313 Studies in Second Language Learning and Teaching Department of English Studies, Faculty of Pedagogy and Fine Arts, Adam Mickiewicz University, Kalisz SSLLT 9 (2). 2019. 313-349 http://dx.doi.org/10.14746/ssllt.2019.9.2.4 http://pressto.amu.edu.pl/index.php/ssllt Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language learning context Richard LaBontee Gothenburg University, Sweden https://orcid.org/0000-0001-5180-7005 richard.la.bontee@svenska.gu.se Abstract This article reports on the final round of piloting for a questionnaire instrument created and designed with the intention of collecting self-report data on adult, beginner Swedish L2 learners’ use of vocabulary learning strategies. The Swe- dish Vocabulary Learning Strategy Survey (SVLSS) version 1.2 is distributed to 182 participants studying Swedish at institutes of higher learning in Sweden. The collected data set is subject to exploratory factor analysis to explore initial interpretations of the underlying constructs of the instrument, and analyzed for content validity and internal consistency. Readability and accessibility of the in- strument is also addressed. Initial findings and interpretations are used to guide the development of a preliminary VLS taxonomy for the SVLSS, as well as sug- gest and perform revisions that will result in the SVLSS 2.0. Keywords: second language acquisition; vocabulary learning strategies; Swe- dish language; questionnaire instrumentation 1. Introduction In studying reported vocabulary learning strategy (VLS) use in second language (L2) learning contexts, questionnaire instruments are often employed to collect Richard LaBontee 314 wide amounts of data in relatively short time frames. These questionnaires have resulted in VLS lists and taxonomy that help to organize the strategies that learn- ers use. Collected data can be interpreted using these classifications in order to examine VLS use patterns for research and diagnostic purposes. However, as pointed out recently by Gu (2018), VLS questionnaires seem to lack in explicit reporting on their creation, design, instrumentation, and validation practices. As a means of exercising transparency in questionnaire instrumentation and taxonomy development, this study aims to perform validation analyses on the Swedish Vocabulary Learning Strategy Survey (SVLSS) instrument (LaBontee, 2016) using collected data from adult, beginner, Swedish L2 learners. Results and analyses provide guidance towards the proposal of a tentative VLS taxon- omy that is, in turn, used to inform revision of the questionnaire item pool and to motivate an updated taxonomy for the next iteration of the SVLSS. 2. Literature review 2.1. Language learning strategies Rubin (1975) described language learning strategies (LLS) as the “techniques or devices which a learner may use to acquire knowledge” (p. 43) in reference to observed behaviors that “good language learners” exhibited during their lan- guage learning. Oxford (1990) offered a classification system for LLS that distin- guished between “direct” strategies which included memorization, cognitive and compensatory strategies, and “indirect” strategies that reflected strategic social and affective approaches, as well as meta-cognitive planning of language learning. More recently, Oxford (2011, 2017) has departed from her six-category LLS taxonomy, instead offering a model that more strongly integrates a self-reg- ulative model of language learning (i.e., Tseng, Dörnyei & Schmitt, 2006). The model classifies LLS into three dimensions, cognitive, affective and socio-cul- tural interactive, that operate interconnectedly on two levels: as meta-strategies that are performed to regulate planning, feelings and strategy use, and as the specific strategies performed with the intention of improving learning or com- pleting a task. Oxford (2017) has attempted a holistic definition of LLS according to a meta-review of LLS research, describing them as diverse in form, purpose- ful, conscious, flexible in use, occurring situated in the contexts they are used in, and as teachable actions or learning behaviors (p. 48). Considering the actual use of LLS, Gu (2018) stresses that strategic learn- ing should be viewed essentially as a problem-solving process that operates in a cyclical fashion as new tasks arise. We analyze the task at hand, analyze our own resources for learning, analyze the context of learning to come up with a Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 315 plan of action, then monitor effectiveness of the plan while adjusting if needed, and then evaluating whether or not our efforts were successful when complet- ing the plan (Gu, 2018, p. 326). 