ensure that the participants received d ifferent questions for each interview The two sets of interview questions were counterbalanced to ensure that a similar style of questions using the same structures was asked during each interview. It was the goal of the investigator to allow the partici pants to speak as much as possible during the interviews. Therefore, if a participant appeared particularly interested in a given topic and spoke about it at length, s/he may have answered fewer questions than another participant. Conversely, if a particip ant spoke less about the topics provided, s/he may have answered more questions than other participants. During the interviews, the investigator did not provide linguistic assistance to the participant s while they were answering questions. However, if a pa rticipant did not understand a question, the investigator repeated the question, and/or rephrased the question to assist the participant in understanding but always maintaining the use of Spanish In some instances, the investigator showed the participant the written question in order to offer further assistance. A digital voice recorder was used to make recordings of the interviews for subsequent analysis, which will be described in the Analysis section, below. W illingness to C ommunicate Q uestionnaires Fo llowing the methodology that has been used in virtually all WTC investigations, (i.e. Yashima, 2000; Baker and MacIntyre, 2003; MacIntyre, Baker, Clment, and Donovan, 2003a; Clment, Baker, and MacIntyre, 2003) participant self rating
90 questionnaires explo ring WTC in English and WTC in Spanish (see Appendix C) were employed in order to determine a quantified value of WTC for the participants. These questionnaires were based on WTC questionnaires designed by MacIntyre, Baker, Clment, and Conrod (2001) to me asure WTC in a variety of situations and contexts. A change was made from the original questionnaire in both reference to the language (then French, now Spanish) and the wording of the Likert scale instructions. Using a five point Likert scale, participant s rated their willingness to participate in a given situation almost always willing to participate in this situation almost never willing to participate in this situation 3 The situations were divided into four main categories: speaking, reading, writing, and comprehension. For example under the speaking category the questionnaire asked participants to rate how willing they would questionnaire was divide d into different possible communication categories in order to explore whether or not participants may be more willing to communicate in one way, writing perhaps, than in another way, speaking for instance. In addition to the Likert scale questions, at the end of the questionnaire there was space for the participants to offer comments in their own words about how willing s/he feels to communicate. This open ended comment section was added to provide deeper insight into what prompted the numeric answers. At the beginning of the study, all participants first completed a WTC in English questionnaire to act as a baseline for how willing participants are to communicate in their first language ( L1 ) Considering how willing some individuals may be when 3 The original questionnaires were for French as a second language and the Liker t instructions stated that
9 1 communicati ng in their L1 may help to explain how willing they are to communicate in their L2. Furthermore, it may help explain if certain individuals are particularly reserved or particularly outspoken. The WTC in Spanish questionnaire followed a pre and post prog ram design in order to help establish how WTC may change over time for these participants. There was only one version of the questionnaire, but due to the nature of the questionnaire and the number of questions it was thought that participants would be unl ikely to remember their original responses and would consider how they felt in that particular moment Language U se Q uestionnaire When seeking to explain why certain changes in linguistic abilities or personal attitudes occur after SA, previous research h as considered the amount of language that was used by the participants during the course of their program (Freed, Segalowitz, and Dewey, 2004; Segalowitz and Freed, 2004; and Magnan and Back, 2007). The Language Contact Profile developed by Freed (1990, 19 95) was the basis for this questionnaire; however, the investigator chose to simplify and develop a new version for the purposes of the current study. It was believed that the original questionnaire was very long and that perhaps it would be too difficult for learners to accurately remember the amount of time spent doing language related activities at the end of their six week period. The new questionnaire (see Appendix D) asked each participant to quantify the number of hours per day s/he spoke Spanish out side of class, with whom s/he spoke (for example, host family members) and for what purpose s/he spoke (for example, to obtain directions or information). Additionally, the questionnaire asked participants to indicate how many hours per day s/he spent spea king English and to whom. The final
92 question on this survey asked the participant to indicate how frequently s/he participated in certain non speaking activities in Spanish, for example, reading Spanish magazines or watching TV in Spanish. In previous res earch (Freed, Segalowitz, and Dewey, 2004; Segalowitz and Freed, 2004; and Magnan and Back, 2007), data were collected one time at the end of the program. In the current investigation, data were collected weekly out of a concern that participants could hav e difficultly remembering how often and how much they communicated during the course of the program. To make this process as simple as possible for participants, the questionnaire was available online through the website www.SurveyMonkey.com. The questionn aire was only available to these selected participants and could not be accessed by outside sources. Reflection B logs asked to write a weekly reflection blog to discuss the ir thoughts and emotions towards communicating in Spanish as well as their experiences of speaking Spanish. By asking them to describe their feelings in their own words the intent was to examine a potential new source of WTC data. While recognizing that t his data is still a type of self reporting, the explanations given in the reflection blogs may allow for a better understanding of the quantitative data obtained through the WTC questionnaires. The blogs were completed on personalized web pages created thr ough the social networking site www.Ning.com 4 These pages were private to the participants and the investigator only. To best allow participants to explain their thoughts and not to concentrate on the language of 4 Ning provided free, private social networking services to groups who signed up through July 2010. Starting in July 2010 they of fer pay only service.
93 expression all blogs were written in Engl ish. Participants were instructed to write 200 to 250 words for each entry. Each week participants were given a different topic to guide their thoughts about communicating in Spanish. In addition, the principal investigator wrote a blog entry for the first reflection topic in order to provide the participants with an example of how to complete this assignment. The weekly topics were as follows: Week 1 How do you feel about using Spanish? In general, when you use Spanish how do you feel? are you excited? n ervous? emotions when you communicate in Spanish. It might be that certain situations put you more at ease and yet other situations make you feel more nervous. For example, when I confident and am pleased that I am able to carry on a conversation; yet, when I well, I get extremely nervous and I fe el that I have to concentrate really hard so like this? Are there times when you feel confident? What else do you feel? Week 2 Difficulty Communicating Think about a time w hen you were trying to speak Spanish with someone, and you had a difficult time communicating. How did this experience make you feel towards Spanish? Did it push you to keep trying to communicate? Or did it make you feel like giving up? Why or why not? How did it affect the next time you tried to communicate? Were you more apprehensive about trying to speak? Or were you more determined to get it right this time? Week 3 How willing are you to use Spanish? When given the chance to use Spanish how willing ar e you to do so? Think about times when you always feel willing to communicate with others in Spanish; why are you more willing to use Spanish then? Next, think about times when you feel very unwilling to communicate. What do you think is holding you back? Week 4 In the classroom versus in the street Many times we behave differently in the classroom than we do outside of the classroom. Do you feel more willing to use Spanish in class than outside of class? Or do you feel about the same towards both situati ons? Or do you feel more comfortable communicating away from the classroom? If you do feel differently,
94 why do you think you are more willing to communicate in one situation over the other? What do you think makes you resist communicating in the other situ ation? Week 5 Dealing with native Spanish speakers Think about your encounters with native speakers who are not your instructor. When given the opportunity to use your Spanish skills with native speakers, do you usually do so? Or do usually tend to avoid interacting with native speakers? Do you wait for native speakers to address you? Or do you feel confident enough to initiate communication? Do you try to keep the conversation going? Or do you prefer to keep the conversation brief and to the point? Pleas e explain your feelings towards interacting with native speakers, keeping in mind certain situations that may make you more prone towards interactions and other situations that may make you shy away from communication. Week 6 Now that week summer program, have your thoughts practicing and learning more about Spanish, in your opinion, how has it affected you? How willing to com municate in Spanish do you feel now? Do you think you will try to seek out more ways to communicate in Spanish now that the program is over? Why or why not? Summary The data collection methods described here are based on previous research as well as new i deas on ways regarding the evaluat ion of collection method was designed to capture various factors that could linguistic abilities and their individual differences. The language background questionnaires were used primarily to create a homogenous group of learners. The oral interviews were used to capture informal speech data. The WTC questionnaires were used to collect a form of quantitative data assessing how willing learners were to use language in a va riety of situations. The language use questionnaires were used to indicate the amount and quality of L2 contact learners had on a weekly basis. Finally, the reflection blogs were used to collect additional data for WTC so that learners could
95 explain what i nfluenced their willingness to use Spanish in their own words. When analyzed together these data sources should provide a clear picture of WTC. Data Analysis In this section, an account of how the data were analyzed is explained in detail. Following the or der in which they were administered the analysis of each data collection method is discussed individually. Oral I nterviews Following research methodology used by Freed (1995), Freed et al. (2004), and Segalowitz and Freed (2004), two 1 minute segments fro and post program interviews were extracted for analysis yielding a total of 4 minutes of tenth minute of both the pre and post interviews. Out of concern for the level of the learners and their abilities to answer more difficult opinion based questions only answers to descriptive questions were used in the data analysis. All speech segments were transcribed by the principal researcher. The speech was unedited and included all recording. The recordings were played back through the computer program Praat which provides a spectrogram of speech allowing for s ilent pauses to be measured from th at spectrogram. The segments were also transcribed by a native Spanish speaking volunteer 5 in order to determine the reliability of the data which resulted in a 97% con firmation of the transcriptions In the 3% of cases in wh ich the raters disagreed, the 5 This volunteer was not a linguist nor a Spanish instructor, but had worked as a secretary and had experience transcribing notes for her employer.
96 In order to determine fluency, each segment was evaluated on seven measures: Rate of speech calculated on the basis of the number of non repeated complete words per minute (u sed by Raupach, 1987; Lennon, 1990; Riggenbach, 1991; Freed, 1995, 2000; Towell et al. 1996; Ejezenberg, 2000; Towell, 2002; Freed et al. 2003; Isabelli Garca, 2003; Kormos and Dnes, 2004; Freed et al. 2004; Segalowitz and Freed, 2004; Segalowitz et a l., et al. 2007; Juan Garau and Prez Vidal, 2007; Garca Amaya, 2008; Llanes and Muoz, 2009). Frequency of unfilled pauses calculated by counting silences which are 0.4 seconds or longer in duration per minute; shorter silent pauses are characteristic of native speech and are not considered to be dysfluent (used by Freed, 1995; Freed et al. 2004; Segalowitz and Freed, 2004; Segalowitz et al. 2004; Trenchs Parera, 2009). Frequency of filled pauses calculated by counting filler sounds like or okay and yeah (used by Freed, 1995; Freed et al. 2004; Segalowitz and Freed, 2004; Segalowitz et al. 2004; Trenchs Parera, 2009). Average length of fluent speech runs calculated by averaging the total number of non repeated complete words produced continuously without interruption by dysfluent pauses (used by Raupach, 1987; Lennon, 1990; Freed, 1995, 2000; Towell et al. 1996; Towell, 2002; Freed et al. 2003; Isabelli Garca, 2003; Kormos and Dnes, 2004; Freed et al. 2004; Segalowitz and Freed, 2004; Segalowitz et al. et al. 2007; Llanes and Muoz, 2009). Frequency of repairs calculated by counting the number of repairs made per minute; repairs incl ude repetition (of a word, syllable or phrase), reformulations, self corrections, and partial repeats (used by Freed, 1995; Freed et al. 2004; Segalowitz and Freed, 2004; Segalowitz et al. 2004). Frequency of L1 or created words used calculated by coun ting the number of L1 words or non words used per minute P roper names using English pronunciation or English filled pause words were not counted (used by Llanes and Muoz, 2009). It was believed that learners at this level would not be familiar with Spani sh equivalents of proper names and the English name could be recognized by NSs of Spanish Clusters of dysfluencies calculated by counting the number of clusters of dysfluencies occurring per minute (used by Freed, 1995; Freed et al. 2004; Segalowitz an d Freed, 2004; Segalowitz et al. 2004). A cluster of dysfluencies is defined as two or more dysfluent elements grouped together. An e xample of a one minute transcribed segment is presented here followed by Table 4 1 showing the analysis of this segment ( the original in Spanish with English
97 pronunciations in italics is presented first, followed by the English translation) The complete set of transcriptions of all participants can be seen in Appendix E. um [3s] um [2s] uh mi [ 0.4 s] viaje peor um fui cuand o yo [3s] uh yo [2s] uh fui a Bahamas con mi con mi familia familia um porque mi mi mam uh [3s] uh se olvid uh or yeah se olvid um [0.5s] que uh un reservacin una reservacin uh [3s] para or para um prxima semana que entonces uh pero [3s] ( laughs ) nos otros program interview) um [3s] um [2s] uh my [0.4s] worst trip um I went when I [3s] uh I [2s] uh went to Bahamas with my with my family family um because my my mom uh [3s] uh forgot uh or yeah forgot um [0.5s] that uh a (masc uline) reservation a (feminine) Table 4 1 Example of speech analysis Rate of s peech Unfilled p auses Filled p auses Fluent r uns Repairs L1/Created w ords Clusters 27.00 9. 00 18.00 2.33 1.00 0.00 7.00 With these measures, various statistical analyses were completed using the statistical program SPSS. For all analyses in this study significance was set at 0.05. First, to determine the differences from pre test to post test for the SA and AH groups individually, a series of t tests 6 were run on each measure mentioned above. Second, a series of univariate analyses of variance were run for each fluency measure to determine if there were any significant differences in fluency be tween the SA learners and the AH learners at the beginning of each program. Third, a series of general linear models for each of the fluency measures were conducted to determine the possible significant changes in fluency over time and the possible signifi cant e ffects of group on fluency measures. Also to consider the possibility that additional coursework was a possible influencing factor for fluency, a second series of tests were run comparing the 6 Due to the overall smaller participant population and the difference in group size, non parametric tests were also run and results were found to be the same with regards to significance.
98 nine SA participants with the six AH participants that we re also enrolled in six credit hours of Spanish. Fin ally, correlations were run using the Pearson Correlation Test between fluency measures and the amount of Spanish used to determine the relationship between fluency and L2 contact. W illingness to Communic ate Q uestionnaires Although the questionnaires look at speaking, reading, writing, and comprehension, it was decided that for the purposes of this study to consider only the speaking portion of the questionnaire. Since the primary linguistic objective of t he current study is to look at oral fluency, it was logical to look at the part of WTC that was more strongly linked to oral communication. F or all questionnaires a score was determined based on the 5 point Likert scale. The speaking section consisted of 1 6 questions for the highest possible total of 80 points (in other words providing a 5 for each communication situation) Data were then analyzed for the WTC in English questionnaire to determine if any outliers exist ed (i.e. if any participant was very un willing or extremely willing to communicate in their native l anguage). To accomplish this, a univariate analys i s of variance w as run Th is analys i s showed that there were no outliers for these participants. Data from the WTC in Spanish questionnaires were then analyzed to determine the changes over time for WTC in Spanish. In order to determine if changes occurred for both the SA group and the AH group from pre to post program, a paired t test was conducted. When comparing the two groups to each other, a u nivariate analys i s of variance w as to determine the differences in WTC in Spanish at pre program time. To look for the changes between the two groups over time, a general linear model was run.
99 Again, considering the possibility that additional coursework m ay have impacted the with the six AH participants that were also enrolled in six credit hours of Spanish. In addition, correlations were run using the Pearson Correlation Test between the WTC in Spanish score and the fluency measures to determine the relationship between WTC and oral fluency. Correlations were also run between WTC in Spanish and the amount of Spanish used to establish the relationship between L2 contact and WTC. Once again statistical analyses were completed in SPSS. Language U se Q uestionnaires For these questionnaires, data were analyzed to compare the amount of language used to the results of the fluency measures and the WTC scores in order to determine th e possible relationship between these factors. For the quantitative data comparison, weekly scores for amount of Spanish and English spoken per day were determined for each participant. The scoring is as follows: 1 point for 0 1 hours spoken, 2 points fo r 1 2 hours spoken, 3 points for 2 3 hours spoken, 4 points for 3 4 hours spoken, 5 points for 4 5 hours spoken and 6 points for more than 5 hours spoken. In addition, an average summer score per participant was determined by simple means of the we ekly scores. Correlations were run in SPSS between the language use scores and both the fluency measures and the WTC scores to determine if any relationships are seen between these factors. In addition, the fluency and WTC scores were compared to the activ ities and interlocutors of the participants to determine if any patterns emerged. Reflection B logs The reflection blogs provided additional data for WTC in Spanish. Following methodology as outlined by Bogdan and Knopp Biklen (2007) a coding scheme was
100 d eveloped for the blog data using the following steps: A t first all blogs were carefully read and main idea s words and phrases used by the participants were identified Next these were used to create the series of codes C odes were developed for six catego of thinking about people, activity, strategy, and relationship and social structure. (These categories are suggested by Bogdan and Knopp Biklen as common classifications fo r coding schema.) The complete list of codes is available in Appendix F. The blogs were then coded by the principal researcher and a volunteer research assistant 7 Any time a word and/or phrase were used by the participants in the blogs the code number was assigned to that line. The coding results of both resulted in a 98% confirmation; the remaining 2% were discarded from the dataset Next, the codes were counted and frequencies of each code were found for both groups of participants and well as for the en tire population of participants. the frequency for the SA group was 1.75 (14 instances appearing for 8 out of 8 participants 8 ), the frequency for the AH group was 1.79 (25 instances appearing for 14 out of 14 participants), and the frequency for the entire population was 1.77. Codes that appeared at least once for all participants in each grouping (SA only AH only and total population) were first considered to be possible common tendencies The frequencie s of these codes were then looked at, and any code with a frequency higher than 1.5 was considered to be a common tendency 7 The research assistant was a Masters student of Spanish at Indiana State University where the principal researcher is currently employed. 8 One of the SA participants did not complete the blogs, reducing the total number of participants from nine to eight.
101 Due to the use of topics for the blogs, it is highly likely that certain codes arose because of the guidelines provided. Nonetheless m any of these codes were also used in other blog entries where they were not prompted to write about that specific topic. Within th e code analysis, it became apparent that that there were some codes that appeared across the entire data set (both SA and A H participants) while others were shared only by the SA participants or only by the AH participants. The codes that were appeared in the full data set were determined to be common tendencies regardless of learning context. ( Several of the situation activi ty and strategy codes were likely to be present for all participants as they described universal settings and behaviors available to all of the participants .) On the other hand, the codes that emerged in only one group or the other were seen as differences between the two groups and two learning contexts Finally the blogs were analyzed for individual differences that were unique to specific participants and affected their WTC in a special way. Finally, the blog data w ere compared to the WTC questionnaire r esults to determine if similarities were seen. Conclusion In this chapter the research questions and hypotheses motivating this investigation were discussed, as well as the participants, data collection methods, and data analysis methods that were used. Re call that it has been hypothesized, based on previous participants would make greater improvements both in oral fluency and WTC in Spanish due to their greater opportunities to c ommunicate in Spanish and interact with NSs. However, it was proposed that any participant who has a large amount of interaction in Spanish may show improvement regardless of learning context. In the next chapter the
102 results of the data analysis are prese nted and discussed, after which point the research questions will be reexamined and answered.
103 CHAPTER 5 RESULTS AND DISCUSSION In this chapter, the results of the data analyses are presented. Quantitative data are presented in the following order: first, analyses of fluency measures of oral interview data; second, analyses of Willingness to Communicate (WTC) questionnaire data; third, correlations between fluency and second language ( L2 ) contact; fourth, correlations between WTC and L2 contact; and fifth, correlations between fluency measures and WTC. Next, additional information discussing the results of the weekly reflection blog data and its relation to the quantitative data is presented. Following the presentation of the results the implications of th e findings are discussed. Quantitative Results and Discussion Fluency M easures for S tudy A broad L earners In this section the results of the fluency measures for the study abroad (SA) group are presented to look for changes over time from the pre program o ral interview to the post program interview Once again the seven fluency measures used in this study were: rate of speech (number of non repeated words per minute), frequency of unfilled pauses (number of silent pauses per minute), frequency of filled pau ses (number of filled pauses per minute), mean length of fluent speech runs (average calculated of all fluent runs per segment), frequency of repairs (number of repairs per minute), frequency of first language ( L1 ) or created words used (number of L1 or cr eated words per minute), and frequency of clusters of dysfluencies (total number of clusters per minute). Table 5 1 shows the means for all fluency measures for the SA group for both the pre and post program interviews. From these simple means it can be s een that the SA group was able to increase the ir speech rate, as well as increasing the average length
104 of fluent runs. Similarly, the learners were able to decrease unfilled and filled pauses as well as the use of L1 or created words. There was almost no c hange in the amount of dysfluen t clusters from the pre to post program interview. Additionally, an increase in repairs was seen over time. Table 5 1 Simple means of fluency measures for s tudy a broad learners Rate Unfilled p auses Filled p auses Length of r uns Repairs L1/Created w ords Clusters Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Mean 45.72 64.50 6.61 5.55 13.50 11.83 3.95 4.88 2.61 4.55 3.94 1.44 5.66 5.27 Standard d eviation 11.72 9.79 2.43 2.94 3.29 3.56 1.09 0.59 1.78 1.55 5 .68 0.84 2.29 1.90 Standard e rror m ean 3.90 3.26 0.81 0.98 1.09 1.18 0.36 0.19 0.59 0.51 1.89 0.28 0.76 0.63 In Table 5 2 the results of the paired T test show where significant changes were seen from the beginning of the SA experience to the end of the program. The significant interactions (marked with *) reveal that SA learners significantly increased speech rate, the average length of fluent runs, and self repairs. The first two measures indicate an increase in fluency. Although not found to be signif icant, there was a decrease seen in filled pauses that is approaching significance, p = 0.079. Once again this would be in the direction of more fluent speech. Although simple means showed a decrease in unfilled pauses and in the use of L1 and/or created w ords, these measures were not found to be significant. Not surprising, due to the small change seen in simple means, the amount of dysfluen t clusters was also not significant.
