Perception changes over time on the usefulness of language learning strategies

Based on the data gathered in this study, it was noted that advanced students seem to have an increasing interest in communication and using the new language for communicative purposes. Autonomy in their language learning process is also observed in advanced students. Although it This longitudinal study describes and analyses the shifts in students’ perceptions on the usefulness of language learning strategies (LLS) when learning Spanish as a foreign language, and explains the roles that these learning strategies play at various moments in time during students’ language learning journey at the undergraduate level. The richness of this investigation lies in the examination of the changing roles and the perceived usefulness of these strategies over a period of 3 years by using a mixed-method approach. Its main contribution is twofold. On the one hand, it offers practical insights within the linguistic field of language acquisition that can be applied to the learning of foreign languages by students with no previous knowledge of them. On the other hand, its longitudinal design not only goes beyond the traditional quantitative approach that offers a limited snapshot view, but also captures changes in strategy use over time while providing details on why strategies were or were not used in the new language at different proficiency levels. The study found some strategies that were initially perceived as useful and helpful at the beginning of the journey were later perceived as neutral, useless or obstructive by the same group of students who had become more proficient in the foreign language. And, conversely, other strategies that were initially perceived as neutral, unhelpful or frustrating, were later perceived as necessary to improve language proficiency, useful and stimulating by the same cohort of students. In this respect, it was found that the perceived role played by these strategies alternated between essential roles to catalytic roles, imperceptible roles, retarding roles and preventing roles, and vice versa. Contribution: The main contribution of this study is that it shows that students’ perceptions on the usefulness of certain LLS over time are influenced by their level of proficiency. This in turn influences, and is influenced by, language attitudes and language learning policy, which fall under the scope of this journal.


Introduction
Longitudinal studies are very useful to capture variation in the use of language learning strategies (LLS) as learners' proficiency develops in the new language.
Undergraduate beginners prefer to get to know the differences that exist between the new language and the language(s) they already know at the start of their learning process. Initially, learners are concerned with the acquisition of a basic vocabulary, pronunciation and an interest in the target grammar. Beginner learners have little to no procedural knowledge about how to learn and the usage of the target language; they also have little declarative knowledge about the new language (Ackerman 2008:445).
Intermediate learners continue to focus on expanding their vocabulary and grammatical knowledge; they also have a fair amount of strategic knowledge, and knowledge of content (such as vocabulary, grammar, etc.). When analysing the use of LLS by foreign students learning English at an American university, Hong-Nam and Leavell (2006:412) found that intermediate learners are increasingly conscious of how a new language is learnt and they gradually become better at selecting the right strategies for specific learning tasks.
Based on the data gathered in this study, it was noted that advanced students seem to have an increasing interest in communication and using the new language for communicative purposes. Autonomy in their language learning process is also observed in advanced students. Although it This longitudinal study describes and analyses the shifts in students' perceptions on the usefulness of language learning strategies (LLS) when learning Spanish as a foreign language, and explains the roles that these learning strategies play at various moments in time during students' language learning journey at the undergraduate level. The richness of this investigation lies in the examination of the changing roles and the perceived usefulness of these strategies over a period of 3 years by using a mixed-method approach. Its main contribution is twofold. On the one hand, it offers practical insights within the linguistic field of language acquisition that can be applied to the learning of foreign languages by students with no previous knowledge of them. On the other hand, its longitudinal design not only goes beyond the traditional quantitative approach that offers a limited snapshot view, but also captures changes in strategy use over time while providing details on why strategies were or were not used in the new language at different proficiency levels. The study found some strategies that were initially perceived as useful and helpful at the beginning of the journey were later perceived as neutral, useless or obstructive by the same group of students who had become more proficient in the foreign language. And, conversely, other strategies that were initially perceived as neutral, unhelpful or frustrating, were later perceived as necessary to improve language proficiency, useful and stimulating by the same cohort of students. In this respect, it was found that the perceived role played by these strategies alternated between essential roles to catalytic roles, imperceptible roles, retarding roles and preventing roles, and vice versa.
has been reported that students at different levels of language proficiency make use of different strategies, Wharton (2000:208) found that more advanced learners usually use learning strategies more frequently and more effectively than beginners. There is a shifting of the role of the teacher from facilitator to interlocutor, and from mediator to facilitator (Hong-Nam & Leavell 2006:412). Green and Oxford (in Griffiths & Oxford 2014:2) also found that higher-level students reported using strategies of all kinds significantly more frequently than lower-level students. However, Takeuchi (in Chamot 2005:123) observed that some learners reported that as their proficiency level advanced, they shifted their use of strategies. The use of compensation strategies, including coining new words and reverting back to the mother tongue, was found by Bedell and Oxford (1996:52) to diminish with higher proficiency.