2.2. Vocabulary knowledge If LLS reflect actions intended to help one facilitate learning any aspect of lan- guage, VLS are used to help facilitate the learning of vocabulary knowledge in a target language (TL). Vocabulary knowledge encompasses a many-faceted and complex system of word-related features that learners attend to when learning and using languages. In order to approach strategy use for learning vocabulary knowledge, some models used to classify L2 word knowledge will be introduced to illustrate what kinds of knowledge strategies can be used to help acquire. Oxford and Scarcella (1994) investigated the schemata development for vocabulary knowledge storage in L2 learners, laying out an outline of “essential knowledge” for knowing a word. Their list illustrates a functional perspective of word knowledge as related to the learner. They list: form (pronunciation, spelling, word parts, morphology), grammatical use (plurality, pre-/affix, sen- tence construction), collocations (syntax of co-occurring words), discourse func- tion in particular situations and contexts, shades of meaning (connotations), and receptive and productive uses. Henriksen (1999) presents a more theoretically-anchored lexical knowledge model that occurs on three mutually interactive dimensions: partial-to-precise knowledge, depth of knowledge, and receptive-to-productive use ability. Partial- to-precise knowledge encapsulates a non-ordinal continuum referring to various registers of use for a lexical item, comprehension of that item, associations to that items, and forms that may be available to a language user. Depth of knowledge refers to the extent in which singular or stringed lexical units are linked in the lex- icon via cognitive networks. Depth accounts for knowledge regarding a word’s morphological, syntactic, and collocational profile, and its potential to carry meaning. The receptive-productive dimension assumes that lexical knowledge commonly moves along a continuum beginning with receptive knowledge (e.g., recognition/comprehension), gradually progressing to productive (e.g., spontane- ous use). However, the receptive-productive dimension should be regarded as ex- isting on a non-precise continuum rather than in dichotomous categories – one can retain lexical knowledge that can be used receptively and/or productively, and that knowledge may increase or decrease in either regard over time. On a func- tional level, Schmitt (2010) explains the dichotomy by stating that receptive knowledge contains meaning recognition and meaning recall, where productive knowledge contains form recognition and form recall. Richard LaBontee 316 Table 1 Nation’s (2013, p. 49) word knowledge taxonomy Receptive Productive FO RM Spoken What does the word sound like? How is the word pronounced? Written What does the word look like? How is the word written and spelled? Word parts What parts are recognizable in this word? What word parts are needed to ex- press the meaning? M EA N IN G Form and meaning What meaning does this word form signal? What word form can be used to ex- press this meaning? Concepts and referents What is included in the concept? What items can the concept refer to? Associations What other words does this make us think of? What other words could we use in- stead of this one? U SE Grammatical functions In what patterns does the word oc- cur? In what patterns must we use this word? Collocations What words or types of words occur with this one? What words or types of words must we use with this one? Constraints on use Where, when, and how often would we expect to meet this word? Where, when and how often can we use this word? Nation (2013) organizes word knowledge into three major categories: form, meaning and use (see Table 1). Form refers to lexical features of spoken knowledge (i.e., phonetics), written knowledge (i.e., orthography), and word parts. Meaning refers to form and meaning (i.e., what does the word form in- form about meaning), concept and referents, and associations (i.e., to related words, to synonyms). Use refers to grammatical functions (i.e., what patterns govern this word), collocations (i.e., words that tend to occur with or near each other), and constraints on use (i.e., register, frequency). The three dimensions of lexical knowledge are fluidly accessed during language learning and use, and are inextricably interconnected. Form, meaning and use knowledge features are expressed in Nation’s model through both receptive and productive examples. 