105 Table 5 2 Paired T tests of fluency measures for s tudy a broad learners Fluenc y m easures (Pre Post) Mean Standard d eviation Standard e rror m ean t df Significance Speech r ate 18.77 12.00 4.00 4.69 8 0.002* Unfilled p auses 1.05 3.67 1.22 0.86 8 0.414 Filled p auses 1.67 2.48 0.82 2.01 8 0.079 Avg l ength of f luent r uns 0.92 0.9 5 0.31 2.90 8 0.020* Repairs 1.94 1.50 0.50 3.86 8 0.005* L1/Created w ords 2.50 5.55 1.85 1.35 8 0.214 Dysfluency c lusters 0.39 1.91 0.63 0.60 8 0.560 *significance 0.05 Following the Output Hypothesis, it is likely that the pressure required by real world (as opposed to classroom) interaction to produce output in a timely manner, assisted these learners in being able to produce more words at a faster rate of speech and thus meet up with the demands of responding to native speakers ( NSs ) The significant improvement in average length of fluent runs indicates that the SA learners were able to produce more words in a fluent string without any dysfluent interruptions suc h as silent or filled pauses. This result also suggests that these learners were capable of having longer stretches of fluency and thus able to create more of a When further considering these temporal features, both speech rate and average length of fluent runs have been found to be positive indicators of more fluent speech when assessing which measures are best for evaluating fluency (Raupach, 1987; Lennon, 1990; Riggenbach, 1991; Freed, 1995, 2000; Towell, Hawkins, and Bazergui, 1996; Ejezenberg, 2000; Kormos and Dnes, 2004). Since self repairs are considered to be a dysfluency because they break the flow of natural speech, this increase indicates a move away from more flu ent speech.
106 However, when looking at the repairs that were made in the post program interviews, 92% of them are correct repairs, while in the pre program interviews, the repairs are only correct 73% of the time. Therefore, although repairs in general may b e seen as a dysfluency, at least these learners were able to make correct repairs the majority of the time and with increased accuracy after their summer experience While it is unknown what caused this shift in behavior, it is possible that as these lear ners spent more time in the L2 community and interacted with NSs, they received additional grammatical information as is suggested by the Input and Interaction Hypothesis. Also of interest here in the results of the SA group was that although there were s ignificant increases in fluency predictors (i.e. indicating improvements), there were no significant decreases in dysfluencies. As with the increase in self repairs, the reason behind this lack of improvement in pauses and L1/created words is unknown. One possible reason for this lack of significant change may be that it is more difficult to speak at a faster rate. Furthermore, when considering the use of L1 words, it mi ght have been that because the participants knew that the interviewer was a native English speaker, and the interviews were completed in a casual style, they felt that it was alright to use English words since they knew that the interviewer would understan d their meaning. It would be useful to see if learners would continue to produce the same amount of English words if the interviewer were a native Spanish speaker that spoke little English. The same can be said of the created words that were used in their speech, since these words were often Spanish versions of English words or some type of created word
107 using English grammatical patterns that the interviewer was likely to understand. For program interview she was trying to sa es muy convenzando where convenzando is a combination of the Spanish gerund pattern for the first conjugation and the second conjugation verb convencer ( to convince ). While it is possible in English to use the geru nd as the object of the copula to create an adjectival phrase, this is not possible in Spanish and the use of the adjective convincente is needed instead. However, as both an English and Spanish speaker, the interviewer was able to recognize the English gr ammar pattern while recognizing the root of the created word and understand this statement. The probability of a NS of Spanish understanding this phrase is not particularly high and most likely would depend on how much interaction that NS has had with lear ners and how familiar s/he is with these types of created words. In all, when considering these dysfluencies, it does appear that, while a non significant decrease, the SA learners may have been able to decrease their dysfluencies sufficiently enough to p roduce a greater average length of fluent runs. When looking further into where pauses (both filled and unfilled) occurred in the speech segments it was hoped that a pattern might be found, (for example perhaps pauses occurred more before and after clauses ) but this was not discovered. Fluency M easures for A t H ome L earners The simple means for the fluency measures for the at home ( AH ) learners are provided below in Table 5 3. For AH learners there is a slight decrease for speech rate, a measure that is rela ted to fluency as indicated by previous research T here is also a slight increase in unfilled pauses, filled pauses, and dysfluen t clusters. There is almost
108 no change seen for the average length of fluent runs, repairs, and the use of L1 and/or created wor ds. Table 5 3 Simple means of fluency measures for a t h ome learners Rate Unfilled p auses Filled p auses Length of r uns Repairs L1/Created w ords Clusters Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Mean 59.71 58. 03 7.10 9.1 0 11.75 12 46 4.76 4.3 9 3.50 3.89 2.96 2.42 5.10 6. 50 Standard d eviation 15.98 12.52 2.06 2.5 0 4.77 4.4 3 1.26 0.82 1.59 1.76 2.44 1.68 2.52 2. 10 Standard e rror m ean 4.27 3.3 5 0.55 0.6 7 1.27 1.1 8 0.33 0.22 0.42 0.47 0.65 0.45 0.67 0. 56 Table 5 4 presents the resu lts of the paired T tests that show significant changes over time in fluency measures. The analyses reveal two significant increases over time: unfilled pauses and dysfluen t clusters. This result is rather surprising in that it is expected that silent paus es and clusters of such dysfluencies would be reduced over time. Rather than showing a shift toward an increase in fluency, the AH learners showed an increase in dysfluent features All other fluency measures show no significant change over time. Table 5 4 Paired T tests of fluency measures for a t h ome learners Fluency m easures ( p re p ost) Mean Standard d eviation Standard e rror m ean t df Significance Speech r ate 1.69 12.17 3. 2 5 0 52 13 0 614 Unfilled p auses 2.0 0 3.2 0 0.8 6 2.3 4 13 0 .03 6 Filled p aus es 0 .7 1 4. 15 1. 11 0 64 13 0 531 Avg l ength of f luent r uns 0 .37 1.09 0.29 1.27 13 0 .22 5 Repairs 0 .39 1.48 0.39 0 .99 13 0 .340 L1/Created w ords 0 .5 4 2.03 0.54 0 .98 13 0 .342 Dysfluency c lusters 1. 39 1. 80 0. 48 2.90 13 0 .0 12 *significance 0.05
109 In the case of the AH context, the findings here indicate that at the time of the second interview these learners had more hesitations in their speech resulting in a style of speech that was choppier. It appears then that the intensive study in the traditiona l participants did have an increase in two types of dysfluencies, overall they appear to show little change in their fluency and were able to maintain factors that have been know n to indicate fluency such as speech rate and the average length of fluent runs. It is unknown why these learners did not show the same level of improvement as the SA learners, but it is possible that their lack of change is due to the number of hours of coursework during the program a possibility that will be discussed in greater detail below Analyses of means were also run for a sub set of the six AH learners who took a second Spanish course, resulting in a total of six credit hours. Table 5 5 prese nts the simple means for this group. For these six learners, there were decrease s in speech rate and in the use of L1/Created words. Increases were seen for unfilled pauses, average length of fluent runs, repairs, and dysfluent clusters. Table 5 5. Simple means of fluency measures for a t h ome learners with two Spanish courses Rate Unfilled Pauses Filled Pauses Length of Runs Repairs L1/Created Words Clusters Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Pre Post Mean 54.58 52.67 7.50 9.33 14.5 0 14.67 3.89 4.00 3.83 5.00 3.33 2.75 6.41 7.50 Standard Deviation 11.94 11.25 1.26 2.80 2.45 4.63 0.35 0.74 1.25 1.26 2.40 1.75 1.43 1.82 Standard Error Mean 4.87 4.59 0.52 1.15 1.00 1.89 0.14 0.30 0.51 0.52 0.98 0.72 0.58 0.74 Table 5 6 shows the res ults of paired T tests for each fluency measure for this sub set of learners. Results find a significant increase in the amount of repairs used by
110 this group. This finding is similar to that of the SA group, who were enrolled in the same amount of credit h ours of Spanish. It is possible that the additional hours of study repairs when they perceived an error had been made. No other significant changes were seen for this group. Table 5 6. Paired T tests of fluency measures for a t h ome learners with two Spanish courses Fluency m easures ( p re p ost) Mean Standard d eviation Standard e rror m ean t df Significance Speech rate 1.92 11.13 4.54 0.42 5 0.691 Unfilled pauses 1.83 3.41 1.39 1.31 5 0.246 Filled pauses 0.17 5.70 2.33 0.07 5 0.946 Avg l ength of f luent r uns 0.12 0.61 0.25 0.46 5 0.664 Repairs 1.17 0.52 0.21 5.53 5 0.003* L1/created words 0.58 1.24 0.51 1.15 5 0.302 Dysfluency clusters 1.08 1.24 0.51 2.14 5 0. 086 *significance 0.05 Unlike the complete group of AH learners a significant change towards more seen here, an increase in repairs, was also present for the SA learners. However no si gnificant changes are seen with speech rate and average length of fluent runs as was found for the SA learners. It appears then that the added coursework may have impacted fluency, but perhaps not as much as it did for the SA learners. Comparing S tudy A br oad L earners to A t H ome L earners Initially a series of univariate analyses of variance were run for each fluency measure to determine if there were any significant differences in fluency between the SA learners and the AH learners at the beginning of the p rograms. The results of these
111 analyses found a significant difference between the groups: AH learners produced a fa ster rate of speech pre program ( p=0.002 ) than the SA learners. These results indicate that initially the AH participants were more fluent s peakers in that one measure, than the SA participants. Analyses of all other measures revealed no significant differences between the groups. Post program, when comparing the SA learners to the AH learners for changes over time, a series of general linear models for each of the fluency measures w as run; Table 5 7 presents the results of these analyses. Of the eight measures, five were found to be statistically significantly different in favor of the SA learners. Results showed that, over time, the SA group increased their speech rate, average length of fluent runs, and repairs significantly more than the AH learners. Of these, two measures, speech rate and average length of fluent runs, have been suggested as indicators of fluent speech (Raupach, 1987; Lenn on, 1990; Riggenbach, 1991; Freed, 1995, 2000; Towell, Hawkins and Bazergui, 1996; Ejezenberg, 2000; Kormos and Dnes, 2004). Correspondingly, the SA group decreased their unfilled pauses and clusters of dysfluencies significantly more than the AH learners These two measures are characteristic of dysfluent speech and therefore, decreasing both is a move towards more fluent speech. No significant differences were seen between the two groups for the amount of filled pauses and the amount of L1 and/or created words used.
112 Table 5 7 Comparison of s tudy a broad learners to a t h ome learners for changes in fluency over time Fluency m easures Type III s um of s quares df Mean s quare F Significance Speech rate 1142.224 1.000 1142.224 15.674 0 .001 Unfilled pauses 25 .574 1.000 25.574 4. 444 0 .04 7 Filled pauses 15.528 1.000 15.528 2. 381 0 138 Avg l ength of fluent run 4.6 47 1.000 4.6 47 8. 48 7 0 .008 Repairs 6.594 1.000 6.594 5.916 0 .024 L1/Created words 10.569 1.000 10.569 1.476 0 .238 Clusters of dysfluencies 8.69 6 1.000 8.696 5. 110 0 .0 35 *significance 0.05 In addition to the above comparison, the SA learners were also compared to the sub set of AH learners who took two Spanish courses during the six week period. As above, a series of general linear models for each of the fluency measures was run; Table 5 8 presents the results of these analyses. Only one significant difference was found between the two groups for speech rate, indicating that the SA learners had a faster speech rate at the end of the program tha n the sub set of AH learners. This indicates that with regards to the other fluency measures the two groups were similar in their fluency abilities. This finding is important because it suggests that the additional three credit hours of study were benefici al to both groups in making improvements in these fluency factors. Furthermore, given that these learners were enrolled in a course designed to improve oral communication it is not entirely surprising that they were able to show gains in fluency. It is pos sible that another course with a different focus, reading perhaps, may not have provided these participants with the same opportunities to advance in their fluency abilities.
113 Table 5 8. Comparison of s tudy a broad learners to a t h ome learners with two Span ish courses for changes in fluency over time Fluency m easures Type III s um of s quares df Mean s quare F Significance Speech rate 770.868 1 770.868 11.31 0.005* Unfilled pauses 15.022 1 15.022 2.35 0.150 Filled pauses 6.050 1 6.050 0.74 0.404 Avg l ength of fluent run 1.192 1 1.192 3.37 0.089 Repairs 1.089 1 1.089 1.45 0.250 L1/Created words 6.613 1 6.613 0.68 0.426 Clusters of dysfluencies 3.901 1 3.901 2.73 0.122 *significance 0.05 These results attest that the SA learners were able to produce more words at a faster pace and create more fluent strings of those words that were unfettered from hesitations than the entire population of AH learners. However, when the SA learners a re compared to the sub set of AH learners who took the same number of credit hours, there is only one significant difference seen: SA learners have a significantly faster speech rate at the end of the program than the AH learners. It appears then that the additional study was an important variable and likely allowed both groups of learners to reach comparable frequencies of dysfluencies and average length of fluent runs. The question remains of why the SA learners were able to show a faster speech rate and is a question that cannot be answered on the data collected in this study. While it is possible that the SA context was the reason behind this difference, there is no evidence seen in this data to support that claim. W illingness to C ommunicate for S tudy A b road L earners In this section the results for the WTC questionnaires are presented for the SA learners to look for changes over time from pre to post program. At the pre program time learners completed both a WTC in English and a WTC in Spanish questionna ire. At
114 the post program time learners only completed the Spanish questionnaire. Initially a univariate analysis of variance was run on the WTC in English questionnaire data to determine if any learners were significantly willing or unwilling to use their native language as compared to the other learners in their group. Results showed no significant differences between the learners with respect to WTC in English. Similarly, a univariate analysis of variance was run on the WTC in Spanish pre program data, an d again no learner was found to be significantly more or less willing to communicate in Spanish than others in their group. Table 5 9 shows the means for WTC in Spanish for the SA group for both the pre and post program questionnaires. The results show an increase for WTC over time Table 5 9 Simple means of W illingness to C ommunicate in Spanish for s tudy a broad learners WTC s ection Mean N St andar d d eviation St andar d e rror m ean Speaking p re 52.55 9 8.84 2.94 Speaking p ost 59.33 9 7.03 2.34 In order t o analyze the changes in WTC for statistically significant differences, a paired T test w as run. The result of th is test is presented in Table 5 10 below. When considering the changes over time for SA learners, results show a significant increase in WTC. T able 5 10 Paired T Test of Willingness to C ommunicate in Spanish for s tudy a broad l earners WTC s ection ( p re p ost) Mean St andar d d eviation Standard e rror m ean t df Significance Speaking 6.77 7.27 2.42 2.79 8 0 .023 *significance 0.05
115 W illingness to C ommunicate for A t H ome L earners Following the same data collection and analysis methods, WTC was also evaluated for the AH learners. As with the SA group, univariate analyses were run for both the WTC in English and the WTC in Spanis h pre program questionnaires in order to determine if there were any individuals who were significantly different from the group. Results of these tests revealed no significant differences, indicating that all learners were equally willing to communicate i n the respective languages. (This is to say that they were equal within each language, not that their WTC was the same for both languages.) Table 5 11 presents the simple means for the changes over time in WTC in Spanish for the AH learners. In general, th ese means show a slight increase in WTC Table 5 11 Simple means of W illingness to C ommunicate in Spanish for a t h ome learners WTC s ection Mean N St andar d d eviation St andar d e rror m ean Speaking p re 54.42 14 12.82 3.42 Speaking p ost 58.35 14 10.43 2.78 Again, as was performed for the SA group, a paired T test w as run to analyze the change over time for WTC in Spanish. Table 5 12 below shows these results. Unlike the SA learners, there is no significant difference seen over time for WTC for the AH lear ners. Table 5 12 Paired T Test of W illingness to C ommunicate in Spanish for a t h ome l earners WTC s ection ( p re p ost) Mean St andar d d eviation Standard e rror m ean t df Significance Speaking 3.92 11.09 2.96 1.32 13 0 .208 *significance 0.05
116 The sub set of AH learners who took two Spanish courses was also considered here, although unlike with fluency no differences were seen here between this group and the entire group of AH learners. The simple means for these learners were 50.67 for pre test and 54. 67 for post test. The results of a paired T test revealed that no significant difference was seen over time (p=0.443). Comparing S tudy A broad L earners to A t H ome L earners At the outset when comparing the SA group to the AH group, a univariate analysis of v ariance was run to determine if there were any statistically significant differences in WTC in Spanish levels between the two groups. The results of this test found no significant difference between the SA and AH learners. In other words, both groups start ed their programs with comparably equal levels of WTC in Spanish. This is important because it shows that while the SA group is a naturally self selected group there is no indication that this group is naturally more willing to communicate in Spanish that To compare the change over time in WTC between the SA and AH learners, a general linear model w as run ; the results of the model are presented in Table 5 10. As can be seen, no significant difference between the SA and AH learners was found. Again, the SA learners were also compared to the sub set of AH learners but no significant difference was seen between the two groups. Table 5 1 3 Comparison of s tudy a broad lea rners to a t h ome learners for change in W illi ngness to C ommunicate over time WTC s ection Type III s um of s quares df Mean s quare F Significance Speaking 22.236 1.000 22.236 0.461 0.504 significance 0.05
117 Th e s e finding s indicate that learning context may have impacted WTC and as learners spent more time interacting in the native speaker community, they became more willing to communicate in Spanish, however this possibility can only be considered when comparing the data here to the more traditional AH setting. Since the possibility of interacting wit h NSs was a reality for the SA group, their WTC may have initially been lower due to their reservations about communication and the realization that it might not be as easy as they originally may have believed. This result is unanticipated due to the fact that SA learners are often described as a self selected group of learners who are more extraverted in the L2 and are more interested in communicating with NSs. Instead, what is possibly being seen here is that the SA learners are demonstrating personal awa reness that their abilities may not be as strong as they originally thought. This lower level of WTC was followed by an increase in WTC although what caused this change is not clear in the quantitative data. Different from the SA learners, the AH learners were not able to show the same significant increase in WTC. However, the important factor here is the already high starting point of the AH learners. At the beginning of the six weeks, these learners reported themselves to be fairly willing to communicate in most situations. In th e AH environment, interaction with NSs appears to be less likely. Because of this, the se learners may have overly estimated their willingness to use Spanish since it was not in their immediate reality. Unlike the SA learners that knew that every day they would be encountering a NS, the AH learners had less contact with NSs and they likely knew that there was little possibility in the near future of increased NS contact.
118 Correlations between D ata S ets C orrelations between data sets were run in order to examine the relationship between the various factors investigated The first series of correlations explored the possible interactions between fluency features and the amount of Spanish used during the program. T he second series of co rrelations examined the relationship between WTC and the amount of Spanish used. The third series of correlations that were conducted sought possible connections between WTC and fluency. Fluency and a mount of Spanish u sed Correlations were run to compare t he fluency measures against the amount of Spanish used during the course of the program for each group of learners separately and for the entire participant population as a whole. For each of these correlations the values from the post program interviews w ere used for the fluency scores. N o significant correlations were seen between the amount of Spanish used and any of the fluency measures. The SA learners reported spending over five hours a day speaking Spanish, and it was originally believed that a relat ionship would be found between this amount of language use and the fluency measures. However, no relationship was seen between this use and their fluency scores. The same correlations were run for the AH group to determine if any significant relationships could be seen between their fluency and use of Spanish. Once again, as was seen with the SA group data, no significant relationship was seen between the amount of Spanish used and the fluency measures. The AH learners reported a lower level of Spanish use, ranging from one to two hours per day, and it was previously believed that a possible relationship could be found between the amount of language used and fluency; yet, no such relationship is seen for th e s e data.
119 Finally, when looking at the possible rela tionships between the total amount of Spanish used during the programs and the fluency measures, correlations were run for all the participants together. Since there was a clear difference in the amount of Spanish used between the two groups, it was though t that perhaps when looking at the entire participant group a relationship would be seen between language use and fluency features; yet again, no significant correlations were found between the among of Spanish used and fluency measures. Given the concepts put forth by the Input and Interaction Hypothesis, and the Output Hypothesis, it would be expected that some relationship would be established between the amount of L2 contact and the fluency improvement seen. When considering the data from the refle ction blogs as well as questions on the language contact questionnaire, it can be seen that the SA learners appear to have made close personal relationships with NSs and that based on these relationships the learners had a large amount of Spanish communication. It is possible that through these relationships, learners were able to spend more time communicating, but this additional the original expectation that learners who spent more hours communicating in Spanish outside of the classroom would make more significant progress in fluency than learners who report lower levels of communication must be rejected. W illingness to C ommunicate and the a mount of Spanish u sed As was done wit h fluency measures, correlations were run between WTC and the amount of Spanish used during the two programs in order to determine the relationships, if any, between these factors. As was mentioned in the previous chapter, the amount of Spanish that was se lf reported by students on a weekly basis; based on
120 the total number of hours spent communicating in Spanish an average score was calculated for each group and for all participants together. As with the fluency measures, correlations were run separately fo r the SA and AH groups, followed by correlations for all participants. For all correlations the WTC score from the post program questionnaire was used As with the fluency features, no significant correlations were found between WTC and the amount of Spani sh used for either group of participants, nor was any significant correlation found for the either population of participants. Correlations between f luency and Willingness to C ommunicate One of the three main goals of this study is to attempt to show a rel ationship between fluency and willingness to use Spanish. To analyze these possible links, correlations were run between the fluency measures and the WTC sections. Again, three sets of correlations were conducted, the first for SA learners only, the second for AH learners only, and the third for the entire group of participants. The results of all three sets of correlations find n o significant correlation for this data between WTC and fluency It appears that based on this data set for these participants th at there is no relationship between WTC and fluency. Additional Information Results and Discussion Additional information for this investigation comes from the personal reflection blogs written weekly by all participants 9 Although the majority of prior r esearch into WTC has relied on questionnaire data, it was believed that the learners should be able to explain their thoughts toward WTC in their own words so that a more complete vision 9 One participant from the SA Group, Student 9, chose not to complete any of the blogs, therefore for this data set the total number of SA participan ts is 8 rather than 9.