Based on the findings, this study was inspired by the abovementioned researchers who reported a shift in the use of LLS apparently influenced by the learners' increasing level of proficiency in the target language. This study therefore aims to answer two important questions: What perceptions do learners have on the usefulness of LLS at various times as they become more proficient in the foreign language? And, what distinct roles do LLS play throughout the students' foreign language learning process? The study does not follow a cross-sectional, quantitative approach that only focuses on quantifying the perceived use of LLS at a specific moment to draw some generalisations of what successful and less successful learners do, and therefore differs from the majority of previous studies. The purpose of this study is to describe and analyse the learners' perceptions on the usefulness of LLS at different points in their development, and to define and explain the distinct roles that LLS play in their own progress as they become more proficient in the new language.

The perceived temporary usefulness of language learning strategies
This study focuses on an intriguing finding that was noticed when conducting longitudinal research on the reported use of LLS, that not all strategies are constantly perceived as useful and helpful throughout the students' foreign language learning journey. Therefore, this study aims to describe and analyse the shifts in students' perceptions on the usefulness of LLS over time and the role that these strategies play when learning a foreign language like Spanish. To date, most studies on the use of LLS (such as Alhaisoni 2012; Bozorgian & Pillay 2013) offer snapshot views of strategy frequency use and fail to provide further details on the specific roles that strategy use plays in learning a foreign language at different points in time. Despite the fact that snapshot-type research has proven weak when investigating changes over time because of the cross-sectional approach, quantitative statistical studies dominate the field (Hajar 2019:239;Kölemen 2021:151). Snapshot views of strategy use fail to provide further details of under what circumstances and why strategies were or were not used; they also prove to be insufficient to capture variation in strategy use. Therefore, longitudinal research seems more appropriate to capture variation and analyse possible patterns of change over time (Dörnyei 2007:79).

Theoretical framework
Based on the work of Oxford et al. (2014:11), Macaro (2006:327) and Ellis (2008:705), the concept of LLS in the study refers to activities, actions or steps consciously and strategically chosen by students to regulate and improve their language learning experience. Griffiths (2008:85-86) identifies six essential features of these strategies: (1) they refer to what students do, and not to what teachers do, which suggests an active approach; (2) initially they are consciously chosen, but eventually some occur automatically; (3) individual factors, contextual factors, and the nature of the learning goal are chosen by learners; (4) their strategic nature implies purposeful, goal-oriented activity on the part of the student; (5) these strategies are made use of by students in an attempt to regulate and improve their own learning; (6) these strategies are used for the ultimate goal of facilitating and consolidating learning.
This study used Oxford's classification of strategies to group strategies according to common functions. According to Oxford (1990:14-16), LLS can be broadly clustered into two groups: strategies that involve direct manipulation of the new language, such as memory strategies, cognitive strategies and compensation strategies; and strategies that help manage the learning process and provide support, such as metacognitive strategies, affective strategies and social strategies.

Factors influencing strategy choice
Multiple factors affect the choice of learning strategies to varying degrees, including the stage of learning and proficiency level, task requirements, degree of awareness, teacher expectations, nationality or ethnicity, age, sex, personality traits, general learning style, motivation level, and purpose for learning the language (Oxford 1989:236). Factors that are believed to affect strategy choice have been clustered by Ellis (2008:711) into contextual factors relating to the situational context of learning, individual factors relating to the learner, and factors relating to sociocultural domains. Strategy selection for Griffiths (2013:10) depends on contextual factors, individual factors, and the purpose for which students learn a language. Strategy choice for Cohen (in Ananisarab & Abdi 2012:14) depends on the learning context, the learners themselves, and the learning task at hand.