2.3. Vocabulary learning strategies Second language vocabulary learning is a crucial factor for language acquisition at every level of proficiency, but is particularly important for beginner learners who must acquire large amounts of words quickly in order to access their new lan- guage. The difficulty and complexity of learning vocabulary requires the choice, and appropriate use of strategies that help and enhance the learning process. Gu (2003) notes that vocabulary learning strategies (VLS) should serve two purposes: to get more knowledge about words, and to be able to use that word knowledge productively. Oxford (2017) defines VLS as “teachable, dynamic thoughts and be- haviors that learners consciously select and employ in specific contexts to improve their self-regulated, autonomous L2 vocabulary development” (p. 244). Exploratory research into language learners’ VLS use has resulted in a va- riety of taxonomy intending to classify different kinds of VLS. These taxonomy Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 317 are realized as strategy lists that have been purposed as questionnaire items to collect data on learners’ reported frequency of VLS use. Gu and Johnson (1996) divided VLS into metacognitive regulation strategies and cognitive strategies on their 108-item Likert-scale Vocabulary Learning Questionnaire (VLQ). Metacog- nitive regulation represented strategic planning and self-regulation of learning, where cognitive strategies were further divided into guessing strategies, diction- ary strategies, note-taking strategies, and memory strategies for both rehearsal and encoding. The VLQ also included a section of items probing learners’ vocab- ulary beliefs regarding their learning situation, experience, and motivations. Re- cently, the VLQ taxonomy organization was updated, grouping components of the cognitive strategy category into initial handling strategies (strategies for es- tablishing new word knowledge), reinforcement strategies, and activation (use) strategies (Gu, 2013). Also, the VLQ item list and construct structure have been revisited with the intention of evaluating and adjusting accessibility to non-Eng- lish L1 participants, construct validity, and item list appropriateness (Gu, 2018). Stoffer (1995) surveyed learners and colleagues for different kinds of VLS, populating a list that comprised her Vocabulary Strategy Inventory (VOLSI), a 53- item Likert-scale questionnaire. Unlike many other VLS lists she performed fac- tor analysis on a data set collected from university-level English speakers learn- ing a variety of L2s in order to investigate the underlying VLS constructs of the VOLSI, rather than presuming categories beforehand. Her taxonomy reflected VLS categories of: authentic language use, creative activities, self-motivation, creating mental linkages, memory strategies, visual/auditory, physical actions, overcoming anxiety, and organizing words. Other questionnaires have sought to collect data on learners’ perceived usefulness of VLS. Schmitt (1997) created a 58-item questionnaire that divides VLS into two categories based on whether or not the strategies were used to discover the meaning of a new word, or to consolidate a word after it had been encountered. Discovery strategies included determination and social strategies, and consolidation strategies included social strategies, memory strategies, cog- nitive strategies, and meta-cognitive strategies. Fan (2003) designed a question- naire that would be used to collect both frequency of VLS use data and learners’ perceived VLS usefulness data. Fan’s 56-item Likert-scale Vocabulary Learning Strategies Questionnaire (VLSQ) closely resembled the VLQ (Gu, 2018; Gu & Johnson, 1996) with regards to overall taxonomy. The VLSQ involved meta-cog- nitive management strategies, source strategies, guessing strategies, dictionary strategies, repetition strategies, association strategies, grouping strategies, analysis strategies and known word strategies. Richard LaBontee 318 Table 2 A taxonomy of vocabulary-learning strategies (Nation, 2013, p. 328) General class of strategies Types of strategies Planning: choosing what to focus on and when to focus on it Choosing words Choosing the aspects of word knowledge Choosing strategies Sources: finding information about words Analyzing words Using context Consulting a reference source in L1 or L2 Processes: establishing knowledge Noticing Retrieving Generating (creative use) Skill in use: enriching knowledge Gaining in coping with input through listening and speaking Gaining in coping with output through reading and writing Developing fluency across the four skills Nation (2013) provides a three-category classification of different types of VLS: planning strategies, source strategies, and processing strategies (see Table 2). A fourth dimension, skill in use, encompasses strategy use but is concerned with the use of vocabulary input and output for both the enrichment of vocab- ulary as well as the development of the four skills – reading, writing, speaking, and listening. Planning strategies are used to choose what to focus on and when to focus on it, for example, choosing words, aspects of word knowledge, and strategies. Source strategies are used to find information about words, for ex- ample, using a dictionary, guessing meaning from surrounding context, or using background knowledge to guess meaning. Processing strategies are used to re- inforce acquired knowledge through the use of noticing, retrieval and genera- tion (productive activation). These VLS classifications are considered fluid de- pending on the actual use of a strategy according to the task being approached, context in which it occurs, and individual using the strategy. 