121 of WTC could be obtained. B y writing about WTC in their own words le arners were able to explain in greater detail the factors that they considered to be part of their willingness to use Spanish, an element that was not able to be captured by the questionnaire data. While both types of WTC were self reported, the addition o f the blog data helped allow us to go beyond the mere numbers of the questionnaires. At the time of this study, only one other study has asked participants to reflect upon their WTC in a more open ended fashion (Kang, 2005); th e current study, therefore, a dds to this newer method of evaluating WTC. To summarize briefly, after the blog data were coded according to the coding scheme, it was found that some codes were common across the entire data set while others were specific to the learning contexts. These will be presented here as common tendencies and differences between learning contexts. Additionally, two unique participants were seen in this data and their individual differences are also discussed here. Common Tendencies Within the common codes seen ac ross the data set, t here were nine common tendencies that of thinking about people that were discussed throughout the blog entries from both the SA learners and the AH learners. Alth ough the codes pertaining to setting, activity and strategy were also commonly mentioned across the data set this was to be expected and these codes will only be discussed in relationship to other codes. ( It should be noted here that these tendencies are p resented in random order and no one tendency was seen as more important than another. )
122 W illingness to Communicate s pecific tendencies Five of the common tendencies found in this data used a reference to WTC and are as follows: WTC depends on situation WT C pertains to mood WTC is need driven WTC is higher when interlocutor does not speak English and WTC is higher when there is a feeling of acceptance The first tendency mentioned here, WTC depends on situation (frequency for SA 1.75, for AH 1.79, for al l 1.77), is more general and indicates that WTC is state like (in addition to being trait like). The participants used this description of WTC to indicate that a variety of factors may impact their willingness to use Spanish in any given moment in time. Wh ile there may be certain factors that can predict more or less WTC, their willingness to use the L2 strongly depends on that particular moment. A participant from the SA group describes this tendency in his first blog posting: In general it depends on the situation, the most uncomfortable thing is when a Spanish speaker talks to me on the street on in an unexpected situation because I'm not as prepared... (SA participant 8, Blog 1) Similar to this quotation, a student from the AH group wrote in his second b log: I think my reaction to communication compl etely depends on the situation. (AH participant 2, Blog 2) This tendency is logical, in that one can expect that any number of variables may influence the willingness to use the language, even if the situation is familiar or habitual. Although the participants discussed the situational nature of WTC in the six different their emotions tied to those situations influenced their communication. Due to this, it is most likely that this tendency arose from the prompt The concept of situations was also used in the WTC questionnaires as learners ranked their WTC based on the situation
123 provided. What is of note here is that the partici pants confirmed that for them WTC can be situational and that they continued to use this description of WTC throughout their blogs rather than describing their own personality traits as factors of WTC. The next tendency WTC pertains to mood (frequency for SA 1.50, for AH 1.57 for all 1.55) indicated that the participants needed to be in a good mood and be willing to put forth the effort necessary to communicate in Spanish. Learners went on to explain that if they are feeling frustrated or tired, they are much less willing to communicate than when they are in a better mood. This finding is not unexpected in that one could assume that using the L2 requires more effort than using the L1, and if a person is in a poor mood the additional effort needed for L2 co mmunication may not seem to be worth the trouble. The following two quotes come from two different SA learners but are quite similar in this respect: My willingness to communicate in Spanish directly pertains to my current mood. If thing or angry about something then it is very difficult to me to speak in another language. Also, if tired or sick I prefer to speak in English speakers of English I have to be i n a very good mood to feel willing to communicate in Spanish. (SA Participant 1, Blog 3) A big thing for me is my current mood. When I'm tired and worn out, it feels like it takes a lot of energy for me to listen and comprehend what others are saying in Sp anish and for me to speak in reply. I get easily frustrated and then don't say much and let others carry on the conversation. (SA Participant 3, Blog 3) While this tendency was expanded upon more by the SA learners, it was also mentioned by the AH learners who shared the belief that a bad mood leads to being less willing to use Spanish: Sometimes I have an urge to speak in Spanish sometimes while at other times I tired or frus ( AH Participant 9, Blog 3)
124 This tendency appeared in three blogs (Blogs 3, 4, and 5) when participants were specifically asked about being willing to communicate in Spanish, however, mood was never mention ed in the prompts. This finding in unique in that mood has not been considered before in WTC research or data collection. For example, in the questionnaires that are widely used there are no situational questions that refer to the is factor was mentioned by various learners in both learning contexts it appears that mood should be considered and in particular when other situational factors may be influencing the mood o f the learner. The SA learners could be considered to be in an emo tionally an d mentally demanding context, and d ue to that context, it is possible that these learners frequently experienced poor moods which impacted their willingness to interact and communicate with others. In another common tendency WTC was shown to be need driven (frequency for SA 2.25, for AH 1.57, for all 1.82) indicating that when a participant needs to accomplish a specific task or obtain a certain item, they are more willing to communicate than in other situations. Again, this can be said to be a logical conclusion because when one has a need to fulfill, and will suffer a consequence if that need is not met, one is more likely to take on the challenge of satisfying that need. This tendency can be seen in quotations from both a SA participant and a n AH participant: I usually initiate conversation with native Spanish speakers because I need directions, to order food, etc. (SA Participant 5, Blog 5) during a trip to Miam i I had to speak to a convenience clerk that only spoke Spanish to get the key to the bathroom .. (AH Participant 8, Blog 4) In general, when discussing this tendency the participants mentioned specific needs, dealing with personal comfort or direct perso nal wishes. Although it is a logical
125 conclusion to state that WTC will be higher when driven by need, it is a factor that is not directly addressed in WTC research. Needs are al lud ed to when asking participants to consider certain situations; on the WTC qu estionnaires used there are situations which are more needs driven than others. Yet when analyzing data and providing results, this needs based WTC is not shown to be an important factor. With these participants, it appears that needs impact their decision to use Spanish and it is possible that if the need is not present, they may choose to not communicate. Also of interest with this tendency is that, for the SA group, needs strongly influenced them in initiating conversations with NSs. This was different f rom the AH participants who did not report needs as a main reason to initiate conversations with NSs. (This difference will be discussed in greater detail in the later section about differences between the two learning contexts.) In all cases, learners fro m both groups mentioned this need with reference to using Spanish in a public place or specifically at school and did not mention This indicates that these learners are more likely to have to use Spanish in public because with their friends they are still able to rely on English. Furthermore, there are clearly times when learners are in public situations and they need to speak Spanish because the other person speaks only Spanish or very little English. This type of need le ads into another common tendency When participants are around other speakers who only speak Spanish they report themselves to be more willing to communicate in Spanish than with interlocutors who know both English and Spanish (frequency for SA 1.50, for A H 1.79, for all 1.68) There was a sense of necessity in this tendency indicating that participants were more willing to use Spanish in these situations because there was no other option available if they
126 wanted to communicate. With other English speakers this same necessity is not present. This can be seen from one of the quotations of the SA participants: I am very willing to speak Spanish around my homestay, because I have no other option. My seora and her two children do not speak any English. When I a m around my classmates, I find it very difficult to speak Spanish. We get along so well, and have so much to talk about, and it is difficult to communicate effectively in Spanish. (SA Participant 5, Blog 3) This tendency also brought with it a sense of poi ntlessness of speaking to another competent English speaker in Spanish. As one of the AH participants expressed: Typically, I feel that I am not very willing to use Spanish in my daily conversations. For one, most people that I would be speaking Spanish to speak English as well Spanish. (AH Participant 2, Blog 3) The SA participants d id indicate though that there we re times when they chose to speak Spanish with non English speak ers in a more positive way. One SA participant stated: I feel pretty comfortable using Spanish to communicate, especially with my host mber to speak in Spanish to people I k now speak English fluently. (SA Participant 4, Blog 3) In a similar way, the AH participants also cited being more willing to use Spanish with Spanish only speakers as rarer opportunity providing them with a valuable c hance to communicate with a NS as indicated by this AH participant: You can only learn so much from a book. For that reason, I love speaking (or important to talk to the m when I can. (AH Participant 1, Blog 5) Unlike the sense of need or obligation to speak Spanish mentioned previously, these two comments indicate a willingness to use Spanish that stems from their willingness to make the most of the opportunities given a nd speak Spanish when they have the chance to do so. There is still a slight difference between these two quotations; in these cases, learning context may play an important role here. For the SA participants there is
127 a wealth of opportunit ies to speak Span ish and they make take it for granted while they are in the SA setting. The SA learner mentions indicating that she knows it is important to speak Spanish while in the abroad context, but yet will still use English a t times. However, AH participants may be more aware of the rarity of communication with NSs and report a willingness to seize that opportunity to communicate. Although both groups of learners report having friends that speak Spanish or are in Spanish class es with them, they do not see these friends as much of an opportunity for speaking Spanish. Although one of the settings in the coding scheme rare for these particip ants to speak Spanish in a casual setting with friends. As part of this tendency speak both languages although in most cases the dominant language of the communication was still English. This thought is conveyed in this quotation from a AH learner: In reality, I use more Spanglish than actual Spanish. My Spanish friends will usually throw some random Spanish phrase into the English conversation subconsciously (like yelling at a basketball game or making a general exclamation) This reliance times when using Spanish is undesirable for these learners Perhaps at this level these learners feel there is too much effort needed to speak Spanish and are not willing to do so when they are in a relaxed setting. There may also be a type of social stigma associated with speaking Spanish outside of the classroom with other English sp eaking friends; learners might feel that others will perceive them in a poor way, as a nerd for example. ponses
128 on the WTC questionnaire, in that, when asked if they would participate in various Spanish speaking activities with other native English speakers they indicate a higher willingness to communicate score. This is in essence the problem of using self reported data because in some instances, such as the questionnaire, participants w a n t to show themselves in the best way possible, yet it other ways they may reveal their actual behaviors. WTC is also seen as strongly dependent on a feeling of acceptance (frequency for SA 2.13, for AH 1.93, for all 2.00) In this tendency learners ment ioned feeling accepted or feeling like part of a group, and that they felt more willing to use Spanish when they were in an environment where they were free to make mistakes without being judged. For some learners this environment was in the classroom, for others it was in a social setting, and for others it could be just about any place. For example one student from the SA group explained this feeling as such: When I speak Spanish in the classroom I feel like my peers are unjudgemental [sic] of my Spanish and am more comfortable speaking with them. I am by no means uncomfortable speaking with native speakers, but I know that they will hear and n otice all of the errors I make. ( SA Participan t 2, Blog 4 ) Another student from the AH group had a similar stateme nt to this, writing: I have always been somewhat hesitant to speak Spanish in class. As ridiculous as Outside of class, I am usually more comfortable speaking Spanish, e specially when speaking with native Spanish speakers. Native Spanish speakers know that I am American and often do not expect me to speak perfect Spanish. This allows me to be more comfortable expressing myself without worring [sic] so much about using per fect grammar. AH Student 14, Blog 4 Furthermore, an important aspect the learners highlighted in this feeling of acceptance was a sense of knowing that there was more focus on communicating in general, and less focus on speaking correctly. If learners felt that they would be embarrassed by
129 errors, they reported feeling less willing to communicate. These situations of acceptance and belonging were also often coded as times when the learners had a strong personal connection to their interlocutor. It is likely that learners will be more relaxed and comfortable around the people they are closest to and would be therefore more willing to take the risks of making mistakes and trying to use Spanish. Like the other tendencies seen here that related directly to WTC, this is a logical result, in that being in an emotionally safer environme nt, one is more willing to try. In the reverse situation, being judged and seen as an outsider, it may make one too unwilling to put forth the effort of speaking the L2 for fear of fa ilure or of portraying oneself in a poor fashion. Non s pecific c ommon tendencies The other four common tendencies found in the blog data do not specifically mention WTC but are related factors that can be said to lead to higher or lower levels of WTC. The se common tendencies are that Comfort level can vary with speaking to NSs NSs are helpful/you can learn from NSs Lack of vocabulary causes problems and Challenges lead to more communication The first tendency mentioned here implies that the participant s feel varying levels of comfort while speaking to NSs and that there may be different factors that influence their level of comfort (frequency for SA 3.13, for AH 2.21, for all 2.55) I n most cases these participants do not state what those factors are, b ut only mention that their comfort changes as they speak to NSs while describing the situations in which they are communicating. T his may be seen in the following quotation from a SA participant:
130 When I have to speak the language to a native, I have varyi ng levels of comfort. conversation or talking with my host family, then I feel slightly uncomfortable. (SA Participant 1, Blog 1) It is possible here that this participan t feels more comfortable in the situation of ordering food because he knows the language that is expected of him and he also is familiar with he must be prepared to di scuss virtually any topic using the necessary vocabulary and grammar that he might not be very familiar with or has not yet learned. In many cases this may be the reason behind why learners feel more or less comfortable when speaking Spanish. Other times t hough, there may not be a lack of knowledge but rather highlighted by an AH participant: speaker. For example, the bus driver of American Etours that I take almost every There are times little nervous and I just (AH Participant 10, Blog 3) It appears that the level of comfort c ould feel a certain level of comfort before they are willing to try to speak Spanish. This level can be unique for each p erson perhaps indicating more of a trait like element of WTC. While at the same time, comfort may change from moment to moment while communicating with the same speaker and thus leading to moment to moment changes in WTC. Further, it is also possible that a person who is very willing to communicate (for example to fulfill a need) could still be in a situation that makes him/her very uncomfortable but will continue to communicate despite their comfort level. W hile comfort level may lead to higher or lower WT C, there may be other factors that are
131 more important. This hierarchy of factors has been seen previously in WTC research and is the basis of the Heuristic Model of Variables Influencing WTC (MacIntyre et al. 1998). Finally, it is possible also that the c oncept of comfort level was triggered by the not the main topic of that blog, nor w as it the topic for any other blogs, however the tendency across several different blogs, however, not just for the first and fourth blogs. Also regarding interactions with NSs, the participants showed a tendency to believe that NSs are helpful to them and that they are able to learn from NSs (frequency for SA 1.50, for AH 1.64, for all 1.59) While this tendency did appear across the six blogs, it was mostly centered o n blog 5 when asked to write about their encounters with NSs. However, in this fifth prompt, no mention was made of the benefits of communication with NSs but rather was focused on the extent and nature of their communication and asked the participants to explain their feelings toward that communication. It appears that one of their feelings is that they will learn from their communication, as is expressed by one of the AH participants: When I speak to native Spanish speakers, I don't feel nervous or hesita nt to of a Spanish speaker than the person, but it makes me a better learner. It gives me the oppurtunity [sic] to learn from a primary source, and it allows me to apply and improve in what I already know. (AH Participant 5, Blog 5) In this quotation, it can be seen that due to the belief that NSs are a source of learning the participants reported that they can feel more willing to communicate in order to obtain more kn owledge about Spanish. When compared to the previous tendency it is interesting that even though the participants report having varying levels of comfort
132 when speaking to NSs they can still view them as a source of knowledge which may make them willing to communicate. This belief was shared by the SA participants as can be seen in the following quotation: I have no concerns or worries when talking to my host mother because she is very patient and understanding and corrects me if I'm wrong, or tells me the proper way to say something. I can learn alot [sic] from her. (SA Participant 3, Blog 2) While this belie f of learning from NSs may lead to more WTC it cannot be said that this is always the case. While the SA participant here does indicate that she is c omfortable when speaking to her host mother and knows that she can learn from her, she does not imply that this makes her more willing to communicate. Therefore, like comfort level, the knowledge that a NSs can impart during communication may not be a stro ng enough stronger variable is inhibiting them from communication at that moment. A tendency that speaks primarily to language skills explains that when there is a lack of vocabulary knowledge, comm unication is difficult and often frustrating (frequency for SA 2.13, for AH 1.50, for all 1.73) This lack of knowledge often led to learners being less willing to communicate because they knew that would not be able to sustain a conversation about certain topics without a wider vocabulary. T his problem is illustrated here by two SA participant s : It is more difficult for me to tell a story in Spanish because my vocabulary is much smaller and I cannot include a lot of details or thoughts. (SA Participant 4, Blog 1) What holds me back is not feeling nervous, but rather not having adequat e vocabulary to express myself. (SA Participant 5, Blog 3) Lack of vocabulary was also expressed by the AH participants as a hindrance to their communication implying that it may lead to unwillingness to speak. Two AH participants expressed this thought in the following quotations:
133 I think the only thing holding me back, beside [sic] needing a larger vocabulary, is my pride. (AH Participant 7, Blog 3) If I really didn't have a ny idea what they were talking about or needed because they were using vocabulary I didn't know, etc I would have to find someone who spoke Spanish as a first language to help. (AH Participant 8, Blog 2) Vocabulary was not menti oned in any of the blog prom pts, but rather came entirely from the participants. In previous WTC research, lack of a particular skill set has not been widely discussed as a variable, however it is believed that learners who are at an overall lower proficiency level will be less willi ng to use their L2 due to the problems they may have with the mechanics of communication. It appears that as learners are able to gain more vocabulary knowledge and expand their ment al dictionary this problem can lessen and may cease to be a n important a f actor in determining if a learner is or is not willing to communicate. Like the other tendencies that have recently been discussed in this section, lack of vocabulary may not be as strong a hindrance and if another stronger element is present learners may overcome this lack and communicate. Such a thought is expressed by one of the SA participants: If I really want to find a way to communicate something then I can find a way to say it. (SA Participant 2, Blog 1) This tendency is also interesting in that it is the only common tendency found here that may impact oral fluency. As participants were speaking in their interviews they had a tendency to use dysfluent elements when they did not know a word or were unsure of a word. (In the instances referred to here learners would expressly say that they did not know the word or as k the interviewer for the word.) This is not to say that there is a direct relationship between oral fluency and WTC here, but merely to show that they can share an influencing factor.
134 The final tendency seen across the participant blogs was that challenges in communication led to an increase in WTC (frequency for SA 1.75, for AH 1.64, for all 1.68) In particular, this tendency was seen in the second blog where participants were expressly asked to discuss a time that they failed in communication. In many cases, challenges may have arisen repeatedly for example in the SA group many learners str uggled with understanding directions to a location, but the challenge may also have been a one time experience. In general, learners came away from this miscommunication with a determination that they could overcome the problem and a willingness to try aga in. One of the SA students describes the feeling of confidence he received upon succeeding where he had previously failed: Last week when I first started asking directions I would have no idea what their [NSs] responses meant. It was disheartening; how ever, I needed to be able to understand them if I was to ever get home again. I asked tens of Spaniards for directions each day and slowly throughout the week I was able to understand more and more of what they were saying. By the end of the week I was abl e to understand almost every word of the directions and even ask for additional directions. It has been a boost of confidence and has motivated me to keep trying to communicate. ( SA Participan t 1, Blog 2 ) Two additional elements are particularly interestin g in this quotation, first is that the learners WTC is need driven here and due to that the participant was willing to continue to try to communicate in order to meet his need of getting home. While the desire to simply understand may also have been influe directions appears to be a stronger factor here. Second, the success of understanding resulted in an increase in motivation to continue trying to communicate with NSs. Motivation is seen in WTC research often as a d to use their L2. This concept appears to be su pported here by this quotation in that
135 successful communication leads to motivation and in turn motivation leads to a willingness to use the L2 more. It also may be th at challenges and overcoming them could lead to an increased motivation in L2 learners, thus indicating that while challenges can influence WTC they may not be as directly related as other factors. The AH learners also expressed the sentiment that challeng es led to additional communication. actually be a risk; my hesitation comes from the fact that I hate to be wrong. [...] It is a very silly attitude to have toward academi cs, languages especially, since the becoming more mature about it though and realizing that I need to overcome this and just speak Spanish. (AH Participant 2, Blog 3) In this recognizes that it is necessary to overcome this dislike, accept that he will make mistakes and know that he can learn from them. It is important to consider that some chall enges may be personally driven and that it may be something that learners have to realize for themselves before they can work to surmount that problem. In general, the sense of continuing to try to communicate is prevalent in the data and may stem from our own American cultural beliefs. From a young age we are taught to always try our T his belief appears to be well ingrained in these learners because they were much more likely to indicate that challenges lead to an increased WTC rather than indicating that they were likely to give up when met with a demanding situation.