The study groups the factors influencing strategy choice into contextual factors relating to the learning task and situation, individual factors relating to the learner, and sociocultural factors relating to the learning context by noting that the 'purpose factor' mentioned by Griffiths can be clustered within the group of individual factors and taking into consideration Cohen's and Ellis' groupings. Because language learners vary considerably in the particular types of strategies they use and the overall frequency with which they employ strategies, it is evident that these factors do not affect students in the same way (Ellis 2008:711).
Although multiple factors affect strategy choice, proficiency level is the only factor that is discussed in this study and its findings. The main reason for this decision is that it is the only factor that has standardised criteria as set out in the Common European Framework of Reference for Languages (2001) and is assessed regularly throughout the students' development. Proficiency level, as an individual factor, has a significant impact on strategy choice and use. Foreign language students at the beginner stage focus on basic pronunciation, the learning of a basic vocabulary and the alphabet; conversely, more advanced learners look to practise the new language in social interactions and therefore look for opportunities to engage. Thus, learning strategies are chosen by both beginners and advanced students to meet their own language learning requirements, taking into account levels of proficiency.

Reported use of language learning strategies as proficiency develops
The following research reports and findings show that there are differences in strategy choice and use among students of different proficiency levels. Lai (2009:273), for instance, found that when studying the perceived use of LLS among first-year students of English in a Taiwanese university, the group of learners who showed higher proficiency reported greater LLS use than the group that showed lower proficiency. While conducting research among Taiwanese university learners of English, Chang and Liu (2013:196) found that while metacognitive strategies were reportedly used most by more proficient students, compensation strategies were reportedly used most often by less proficient students.
It is reported in studies conducted in different scenarios and contexts (Alhaisoni 2012:122;Bruen 2001:221;Chang & Liu 2013:196;Griffiths 2003Griffiths :373, 2007Hong-Nam & Leavell 2006:400;Lai 2009:255;Peacock & Ho 2003:182;Wharton 2000:205-206;Yilmaz 2010:686) that compared to students with low proficiency, students with high proficiency generally use strategies more frequently. According to Chamot (2004:18), differences between less proficient and proficient students seemingly relate to the number and range of strategies used, how the strategies are applied to the task, and the appropriateness of the strategies for the task. These studies also show that more proficient learners report (1) more frequent strategy use, (2) more types and a greater variety of strategy use, (3) to be seemingly better at choosing the appropriate strategy for the task at hand, and (4) applying the strategies to the task as distinguishing themselves from less proficient learners.
The relationship between reported strategy use and proficiency, although positively correlated, is not so simple or straightforward as studies apparently suggest. It is important to note that results cannot be generalised, although the abovementioned studies show that more proficient students report using more strategies. In fact, other researchers have reported either negative correlations, or no correlation, between the level of proficiency and the use of LLS (Lai 2009:258).
It was found that more proficient students reported using fewer communication strategies than less proficient students; however, the more proficient students used strategies more effectively than the less proficient students. This was noted in Chen's (1990:178-179) work that was conducted among Chinese students. Chen also found a positive correlation between communicative effectiveness and proficiency level. Research that was conducted among university students at Jiangxi Normal University in China by Rao (2006:491) observed that high-and low-level students do not use strategies as frequently as students with moderate proficiency. Green (in Bedell & Oxford 1996:22) found that 'higher proficient students used more strategies overall than lower proficient students, but mid-proficient students used more strategies than either high or low proficient students, thus creating a curvilinear pattern'. Hong-Nam and Leavell (2006:399) also found 'a curvilinear relationship between strategy use and English proficiency, revealing that students in the intermediate level reported more use of learning strategies than beginning and advanced levels'.
These findings show the need to investigate the matter more closely by using different approaches, although, at first glance, all these results might look inconsistent. Hence, the need to use qualitative methods alongside quantitative methods to better understand the reasons behind the use of learning strategies by more and less proficient learners.

Research method and participants
This longitudinal study made use of quantitative and qualitative techniques to collect data from seven undergraduate students at different points in time over a period of 3 years. The participants, who started studying Spanish with no previous knowledge in their first year, successfully completed the third year in 2016 with an upperintermediate level B2.