2.4. Questionnaire validation The use of questionnaire instruments as elicitation tools offers up a relatively quick and efficient method for collecting large scale self-report data on learners’ LLS use (O’Malley & Chamot, 1990). These data can then be used to examine strategy use patterns amongst groups of learners through a variety of analytical approaches (e.g., Gu & Johnson, 1996; Fan, 2003). However, the most popular VLS taxonomies offered through VLS lists or questionnaires have not reported (published) validation procedures, as pointed out by Gu (2018) in a recent update to the VLQ. Gu (2018) explains that questionnaires used to collect VLS use data assume that “strategies are latent sets and episodes of behavior that can be observed and described” (p. 328), and are represented by item statements on the question- naire. A Likert-scale style questionnaire asks participants to respond to these Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 319 statements with how often they perform the represented strategic behaviors, then in turn uses collected responses to illustrate strategy use by individuals or groups of learners. Although perfect representation of the construct reflected by a questionnaire is near impossible, best representation of the intended VLS con- structs with the statements presented on a questionnaire should be a goal of val- idation procedures. Gu (2018) points out that systematic review of statements in the item pool should be performed to ensure relevance and representativeness of the target construct(s), items constituting a single scale should represent con- sistency amongst themselves, and interpretation of collected scores should be used appropriately and with the intended population (p. 328-329). Dörnyei and Csizér’s (2012) chapter on questionnaire creation, use, and validation offers some guidance on approaching validity and reliability evalua- tion. They caution that item generation and/or sampling should be considered carefully with regard to the constructs that are intended for investigation. This requires a particular attention to the way statements are crafted, as “the word- ing of the questions assumes an unexpected amount of importance . . . [and] can produce radically different levels of agreement or disagreement” (p. 76). When generating questionnaire items, Dörnyei and Csizér (2012) suggest the use of exploratory data gathered from respondents using qualitative methods to generate item pools, or borrowing questions from other established ques- tionnaires. When writing and revising the item pool, short, simple, unambigu- ous, single-question items that reflect the core construct studied are best. Fur- ther, following data collection, factor analysis and calculation of a Cronbach’s alpha coefficient are suggested as methods for exploring and interpreting po- tential underlying factors that influence the interrelated correlations between items on an instrument, and for determining item homogeneity and reliability in a supposed thematic group, respectively (p. 85). 2.5. Swedish VLS context Most studies exploring VLS use in the Swedish L2 context have done so in the context of public primary and secondary school Swedish L2 programs. For exam- ple, the STRIMS-projektet, a Swedish research initiative during the 1980’s and 90’s was concerned with performing exploratory research into the LLS used by stu- dents in classroom contexts at primary and secondary school levels (Tornberg, Öman, Bergström, & Håkanson, 2000). Other studies have gathered data concern- ing LLS or VLS use by young-to-adolescent-aged students in the Swedish school system as a component of establishing group or individual learning profiles of stu- dents (Allestam, 2007; Magnusson & Öggesjö, 2013; Malmberg, 2000; Wareborn, 2004). Findings of these studies suggest that younger Swedish L2 learners prefer Richard LaBontee 320 the use of memorization strategies over more creative (production-based) strate- gies, and that the use of strategies is influenced by contextual factors related to interest and relevance for word learning (i.e., multimedia, social media). In the adult Swedish L2 learner context, Granberg (2001) used Oxford’s (1990) 50-item Strategy Inventory for Language Learning (SILL) coupled with inter- views to collect LLS use data from a single learner as part of a longitudinal, qualita- tive case study illustrating the Swedish language learning experience of the adult immigrant to Sweden. Sandh (2013) also used the SILL to investigate the VLS use of two groups of adult, mixed-proficiency Swedish L2 learners. The SILL is intended for exploring LLS, but was adapted by Sandh to investigate VLS use. Collecting data with an instrument that was not designed to collect a certain type of data, and in a con- text that was not originally intended for that instrument, can result in unreliable, misrepresented data (Dörnyei & Csizér, 2012). Her findings suggested that meta- cognitive and social strategies were most popular for vocabulary learning, while af- fective strategies were not often used. Learning profiles were proposed according to collected data that divided learners into expert vs. novice language learners, groups according to time spent studying, L1s and motivations. 3. The study To the knowledge of this author, no large-scale research efforts have been under- taken to explicitly explore what VLS are used by adult, beginner Swedish L2 learners studying at Swedish institutes of higher education. Findings from such research would help to establish a starting point for better understanding learner groups in the adult Swedish L2 learner context, and could be used to develop reflective and diagnostic tools for educators in the field. In order to facilitate such research, a new instrument, the Swedish Vocabulary Learning Strategy Survey (SVLSS), has been created with the explicit intention of being developed for use and distribution with adult, beginner Swedish L2 learners. As a final part of an ongoing pilot process for the SVLSS (LaBontee, 2016; LaBontee, 2018), this study represents efforts to evalu- ate the SVLSS (version 1.2) with regard to accessibility, construct validity, and con- tent validity. The primary question raised in this study is: 1. What VLS constructs seem to be represented by the SVLSS instrument? This question will be addressed through the use of exploratory factor anal- ysis and subsequent interpretation of any factor groups found. A VLS taxonomy for the SVLSS will be proposed from these findings. Using this proposed model, a critical examination of the questionnaire items as related to their respective VLS groups will be performed: Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 321 2. Are all items on the SVLSS instrument relevant to and representative of the proposed VLS taxonomy? 3. Is the SVLSS instrument accessible to a range of English L2 users? Any issues arising from investigation of the above questions will be ad- dressed and adjustments will be made to the SVLSS instrument. The findings of this study, alongside instrumentation considerations related to VLS taxonomy occurring in the literature (LaBontee, in press), are applied to the SVLSS instru- ment, resulting in adjustments to its conceived VLS taxonomy and item pool. 3.1. The SVLSS (version 1.2) Creation and development of the SVLSS were originally driven by the lack of an instrument designed explicitly to collect self-report data on adult, Swedish L2 learners’1 VLS use. Creating an instrument for specific use in the planned context (rather than adapting another instrument) can help to better ensure that the data collected is the data intended to be collected, and that the instrument con- tent is relevant to the surveyed demographic (Dörnyei & Csizér, 2012). The VLS use data collected by the SVLSS will be used to explore vocabulary learning pat- terns for groups and individuals learning Swedish as a L2. This information is intended to be used both in research contexts and for pedagogic purposes. The SVLSS instrument used in this study represents the third iteration of the piloting process. The initial SVLSS 1.0 was the result of item pool generation based on content-analysis of qualitative data collected from the target audience that used semi-structured interviews paired with a vocabulary learning task (LaBontee, 2016). The SVLSS 1.0 was piloted for general accessibility regarding reading ease and item pool appropriateness, and was adjusted according to participant response data, and one-on-one feedback, resulting in the SVLSS 1.1. This iteration was piloted again using the same approach, and further refined for readability and item fit, resulting in the iteration used in the study at hand, the SVLSS 1.2. The SVLSS 1.2 was distributed online using the Google Forms™ survey platform. The instrument uses Likert-scale response to statements ranging from 1 = Not true of me, to 5 = True of me. The SVLSS 1.2 contains 74 item statements written in