136 Differences betwe en S tudy A broad L earners and A t H ome L earners There were several differences seen in the blogs between the two groups. There were four tendencies seen in the SA blogs that did not appear with great frequency in the AH blogs (one or two learners may have al so mentioned the topics but they were in the minority of the AH group). Of these four tendencies two were beliefs held by the participants: that Lower confidence leads to lower WTC and Some NSs are impatient and do not wish to talk to foreigners T he othe r two tendencies we re concerned with topics discussed: SA participants wrote about specific speech activities more than AH participants and SA participants referenced specific NS relationships more than AH participants Perhaps because they were experienci ng lower confidence in their communication abilities while abroad, the SA learners pointed out in their blogs that when their confidence was lower their WTC was consequently lower (frequency for SA 1.50, for AH 0.28) This is expressed in the following quo tation: what I want to eat because I am not loud. It makes me less willing to speak Spanish in restaurants... (SA Participant 7, Blog 1) When confidence is lower, it is logical that one would avoid speak ing when possible This is an expected result as similar findings have been show n in previous WTC research where confidence is closely related to WTC and learners who reported having lower confidence in th e L2 reported being less willing to use the L2 Although low confidence is mentioned by a few of the AH learners, collectively they do not appear to have experienced the same moments of lower confidence resulting in lower WTC. It is possible that these par ticipants were naturally more confident in their abilities tha n the SA learners although it seems more likely that this difference is related
137 to the learning contexts. Since the AH participants had used Spanish principally in an academic setting they wer e perhaps not as accustomed to the pressures of speaking Spanish in a real word setting and were not as acquainted with situations that might cause them to have lower confidence. It is also possible that without an extended period of real world tests of t heir abilities, these learners may never experience a decrease in their confidence if they continue to do well in Spanish courses and progress to other more advanced classes. The SA learners, however, did experience this real world testing and as a result, found themselves in situations that lowered their confidence in their speaking abilities and seemingly influenced their WTC in Spanish. Also due to their authentic target language experiences, the SA learners had the opportunity to interact with a wider v ariety of NSs than the AH learners. In particular, the SA learners described NSs differently than the AH learners. While, as mentioned above, both groups of participants found NSs to be helpful and a source of knowledge, the SA learners also described some NSs as being impatient and reluctant to speak to foreigners (frequency for SA 1.75, for AH 0.29) This perceived attitude seems to have speak with NSs who appeared disinter ested in communication or impatient with the learners. The following quotations highlight this tendency : It is very difficult to speak in Spanish to natives also because they seem very impatient with me and begin to speak rampantly and I do not feel like m aking a fool out of myself. (SA Participant 1, Blog 3) ...it was frustrating that the native speakers weren't very cooperative and helpful. This made me uninterested in speaking Spanish. (SA Participant 8, Blog 2) I feel like the restaurants and public sto res have been the hardest source for me to speak Spanish because they always seem to be in a hurry or frustrated, so they don't even try to listen or help me communicate. [...] Now I avoid sales people or situations like that altogether. (SA Participant 3, Blog 2)
138 This tendency is important to consider in WTC research, because so often the focus is on the learners and their willingness to use the L2, and not the potential interlocutors. If other speakers have no interest in communicating with learners it wi ll not matter how willing to communicate the learner is, communication is not likely to occur. Even in situations that would normally prompt these learners to be willing to use Spanish, a good mood or a personal need, can be subverted by a disinterested in terlocutor. This disinterest appears to result in lowered WTC when the learner finds themself in a similar situation. As Participant 3 pointed out, after her poor experiences she had a tendency to avoid the situation all together. Encounters with impatient or disinterested NSs may be particularly disheartening to SA learners because of an underlying assumption that they will have unlimited opportunities to speak Spanish while they are abroad. When learners encounter these types of NSs they must attempt to n ot let themselves be overly discouraged and to keep on trying to engage other NSs in conversation if they wish to maximize their time spent abroad. However this practice may be much easier said than done. What is also of interest here is the defensive mann er shown by the SA learners; the SA learners may be deflecting the problems they were having while communicating by placing blame on the impatience and attitudes of the NSs. Perhaps rather than accept their own shortcomings in speaking Spanish, the SA lear ners describe the problem as out of their control and a result of the NSs. It is possible that the AH learners had experienced impatient or disinterested NSs at some point in their Spanish speaking career, however they did not report this in their blogs. A H learners had a tendency to view NSs in a more general sense and did not distinguish a separate group of NSs like the SA learners. There are two possible reasons for this, first, when the AH learners
139 discussed their interactions with NSs the majority of t hese interactions occurred in the United States where the NSs were often the foreigners in this situation and they may have been more sympathetic to someone learning a language. Second, in many of the reported interactions the AH learners were at work in s ervice jobs where they were helping the NSs and perhaps this situation influenced the NSs to be more willing to continue communication. The last two tendencies about specific speech activities and specific NS relationships, appear to be strongly linked t o the learning context in that the SA context provided greater opportunities for both of these. Since the SA participants were using Spanish in a variety of settings and with a variety of interlocutors it unsurprising they completed a larger variety of spe ech activities, such as ordering food (frequency for SA 1.50, for AH 0.36) asking for directions (frequency for SA 1.88, for AH 0.0) getting help (frequency for SA 1.75, for AH 0.71) and talking with their host families (frequency for SA 2.25, for AH 0. 0) As SA learners discussed the various topics in the blogs these activities were used as specific examples to reinforce their point. This can be seen in the following quotations: Like [ SA participant #4 ] my homestay family does not speak English either. I am always speaking Spanish with the family, my use of Spanish is critical in planning events and meals, and for following the rules of the household. (SA Participant 2, Blog 3) Most of the time others initiate the conversation, unless there is something specific that I need, such as directions, information, or to order in a restaurant. (SA Participant 3, Blog 5) The SA learners have personal experiences to refer to when discussing their WTC and the likelihood of being willing to communicate in a similar s ituation. Based on this previous experience the SA learners may have a more realistic view of how willing they
140 might be to use Spanish and are, perhaps, likely to consider times in which they might not wish to use Spanish. The AH learners did not demonstra te the same tendency to discuss specific speech activities in their blogs, suggesting that perhaps they did not have the same type of specific experiences and could not relate to potential problems in the same way. It is also possible, however, that the AH learners simply chose to take a more general approach when discussing these topics and did not feel a need to use specific examples. The SA learners also discussed specific relationships they had developed with NSs and how these relationships were benefic ial to their use of Spanish. In the fifth blog learners were asked specifically about communication with NSs, so it is likely that this tendency stemmed partly from that prompt. However, throughout their blog entries the SA learners discussed NSs even when not specifically prompted to do so. In many cases these NSs were members of the host family, in particular the host mother, but in other instances the NSs had developed friendships outside of their homestays. Two of the SA participants discuss their relat ionships with NSs in the following two quotations: I speak to my seora on a regular basis and she is pretty easy to hold a conversation with, but I do have a hard time when I want to tell her a story but I know some of the words or how to explain th em. Her husband is very helpful because he always corrects my grammar and I really appreciate him taking the time to fix what I say wrong, and explain the correct usage of the words/grammar. (SA Participant 4, Blog 2) There is this one bar called Srs Patat as that I feel very comfortable talking with the bar tender, David. He knows a little bit of English and works on that with me but also helps me with my Spanish. With people like David where I know they want to help me and are patient, I have no problem ke eping the conversation going and even enjoy it when I am doing well. (SA Participant 3, Blog 5) In these instances where SA learners wrote about developing relationships with NSs they also indicated that they were more willing to communicate with these ind ividuals.
141 The learners knew that these NSs were interested in speaking with them, would be patient with them, and would also help them with communication. The AH learners for the most part did not cite specific NS relationships in their blogs and in genera l spoke about their interaction with NSs as one time occurrences (This is not to say that none of the AH participants wrote about specific NS friends, there were three that did do this, but they were the only ones in that group.) Again this tendency appea rs to be a result of the learning environment. The SA participants were living with NSs during their program and were able to have daily interaction with those NSs allowing for them to develop personal relationships. These learners also had the opportunity of spending time in the L2 community and due to the length of their stay were able to develop at the least an acquaintance with other NSs. The AH participants were not able to have this same access to NSs during the course of the program (except of course with their classroom instructor) and so it was less likely that these learners would establish a relationship with a NS during this time frame. Although when considering WTC for the AH participants, the lack of closer personal relationships does not appe ar to have impacted their desire to communicate with NSs. As with the SA group, the AH learners also had three tendencies th at appeared across their blogs but were not see n in the SA blogs. Two of these tendencies were perspectives held by the participants about language and were Love for Spanish and Errors make one look bad. The third tendency was the strategy of Initiating conversations with NSs, which the AH participants reported doing more than the SA participants. In many of their blog entries, the AH learners expressed their love for
142 Spanish; often this was the first statement they made in their blogs. This can be seen in the following two quotations: I absolutely love Spanish and being able to communicate with different people. (AH Participant 1, Blog 1) In general, I love the Spanish language and a great deal of the cultural aspects and traditions of Hispanic people. (AH Participant 8, Blog 1) This tendency does not appear to have come directly from the blog prompts in that love for Spanish is never e xpressly stated in these instructions. Nonetheless, this seems to be in response to the question which asked them to write about how their emotions while using Spanish. Although these participants went on to more precisely define their emotions while using Spanish, these learners also felt it necessary to describe their emotional contact to Spanish in general. This tendency was only seen in the AH blogs (frequency for SA 0.0, for AH 1.57) and although it is possible that the SA learners also love Spanish, they did not express this directly in their writing. This love for Spanish can be said to fall with the realm of attitudes towards the language, implying a strong positive attitude. In some instances, as can be seen the first quote, this love for the langu age can also translate to a love of the ability to communicate. It is possible then that if these learners love having the ability to communicate there is a chance that they are also more willing to use their ability when given the opportunity. This strong connection with the language may be one of the reasons the AH learners ranked themselves at such a high level of WTC in the pre program questionnaire. The second tendency seen for the AH participants was the belief that Errors make one look bad (frequenc y for SA 0.50, for AH 1.57) These learners were concerned that when they made errors in Spanish, other s would perceive them as poor Spanish
143 speaker s This thought is explained by an AH participant when she discusses being in the classroom: om where I know everything I say is being monitored for grammatical errors. In most cases in the classroom everything is put on the spot and you have to react so quickly to a question answered [sic] and that's hard for diot making errors. (AH Participant 12, Blog 4) For these learners, errors were not directly mentioned with regards to WTC but they did express the idea that errors make them uncomfortable. Given the previous discussion about the importance of comfort leve l when speaking with NSs, it is possible to consider that errors lead to a lower comfort level and thus a lower WTC. The SA group did not seem to be as concer n ed about errors and how they were perceived Although a few of these learners did mention errors and their negative connotations, as a group they seemed to accept that errors were going to happen and did not appear to allow errors to influence their WTC Being abroad likely made the SA group realize that trial and error were essential for communicatio n, whereas the AH group was in a classroom setting that had an emphasis on accuracy strongly suggesting that this tendency is a product of program setting. The final tendency seen for AH group is the strategy of Initiating conversations with NSs (frequenc y for SA 1.00, for AH 1.64) Once again, it is most likely due to the prompt provided for the fifth blog that the learners chose to use this exact wording. Of interest here though, is not that they used this wording, but that the AH participants stated tha t when in the presence of NSs they were highly likely to initiate conversations. These learners were interested in seizing the opportunity to interact with these NSs apparently due to the rarity of this event. One participant explains it as such:
144 When I do encounter a native speaker I tend to initiate the conversation in Spanish letting them know that I also speak their language. Once the speaker knows that I can speak Spanish they tend to keep the conversation going, maintaining basic talk. Overall I pur posely initiate a conversation in Spanish in order to practice my skills and so that they can correct me. (AH Participant 9, Blog 5) In this quotation it can be seen that this participant is willing to initiate conversations with NSs to specifically fulfil l the purpose of practicing Spanish with that speaker and obtaining valuable feedback about her abilities. This tendency is directly tied to WTC in that by initiating conversations with NSs, these learners are showing that they have a high level of WTC. Th is tendency to initiate conversation may also be one of the reasons why this group showed high scores on the WTC questionnaire, as many of the situations described on the questionnaire would require the learner to initiate the conversation. Conversely, the SA learners appear to only initiate conversations with NSs when it is absolutely necessary, in a needs based situation for example. These participants indicated that they would rather take a passive approach and allow the NS to begin any possible communic ation. This tendency may stem from their belief that some NSs are impatient and are not interes ted in speaking with foreigners, or may simply be because it is easier to take the passive role in a SA setting. Based on their past problems with th ose situatio ns, the SA learners may have adapted their behavior to wait for the NS to initiate the conversation and thus show their interest in interacting with the learner. Specific I ndividual D ifferences While analyzing the blog postings for individual differences t here were two individuals who stood apart from the rest of the learners. One was from the SA group, SA Student 6, and the other was from the AH group, AH Student 4. These two individuals were unique in that they described themselves as almost completely fe arless
145 about communicating in Spanish and, essentially, portrayed themselves as almost always willing to communicate in any given L2 situation. Based on their blogs, both of these learners appear to be highly motivated to learn Spanish, and have low levels of anxiety towards the language. These two factors motivation and anxiety, have been considered to be indicators of WTC in previously designed models by MacInytre and Charos (1996) and MacInytre et al. (1998). This appears to be supported in the data col lected for both of these learners. While the statistical analyses did not find these individuals to be outliers in the quantitative data, their comments in the blogs were unlike any of the others in their groups. The student from the SA group did describe herself as somewhat nervous at first, but this quickly changed in her blog postings. While like other learners in the SA program she spent a lot of time speaking with her host family, she also sought out NSs whenever possible. The way in which she describ es these encounters seems to suggest that for the most part NSs were patient and willing to communicate with her, but it is also suggested that she was more aggressive at taking the opportunity to communicate than other students in the program. This can be seen when she describes speaking to a mailman: I find it becoming easier for me to talk to strangers and the like the more I stay in Spain. The other day I confronted a local postman for some information about mailing packages. It was really interesting. I learn [sic] that I still had a limited Spanish vocabulary about mailing things, but I was still able to communicate quite effectively. (SA Student 6, Blog 3) It appears, from this posting, that the student is willing to challenge herself beyond her abil ities, whether she may be cognizant of it or not. In addition, although she learned that her vocabulary was lacking in certain ways, she is still able to see the benefits in ther
146 encounters with NSs indicate that in most situations she was able to be successful and at the very least, was inspired to continue to keep trying to communicate with anyone she could. In her final blog post her desire to continue to seek out ways to i nteract with Spanish are very strong, as described by this quotation: I know I want to come back to Spain and travel around with friends using my Spanish. When I return to the United States, I am going to try my best to watch Spanish T.V. or listen to Span ish everyday. If that means one hour of Univision one and a two minute song on Youtube the next day, I will do anything possible to keep myself from falling off my current Spanish plato [sic]. ( SA Student 6, Blog 6 ) In this posting it can be seen that this student has come up with specific ways to keep herself in touch with Spanish, doing all that she can to maintain her current level of Spanish skill. While other students in the SA group expressed strong feeling towards maintaining contact with Spanish, th clearly distinguished her from the others. It is important to note here that although this learner was very willing to speak Spanish, she was also the least fluent speaker in her program group as judged b y the quantitative fluency measures discussed above; further, she was one of the least fluent speakers overall for the entire participant population (there was only one other learner with lower scores in the AH group). Therefore, this participant appears t o support the finding of the quantitative data that there was no relationship seen between WTC and fluency. In other words, a learner may be very willing to use Spanish but still exhibit a lower level of fluency. The student from the AH group had a similar sense of willingness to always use Spanish and interact with any Spanish speaker; however, his high level of WTC appears to come more from being extremely comfortable with himself and his abilities He shows a high level of confidence and a low level of a nxiety creating the ideal
147 context for high WTC Even in his first posting about communicating in Spanish, he seems almost blas about it, as can be seen in this quotation: Spanish is just another form of communication and speaking. As a result, it doesn't it. Generally, when conversing with a Spanish speaker, they are willing to help and or understand if someone is learning and generally appreciate the fact that someone is tr me nor am I ultimately nervous. ( AH Student 4, Blog 1 ) This learner did report having many Hispanic friends and frequently speaking to them in Spanish. In addition he writes about growing up in a community in South Florida where he heard Spanish on a daily basis. 10 For him, communicating in Spanish is nothing to get excited about nor anything to worry over. The only time he expresses being less willing to communicate is when his interlocutor speaks English as he describes here: have lived here a while, there [sic] are more afflu ent [sic] with [English] than I am than English, though the context can be a limiting factor, such as the aforementioned situation. Otherwise, why the hell not? ( AH Student 4, Blog 3 ) This posting again shows his casual attitude towards using Spanish as a means of communication. It is also evident in his blog posting that, unlike the other learners, he does not view communication in Spanish as needing additional effort. Altho ugh this learner writes that he is aware of his shortcomings and a need to continue working to improve his linguistic skills, he does not let any doubts about his abilities impact his willingness to use Spanish. Like the SA learner, this participant may be very willing to communicate and yet did not show a higher level of fluency than other participants. His fluency measures place him in about the middle of both groups, and he showed a 10 Since these were not family members and he indicated on his form that it had been seven years since he lived in that community he was not eliminated from the study.
148 decline in all measures in his post program interview (significant for s peech rate, average length of fluent runs, unfilled pauses, and dysfluency clusters) Again the results of a lack of interaction between fluency and WTC appear to be supported by this Conclusion When first looking at the results of the data analysis it appears that the SA context had a positive influence on fluency over the six week int ensive study period. T he AH context seems to have allowed learners to maintain their fluency on several measures (but not all) yet overall does not appe ar to have influenced significant gains in the factors examined here However when comparing SA learners to a sub set of AH learners who were enrolled in the same amount of credit hours, it appears that the additional coursework may have been a greater in fluence than learning context. The SA learners were still able to show a faster rate of speech overall, but the source of this improvement is unknown as no correlations could be shown between the amount of Spanish used and fluency features. The expectation that learners who spend more time interacting in the L2 community will show greater gains in fluency does not appear to be supported by this data. For WTC, again at first the SA context appears to have had a positive influence on WTC when considering the results of the quantitative data. Yet when the SA group was compared to the AH group (and the AH sub set) statistically there was no difference seen between the two groups When exploring the blog data it can be seen that the SA and AH learners share many of the same beliefs with regard to WTC and the factors that will impact their willingness to use Spanish. In addition, there were differences seen between the two groups. The SA learners appeared to show
149 differences that allowed them to be more specific w hen discussing situations that lead to lower WTC, and as well, referenced specific speech activities and personal relationships rather than showing a tendency to generalize. The AH learners showed differences that allowed them to discuss general instances that promoted communication and appear to have led to higher WTC. Finally the blogs revealed two unique participants with very high WTC whose high motivation and lower anxiety levels appear to strongly impact their willingness to use Spanish. A lthough inc reases could be seen for both fluency and WTC, no relationship was found to exist between the two f actor s. If anything it appears that regardless of their fluency abilities, the learners from both groups seem generally willing to communicate. This notion i s also illustrated in the reflection blogs of the two unique participants who despite begin very willing to communicate did not show the highest levels of fluency.
150 CHAPTER 6 CONCLUSION T his chapter offers some concluding remarks to summarize and contextu alize the findings of the current study. First, the research questions are answered and the ways in which this study supports and adds to previous literature about fluency and Willingness to Communicate ( WTC ) are discussed Next, limitations to the study a re presented and explained. Finally, plans and suggestions for future research are given. Answers to Research Questions Question 1 Fluency R esearch How does intensive short term (six weeks) language study relate to fluency in Spanish? How does oral fluency differ among learners in an at home context versus those in a study abroad context? For certain learners, intensive short term language study was shown to have a positive impact on oral fluency by improving specific temporal features of speech. The study abroad ( SA ) learners were able to show a significant improvement in speech rate and the average length of fluent runs indicating a shift towards more fluent speech and supporting the findings of previous research (Lennon, 1990; Freed, 199 5; Towell et al., 1996; Towell, 2002; Isabelli Garca, 2003; Freed et al., 2003; Freed et al., 2004; Juan Garau and Prez Vidal, 2007; Garca Amaya, 2008; Llanes and Muoz, 2009 ) I n the case of at home ( AH ) learners the intensive period of study does not appear to have had much of an impact on oral fluency in that significant improvements were not seen. Significant increases were found for hesitations indicating a shift away from mo re fluent speech. However, this does not necessarily mean that the intensive study had a negative impact on fluency. Given that only two minutes of the data were analyzed, it would be rather
151 rash to make such a strong statement. This methodology of analyzi ng two minute long samples has been fairly common in prior research (Freed, 1995; Freed et al., 2003; the hope of obtaining comparable results. Yet, perhaps it is nece ssary to look at larger segments of data before being able to make a more decisive conclusion. When comparing the SA group to the AH group, it was found that the SA learners made significantly greater improvement in amount of words, speech rate, the averag e length of fluent runs, unfilled pauses and clusters of dysfluencies. Overall, t hese findings support prior research that has also found an advantage for the SA context over the AH context in improving oral fluency (Freed, 1995; Freed et al., 2003; Freed et ) However, when comparing a sub set of the AH learners who were enrolled in the same amount of credits as the SA learners (six hours : three in a grammar course and three in an intensive oral commu nication course ), only one significant difference was found : a faster speech rate for the SA learners. This finding indicates that perhaps additional study in particular study with a focus on oral communication, can lead to gains in oral fluency regardles s of learning context In the prior research mentioned above, the amount of credit hours is not considered, but may be an additional reason behind the improvements in fluency. In many language based SA programs, learners have the opportunity to take a larg er amount of credits in their second language ( L2 ) than they normally would in an AH setting, and based on the results seen here in future studies these additional credit hours should be considered.
152 Nonetheless, the SA learners from the current study did manage to show an improvement over all of the AH learners in speech rate. Logically, this would lead one to believe that learning context might play a role in this improvement. Furthermore, this idea appears to support the Input and Interaction Hypothesis and the Output Hypothesis given the increased opportunity for interaction with NSs and the chance to produce output available in the SA context. However, there is no evidence in these data to support or deny that this thought While previous investigations (Isabelli Garca, 2003; Segalowitz and Freed, 2004; Segalowitz et al., 2004; Freed et al.,2004) were able to show that learners who spent more time interacting with native speakers ( NSs ) and using their L2 made greater improvement in fluency, the current study was not able to replicate these findings. This indicat es that interaction with NSs or producing output does not appear to have had much effect on fluency for these participants What then is the reason for this difference? It is possible that the qu ality of the communication can explain the different findings The SA learners did report spending many hours of extended conversations with their host family on their language use surveys and reference these conversations in their reflection blogs. Perhap s, through longer spontaneous conversation that broke away from the formulaic speech of service encounters or greetings these learners were able to develop skills that enabled them to increase their rate of speech. Unfortunately for this data set, the AH l earners had a tendency to leave that question blank on the questionnaire. We might assume then that the AH learners did not participate in extended conversations they simply did not report their extended conversations, and there is not a point of comparison to the SA group. It does appear useful though to find ways in which to
153 capture this type of quality of language use and perhaps through a more reliable source. In addition to collecting controlled interviews designed to elicit learner speech and/or measure other elements (such as proficiency), data collection may need to move towards natural data as means of better understanding how learners use language while abroad and how it may impact their oral fluency. Finally, it is enco uraging to find that learners were able to make some gains in fluency in such a short time frame. With the exception of Llanes and Muoz (2009) (with a month long SA period), the above mentioned studies investigated the speech of learners who had spent a m inimum of a full semester (13 to 16 weeks) abroad. It appears then from both the current study and the study of Llanes and Muoz that fluency can be positively affected early on in the immersion process and that improvements in oral fluency can happen rela tively quickly. However, it may be that the shorter time frame of the current study is a reason behind one of the differences between these results and the results of previous investigations. Unlike the semester long or year long investigations, no signif icant change was seen in the amount of unfilled and filled pauses for the SA learners in the current investigation. Once again, this is similar to Llanes and Muoz, and appears to indicate that, in order to reduce the amount of hesitations in speech, learn ers may need to spend more time in the immersion setting or, based on results seen here, take additional course hours Future studies should continue to look at shorter term SA programs in order to see if these findings can be duplicated.