At the beginning, a general questionnaire was administered to collect basic information from the participants, and to find out their motivation to study Spanish and if they had previous experience in learning other languages. The Strategy Inventory of Language Learning (SILL), developed by Oxford (1990), was administered in the first year and recurrently in the first and second semesters of the second and third years. Interviews and in-depth descriptions of the use of strategies were also used to better understand the quantitative data collected by Oxford's SILL. There has been an ongoing concern, highlighted by White, Schramm and Chamot (2007:93), about the limitations of using only selfreported and quantitative instruments to access learners' mental processing.
Likert-scaled ratings are used by Oxford's SILL for each item. It is structured around six strategy factors; each of these factors is represented by a specific set of strategy items: items 1-12 represent memory strategies; items 13-40 represent cognitive strategies; items 41-48 represent compensation strategies; items 49-64 represent metacognitive strategies; items 65-71 represent affective strategies; and items 72-80 represent social strategies (Hsiao & Oxford 2002:373). A self-reported generalised picture of individual learners' typical strategy use was provided by using Oxford's SILL (Oxford 1999:114). Frequency of perceived strategy use was reported by students through choosing for each strategy item the value that best matched their frequency of use on a scale from 1 (never or almost never) to 5 (always or almost always). Three class intervals of range 1.33 were established to classify perceived frequency of use of strategies, considering that the scale range was 4. The class intervals were the following: low frequency use from 1.00 to 2.33, middle frequency use from 2.34 to 3.66, and high frequency use from 3.67 to 5.00.
The reason why Oxford's SILL (a quantitative instrument) was used in this study was because it has been reported to be the most frequently utilised instrument for assessing LLS use (Dörnyei 2005:181;Nemati 2013:33). It has undergone notable revisions since the 1990s and has been translated into multiple languages. It is currently acknowledged by Nisbet, Tindall and Arroyo (2005:101), Putri and Fatimah (2021:193-194), and Alfarisy (2022:91) as a very popular instrument that is widely used around the world because it is comprehensive and easy to employ for identifying perceived strategy use of language students, and 'it has been extensively checked for reliability and validated in multiple ways' (Nisbet et al. 2005:101).
Two types of interviews were used to collect qualitative data. Firstly, a preliminary, non-structured general interview, with the purpose of building a relationship with the student, getting to know them, and verifying the information provided in the general questionnaire. Secondly, four indepth interviews to generate data concerning the factors associated with the persistent use of specific strategies and the significant differences in their use over a specific time period. All the data collected helped to describe and analyse the changing roles in the use of LLS.

Findings on reported frequency of strategy use
The data gathered by Oxford's SILL showed that students in the first year were making more frequent use of strategies that regulated their own learning process (metacognitive strategies) than strategies that fostered interactions with other people (social strategies) and strategies that enabled them to use and manipulate the language being learnt (cognitive strategies). The data also showed that in the second and the third years, the reported perceived use of metacognitive strategies was no longer the strategy category with the highest perceived use. Students in the second year reported using more social strategies than the other types of strategies, and in the third year, more cognitive strategies than the remaining strategies.
Considering that Oxford's SILL uses Likert-scaled ratings from 1 (which means never or almost never) to 5 (which means always or almost always), Table 1 shows the average ratings of the reported use of LLS by the student cohort towards the end of each stage.
Interviews provided rich data that helped better understand the quantitative findings mentioned above. Qualitative techniques were critical to identify and explain patterns of change in LLS use over time as the students developed proficiency in Spanish. The dialogues below are edited excerpts of different interviews that the researcher held with the participants. The numbers are used to differentiate between students and the letters indicate the specific proficiency level of students. Thus, 'a' refers to A level, 'b' to B1.1 level, 'c' to B1.2 level, 'd' to B2.1 level, and 'e' to B2.2 level: 'I see that you have increased the use of the cognitive strategy item 'I watch TV shows or movies or listen to the radio in the new language' more than in the past. Can you explain why and how you do it?' (Researcher) 'I listen to a lot of music, and I watch short movie clips online just to familiarise myself with the language.' (Participant 1b) 'So, why didn't this strategy help you so much in the first year, and now in the second year is helping you more? Why is it so?' (Researcher) 'Because in the first year I just couldn't comprehend most of what was being said. Now that I have a better understanding of the language, I can understand most of it.' (Participant 1b) These responses show that first-year students were not making use of certain cognitive strategies such as watching or listening to media in the new language due to the fact that their level of proficiency did not allow them to do so. It was only as from the second year that students improved in their understanding of TV shows, movies and radio broadcasts. This was consistent with previous research findings, showing that strategy choice was affected by the student's level of proficiency at different points in time. What was probably a frustrating learning activity in the first year progressively became an exciting activity in the second and third years.