English. All items were written with the intention of eliciting self-re- port data from Swedish L2 learners on their strategic vocabulary learning behav- iors. In the current and previous iterations of the SVLSS, items were separated into 7 sections. Each section represented initial grouping themes during the con- tent mapping of strategic behaviors as part of the item pool generation process 1 Heretofore referred to as “the target audience” or “target demographic.” Richard LaBontee 322 for the SVLSS 1.0. These arrangements were used to organize the SVLSS with the intention of contributing to ease-of-use and content comprehension for partici- pants by grouping similar items into sections. The groupings were thematic ac- cording to their perceived VLS types: production strategies, investigation strate- gies, lexical-associative strategies, memory strategies, and motivation/planning strategies. Though these groupings were used to guide item list arrangement on the SVLSS 1.0, they were not displayed on the actual questionnaire. A detailed list of items listed on the SVLSS 1.2 can be found in Appendix A. English language was used for this questionnaire due to its status as lingua franca, and due to its use in Sweden as a functional language of higher educa- tion. As all participants of this study were studying at Swedish institutes of higher education, they were all required to have (and reported having) interme- diate-to-high levels of proficiency in their command of English language, and all felt comfortable with it as a mediating language with which to complete the SVLSS. Distribution of the survey to participants was administered by teachers of Swedish as a L2, language learning program leaders, administrators at the various institutes, and by the author, in-person. Students were asked to fill out the SVLSS outside of class and at their own pace. 3.2. Participants and collected data Participants in this study were adults over age 18, living in Sweden, and studying Swedish as a second language at various institutes of higher education2. Partic- ipants were recruited from A1, A2 and B1 Swedish language classrooms, and all identified as “beginner learners” when taking part in the study. All participants volunteered their time to fill in the questionnaire, establishing a convenience sample for the study. The final number of participants who responded to the SVLSS 1.2 after data cleaning was 182. A demographic survey preceded the SVLSS 1.2 instrument, probing partici- pants for background and individual differences information. Participants’ levels of education, their degree of multilingualism, their age, the time they have spent learning Swedish, the time they have spent living in a Swedish language dominant environment, and their native and other languages spoken were all surveyed. Col- lected demographic information for the participant sample is outlined in Table 3. 2 Sixteen institutions of higher education with Swedish as a Second Language instruction programs across Sweden. For privacy purposes, the institution names are not listed. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 323 Table 3 Participant demographics information Age group* Time spent studying Swedish Time spent in Swedish language environment Education 18-23 63 0-3 months 96 0-6 months 93 High School diploma 39 24-29 65 4-6 months 22 7-12 months 45 Bachelor’s 66 30+ 54 7-12 months 32 13+ months 43 Master’s 50 13+ months 30 PhD 27 Multilingualism Native languages Monolingual 7 Germanic 72 (English, German, Dutch, Swiss) 1 additional language 75 Romance 40 (Italian, Spanish, Portuguese, French, Catalan, Roma- nian, Basque) 2 additional languages 52 Slavic 24 (Russian, Ukrainian, Polish, Czech, Serbian, Slovenian, Bosnian, Bulgarian, Croatian) 3+ additional languages 48 Other 46 (Persian, Urdu, Dari, Finnish, Hungarian, Greek, Can- tonese, Chinese, Thai, Latvian, Lithuanian, Japanese, Thai, Vietnamese, Turkish, Indonesian, Arabic) Note. Total sample N = 182. Participants were mostly younger adult learners (under age 30), and the majority of them spent less than 12 months learning Swedish or living in a Swe- dish-language-dominant environment. Participants exhibited a wide range of multilingualism where, in fact, monolingual speakers were the overwhelming minority. Native languages reported were mostly Germanic or Romance lan- guages, but were considerably diverse across the entire sample. 