154 Question 2 W il lingness to C ommunicate R esearch How does intensive short term (six weeks) language study relate to willingness to communicate in Spanish? How does willingness to communicate differ among learners in an at home context versus those in a study abr oad context? The answer to this question is similar to the first: WTC appears to be positively impacted by SA showing an increase in WTC yet does not seem to be as strongly influenced by AH study which showed maintenance of WTC The findings of this study support the findings of Yashima and Zenuk Nishide (2008) that WTC increases over time for students who participate in SA and is maintained over time for students in a traditional AH setting. This is important because it is the only other study to assess c hanges in WTC over time (at the time of the current study) Furthermore, when it is considered that the participants from Yashima and Zenuk Nishide (2008) participated in a ten month SA program, it is impressive that the SA learners from this study were ab le to show similar results after only having been abroad for six weeks. The majority of previous WTC research has focused on surveying large groups of learners one time to determine the different levels of WTC that can be found based on learning context, situation, sex, and age. There are only a few studies (Yashima, Zenuk Nishide and Shimizu, 2004; Yashima and Zenuk Nishide, 2008; Kang, 2005; Cao and Philp, 2006; Cao, 2006; Peng, 2007) that have used different approaches to collecting and analyzing data. Of these, only two have considered the role that SA may play in Nishide, and Shimizu, 2004; Yashima and Zenuk Nishide, 2008) and only Yashima and Zenuk Nishide (2008) has analyzed changes over time in WTC. Furthermor e, Yashima and Zenuk Nishide (2008) is the only study to investigate whether or not WTC can be seen to impact linguistic abilities. From
155 a methodological standpoint alone, the current study adds to the previous literature in that it introduced a new way to collect WTC data through personal reflection blogs and used a pre and post program WTC questionnaire design. Although the blogs are self reported data, they go deeper than the traditional WTC questionnaire allowing the participants to express their own i deas on this topic. Additionally, the current study also performed analyses to determine if a relationship could be found between WTC and fluency. The additional data collected in the blogs helped add to our understanding of WTC tive through the appearance of several common beliefs as well as certain belief s that may be linked to learning context. When considering tic Model of Variables Influencing WTC (see Figure 3 2 p. 66 ) it appears that severa l of the belief variable categories By adding new information to this existing model, it is hoped that a more complete picture of WTC will be shown and that the new findings of this study can be included in this mo del. In Layer VI under Intergroup Climate it is possible to add the belief of WTC being higher when the group is accepting and non judgmental Here though it might be that Intergroup Climate is placed too far away from WTC in the original model. The inform ation provided by the learners blogs suggest that this variable may be more strongly linked to WTC and therefore should be raised up to a higher layer. In layer V under Social Situation it appears logical to add the belief that WTC depends on situation. In this instance since the belief so closely mirrors the original concept of Social Situation, the placement in the model appears to be sufficient. In that same layer,
156 under Communicative Competence, it is possible to include the both the belief that a lack of vocabulary causes communication problems and that errors create a poor image Since th e s e belief s did not appear to have as direct a connection with WTC as other beliefs, they seem to be placed at a suitable level in the model. In Layer IV under L2 Self Confidence that belief that lower confidence leads to lower WTC can be added. There are several beliefs seen in the data that do not appear to easily fit under the existing categories and it appears to be necessary to add additional spaces for these belie fs. In Layer III, directly beneath WTC (Layer II), it would be beneficial to add a space for Need and a space for Mood. Both of these, were reported by the participants as being strongly connected to WTC and therefore should be closer to Layer II than othe r categories. In Layer VI, a space could be added for challenges, including the belief that the desire to overcome challenges will lead to more communication. This does not seem to be a variable that relates as directly to WTC as other variable and it appe ars to fit into the theme of that level Social and Individual Context. Finally, it appears that a layer may need to be added to the model to include attitudes towards the other speaker. This new layer could be comprised of two variables: desire to speak t o a L2 only speaker and perception of other speaker. The first variable in this new layer would consider the belief that a learner is more WTC when the other speakers can only communicate in the L2. The second variable would take into account the belief th at there are other speakers who may be impatient with learners or disinterested in speaking with learners.
157 Question 3 Relationship between Fluency and Willingness to Communicate What is the r willingness to communicate in Spanish ? The final element of WTC that was considered in the current study was a possible relationship between WTC and oral fluency. However, the results found here were not able to find a connection between the two This re sult is different from that of Yashima and Zenuk Nishide (2008) who were able to show a connection between WTC and linguistic ability, specifically language proficiency. Of course it is important to remember here that oral fluency is not equivalent to lang uage proficiency. It was originally hypothesized that learners who were more willing to communicate would demonstrate more fluent speech. However, the results of this study do not support this hypothesis. Based on the results shown here, it is possible tha t learners do not consider their own abilities when assessing their WTC, and while they may be lacking by linguistic measures this does not mean that they are lacking in their willingness to use the L2. In searching for a possible relationship between WTC and fluency, further rese arch is needed that can look at both of these factors in multiple ways. It may be that temporal features of fluency do not show a connection to WTC, but other features which pertain to a more complex definition of fluency might. Ad ditionally, if other data is collected on WTC that incorporates different viewpoints, outsider observation as well as self reported data for example, this may assist with finding the possible connection. Overall, the current research increases our understa nding of the concept of WTC and adds to the existing literature through the use of new methods, the inclusion of the effects of time on WTC, the inclusion of the effects of short term SA on WTC, and the search for a possible connection between fluency and WTC.
158 Limitations As with any type of investigation, there were limitations to this study. Three key limitations were identified for this investigation as a small participant population, the methods for assessing WTC, and the method for assessing language use. While these limitations did not prevent the investigation from obtaining viable results, it is believed that as limitations can be overcome, better data and additional explanations of results can be obtained. As with many other investigations that eva luate learners from a single SA program, the study must be limited to the number of students that choose to participate in a given semester/session. For this study, only ten students at this level elected to participate in SA. Fortunately, through monetary compensation, it was possible to limit participant loss (only one student from the program chose not to participate). A similar situation was true of the AH group. Data were collected during the summer semester, and as such there were fewer students who e lected to take summer courses than would be enrolled in a Fall or Spring semester In this instance, providentially, the researcher was able to work with the instructor of the AH course to make the data collection part of their participation grade, which o nce again assisted in diminishing participant loss. Additional participants would always be welcome in a study of this type to ensure that the results can be related to a wide variety of individuals; in other words, the greater the sample, the greater the chance that the results will be repeated with other participants. It may be that the nine SA participants were particularly communicative and worked harder at improving their fluency than other students; with a greater population size the results could car ry more weight.
159 One of the most salient limitations to this study was the use of only self report data for measuring WTC. Previous recent research into WTC has begun to use observational WTC data along with the self report data in order to have a more comp lete picture of how willing learners are to use their L2. It was initially hoped that observational data could be included in this study, but unfortunately the opportunities to observe learners in this investigation were very limited, especially the learne rs in the SA environment. It was considered at one time to attempt to use some type of recording device to capture learner behavior in certain situations, but the learners would have to be recorded in a variety of settings in an unobtrusive way. It was hop observed in the classroom; however, this data would have only been useful for that particular situation. Moreover, this type of data collection required instructor participation which was not forthcoming and so was not able t o be realized. For future studies, real world setting, in order to go beyond the data we are able to collect now. Perhaps it would be possible to collect data from the home stay families to discover their point of view of how willing the learners appear to be to communicate with them in Spanish. It might also be possible to collect data a t a social event including both learners and NSs where both self observations and NSs obs ervations of learner behavior could be obtained. For example, the social event could be hosted by a Spanish focused student organization and there could be time for open socializing and then other more structured activities that could be both fun and usefu l for collecting data. It is hoped that there may be other ways of surveying learners in the moment as they are in the L2
160 context. Perhaps through using the messaging t hrough cell phones or Clickers learners could be contacted briefly and quickly to obtain moment to moment WTC data. To accomplish the above mentioned social event data collection, each speaker could have an unobtrusive recording device that could capture b oth natural speech data and observe WTC. It may be necessary to take a risk on some of these newer methods in order to find what may or may not work. This type of limitation will always occur with any type of personal cognitive difference in that researche rs must rely on self reported data to a certain extent. The next limitation discussed here i s the assessment of the amount of language used during the programs. Like the WTC data, the amount of Spanish used was dependent on learner reports. It was hoped th at by assessing the amount of Spanish used on a weekly basis and using an online survey (see Appendix D ), learners would be able to easily report their amount of communication. However, there were several learners who forgot to fill out parts of the survey on a regular basis, despite regular reminders to complete the entire survey. This meant that only parts of the survey could be used in data analysis. Additionally, learners tended to spend very little time completing the survey and usually put the same nu merical amounts for each week. If such a survey is used in a future study, changes sho uld be made, perhaps asking for learners to describe their communication in their own words to better understand the extent and type of contact that learners have In add ition if a method can be developed to observe the amount of language used by the learners, this would be welcome. In general, limitations are always expected in an investigation and there are no methods which can be considered to work perfectly every time. However, if these three
161 limitations are addressed in future studies, researchers may be able to obtain more detailed data and results. In general in second language acquisition ( SLA ) research a larger participant population may show different results that can be applied more generally to other learners For WTC data, we will be able to have a more complete picture of WTC with observational data and additional methods of data collection. For language contact, we will be able to better understand how learner s are interacting with NSs and hopefully can find the relationships between that language contact and linguistic abilities. Finally, through participation by administrators and instructors, researchers will have more access to a wider variety of data colle ction methods. Future Studies There is still much we do not know about SA, fluency and WTC. There is an almost endless list of questions that can be addressed in future studies on these three topics. In continuing with SA research it is important to contin ue to explore a variety of programs that are available to university and other types of learners today and to follow tendencies in SA to see where research will be most beneficial. It is likely that short term programs will continue to gain in popularity and must continue to be researched so that participants in these programs will have a better idea of what to expect and what they can truly hope to achieve. With fluency research it is important to consider the possible variables that can be attributed tow ards helping a learner develop and improve on temporal fluency measures. As was seen in th e s e data, while learning context may be part of shaping fluency, the amount of time spent studying language on a daily basis also appears to impact fluency. Likewise, certain instructional methods may be more beneficial to improving fluency than simply participating in SA. In addition, based on the current
162 results, something more may be occurring during SA that has not yet been considered. Attempts to flush out what th ese variable s might be may assist future researchers in better understand what elements contribute to improvements in fluency. As was mentioned above in the limitations section, future investigations should strive to continue to develop new methods to coll ect data for WTC and for language use. As new technology becomes more readily available, it may be incorporated in observing WTC and/or language use in more settings and in more ways tha n previously possible. In order to test out these new methods, smaller pilot or case studies may need to be conducted before testing them on a larger population of participants. It is important to keep in mind that even in small scale studies important data can be collected that will assist other larger studies in exploring these topics. Conclusion WTC as they participated in short term intensive language programs, one in a SA setting and the other in a more traditional AH setting. Based on prev ious research in these areas it was expected that learners in the SA context would demonstrate larger gains in both fluency and in WTC than the learners in the AH context. However, the results found that the SA context while seemingly beneficial to both fa ctors investigated here, may not be as influential as once believed. While SA learners were able to show gains in fluency and WTC, their results were not drastically different from the AH learners. An important factor to emerge from these results is that i t might be additional SA environment. The use of reflection blogs also provided new information about WTC and allowed for new beliefs to be considered that had not p reviously discussed in other
163 research. While the original hypotheses were not confirmed here, this study is still able to contribute to the body of literature on fluency, WTC, and SA, and provides new information that may help researchers to better underst and second language acquisition.
164 APPENDIX A LANGUAGE BACKGROUND QUESTIONNAIRE Name: __ ______________________ Program (circle one): Santander Study Abroad / On Campus at UF Do you consider your first language to be English? Y / N If no, what langu age do you consider to be your first language? _________________________ Do you consider yourself to be bilingual (i.e. grew up speaking 2 or more languages)? Y / N If yes, what languages do you speak? In grades 9 12 I took Spanish (circle all that are applicable): 1 2 3 4 5 Have you taken any standardized Spanish tests? Y / N If yes, can you remember your score? SAT II _____ AP ______ IB ______ CLEP ______ Other ____ ________ Did you take Spanish or have any Spanish instruct ion prior to grades 9 12? Y / N If yes, please explain Have you taken Spanish at UF? Y / N If yes, please list the courses you have taken Have you taken Spanish at any other colleg e or university? Y / N If yes, please list the courses you have taken Do you have any family background with Spanish? Y / N If yes, please explain Have you ever lived or socialized extensively with relatives, or anyone, who only speaks Spanish with you? Y / N If yes, please explain Have you ever lived or spent time in a Spanish speaking country for more than one week? Y / N
165 If yes, please expla in Have you studied any other foreign language? Y / N If yes, please indicate which language and the years studied
166 APPENDIX B ORAL INTERVIEW QUESTIONS Preguntas para la entrevista A Puedes darme una descripcin personal? o De dnde eres ? o Qu estudias en la universidad? por qu? o Cmo eres t ? o Qu te gustara hacer en el futuro despus de graduarte? Qu te gusta hacer en tu tiempo libre? Qu actividades haces en la universid ad? Cul fue el peor viaje que hiciste? Qu te pas? Regresaras a ese lugar en el futuro? por qu s o no? Cul es tu cena preferida? Tpicamente qu tipo de comida prefieres? Qu msica te gusta escuchar? Tienes un/a cantante favorito/a o un g rupo favorito? Si pudieras tener una habilidad sobrenatural cul querras? Votaste en la ltima eleccin nacional? En tu opinin es importante que los jvenes voten? por qu? Qu vas a hacer este otoo? Qu hiciste para celebrar tu cumpleaos a l os 16 aos? Cul es tu da festivo favorito? por qu? English translation: Can you give me a personal description? o Where are you from? o What are you studying at school? why? o What are you like? o What would you like to do in the future after graduation? What do you like to do in your spare time? What activities do you do at school? What was the worst trip you took? What happened to you? Would you ever go back to that place? Why or why not?
167 What is your favorite dinner? Typically, what type of food do yo u prefer? What music do you like to listen to? Do you have a favorite singer or group? If you could have a super power what would you want? Did you vote in the last national election? In your opinion, is it important for young people to vote? why? What are you going to do this Fall? What did you do to celebrate your sixteenth birthday? What is your favorite holiday? why? Preguntas para la entrevista B Cmo es tu familia? o Cuntos hermanos tienes? o Dnde viven ellos? o Cmo son tus padres? Dnde tra bajan? o Vives cerca o lejos de tus otros parientes (abuelos, tos, etc.)? Cul es tu pelcula favorita? Por qu es tu favorita? Qu vas a hacer durante las prximas vacaciones de primavera? Qu hiciste para La Noche Vieja el ao pasado? Fuiste a un a fiesta? Hiciste algo especial para celebrar? Con quin estuviste? Si pudieras viajar a cualquier pas en el mundo sin preocuparte del dinero Adnde viajaras? Cul es tu clase favorita de la universidad? por qu? Te gusta leer? Normalmente, qu lees? Tienes un libro preferido? En tu opinin, es mejor que el gobierno ayude las empresas o que ayude la gente normal durante este periodo de recesin? Qu te gusta hacer durante las vacaciones? En el futuro, te gustara vivir en una ciudad g rande o en un pueblo pequeo?
168 English translation: What is your family like? o How many siblings do you have? o Where do they live? o What are your parents like? Where do they work? o Do you live close or far from your other relatives (grandparents, aunts and u ncles, etc.)? What is your favorite movie? Why is it your favorite? What are you going to do for the next Spring Break? something special to celebrate? Who were you with? I f you could travel to any country without worrying about money, where would you go? What is your favorite class at school? why? Do you like to read? Normally, what do you read? Do you have a favorite book? In your opinion, is it better for the governmen t to help companies or help regular people during this recession? What do you like to do on vacation? In the future, would you like to live in a big city or a small town?
169 APPENDIX C WILLINGNESS TO COMMUNICATE QUESTIONNAIRES W illingness to C ommunicate in English Questionnaire Name: _____________________ Program? Santander / On Campus at UF This questionnaire is composed of statements concerning your feelings about communication with other people, in English. In the space provided, please indicate th e frequency with which you choose or would choose to speak in English in each situation using the following numerical scale. You may not have encountered all of these situations, but given how you feel in other situations, and what you know of yourself, pl ease do your best to judge how willing to communicate you would be. Remember there is no right or wrong answer to these questions/situations, this survey is just to look at how and when people choose to communicate. 1 = almost never willing to communicat e in this situation 2 = somewhat willing to communicate in this situation 3 = willing to communicate half of the time in this situation (50 50) 4 = usually willing to communicate in this situation 5 = almost always willing to communicate in this situation Speaking in English 1. Speaking with a group about your summer vacation. ______ 2. Speaking to your teacher about your homework assignment. ______ 3. Speaking one on one with a classmate about your plans for tonight. ______ 4 A stranger enters the room you are in, how willing would you be to have a conversation if he talked to you first? ______ 5 You are confused about a task you must complete, how willing are you to ask for instructions/clarification? ______ 6 Talking to a friend while waiting in li ne. ______ 7. Talking to a stranger while waiting in line. ______ 8 How willing would you be to be an actor in a play? ______ 9 Describe the rules of your favorite game. ______ 10 Play a game, for example Jeopardy ______ 11. Give a presentation in cla ss. ______ 12. Give a presentation to a group (outside of class). ______ 13. Order food in a restaurant. ______ 14. Report your stolen wallet to the police. ______ 15. Ask a salesperson in a store for help. ______ 16. Talk to a stranger you find attractive at a bar. ______ Reading in English 1. Read a novel. ______ 2. Read an article in a paper. ______
170 3. Read letters from a non native speaker pen pal. ______ 4. Read personal letters or notes written to you in which the write r has deliberately used simple words and constructions. ______ 5. Read an advertisement in the paper to find a good bicycle you can buy. ______ 6. Read reviews for popular movies. ______ 7. Read a blog written by a non native speaker. ______ 8. Read a blog written by a native speaker. ______ Writing in Englis h 1. Write an advertisement to sell an old bike. ______ 2. Write down the instructions for your favorite hobby. ______ 3. Write a report on your favorite animal and its habits. ______ 4. Write a story. ______ 5. Write a letter to a friend. ______ 6. Write a letter to a friend who is learning English. ______ 7 Write a newspaper article. ______ 8. Write the answers to a fun quiz from a magazine. ______ 9 Write down a list of things you must do tomorrow. ______ 10. Participat e in an online chat with both native and non native speakers. ______ 11. Write a blog. ______ 12. Write an essay comparing two short stories written in English. ______ Comprehension in Engl ish 1. Listen to instructions and complete a task. ______ 2. Bake a cake. ______ 3. Fill out an application form. ______ 4. Take directions from a native English speaker. ______ 5. Watch a movie ______ 6. Listen to announcements made over a loud speaker (as in an airport or train station). ______ If there is anyth ing else you would like to tell me about your abilities and/or desires to speak English please feel free to explain in the space below:
171 W illingness to C ommunicate in Spanish Questionnaire Name: _____________________ Program? Santander / On Campus at U F This questionnaire is composed of statements concerning your feelings about communication with other people in Spanish and about completing various activities in Spanish. In the space provided, please indicate the frequency with which you choose or wou ld choose to use Spanish in each situation using the following numerical scale. You may not have encountered all of these situations, but given how you feel in other situations, and what you know of yourself, please do your best to judge how willing to com municate you would be. Remember there is no right or wrong answer to these questions/situations, this survey is just to look at how and when people choose to communicate. 1 = almost never willing to participate in this situation 2 = somewhat willing to p articipate in this situation 3 = willing to participate half of the time in this situation (50 50) 4 = usually willing to participate in this situation 5 = almost always willing to participate in this situation Speaking in Spanish 1. Speaking with a group of Spanish class students about your summer vacation. ______ 2. Speaking to your teacher in Spanish about your homework assignment. ______ 3. Speaking one on one with a classmate about your plans for tonight. ______ 4 A stranger enters the room you are in, how willing would you be to have a conversation if he talked to you first? ______ 5 You are confused about a task you must complete for class how willing are you to ask for instructions/clarification from your teacher ? ______ 6 Talking to a friend while waiting in line. ______ 7. Talking to a stranger while waiting in line. ______ 8 How willing would you be to be an actor in a play in Spanish ? ______ 9 Describe the rules of your favorite game to a native Spanish speaker friend ______ 10 Play a game in Spanish for example Jeopardy ______ 11. Give a presentation in Spanish class. ______ 12. Give a presentation to a group of native Spanish speakers. ______ 13. Order food in a restaurant while in a Spanish speaking country. ______ 14. Report your stolen wallet to the police while in a Spanish speaking country. ______ 15. Ask a salesperson in a store for help while in a Spanish speaking country. ______ 16. Talk to a stranger you find attractive at a bar. ______ Reading in Spanish 1. Read a novel. ______ 2. Read an article in a paper. ______ 3. Read letters from a native speaker pen pal. ______
172 4. Read personal letters or notes written to you in which the write r has deliberately used simple words and constructions. ______ 5. Read an advertisement i n the paper to find a good bicycle you can buy. ______ 6. Read reviews for popular movies. ______ 7. Read a blog written by another Spanish student (someone learning Spanish). ______ 8. Read a blog written by a native Spanish speaker. ______ Writing in Sp anish 1. Write an advertisement to sell an old bike to be printed in a Spanish newspaper ______ 2. Write down the instructions for your favorite hobby to be read by a native Spanish speaker friend ______ 3. Write a report on your favorite animal and i ts habits to be read by your Spanish teacher ______ 4. Write a story to be read by any Spanish speaker ______ 5. Write a letter to a friend who is also learning Spanish ______ 6. Write a letter to a friend who is a native Spanish speaker. ______ 7 W rite a newspaper article intended for native Spanish speakers ______ 8. Write the answers to a fun quiz from a magazine. ______ 9 Write down a list of things you must do tomorrow. ______ 10. Participate in an online chat with both native and non native speakers. ______ 11. Write a blog in Spanish to be read by any Spanish speaker. ______ 12. Write an essay in Spanish comparing two short stories written in Spanish to be read by other students and by Spanish teachers. ______ Comprehension in Spanish 1. L isten to instructions given to you by your Spanish teacher and complete a task. ______ 2. Bake a cake if the instructions were not in English. ______ 3. Fill out an application form. ______ 4. Take directions from a native Spanish speaker. ______ 5. Wat ch a Spanish movie without English subtitles ______ 6. Listen to announcements made over a loud speaker (as in an airport or train station). ______ If there is anything else you would like to tell me about your abilities and/or desires to speak Spanish please feel free to explain in the space below:
173 APPENDIX D LANGUAGE USE QUESTIONNAIRE 1. Name: _______________ 2. Please select your Summer program: Santander / UF On Campus 3. On average how many hours did you spend speaking, in Spanish, outside of class with native or fluent Spanish speakers per day this week? 0 1, 1 2, 2 3, 3 4, 4 5, 5+ 4. Who did you speak to, in Spanish, outside of class? For approximately how long did you speak to them each day? Classmates/Friends from UF 0 1, 1 2, 2 3, 3 4, 4 5, 5+ Host fa mily 0 1, 1 2, 2 3, 3 4, 4 5, 5+ Instructors 0 1, 1 2, 2 3, 3 4, 4 5, 5+ Other American students 0 1, 1 2, 2 3, 3 4, 4 5, 5+ Other non American students 0 1, 1 2, 2 3, 3 4, 4 5, 5+ Other speakers (explain below) 0 1, 1 2, 2 3, 3 4, 4 5, 5+ 5. How ma ny hours per day did you use Spanish outside the classroom for each of the following purposes? to clarify classroom work 0 1, 1 2, 2 3, 3 4, 4 5, 5+ to obtain information 0 1, 1 2, 2 3, 3 4, 4 5, 5+ for superficial or brief exchanges 0 1, 1 2, 2 3, 3 4, 4 5, 5+ extended conversations 0 1, 1 2, 2 3, 3 4, 4 5, 5+ other (explain below) 0 1, 1 2, 2 3, 3 4, 4 5, 5+ 6. How many hours per day did you spend doing the following activities? speaking English to speakers of Spanish 0 1, 1 2, 2 3, 3 4, 4 5, 5+ spe aking English to speakers of English 0 1, 1 2, 2 3, 3 4, 4 5, 5+ communicating in English to people back home 0 1, 1 2, 2 3, 3 4, 4 5, 5+ 7. How many hours per day did you spend doing each of the following activities in Spanish, outside of class? reading Spanish newspapers 0 1, 1 2, 2 3, 3 4, 4 5, 5+ reading novels in Spanish 0 1, 1 2, 2 3, 3 4, 4 5, 5+ reading Spanish magazines 0 1, 1 2, 2 3, 3 4, 4 5, 5+ reading schedules, menus and/or announcements 0 1, 1 2, 2 3, 3 4, 4 5, 5+ reading electroni c sources in Spanish (internet) 0 1, 1 2, 2 3, 3 4, 4 5, 5+ watching TV in Spanish 0 1, 1 2, 2 3, 3 4, 4 5, 5+ listening to the radio in Spanish 0 1, 1 2, 2 3, 3 4, 4 5, 5+ watching Spanish movies 0 1, 1 2, 2 3, 3 4, 4 5, 5+
174 writing homework as signments 0 1, 1 2, 2 3, 3 4, 4 5, 5+ writing personal notes or letters 0 1, 1 2, 2 3, 3 4, 4 5, 5+ writing email 0 1, 1 2, 2 3, 3 4, 4 5, 5+ 8. Do you have any final comments about your use of Spanish for the week?