Between first year and second year, a normal developmental progression occurs; the students' expansion of knowledge of 'Why is it so important for you to practice the language now and not in the past?' (Researcher) 'Because I've realised that I need to practice the language verbally so that I can become more comfortable when speaking to people. I want to be able to answer questions in Spanish in class but in order for me to do that I need to be comfortable and not get so nervous. I want to practice as much as I can.' (Participant 2b)

[Another student]
'And why did you report low use of the cognitive strategy item 'I say or write new expressions repeatedly to practice them' in the past, and now you report higher use?' (Researcher) 'I think it was … I was a bit too shy to say the things mostly so now I've decided to stop being so shy and recognise the need to actually practise.' (Participant 5c) However, as genuine interactions with proficient or native speakers of Spanish are scarce in South Africa, it was found that students who did not have Spanish-speaking friends ended up interacting with each other. Thus, contextual factors relating to the situational context of learning (as pinpointed by Ellis 2008; Griffiths 2013), which characterise the teaching of a foreign language, increased the perceived use of the metacognitive strategy where students actively look for language partners. This eventually caused second-year students to start seeing their classmates as potential language learning partners to practise the language with in order to improve and consolidate their learning experience (Griffiths 2008): 'As for the metacognitive strategy item 'I actively look for people with whom I can speak the new language', the information you gave us shows that you have increased the use of this strategy. Can you explain how you do it and why?' (Researcher) 'I used to refrain from interacting with my classmates in the past. Now I interact more freely with them and we try to use the language when speaking together … In the past … In the first year … you are a new student and everything was stressful … You are mostly nervous all the time so this year you are accustomed to everything so you are more relaxed ….' (Participant 1b) As the need for using the language to communicate with others became stronger in the second and third years, practising with a classmate became more helpful, but it was not fully rewarding. 'The sense' that the students were benefiting from practising with each other was still not so strong: 'As for the social strategy item 'I have a regular language learning partner', you indicated that you didn't use it. Your reported use was very low in the 1st SILL and now you are using it more. How is this happening?' (Researcher) 'I have a friend … Yes. We get together most of the time and then we talk in Spanish. It is helpful but not a lot … We get the sense [that] we are both learning.' (Participant 1b) Despite the limitations in interacting with native or competent speakers, there seems to be an interesting difference shown in the data between focusing on the new language (where the language itself is the object of study) and focusing on communication (where using the language for communicative purposes is the object of all endeavours). It appears that students in the first year were more inclined to focus on the language and study it as the object of all their endeavours, whereas students were more inclined to not only study the language in the second year, but also to use it for communicative purposes. The shift of emphasis from studying for the sake of completing the course to being able to use the language in a communicative setting is crucial and necessary in the development of language proficiency. The below student reflects on this and acknowledges that more practice was needed to become more confident in learning and using the new language: 'Well we've set up … my friends and I … we meet up and then we speak Spanish together … It helps to practice speaking because we do a lot of reading and practice in class, but uhm to actually speak the language and to get comfortable with it you need to practice it more. So, we just … we do that more often.' (Participant 2b) It was reported that the large amount of information given in the second year prevented students from creating associations between new material and what students already know. Some students perceived that the memory strategy item that deals with associations was confusing them instead of helping. This confusion was possibly happening because the associations were no longer linear or direct, involving concrete and simple words or expressions, but rather abstract and incorporating more complex structures: 'Why have you reported a decrease in the use of the memory strategy item 'I create associations between new material and what I already know?' (Researcher) 'Uhm usually because now that we are in the second year we get so much new information to learn that I just sort of write notes and I don't really create associations anymore I feel that the content is a lot more and that creating associations sometimes make me a little confused …' (Participant 2b) The realisation that students need to try to understand the language without translating it word-for-word into their own language and should learn the language more independently seems to be a major characteristic that distinguishes second-year students: 'As for the cognitive strategy item 'I try to understand what I have heard or read without translating it word-for-word into my own language'. Why is this so important to you?' (Researcher) 'Uh it's to learn the language more independently so that you don't have to rely on your own language to understand so you to learn it better.' (Participant 5c) 'Do you find that this is becoming one of your major concerns now that you want to understand the language itself better?