175 APPENDIX E INTERVIEW TR ANSCRIPTIONS Silent pauses are indicated with brackets and the amount of time of the pause, for example [2s] means a two second silent pause Silent pauses which were less than 0.4 S tudy A broad Pre Program Inter views SA 1 5:00 porque es muy interesante uh a m tambin uh viajara a Espaa y voy a hacer esta uh este verano pero [0.4s] yo quiero ir a Australia porque interesante y bonita uh yo quiero ir a Berlin y [0.5s] um la pared de Berlin um po rque no no s porque quiero ir a Alemania yo quiero um yo he viajado a Europa pero nunca uh fui a Alemania y es [0.4s] yo quiero ir 10:00 pueda uh a Miamiii y Georgia si estoy es estoy aqu um pero generalmente me gusta relajarme y uh yo fui a uh a Ing Inglaterra um cuando tena [0.4s] seis aos pero me acuerdo no me acuerdo muy bien porque uh estaba muy memorable fue muy y [3.2s] yo conduc con mi familia a California de Nueva York uh y y um es fue muy [2.7s] uh [2s] largo y uh es estaba uh fue muy m emorable tambin SA 2 5:00 um [4s] fuiii a una fiesta de mi amigo Joni y [0.4s] um yo juegu um videojuegos [0.4s] como RockBand y [2.1s] um disfrutarse con amigos de diferentes universidades y fue divertido um [3s] yeah uh todo y estaba en uh Wellingto n dondo yo vive y uh [4s] vimos unas pelculas de accin uh las pelculas de James Bond y es muy hermosa y exclusivamente uh [0.6s] muy misterio 10:00 um [5 .1 s] un otro tiempo porque fue muy divertido y um no hay [6 .4 s] ( laughs) [2s] [2s] uh yeah uh oh! no hace nieve [ 0.4 s] nunca en Florida [2 .1 s] ( interviewer: fue difcil esquiar? ) uh uh [2s] um estoy uh muy uh [6s] careful ( whispered ) um [5s] estoy preparado cuando [0.5s] uh tratar de un otro cosa nueva y uh yo sabe um los riesgos SA 3 5:00 um um [4 .5 s] leo uh para um para mis clases [1 .4 s] um pero um acabo de uh leer Twilight ( laughs ) um y um me gusta me gusta Twilight y uh s uh los cuatro [0.4s] uh todos uh libros le s um durante u m [1s] durante Navidad [2s] y uh porque tengo mucho tiempo en Gainesville, Georgia [2.6s] so para leer 10:00 s s y um fue uh cuatro cuatro pginas [3s] y uh oh mi uh mi familia um [0.4s] le gusta uh ejer ejercicio y um le gusta um montar en bicicleta s y um correr y um puedo en uh el verano uh este verano despus de Espaa um mi madre mi padre y yo um vamos a Wisconsin para biciclar um [3.2s] el todo uh [4s] gate ( laughs ) lo siento uh so s ( interviewer:y esto qu es? ) oh s uh se llama BRAG en ingls Bike Ride Across Georgia um ca cada verano [2s] uh mi familia y yo [2s] uh lo [1s] hicimos
176 SA 4 5:00 canciones y Muppets? uh yyy [3s] es [1s] y uh [3s] pa [0.5s] pasamos? pasamos tiempo con l ( interviewer: qu comiste? ) uh [1s] de Mxico oh uh de? uh [2.5s] mucha comida ( laughs ) um e en Mxico [2s] uh tiene un cena uh [2s] se llama uh [3.1s] uh [2.9s] es mi favorito uh [3.4s] gorditos? s con tostado yyy pollooo y col uh lettuce oh lechuga uh tomatos y uh caballos? 10:00 uh mi familia que [3s] vi vive en la cerca viven aqu pero uh um pero no todos y uh [4s] tuve una fiesta [0.5s] con mis amigos [2s] ( interviewer: qu hicieron Uds.? ) uh s uh [5.6s] bai bailar yyy comer [0.4s] y [6.5s] y uh yo creo es todo pero fue divi muy divertida y [2s] uh [2s] y muchas de mis ami muchos de mis amigos uh [2s] fue uh fue a la fiesta SA 5 5:00 [3s] mi uh amiga [5s] digame uh [2.4s] que un un otra pelcula uh en [2s] er el ao [3s] de [5.8s] pro prximo ao s yeah uh fui uh New York City uh er uh nueva c iudad y uh [3s] uh con mi amiga Erin y um no puedo ir a Times Square [1.6s] pero uh [3.1s] fuimos a restaurante [3s] uh no uh recuerdo uh el restaurante pero [2s] muy [1s] divertida? 10:00 uh dos erma um dos veranos pasado uh mi amiga Erin uh de New York City er nueva ciudad um [3.4s] fuimos a la Paris y London y um solo [0.4s] uh Erin y yo [2s] y muy divertido ( laughs ) um [4.8s] y mi familia y yo quiero ir um again me gusta [2 .1 s] travelling [4s] uh s o or unos viajes [3 .4 s] SA 6 5:00 s yo ir pero p roblamente en e nero o octubre no en diciembre o julio pero dos meses son muy um popular y um ok [2.2s] um querra [3.5s] oh lord [2s] uh [3s] vuelo [1.9s] volar quiero velar uh [1s] a uh Australia [3.4s] muy uh divertido y me gusta ( laughs ) y um [2s] si pu diera muchas pases en el mundo y no [0.4s] pagar por un boleto ( laughs ) y en escuela? uh vive vivie vivira en casa uh con tres amigas 10:00 uh no me gusta la msica rock pero es oh porque es muy fuerte y [2s] no me gusta oh um cuando era [2s] once ao s y um fui a NSync es mi primer concierto um el ao pasado um fui a Brad Paisley [0.4s] con mis amigas um no ir a um no voy a muchos conciertos pero no tengo mucho dinero um [2.4s] ( laughs ) uh NSync fue muy popular cuando era nia SA 7 5:00 no no uh li e lie lie is it lie? lie? leer? wait is that correct? el e lie? libros luego well yeah la ao abeja uh es uh asinde uh asinde este noche con mi padres ( laughs ) con mi padre y mi madre um fui a Ruth Chris [2s] uh para la cena entonces mir con mi uh per [ 0 .4 s] mi enfrente a XXX uh yeah mi padres uh la pelcula uh Four Christmases entonces celebro uh celebramos uh en [3.4s] el cayendo de uh globos
177 10:00 un uh h aparoto natativo uh s [ 0.4 s] um sabe [ 0.4 s] que um sus um pues teji pues es oh pues la oportunidad de uh interact con un uh hablar netativo es muy porque importante para aprender un lingua por lo generalos [0.4s] la oportunidad es ok you know a Spainard listening and speaking Spanish yeah o k pues yo tiene muchos um interestes en las linguisticas las um [2s] la um la structura de lengua SA 8 5:00 um [3s] um [2s] uh mi [ 0.4 s] viaje peor um fui cuando yo [3s] uh yo [2s] uh fui a Bahamas con mi con mi familia familia um porque mi mi mam uh [ 3s] uh se olvid uh or yeah se olvid um [0.5s] que uh un reservacin una reservacin uh [3s] para or para um prxima semana que entonces uh pero [3s] ( laughs ) nosotros fuimos 10:00 um me gusta um la msica de um Frank Sinatra y uh y especialmente [2.4s ] Michael Buble ? ( laughs ) es mi um me gusta Michael Buble um mi mi hermana y uh y yo uh fuimos a un concierto de Michael Buble en uh Orlando um dos aos [0.4s] pasados? fue muy bien Michael Buble es uh muy cmico [2.7s] s so s [1s] s s me gusta uh su u h [0.4s] su voz [2s] SA 9 5:00 y quizs quiera hacerlo en Latinoamerica pero um quiero hacer algo [0.5s] en un un oncolo cancer tambin tienen buena msica all ... ( interviewer: te gust a la msica? ) s me gusta rock country y algo rap pero uh no jazz o blues o algo como esto uh ( laughs ) no s nunca pienso sobre eso um [5s] quiero muchas uh mi madre dira que yo tengo demasiados 10:00 La Navidad porque mucha familia y uh [2.8s] tradition tradicin no s pero uh [2s] el tiempo del ao es muy festiva y um [4.3s] a ahora especialmente porque vivo aqu en Gainesville puedo viajar a mi casa y pasar tiempo con mi familia [3.1s] uh [4s] siempre paso tiempo con mi familia a veces uh mi familia [0.4s] va a las montaas uh cabin despus de Navidad con mi uh la familia de mis tas [4s] ( interviewer: qu montaas? S tudy A broad Post Program Interviews SA 1 5:00 um es un restaurante de comida de um de delis uh mucha carne y es ms o menos todo p a ra almorzar almorzar [1s] pero puedo cenar aqu o all all ( laughs ) mi estomac uh el estomago no puede aguantar la comida uh ( laughs ) aqu quiero tener una um ensalada no me gusta c omer tanta carne es demasiado y no quiero uh ver ms jamn despus d e um este viaje uh [2s] entonces [1s] no me gusta esto
178 10:00 a ocho de M ustang uh nuevo yo he uh yo choqu uh s yo choqu el coche uh ( interviewer: s? ) ( new question from interviewer about plans for this Autumn ) este otoo um voy a regresar a Gainesville para tomar clases um tambin yo voy a estudiar para los[1 s] MCATS y tengo que ir buscar un trabajo um voluntario para mejorar mi res mi um aplicacin de la escuela de medicina [3.2s] ( interviewer: vas a tomar el examen este ao? ) no pero unos aos despus um tengo que tomarlo en la primavera y si yo tuviera to marlo un una vez ms SA 2 5:00 hmm [2.4s] um me encanta Madrid [0.4s] y quiero um pasar ms tiempo all y um la vida de noche all es muy divertido y hay muchos lugares para comer y um beber laughs ) yo tom el metro y apren apren de mucho de um transportacin pblico o pblica y um fui al Palacio Real y um la Prado fue muy era muy interesante um la pintura de um [3s] la pinutra ms famosa del museo ( snaps fingers ) um Las Meninas um he visto en libros de historia y era muy guay 10: 00 mucho importante es la vida y como como funciona um la vida social del un de del um fuentes en los estados unidos y um [2.4s] no es tan um etnocentric centrico como las noticias de estados unidos y yo siento como um aprender los dos um [1s] los interviewer: lees el peridico? ) no no he ledo um no las noticias que um que vi en la casa o que vi en la casa son solamente gente famosa siempre SA 3 5:00 uh en este viaje mucha uh de uh la comida es muy frita frita comida y um comida frita y um me uh [1s] me fa hace falta la ensalada mucho me gusta mucho ensalada pero aqu me gusta el espageti [1s] y um de mi seora y tambin en los estados uni dos me encanta uh los espagetis y um es mi favorito pero uh tambin me gusta um ir a uh a restaurante de uh Chino [0.4s] s pero uh no es no es uh no est bien para mi salud pero me gusta [1.1s] s pero uh pienso uh que creo que uh la los restaurantes de C hino uh son mis favoritos para comer afuera de casa 10:00 mi casa en Georgia en los estados unidos y um [1s] uh pasa um solamente dos semanas con mis padres y um desperterte um voy a um volver a Gainesville um Florida para encontrar un viaje or un viaje ? un trabajo uh porque necesito dinero ( laughs ) y um pues en el otoo uh escuela en uh en fuerte s y tengo um mis clases y pero um no uh no puedo espe esperar es no puedo um uh uh pero um estoy muy uh emocio emocionada porque um quiero ver mis amigas y mis amigos [0.4s] tambin hasta SA 4 5:00 ah [5s] yo trabaj s en un campo con nios y mi hermano pero s uh hasta dos de la maana ( laughs ) no fue muy muy divertido pero es muy bien porque es uh ms dinero como normal yyy es divertido me gusta porque los ninos y fireworks es muy
17 9 invierno y mayo uh [3s] hace seis aos ahora uh yo creo o ms uh me encanta es diver divertido 10:00 uh entend y uh y oh pero mi favorito [ 1s] pelcula uh yo no s como se dice pero XXX porque origi ori originalmente era un [2s] yo [1s] tena una especializacin especializacin en uh exercise pathology [3.4s] y ooh uh me gusta leer mucho pero no tengo tiempo ahora uh como cuando un nio ni a uh so le mucho mucho mucho pero ahora no tengo tiempo uh [3.3s] pero mi libro favorito es se llama [3s] uh ( laughs ) [3.1s] Everything You Love SA 5 5:00 [2s] um [1s] es difcil para [1s] uh decir porque normalmente uh me disfruta [4ms] me disfrutan todas las vacaciones porque [0.4s] si hay um por ejemplo mi familia uh cada [0.6s] verano um iba al un al playa en [1s] uh en los estados unidos y un verano um hay problemas con el tiempo porque um [0.5s] huracn or y entonces necesitbamos salir pero est bien porque antes uh de salir eran muy divertido pero no pienso que or pienso que 10:00 diecisis um [4.1s] voy or uh fui al restuarante con mis amigos y el restaurante uh se llama [1s] y despu s uh [1.3s] ten a or tuve un fiesta en mi casa [0.5s] con mis amigos y [2s] mi cumpleaos de diecisis no [1s] uh era [1s] tan interesante interviewer: por qu? ) porque [1s] hay um uh menos reglas despus de esa dieciocho or hay muchas habilidades SA 6 5:00 es quiamos or [1s] esquiamos todo el da y um estab a mos muy cansad cansado um mi en la televisin miraron la [0.4s] bolsa en Nueva York y um yo beb un poco de cham champin para nada ms ah um me gustara [0.5s] uh ir al Italia porque um e s yo o es un [0.5s] es un pas muy bonita uh soy catlica y me gustara um ir al Vatican [0.6s] y uh no hablo mu no hablo un nada Italia pero es pero 10:00 uh [1s] no s uh por um cuando voy a a ir escuela de leyes uh esperando um me me gustara vivi r en un ciudad grande a Boston o Nueva York pero despus cuando um tendra muchos aos no s um ahora mi familia viv um al lado de la ciudad en los suburbs y me gusta muy calm like no tengo uh opiniones fuertas en realidad me gusta me gusta murmurs ) SA 7 5:00 um [3 .1 s] aggravate agravado? um cmo se dice aggravating ? [2 .5 s] es es muy um [2s] animado annoying [2s] no recuerdo los palabras en espaol ( laughs ) ( cab ) ah s um [3 .8 s] pues ahora XXX un um [1 .1 s] mi primera um semana aqu en Santander um tengo un um [2s] sh uh shock cultural um mal decir para m es una cosa muy considerado um no est conecta uh n
180 no est contectado con mis amigos no veo nunca nm ero no tengo una um mobi mobil mobiiii mobil? no mobi or no la celular no tengo nada no [0.5s] no [0.4s] no s las suenas de mis amigos 10:00 s siemp uh pienso es muy importante para leer en el tiempo libre pero uh no um [2s] tengo mucho yo um [2s] digo que [1s] pues um [2s] yo dira que [3s] en el futuro tendra ms tiempo um [2s] um para leer um de las noticias y las um novellas las la las noveles las novelas ms porque yo um quiero um [3s] um pues es importante pero si no quiere si no quiere um el convencio SA 8 5:00 mi hermana y su [2.1s] uh esposo entonces per um [3s] porque en diciembre noso er [2s] uh ellos um [1s] er no [0.4s] ellos estn [2s] no [0.5s] ellos [2.5s] va van a encasarse ( laughs ) s um [2.4s] s eh [2s] en un posiblemente vi ajar a mi casa ( laughs and sighs interviewer: vas a verlos pronto? ) no uh no porque [1s] ah well la la la [3.4s] la pasado primavera yo [1.1s] yo fuimo uh yo y mi madre y mi hermana fuimos a Ketucky uh visi para visi 10:00 [2s] hmm Costa Rica porque um [2.1s] yo viajo yo [2.1s] via viajo a a Costa Rica porque um [4s] um yo [1.4s] yo estaba un um [2s] misionerio en Costa Rica por um dos semanas um y entonces me encanta Cosa Rica y los nios aqu or all um y [5s] s es ellos tienen un [1s] un es um un sitio especial en mi corazn y yo quiero um volver a Costa Rica SA 9 5:00 ( interviewer: s ) s ( laughs ) es to es lo que yo pensado pero no estaba seguro um estaba en Atlanta [2s] en durante este semana en el ao pasad o para una [0.5s] conferenca uh con Campus Crusade for Christ y [1s] uh tenemos una fiesta de baile y los XXX todo hasta de once y media a las [1.1s] hasta doce doce y media esta noche y [1.3s] baila bail mucho ( laughs ) yo bail mucho um globos de gomo y cayeron del ceiling ( laughs ) y um globos de gomo y otras cosas beb 10:00 entonces s um [4s] no ( laughs ) no uh tengo una clase favorita en particular pero [2s] uh me gusta [2s] el espaol por supuesto y es importante que yo estudie espaol uh [1s] par a mi futuro porque quiero trabajar en [0.5s] con gente del de los pases en um [1s] Centroamrica y bien si yo puedo [2s] podr ayudar a todas la todos la gente necesito hablar espaol es uh [1s] quiero entenderlos um no estara posible para hacer todo eso sin estudiar ms uh para aprender A t H ome Pre Program Interviews AH 1 5:00 ba mos a Maine y praticamos uh [2 .6 s] la vela la el surf de vela mucho y es like Maine algunas das solo West Palm hay [4ms] epocas olas
181 solo necesita el viento p ero Maine es mejor 10:00 s mucho no tengo miedo de [4ms] altura pero like algunas personas [4ms] s y ellos [2.4s] son fueron [2s] muy [5ms] loca teatro fui muy interesante [4ms] y la vista fue interviewer: y la ciudad? ) well probablemente voy a vivir en la ciudad en una ciudad grande porque el FBI con el FBI BI tengo que ir con ellos quiere y la I mean me gusta mucho las ciudades y [5ms] los las ciudades pequeos los pueblos s on buenos para like AH 2 5:00 Problamente cuando era un nio um mi familia [2s] um [2.3s] fuimos a un parque nacional eeee uh no me gusta um flustered um [2.1s] cosas rpidas y uh [2.7s] um cua ndo era un niooo no me gusta aguaaa uh los ros y mi padre um [3.3s] entraeme? en una ro en uh like a tube y [2.2s] no me gusto uh gust [3.9s] y uh ahora Yellowstone el parque parque de Yellowstone no s pr oblemente siete u ocho so 10:00 [3.4s] um un semester so ISA so um pero es um con la universidad [5.6s] ( interviewer: vas a to mar clases? ) uhh no s y vivo uh con una familia uh host family [2 .1 s] s [2 .8 s] so [2.4s] uh y viajar uh por todo toda del uh del Europa uh espero AH 3 5:00 oh okay no tengo un peor viaje ( interviewer: no ? ) no um yo yo vive en Iowa para um para tres aos y [2s] es muy malo ( laughs ) y aburrido pero [4s] ( interviewer: y el mejor viaje? ) um mmm [2.3s] hay mucha no s me gusta todas ( laughs ) um yo fue a Colorado y yo uh caminar en los um [1.5s] y o camino [1.8s] camin uh en las montaas uh con mi familia y yo fue uh [2.3s] um Washington DC um para un uh la classroom de presidenciales uh [2.3s] en la escuela secundaria it was s me gust mucho y um yo recibi mucho correos [1.1s] tambin 12:00 d iecisis diecisis oh um ah [2.2s] mi padres uh [2.8s] da dame un [3.2s] um un dos noches en el hotel por la playa con mis amigos y tenemos una fiesta para la playa uh en Cocoa Beach uh [3s] pues yo uh estoy solamente diecisis ( laughs ) um pero um uh um n adamos ( laughs ) [2.1s] uh mis padres estaba all um pero en um en um uh otro room [0.8s] y um no en um en mi uh en my room AH 4 5:00 y mi amigo y su familia en Medelln yyy pues y de all viajamos a Bogot y poco ciudades XXX y muy bonito y toda la gent e es super amable [ 0.4 s] es increble y es y la misma edad que yo pero ha viajado a Japn por uh un ao y por eso no ha