'Yes.' (Participant 5c) In connection with a compensation strategy, it was found that there is an interesting progression of how students' perception of some strategies changed over time, referring to the making up of new words when the correct ones are not known.
http://www.literator.org.za Open Access The following remarks were made by some students during the first semester and even during the second semester: 'I just don't think it's going to be helpful to make up new words if I don't understand a word it will just lead me into the wrong direction I guess.' (Participant 1b) 'No, I just don't think it's rational to try make up words for yourself so I rather like go into a dictionary and look for the word itself yeah.' (Participant 8b) 'Uhm I'd rather not make up words uhm I'd rather learn the correct words and uhm instead of making them up otherwise it's not correct.' (Participant 4c) A slight change of perception was noticed in respect of this compensation strategy during the second semester of the second year. In order to avoid breakdowns in communication, some students started to think that it was perhaps worthwhile to make up new words: '[The data] shows that you've increased the making up of new words when the right ones are not known. Why is it that it was not relevant to you and now it is becoming more relevant?' 'Because I still want to say something and I don't always know the right word and I uhm so I just say a word and then if it's wrong then someone can help me to correct it but usually it's right because there are a lot of similarities in English and in Spanish so you can just …' (Participant 7c) Second-year students reported an increase in the frequency of use of the cognitive strategy related to reading for pleasure in the new language. They started to realise that their newly acquired vocabulary allowed them to read more fluently in Spanish than previously: 'Yeah, there is also an increase in reading for pleasure in the new language.' (Researcher) 'Uh, yes, because now that my vocabulary is a bit more broadened I feel I can understand better so I don't have to sit and I don't feel like I have to … I don't feel I have to sit with a dictionary and look at every single word I understand the general idea of the passage.' (Participant 4b) This student no longer needed to look up every word in the dictionary or glossary (cognitive strategy) during reading activities, implying that the student probably decreased the frequency of use of dictionaries when reading in Spanish, while simultaneously increasing the frequency of reading for pleasure (cognitive strategy).
Students further reported that the development of their vocabulary made them consult the dictionary less frequently when they were reading texts in Spanish in the second semester, as they were now reading without looking up each unfamiliar word: Third-year students reported that the strategies they most frequently used were cognitive, social and compensation strategies as shown in the data. No memory strategies were included in this top list however, contrary to what happened previously.

Discussion of findings
The realisation that learning does not only happen in the classroom, but also during activities outside the classroom, seems to make first-year level students proactive in planning and arranging their own language learning process. At second-year level, students start to realise the 'social nature of learning and the key role that teacher and students' peers play in facilitating learning for an individual' (Pearson & Cervetti 2015:7). The formation of practice groups is encouraged by the need to engage in interactive activities with more capable peers; this emerges as a practical alternative to counterbalance the lack of opportunities to practise the foreign language outside the classroom. The newly acquired upper-intermediate level of proficiency seems to empower third-year level students to engage more with strategies related to the 'ways of using' the language than with strategies related to the 'ways of learning' the language.
From the data collected in the interviews, it was evident that certain cognitive strategies such as watching movies, listening to music and reading for pleasure were not preferred by beginner students, but progressively became more useful as their proficiency developed. Other cognitive strategies such as repeatedly saying or writing expressions became more regular as beginner students expanded their vocabulary and learnt more grammatical structures. It was also evident that social strategies, such as practising the language with others, were not the main concern of beginner students until they progressed to the second year. However, contextual factors -related to the fact that students were only exposed to the new language within the classroom and could not easily find proficient speakers of the new language in their surroundings -caused students to reconsider their classmates as potential language learning partners outside the classroom. Another observation from the data is that the perceived use of certain compensation strategies, such as making up new words, increased over the time period. With beginner students reluctant to make up new words, whereas more advanced students were willing to make up new words to avoid communication breakdown.