182 empezado en la universidad [1s] el ao pero [1.4s qumica y [2.3s] mm [2.6s] no es que [2s] trato de [1.7s] con qu? entender [2s] y gustar lo que estoy haciendo 10:00 en ingls traje lo mismo pero tengo la oportunidad de ver un poquito ms de diversidad y leer cosa s cmo qu? que proveer esta um XXX [4.2s] que son con qu? [3s] te ayudan en mis estudios as por las ciencias [5.3s] y en el peridico en revistas y nivel de [1.5s] escritu ra que es super [3s] puede escribir a todas las cosas [2.1s] cmo qu? super bien [2.6s] para que [2.8s] para que sabes exactamente de que est hablando AH 5 5:00 uh hablamos mucho [3.6s] y [2.1s] es as pero es [1s] una um emocional huh [2s] es una ti em una uno tiempo um feliz so sweet uh [2.8s] um para comida [2.4s] um [1.3s] me gusta las pescas um el area de pescas me gusta mucho [0.4s] so me gustara um [1.8s] s las pescas y la familia de las um las pescas como um yo no s la palabra de shrimp um [1.2s] es mi favorat shrimp 10:00 similare? cosa similare y hmm [3s] todo yo no s pero um [4.2s] um las um el cumpleaos de Jess porque es um [4.1s] efect um efecte el mundo en mi opinin so me gusta [1.6s] este uh est um [2s] what do we um ( clears throat ) [2 .2 s] cambiamos um [3 .6 s] los [6 .4 s] I forget how to say presents aah los uh [4 .1 s] los las cosas ( laughs ) [1 .3 s] los cosas de um una mensaje y um [2s] um vamos a las casas de mi familia AH 6 5:00 um quedaba mos en el hotel pero um [3.2s] el cridad um estaba gratis [2.5s] todo todo el da s muy bien y um en el futuro s re regreso pero [3s] um ahora no me recuer no re no recomiendo ah [3.4s] la ah Tampa uh para el producto [1.5s] para el producto [3.4s] ( i nterviewer: por qu? ) um los restaurantes? um yo prefiero um [4s] comidas con um vitamine pero soy de nuevo orleans y muchas las comidas son [2s] uh spicy y um todas las comidas um que tienen [2s] uh un sabor 10:00 no s um pero [2.6s] todas las Super Heros um [3.1s] tiene um [3.8s] muchas cosas de que quiero [3s] s tie tengo un favorito pero es uh Captain America uh Captain America um tiene la hablidad que [2.1s] um guardar el mundo [2.6s] y um [2.4s] ayuda todos todas las personas [2s] circulando el mundo [1s] n no es muy fuerte pero Superman no es puedo ayudar todas las personas... s porque si [3s] tengo el poder de ayudar personas o [2.1s] las cosas de social AH 7 5:00 dos aos ( laughs ) dos aos uh y con mi uh hermano uh y [2.1s] conducimos uh [4.5s] en el [4.1s] el la costa oeste uh [4.3s] me gust pero [4.6s] uh [3.2s] me enoj con mis padres y hermana XXX porque [5s] mi [2s] mi edad o uh no s [2.7s] pero y si si pudiera
183 10:00 hmm [3s] volar o [3 .8 s] hablar con animales ( interviewer: s ? ) s volar pues no no no um [4s] no no no olvi olvida olvide esas cosas um quiero l [2.5s] telekenesis s si tienes uh telekenesia no puedo recordar la palabra palabra uh como Jean Gray en XMen s? uh porque ella puede hacer todas las otras AH 8 5:00 y tambin con guitarras electricas y stuff y entonces um [1s] la mezcla parece muy interesante um porque [1.6s] um muuuchas gente puede conectar por la musica de su banda [0.4s] y me gusta mucho y tanto [5.2s] qu ms? um [2s] wow [4s] una vez um me um m e gusta jugar ftbol tambin mi hermana uh eh ella le gusta jugar y um un fin de semana mis pardes uh nos uh uh well entonces pa pasamos tiempo en un hotel en /kawa/ pienso? y um [1s] para todo el fin de semana 10:00 es un pueblo muy pequeo que se lla ma Itri es en um [2.9s] um al lado de pueblo XXX porque um [2s] uh tiene un castillo en las montaas y tambin la inglesia en que uh [1s] en que se um [2.6s] casaron mis abuelos y entonces um [1s] por [1.2s] um por das pasaje por las montaas o sea y me gust pero [2s] en Roma AH 9 5:00 para um el ao nuevo? okay qu yo hice el ao pasado? oh um [4s] no recuerdo pero pienso que yo fui a una fiesta y um mire la tele um [2s] um haba un programa prog program programa de Nueva York eh um es un programa especial um tambin haba fireworks um que yo vi y um fui a la playa donde hay uh haba fireworks y fui a una fiesta a la playa s [2.4s] y [2.6s] nada 10:00 um una ciudad gra nde porque tiene muchas oportunidades y mucha gente y trabajar y quiero um um permanecer con mi familia y ayudar muchas personas trabajo en un hospital um viajar tambin ser enfe um que yo um [2s] estudio all pero si yo um pero es posible que yo voy a um a un estado en el norte para estudiar AH 10 5:00 um [8 .7 s] um [3 .4 s] durante la vacacin uh [4 .1 s] estuve en en la casa con mi famili a qu fue todo ? ( interviewer repeats question ) oh um [3s] me gust The Game en BET ( laughs ) porque um es sobre ftbol y yy el hombre es muy muy guapo ( laughs ) 10:00 veo los um artifacts um um [5s] hmm no tengo un clase favorito ( laughs ) pero pero ahora me gust me gusto mi clase um [2.4s] uh SPN 2240 porque um [3.1s] tiene mucho uh eje rcito en grupos y la profesora es um muy simptic a [4s] hablan mucho a la clase y con um con [2s] partners? um y leen [2.6s] uh escrito del libro [4.3s] s
184 AH 11 5:00 yeah s vacaciones de verano y uh [3s] solo uh um quedaba quedamos en la uh ah XXX um bamo s a Cape Cod cada [1s] um cada s cada a o pero um [2s] n no hacia l una casa um un t imeshare y mi mi madre conoce una uh una persona que tiene un t imeshare s] ( interviewer: Te gusta la playa? ) s uh me gusta ir a la playa porque en Conneticut no hay muchas playas bienes so entonces cuando podemos ir a Cape Cod es muy bien 10:00 museos catedrales y fuimos en uh [3.1s] uh um fuimos con guas alrededor del c iudad y tambin fuimos a um Florence para un da y [3s] otras cosas como that oh s s um vis vista um tambin uh uh Tower of Pisa uh pero uh no mi madre no le gusta uh Colosseum [2s] muchos muchos muchos lugares [2.4s] ( interviewer: en Roma? ) s quedamos en Rome para muchos das fue un poco aburrido [1s] en el uh terminacin pero me gust [2s] uh fuimos en [1.3s] un pocos um como dos grupos de turistas para like un da a la vez pero mi familia y yo fuimos like um caminam os AH 12 5:00 para co comer y y caminar uh caminaba con mi hermana yyy uh voy a [2s] XXX uh ere um ere estaba miedo [2 .4 s] y um [2s] algo [3s] um divertir [3 .6 s] es posible pero [2s] um yo no s ahora porque es muy muy uh gast gasto? uh oh y me gusta um [3s] camarones y um langosta pero es la comida um mariscos y tambin me gusta arroz y frijoles pero mi comida favorita favorito es camarones 10:00 um [3s] mi cantante favorito es Michae l Jackson porque me gusta la msica y [1.4s] um l es un [1.1s] muy generoso individual um con um por ejemplo um l [1s] ha fuido a [2.1s] conciertos no pero [3s] um [1.6s] me g ustara pero no um [3.2s] el los boletos? tickets es muy gasto AH 13 5:00 pero um en uh [2s] en um en [4s] uh el semestre pasado yyy el seme uh semestre antes del lo um s he cocinado mucho porque uh vivo en un apartamente y no tengo un plane de [0.4s] comida aqu en [0.4s] campus y as um [2.1s] o y no tena mucho dinero para uh comer afuera y as um [2s] y a m me gusta cocinar tambin so [3s] y um para um uh divertida um [2.3s] tuvo una msica um [2s] uh me gusta hip hop um [2.6s] uh un poquito de ro ck um [2s] 10:00 diecisis? um no no he uh no he hecho mucho pero um [3s] um solamente a mi casa con uh mis con mi familia y mis uh amigos muy buenos li um como [2s] tres o cuatro amigos [1.4s] solamente en una casa uh? normalmente um [2s] uh tuve bi z cocho y um [2s] normalmente pasabamos tiempo con uh con cada uno [3.2s] uh jugamos juegos o y comimos mucha comida probablemente la pizza
185 AH 14 5:00 nosotros caminamos oh caminar ( laughs ) caminamos por la ciudad y um [2s] fuimos de compras y [2.6s] come mos uh [4.1s] comemos la cena pero um nosotros um slo estamos en la cuidad por un da nada ms so fuimos a um Buffalo or Nueva York y Boston y uh [2s] um Cape Cod tambin en verano ( laughs ) no no no el invierno ( laughs ) pero haca um un poquito fro [4.5 s] llevo una chaqueta 10:00 diecisis aos? uh mia una de mis amigas mejores tiene la misma cumpleaos de yo y nuestros um padres [2s] uh [3s] tuvieron una fiesta para ella y yo con todas de barbecue y um nadamos e n la piscina y [2s] pasa tiempo con todos [1.2s] es um el catorce de febrero si pues en Sarasota no hace fro um en febrero y podemos tener la uh fiesta outside en el backyard y nosotros tuvimos uh suerte porque no llueve ese da A t H ome Post Program Inte rviews AH 1 5:00 ahora se escuchando um [2 .6 s] muchos tipos pero mucho uh [1s] alternativo y [2.4] tambin un poco de reggaetn y Lus Mandel y un poco de msica espaol pero [1s] uh me gusta Jimmy World y [2s] no s Linkin Park hay hay tantos favoritos [1.3s] febrero fui um un concierto de Less Than Jake es un uh es un banda de un grupo de Gainesville [1s] y pero ahora son populares en muchos estados pero [2.8s] empecemos en empecieron en Gainesville um s uh 10:00 nationals nacionales para [1s] l ike todo el ao y [2.4s] cual mi uh [2s] mi compaera y yo uh cualifiquemos para nationals y [1s] so estaba ella y por eso no pude um ( interviewer: quin gan? ) no nosotr os pero uh [1.3s] pero un equipo uh cerca de nuestra escuela y por eso es todo en la los estados unidos like cada all like un mil equipos all so [3s] well no he no um estaba parte del [0.4s] uh debate pero uh hablar como [2s] um [3.6s] actuar like actore s AH 2 5:00 Santa Bernadette um [2.2s] fue all y uh [1s] um [2.4s] Mary um apareci a su a ella all [5s] uh es [1s] muy impresive y sea grande y uh [2s] um [3s] es uh? uh un [0.4s] un catedral pero pequeo no es um tan grande como los otros cate cat edrales [4s] hay uh hay um spring como un ro underground y uh [2s] um [2.4s] es es decir uh [3s] hacer un remedio por muchos enfermedades [4.2s] so 10:00 so s es es difcil pero uh [1s] algunos de mis amigos um hablen creole [0.4s] as no es mucho muc ho problema [3s] ( new question from interviewer about favorite class ) uh mi clase favorita [2s] um problema [2.4s] problemente es uh oh fue [2.1s] el clase deee [2.4s] hablando speech [1s] en mi uh primer ao [4s] uh [1s] apren aprendimos uh hablar ms bi en y uh [4.4s] no s sencillo hablar [3s] so umm [2s] creo que no busco un trabajo de [0.4s] hablar mucho uh [3s] um [4s] que no necesito
186 AH 3 5:00 el ao pasado yo fui a [1s] donde [2s] um [3.5s] where did I go oh yo uh yo fui a un parque con [1s] much as um fireworks con mi familia y amigas y tenemos una barbecue or cookout y [4s] um yo disfrut [2.4 s ] uh [2s] um no s no hay [1 .6 s] mucho um [3s] problamente um [2s] barbecue [3s] maybe ? um [2s] jugar disco y um [2.4s] la el mar con um [1s] uh no s la p alabra en ingls but 10:00 yo quiero vivir en una c casa [2.4] medio no no muy grande porque yo no quiero limpiar um pero no muy pequeo porque yo quiero un um un gran familia [1s] uh per um me gusta las ciudades como Nueva York y y um Orlando tambin pero no Gainesville no [2.3s] no Iowa um porque no hay um muchas personas y las personas que viven en lugares como este estos hay um muy simplistica en en sus um [0.8s] bien um bien fondo y um muy [3.2s] uh resistaaas s resistas a sus ideas y [2.3s] uh [ 3s] yeah AH 4 5:00 [2s] uh [2.3s] normalmente cosas con arroz o pasta [1.2s] est fcil ( laughs ) tengo problemas con cosas ms difcil em ( very very soft ) [9.4s] ( interviewer asked him to repeat ) usualmente cocino arroz o pasta para cenar y otra pregunt a [0.4s] cual era? ( interviewer repeats question ) em todos es que [2.4s] no tengo prefir preferencia es no tengo plata para hacer para esto y salgo con amigos [1s] cuando ah puedo ah ellos lo pagan [4s] 10:00 [3.2s] diecisis creditos [3s] no s cuantos [3.4s] tengo que pensar ( laughs ) pero tengo muchas clases importantes para para la escuela [1s] de la medicina [2.6s] y por eso probablemente [2s] estar estudiando todos lo jugar volibl probablemente pero no ms que eso y [4s] ( interviewer: dnde juegas? ) um [3.2s] donde puedo jugar ( laughs ) [5s] s s [5s] s por un ao antes de esto corr interviewer: s? ) s [4s] ( interviewer: qu descubrieron? ) los dos deportes [2.4s] los dos [3s] yo quiero dentro pero XXX depende de cuando tengo tiempo [3s] y quien AH 5 5:00 [3 .1 s] hm um el ao pasado comooo? ( interviewer clarifies ) en diciembre um solamente um [3s] solamente um de um [3s] feliz navidad um celebramos o celebr um bremos de cumpleaos de Cristo? ( interviewer clarifies that the question was about ) oh okay um fumios a la iglesia y uh [1 .2 s] cant en um el coro en mi igles ia um a las what was it? no era las um doce en la noche y [3s] todo bien s um fui um a la casa de mi amiga mejor
187 10:00 el profesor Muhammed es uh [1s] es muy cmico so s um s s um [1s] los partes de [2s] muy pobre no muy rico porque um [1s] quiero ver las personas um de necesi que necesi um que necesita um muchas ayudar y s [4s] ( interviewer: necesitas leer mucho? ) uh no no uh en realidad no ( laughs ) um no leo mucho pero cuando um leo un libro es de  um gospel ( pronounced with Spanish accent ) o de hmm AH 6 5:00 no um mi familia y yo salimos a Houston um el da antes uh uh con un uh amigo de um la familia que uh [2s] que vive en Houston s y quedabamos uh para um tres semanas y entonces um [1.2s] mi mis padres uh [1s] uh aquil aquilan um un a casa en [0.6s] uh Houston uh tena uh mi cuarto solo y el [1s] uh [1s] el un otro un otro da uh [1s] uh [1.2s] uh necesitaba um 10:00 uh no pero uh fui en el clase y tena um un exam en en el da de mi cumpleaos ( ) oh um la Noche Buena um no recuerdo que pas uh no est uh [1s] borrachero pero um pienso que um estaba enfermo y um quedo en la casa [2s] pero mis mi familia u m [2s] fu fui a a iglesia y um uh [2.4s] el prximo da [3.5s] fuimos a uh por Ocala uh vio las uh [0.5s] uh las los equis los uh farms uh [1s] uh en AH 7 5:00 um uh un poquito uh he volado [2s] tres veces antes con or con un piloto ( interviewer: es d ifcil? ) uh no no es fcil como conducir un coche uh [2s] s ms ms uh no s dials y gauges que butones [3.2s] y oh en la tarde uh [3s] oh um [2.9s] era una fiesta casa de m mi amigo Victor y tuvimos [2s] dos o tres kegs [0.4s] de cerveza y [1s] disparam os uh [1s] jetes fireworks 10:00 oh me gusto [2s] um pues me gusta Rio de Jane i ro pues s um [1s] tambin Venezia si Venice ah las montaas hay apren aprendimos de [1.5s] um cmo se dice no lo puedo recordar el nombre um [3s] b barrios pobres no pre uh hay una palabra favelas uh que que es la palabra que han us [1.8s] uh reas donde personas viven que son muy pobres [2.9s] um pero tambien hay [1s] hay la playa y [1s] edificios grandes [3s] ( new question from interviewer about favorite books ) acabo d e leer uh [2.7s] Deception Point por Dan Brown uh me gusta leer uh ficcin de ci ciencia ciencia ficcin um [1.1s] y tambin AH 8 5:00 y no s despus de este ao espero que um [3.1s] co uhhh [2.7s] comp comprar un [1s] o sea [0.5s] vivir en un un a partamento uh cerca de campus no s pero [1s] s [4s] um [3s] parece que estos preguntas son los mismos de antes ( laughs ) ( interviewer assures him they are different ) um [2s] pues oh no s uh es que hmm [2s] hmm no puedo darme cuento de um [1s] de un un es o que um [1s] porque [2s] ah XXX la escuela es considerado algo que um [2s] uh en que voy a trabajar
188 10:00 no s ( laughs ) uh [3s] es posible s no s um [1s] es posibl e que estamos jugando al [0.5s] cualquiera deporte del momento um basquetbol [1s] no s pero um s uh recordo que [1.2s] um mis amigos y yo estbamos en juntos en una casa pero no s um [2s] estbamos um mirando uh [1s] una banda que estaba tocando en en la televisin no recuerdo pero um [3s] fue uh ms [1.4s] sencillo pero [4s] AH 9 5:00 um uh no s ( laughs ) uh no [3s] um yo vivo en un apartmento y a veces yo cocino pero um mucha de uh del tiempo yo voy al restaurante [0.5s] s [0.8s] con mis amigas um [1.6s] no tengo una restaurante favorito pero me gusta Olive Garden y me gusta comer sandwiches tambin [3s] s umm [2s] hmm [1s] uh [4.2s] quiero uh poder comer ( laughs ) afuera ms en Downtown porque yo pienso que es muy divertido y puedo ver los amigos y puedo pasar tiempo con ellos y [1.5s] s [1s] y pienso que es divertido en los resta urantes 10:00 mucha de mi familia um viven aqu en Florida [2.8s] yeah hmm [1s] um [0.5s] para m me gustan los das cuando no tenemos escuela [1.9s] s [1s] um como la XXX pero hay otros tambin como mi cumpleaos porque um [1.4s] um mis amigos me um d an [1s] un regalo ( laughs ) s [3.3s] oh um yo no hazo or no hago mucho porque mi familia uh no creen en um los festivos pero hay uh festivos de um [1s] uh India que nosotros celebramos um hay festivos de color y baile s AH 10 5:00 [4.2s] oh cuando era un nio una nia um mi hermana y yo um fue a Nueva York y [2s] era durante um [2s] l el [1s] winter [1s] como se dice winter? ( interviewer ) oh okay durante uh [2s] the winter um [1 .1 s] y [1 .8 s] no me gusta porque um teni mucho snow y [8s] hace [1.4s] hace um [1s] muy fro or mucho fro en Nueva York 10:00 escuchamos a gospel y um [3.7s] uh siempre [2s] uh vamos a la iglesia um [1.9s] todos los domingos entonces cuando um estoy estoy triste um quiero escuchar gospel es mi fa vorito um [3.2s] mi canci n favorito es de um Kirk Franklin y porque um [1s] est muy um animada sobre gospel s y um cuando yo escucharlo cuando lo escuch um [4.7s] yo yo siento AH 11 5:00 oh [1s] uh s um [1.4s] yo fui uh trabajando en Disney World y uh [2.2s] fui trabajando entonces uh no hizo or no hice mucho pero despus del trabajo fui a una yeah estaba en Disney World para el cambio del ao y uh yo trabaj en uh [1.5s] en en el hotel All Star Sports pero en en la cafeteria uh y oh yo yo vi el ball interviewer: trabajaste en otros das festivos? ) uh s y la Navidad y [1.5s] la la or el Da de gracias um [1s] fueron muchas personas en en el hotel
189 10:00 uh aprend q ue um las esterotpicas de frica no son ciertas [0.6s] porque uh muchos de las problemas de frica um son de [0.7s] otros pases como la Inglaterra y new question from interveiwer about reading ) uh s me gusta leer mucho uh [1s] mis libros favoritos probablemente son los de Harry Potter porque creo que es una cuenta muy bien y me gusta la fantasia pero uh tambin me gusta Tom Clancy [1.2s] uh Dan Brown Michael Crichton uh me gusta el libro Jurassic Park el libro el la pel cula es uh un poco largo me gusta las or los libros de Jurrasic Park um le ms en el colegio de ahora pero [1s] no he ledo mucho [2.1s] um le el un libro AH 12 5:00 mi novio y yo uh [5.8s] uh estaba en [2s] uh en Gaineville s eran aburrido [4s] um vacaciones? [2s] prefiero fui a un lugar especial como [2s] hmm [2s] como Disney World o como un lugar um [1.2s] en vez de este los es estados unidos ojal que uh [1s] uh tuviera dinero para uh [1s] para un vacacin una vacacin s [2s] 10:00 como Ja mes Patterson uh [3s] structure los libros s es muy es interesante [1.3s] es um delito y crimen um una mujer que fue un investago investigadora uh era un crimen tambin y uh fue [2s] hmm muertiendo? or killing a la otras personas pero la gente uh no sabe n quien quien es y y al final um el libro [1s] uh [1s] uh un de la uh el inves los investaga investigadores uh encon tr la or el crimen AH 13 5:00 s es uh s es interesante mmm [2s] porque [3s] yo uh en el uh en la un el empiezo [2s] o um [1s] a pri mer yo no pienso que [1.8s] yo pienso que va a ser aburrido pero es interesante [3s] ( interviewer: qu hiciste? ) umm [3s] nada en particular um pas pas tiempo uh [2.4s] qued en casa pero mis amigos um nosotros estn estbamos en mi casa solamente con l os uh [3s] fireworks um y [1s] vemos televisin y comemos ( laughs ) y tpicas cosas 10:00 um leo yo no leo mucho um [1s] no tengo tengo tiempo pero no paso uh mi tiempo leyendo pero um no leo uh novelas pero leo muchas um revistas um como re revistas de naturaleza y de um [1.2s] de um salud como y cualquiera um pero tengo un libro favorito es y es um uh The Outmist por uh Pablo Cuelo uh y leo el peridico que es est um que yo tien um [1.6s] tengo que leer para un clase de eventos de ahora entonces um [1s] quiero leer pero no hay tiempo ( laughs ) [1.5s] y los libros uh de clase que nosotros leemos tambin son grandes y um quiero tiempo libre ms ( laughs ) AH 14 5:00 um [3.1s] estuve en North Carolina problemente porque mi padre acabo de mov er all y um mi mam y mi hermano y yo uh vayamos all para estar con ello con l para el ao nuevo uh [1s] comemos con um mi to y ta que viven cerca de all y [2s] fue muy calma hm no hicimos mucho [3s] despus de la medianoche [2s] pero no ms despus [2s] pues la una maybe