Conversely, the reported perceived use of certain memory strategies, such as creating associations between new content and what students already knew, decreased from beginner students to more advanced learners. This showed that mental associations became more subconscious and automatic at higher levels of proficiency making the perceived use of this strategy less frequent. Certain cognitive strategies such as translating text word-for-word to understand the meaning were also reported to be made use of less frequently. This is because more advanced students learn to derive meaning from the context compared to beginner students who normally depend on their dictionary.
In order to classify and comprehend the above-mentioned findings and to be able to discuss the role that the use of LLS plays in learning a foreign language like Spanish, it is important to remember not only the definition but also the main purpose that lies behind the concept of LLS. As indicated earlier, strategies are considered to be activities, actions or steps strategically chosen and consciously taken by students to regulate and improve their language learning experience. The ultimate goal is to improve the learning experience of the new language. Taking into consideration the findings mentioned above in the previous section -that students' level of proficiency determines to a certain extent the perception they have on the usefulness of each LLS throughout their three-year language learning journey -it is possible to understand why it is that LLS are sometimes perceived to produce desirable effects, and other times perceived to produce neutral effects or undesirable effects. Based on these perceptions and considering the students' level of proficiency at different points in time, a tentative framework that describes the multiple roles that LLS can play is proposed below. The implicit question that this framework seeks to answer is: What distinct roles do LLS play at different points in time of the new language learning process: • Essential role: Students perceiving the use of certain strategies as indispensable and necessary to develop proficiency in the new language. • Catalytic role: Students perceiving the use of certain strategies as facilitating and/or speeding up the development of proficiency. • Imperceptible role: Students perceiving the use of certain strategies as having no effect on the development of their proficiency. • Retarding role: Students perceiving the use of certain strategies as slowing down their further development of proficiency in the new language. • Preventing role: Students perceiving the use of certain strategies preventing them from further developing their proficiency.
Although it is assumed that LLS play mostly supporting and constructive role, being dynamic in nature, this study found that the perception is that the benefit of LLS use changes over time according to changing levels of language proficiency. To illustrate this, the following examples show the roles that LLS can play at different points in time of the development of language proficiency: • The following LLS can play an essential role for beginners: 'using reference materials such as dictionaries and glossaries', 'imitating the way native speakers talk', and 'reading in the new language'. In the case of students at both intermediate and advanced levels, 'actively looking for people with whom to speak the new language' (Oxford 1990). • The following LLS can play a catalytic role for beginners: 'looking for patterns in the new language', 'planning goals for language learning'. In the case of students at intermediate and advanced levels, 'watching TV shows or movies' or 'listening to the radio in the new language', and 'trying to think in the new language' (Oxford 1990). • The following LLS can play an imperceptible role for all levels of students: 'keeping a private diary or journal where they can write their feelings about language learning', 'physically acting out a new word to remember', and 'giving themselves a tangible reward when they have done something well in their language learning' (Oxford 1990). • The following LLS can play a retarding role for beginner students: 'reading without looking up every unfamiliar word', while their vocabulary is small. In the case of students at intermediate or advanced levels, 'listing known words that are related to a new word and drawing lines to show relationships', because their pace of work slows down, as the students' vocabulary by this time is larger and contains more complex, polysemic terms, making it a time-consuming activity (Oxford 1990). • The following LLS can play a preventing role for beginner students: 'watching movies or listening to radio' can be a frustrating experience if not done appropriately and at the right time. In the case of students at intermediate or advanced levels, 'looking up every unfamiliar word' can impede the evolution of compensation strategies, such as 'learning to guess the general meaning by using clues that can be found in the text' or 'finding a different way to express an idea when one cannot think of the right expression' (Oxford 1990).

Conclusion
The researcher found a lack of practical and descriptive roles that explain how students at different levels benefit from LLS according to their perceived usefulness at different stages of the students' development, and therefore created the proposed framework that refers to the categorisation of the different roles that strategies can play. This framework adds a new dimension and provides valuable information to quantitative studies from a practical point of view, as these have been the dominant type of study in LLS research. The proposed framework will allow students to report not only the frequency of strategy use, but also the perceived usefulness of each strategy at different moments of the development of their proficiency in the new language. By adding a second dimension to the quantitative data, this can address one of the weaknesses of snapshot-type studies. The five roles defined in the proposed framework can be easily coded as nominal data to enrich statistical analyses.