190 10:00 s um leo or le um oh A Walk to Remember y The ( sighs ) u m [4s] creo que es divertido [0.5s] leer las novelas romnticas y [1s] de personas normales [1s] porque es ms realistico y no me gus tan libros de fantasia con gente que uh [2s] son [1s] como se dice magical? yo no s pero um [1s] ellos [1s] pudieran um o podran ( laughs )
191 APPENDIX F CODING SCHEME FOR BLOGS Situation 1. Public 2. School/Classroom 3. Host Family 4. Friends (in private) Perspective s held by participants about language 5. Love Spanish 6. Spanish is fun 7. W illingness to C ommunicate (WTC) depends on situation 8. WTC pertains to mood a. Poor mood leads to lower WTC (frustrated, angry, tired) b. Good mood leads to higher WTC (happy, well rested, relaxed ) 9. Classroom is a safe place 10. Errors make one look bad 11. Out of classroom comfortable because there are no grades 12. WTC dependent on patience of the interlocutor 13. More WTC at end of program 14. Desire to seek out new ways to communicate in Spanish in the US 15. Hope to c ontinue progress, not decline 16. Lack of vocabulary causes problems, is frustrating 17. Feel that they are improving every day 18. WTC is need driven 19. Speaking English is a way to relax 20. More WTC when accepted and not being judged 21. WTC is based on previous success 22. More confident speaking Spanish at end of program 23. More aware of areas that need improvement at end of program 24. Easier to listen and understand than speak 25. Challenges lead to more communication (idea of overcoming it by keep on trying) 26. Grammar improved by the end of the program 27. Lower confidence leads to lower WTC 28. Feel little or no change at end of program 29. Some native speakers ( NSs ) 30. Less need to speak Spanish with other E nglish speakers 31. Teacher is patient, allows learners sufficient time to answer/speak 32. Comfort level can vary when speaking to NSs 33. Enjoy communication with NSs when patient 34. 35. More WTC with people they know bett er
192 36. 37. NSs are helpful/you can learn from NSs 38. Important to take advantage of speaking to NSs Activity 39. Ordering food 40. Talking with host family 41. Asking for directions 42. Getting help 43. Speaking Spanish in public places 44. Speakin g English with friends 45. Using Spanglish 46. Using Spanish for fun Strategy 47. Ask many people 48. Wait for NSs to continue conversations 49. Wait for NSs to initiate conversation 50. Repeating oneself 51. Initiate conversation with NSs Relationship and social structure 52. Relatio nship with Host mother 53. Relationship with Host family in general 54. Relationship with other host family member (not mother) 55. Relationship with other student 56. Relationship with NS (not in host family)
193 LIST OF REFERENCES ACTFL (1999). ACTFL Proficiency Guideli nes (Speaking). http://actfl.org/i4a/pages/index.cfm?pageid=4236 AIFS (2009). President Obama and AIFS Encourage More Students to Study Abroad in China. http://www.aifs.com/media/pr/2009_11_college_china.asp Allen, H .W. & Herron, C. (2003). A mixed meth odology investigation of the linguistic and affective o utcomes of summer study abroad. Foreign Language Annals 36 (3), 370 385 Baker, S.C., & MacIntyre, P.D. (2003). The role of gender and immersion in communication and second language orientations. Lan guage Learning, 53 (1), 65 96. Barron, A. (2007). "Ah no honestly we're okay:" Learning to upgrade in a study abroad context Intercultural Pragmatics, 4 ( 2 ) 129 166 Barron, A. & Warga, M. (2007). Acquisitional pragmatics: Focus on foreign language learn ers. Intercultural Pragmatics 4 (2), 113 127. Bavelas, J.B. (2000). Nonverbal aspects of fluency. In H. Riggenbach (Ed.) Perspectives on Fluency Ann Arbor: The University of Michigan Press. (91 101). Blake, C. (2009). Potential of text based internet cha ts for improving oral fluency in a second language. The Modern Language Journal, 93 (2), 227 240. Bogdan, R.C. & Knopp Biklen, S. (2007). Qualitative Research for Education: An Introduction to Theories and Methods Boston, MA: Pearson Education, Inc. Cao, Y., & Philp, J. (2006). Interactional context and willingness to communicate: a comparison of behavior in whole class, group and dyadic interaction. System, 34 (4), 480 493. Cao, Y. (2006). Temporal fluctuation in situational willingness to communicate in a second language classroom. New Zealand Studies in Applied Linguistics, 12 (2), 1 16. Clment, R., Baker, S.C., & MacIntyre, P.D. (2003). Willingness to communicate in a second language: the effects of context, norms, and vitality. Journal of Language and Social Psychology, 22 (2), 190 209. Churchill, E. & DuFon, M.A. (2006). Evolving threads in study abroad research. In M.A. DuFon and E. Churchill (Eds ) Language Learners in Study Abroad Contexts Clevedon, England: Multilingual Matters Ltd. (1 27).
194 Coll entine, J. (2004) The effects of learning contexts on morphosyntactic and lexical development. Studies in Second Language Acquisition 26 ( 2 ), 227 248 Cohen, A.D. & Shively, R.L. (2007). Acquisition of requests and apologies in Spanish and French: Impact of s tudy a broad and strategy building intervention. The Modern Language Journal, 91 (2), 189 212. Crystal, D. & Varley, R. (1993). Introduction to Language Pathology. London: Whurr. Cubillos, J.H., Chieffo, L., & Fan, C. (2008). The impact of short term s tudy abroad programs on L2 listening comprehension skills. Foreign Language Annuals, 41 (1), 157 185. De Raad, B. (1992). The replicability of the Big Five p ersonality dimensions in three word classes of the Dutch language. European Journal of Personality, 6 (1), 15 29. Derwing, T.M., Munro, M.J., & Thomson, R.I. (2007). A longitudinal study of ESL Applied Linguistics, 29 (3), 359 380. Dewey, D P. (2004) A comparison of reading development by learners of J apanese in intensive domestic immersion and study abroad contexts. Studies in Second Language Acquisition, 26 ( 2 ), 303 327 Daz Campos, M. (2004). Context of learning in the acquisition of Spanish second language phonology. Studies in Second Language Ac quisition, 26, 249 273. Ellis, R. (1994). The Study of Second Language Acquisition. Oxford: Oxford University Press. Ejzenberg, R. (2000). The juggling act of oral fluency: A psycho sociolinguistic metaphor. In H. Riggenbach (Ed.) Perspectives on Fluency Ann Arbor: The University of Michigan Press. (287 314). Fillmore, C.J. (1979, 2000). On fluency. In H. Riggenbach (Ed.) Perspectives on Fluency Ann Arbor: The University of Michigan Press. (43 60). [Reprinted] Freed, B.F. (1990). Language learning in a st udy abroad context: The effects of interactive and non interactive out of class contact on grammatical achievement and oral proficiency. In J. Alatis (Ed.) Georgetown University Round Table on Language and Linguistics 1990: Linguistics, Language Teaching a nd Language Acquisition the Interdependence of Theory, Practice, and Research. Washington, D.C.: Georgetown University Press. Freed, B. F. (1995). What makes us think that students who study abroad become fluent? In B. Freed (Ed.) Second Language Acquisi tion in a Study Abroad Context. Amsterdam, The Netherlands: John Benjamins Publishing Co. (123 148).
195 Freed, B F., Segalowitz, N & Dewey, D P. (2004) Context of learning and second language fluency in French. Studies in Second Language Acquisition 26 (2 ), 275 301 Freed, B F., So, S & Lazar, N A. (2003) Language learning abroad: How do gains in written fluency compare with gains in oral fluency in French as a second ADFL Bulletin 34 ( 3 ), 34 40. Gardner, R.C. (1985). Social Psychology and S econd Language Learning: The Role of Attitudes and Motivation. London: Edward Arnold Publishers. Gardner, R.C., Day, J.B., & MacIntyre, P.D. (1992). Integrative motivation, induced anxiety, and language learning in a controlled environment. Studies in Seco nd Language Acquisition, 14 (2), 197 214. Garca Amaya, L. (2008). The effect of topic on rate of speech. Estudios de ling stica inglesa aplicada, 8, 117 150. Gass, S.M. (1997). Input, I nteraction, and the S econd L anguage L earner Mahwah, N.J.: L. Erlbau m. Geeslin, K. & Guijarro Fuentes, P. (2005). The acquisition of copula choice in instructed S panish: T he role of individual characteristics. In D. Eddington (Ed.), Selected Proceedings o f t he 6th Conference o n t he Acquisition o f Spanish a nd Portuguese a s First a nd Second Languages Sommerville, MA: Cascadilla Press (66 77). Hassall, T. (2006). Learning to take leave in social conversations: A diary study. In E. Churchill & M.A. DuFon (Eds.) Language L earners in S tudy A broad C ontexts Buffalo, NY : Multiling ual Matters. (31 58). Howard, M. (2005). On the role of context in the development of learner language: insights from study abroad research. ITL Review of Applied Linguistics 147 148 (1 20). Isabelli Garca, C. (2003). Development of oral communication ski lls abroad. Frontiers: the interdisciplinary Journal of Study Abroad, 9, 149 173. Isabelli, C & Nishida, C. (2005) Development of the Spanish s ubjunctive in a n ine m onth s tudy a broad s etting. In D. Eddington (Ed.), Selected Proceedings o f t he 6th Confere nce o n t he Acquisition o f Spanish a nd Portuguese a s First a nd Second Languages. Sommerville, MA: Cascadilla Press (78 91). Juan Garau, M. & Prez Vidal, C. (2007). The effect of context and contact on oral performance in students who go on a stay abroad. V IAL, 4 117 159. Kang, S.J. (2005). Dynamic emergence of situational willingness to communicate in a second language. System, 33 (2), 277 292.
196 Kasper, G. & Schmidt, R. (1996). Developmental issues in interlanguage pragmatics. Studies in Second Language Acq uisition, 18, 149 169. Kasper. G.& Rose, K, R. (2002). Pragmatic Development in a Second Language Oxford : Blackwell. Koponen, M. & Riggenbach, H. (2000). Overview: Varying perspectives on fluency. In H. Riggenbach (Ed.) Perspectives on Fluency Ann Arbo r: The University of Michigan Press. ( 5 24) Kormos, J. & Denes, M. (2004). Exploring measures and perceptions of fluency in the speech of second language learners. System, 32, 145 164. Krashen, S. (1985). The Input Hypothesis: Issues and Implications. New York: Longman. Lafford, B.A. (1995). Getting into, through and out of a survival situation: A comparison In B. Freed (Ed.) Second Language Acquisition in a Study Abroad Co ntext. Amsterdam, The Netherlands: John Benjamins Publishing Co. ( 97 122 ). Lennon, P. (1990) Investigating fluency in EFL: A quantitative approach. Language Learning, 40, 387 412. Lennon, P. (2000). The lexical element in spoken second language fluency In H. Riggenbach (Ed.) Perspectives on Fluency Ann Arbor: The University of Michigan Press. (25 42). Llanes, A. & Muoz, C. (2009). A short stay abroad: Does it make a difference? System, 37, 353 365. Lord, G. (2006). Defining the indefinable: Study abr oad and phonological memory abilities. Selected Proceedings of the 7 th Conference on the Acquisition of Spanish and Portuguese as First and Second Languages, cd. Carol A. Klee and Timothy L. Face. Somerville, MA: Cascadilla Proceedings Project. (40 46). Ma cIntyre, P.C. (1994). Toward a Social Psychological Model of Strategy Use. Foreign Language Annals, 2 7, (2) 185 195. MacIntyre, P.D., Baker, S.C., Clment, R., & Conrod, S. (2001). Willingness to communicate, social support, and language learning orientati ons of immersion students. Studies in Second Language Acquisition, 23 (369 388). MacIntyre, P.D., Baker, S.C., Clment, R., & Donovan, L.A. (2003 a ). Talking in order to learn: willingness to communicate and intensive language programs. The Canadian Modern Language Review/La Revue canadienne des langues vivantes 59 (4), 589 607.
197 MacIntyre, P.D., Baker, S.C., Clment, R., & Donovan, L.A. (2003b). Sex and age effects on willingness to communicate, anxiety, perceived competence, and L2 motivation among junior high school French immersion students. Language Learning, 53 (Supplement 2), 137 165. MacIntyre, P. D., & Charos, C. (1996). Personality, attitudes, and affect as predictors of second language communication. Journal of Language and Social Psychology, 5 3 26. MacIntyre, P.D., Clment, R., D rnyei, Z., & Noels, K. (1998). Conceptualizing willingness to communicate in a L2: A situational model of L2 confidence and affiliation. The Modern Language Journal, 82 (4), 545 562. Magnan, S S & Back, M (2007). Socia l Interaction and Linguistic Gain during Study Abroad Foreign Language Annuals, 40 (1), 43 61. Marriott, H. (1995) The acquisition of politeness patterns by exchange students in Japan In Freed, B. (Ed.), Second Language Acquisition in a Study Abroad Co ntext. Amsterdam, The Netherlands: John Benjamins Publishing Co. (197 224). Matsumura, S. (2007). Exploring the aftereffects of study abroad on interlanguage pragmatic development Intercultural Pragmatics, 4 ( 2 ) 167 192. McCroskey, J.C., & Baer, J.E. (19 85). Willingness to communicate: The construct and its measurement. Paper presented at the annual convention of the Speech Communication Association, Denver, CO. [Cited in McCroskey & Richmond (1990)] McCroskey, J.C., & Richmond, V.P. (1990). Willingness t o communicate: Differing cultural perspectives. Southern Communication Journal, 56 72 77. Teaching German, 37 (1), 1 9. Phonological memory predi cts second language oral fluency gains in adults. Studies in Second Language Acquisition, 29 557 582. Open Doors. (2009). Open Doors: U.S. Students Studying Abroad. http://opendoors.iienetwork.org/?p=150651 Pawley, A. & Syder, F.H. (2000). The one clause at a time hypothesis. In H. Riggenbach (Ed.) Perspectives on Fluency Ann Arbor: The University of Michigan Press. (163 199). Peng, J. (2007). Willingness to communicate in the Chinese EFL classroom: a cultural perspective. In J. Liu (Ed.) English Language Teaching in China New York: Continuum International Publishing Group.
198 Raupach, M. (1987). Procedural learning in advanced learners of a foreign language. In J. Coleman & R. Towell (Eds). The Advanced Language Learner London: CILT. (123 155). Regan, V. ( 1995). The acquisition of sociolinguistic native speech norms: Effects of a year abroad on second language learners of French. In B. Freed (Ed.) Second Language Acquisition in a Study Abroad Context. Amsterdam, The Netherlands: John Benjamins Publishing C o. (245 267). Riggenbach, H. (1991). Toward and understanding of fluency: A microanalysis of nonnative speaker conversations. Discourse Processes, 14, 423 441. Sallinen Kuparinen, A., McCroskey, J.C., & Richmond, V.P. (1991). Willingness to communicate, co mmunication apprehension, introversion, and self reported communication competence: Finnish and American comparisons. Communication Research Reports, 8 55 64. Schmidt, R., & Frota, S. (1986). Developing basic conversational ability in a second language: A case study of an adult learner of Portuguese. In R. Day (Ed.) Talking to Learn: Conversation in Second Language Acquisition. Rowley, MA: Newbury House. (237 326). Segalowitz, N. (2000). Automaticity and attentional skills in fluent performance. In H. Rig genbach (Ed.) Perspectives on Fluency Ann Arbor: The University of Michigan Press. (200 219). Segalowitz, N. & Freed, B. (2004). Context, contact, and cognition in oral fluency acquisition: Learning Spanish in at home and study abroad contexts. Studies in Second Language Acquisition, 26 173 199. Segalowitz, N., Freed, B., Collentine, J., Lafford, B., Lazar, N., & Daz Campos, M. (2004). A comparison of Spanish second language acquisition in two different learning contexts: Study abroad and the domestic cl assroom. Frontiers: the Interdisciplinary Journal of Study Abroad, 10 1 18. Swain, M. (2000). The output hypothesis and beyond: Mediating acquisition through collaborative dialogue. In J.P. Lantolf (Ed.) Sociocultural Theory and Second Language Learning. Oxford: Oxford University Press. (97 114). The Canadian Modern Language Review, 50, 158 164. Taguchi, N. (2008). Building language blocks in L2 Japanese: Chunk learning and t he development of complexity and fluency in spoken production. Foreign Language Annals, 41 (1), 132 156. Tanaka, K., & Ellis, R. (2003). Study abroad, language proficiency, and learner beliefs about language learning. JALT Journal, 25 (1), 63 85.
199 Titelman, G. Y. (1996). Random House Dictionary of Popular Proverbs and Sayings New York: Random House. (154). Towell, R. (2002). Relative degrees of fluency: A comparative case study of advanced learners of French. IRAL, 40, 117 150. Towell, R., Hawkins, R., & Ba zergui, N. (1996). The development of fluency in advanced learners of French. Applied Linguistics, 17 (2), 84 119. Trenchs Parera, M. (2009). Effects of formal instruction and a stay abroad on the acquisition of native like oral fluency. The Canadian Moder n Language Review, 65 (3), 365 393. Wennerstrom, A. (2000). The role of intonation in second language fluency. In H. Riggenbach (Ed.) Perspectives on Fluency Ann Arbor: The University of Michigan Press. (102 127). Yashima, T. (2002) Willingness to communi cate in a second language: The Japanese EFL context. The Modern Language Journal, 86 (1), 54 66. Yashima, T., Zenuk Nishide, L., & Shimizu, K. (2004). The influence of attitudes and affect on willingness to communicate and second language communication. La nguage Learning, 54 (1), 119 152. Yashima, T., & Zenuk Nishide, L. (2008). The impact of learning contexts on proficiency, attitudes, and L2 communication: Creating an imagined international community. System, 36 (4) 566 585.
200 BIOGRAPHICAL SKETCH b usiness o ption from Pennsylvania State University. She continued on to receive a Masters of Arts in Spanish with a concentration in l inguistics from the University of Florida. Melanie c ompleted her doctorate in Spanish with a concentration in l inguistics in December 2010, language acquisition of Spanish. Melanie is currently an Assistant Professor of Spanish and Linguistics at Indiana State University.