4. Analyses and findings 4.1. Factor analysis and interpretations Exploratory Factor Analysis (EFA) is a statistical tool that is often used to explore the dimensionality of underlying constructs in questionnaire item lists. In L2 strategy research, it has been used to evaluate construct validity for several in- struments such as Oxford’s (1990) SILL (Park, 2011), the VOLSI (Stoffer, 1995), and the VLQ (Gu, 2018). EFA is explicitly used for exploratory investigations into underlying instrument constructs. As Osborne (2014) puts it, “EFA is an explor- atory technique. As such, it should not be used, as many researchers do, in an attempt to confirm hypotheses or test competing models” (p. 6). EFA is used3 here to provide guidance for two facets of instrumentation. First, it is used to examine the relationships between items on the SVLSS to help determine and interpret the dimensionality of VLS constructs that they represent. Second, the statistical relationships revealed by the EFA will be used to initially inform and guide revisions to the SVLSS 1.2 item pool. 3 EFA was performed using the Statistical Package for the Social Sciences (SPSS) program. Richard LaBontee 324 A Kaiser-Meyer-Olkin measure of sampling adequacy returned a result of 0.749, indicating that the sample size for this study holds “middling” acceptabil- ity for factor analysis (Cerny & Kaiser, 1977). EFA is performed here using princi- pal axes factor extraction with a varimax rotation. Principal axes factor extrac- tion allows for a choice of initial estimate of communality, and varimax rotation maximizes the variance within an extracted factor, making larger loading scores larger, and smaller scores smaller (Osborne, 2014). A total of 22 factors were found with initial Eigenvalues greater than 1, accounting for 69.50% of the variance. A table outlining the initial total variance explained for the SVLSS 1.2 can be viewed in Appendix B. As the returned 22 factors failed to display clear breaks in explanation of factor variance, scree plot- ting was consulted to provide further guidance. The number of data points above the “elbow” in which the slope of the scree plot curve changes markedly can provide a good estimate of the ideal number of factors to retain (Osborne, 2014, p. 18). The Scree plot elbow-bends found in Figure 1 suggest testing for either a 3- or 6-factor level solution for extraction and analysis. Figure 1 Scree plot Rotating component matrices can be read to evaluate loading values in order to assess questionnaire items with regard to their contribution(s) to the factor constructs. Rotated component matrices were run to assess VLS item Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 325 loading scores. At the six-factor level, common features between items with ad- equately high loading scores were observed more clearly than at the three-fac- tor level, leading to acceptance of the six-factor extraction. Factor loadings were cut off at the 0.3 level in order to control for initial fit. The six-factor rotated factor pattern can be viewed in Appendix C. All items were grouped together according to the factor they scored high- est on, though it should be noted that cross-loading was observed for most items. Some items did not necessarily fit the conceptual “theme” of the factor group that they scored highest on. This result was somewhat expected due to the still-nascent organization of the SVLSS item pool, and exploratory nature of EFA. Those items that did not obviously fit are discussed in connection with their respective factor groups below. It bears repeating that the purpose of the EFA here is not to confirm previously established questionnaire constructs, but to explore possible interpretations of the constructs emerging from analysis. Table 4 Six-factor extraction interpretation Factors and titles Rotated Eigenvalue % of variance # of items 1. Memorization strategies 6.143 8.301% 17 2. Depth enhancing strategies (Use) 4.929 6.661% 12 3. Association-based strategies 4.926 6.657% 14 4. Depth enhancing strategies (Sources) 4.297 5.806% 12 5. Self-regulation and reflection strategies 4.257 5.752% 11 6. Lexical information strategies 3.923 5.301% 8 Initial evaluation of item groupings offered six thematic categories related to similarities between the types of VLS that clustered onto each factor. These item groupings were titled according to their core characterizations: memoriza- tion strategies, depth enhancing strategies through use, association-based strat- egies, depth enhancing strategies through sources, self-regulation and reflec- tion strategies, and strategies that use lexical information (see Table 4 above for categories). Item lists4 for each factor and interpretations for preliminary item grouping themes are reasoned below following corresponding tables that pre- sent the items included in each list (see Tables 5-10). Factor one accounted for 8.3% of the variance with 17 items loading onto it. Many of the higher scoring items grouped on this factor reflected strategies that demonstrate the use of repetition to better retain already acquired or en- countered word knowledge (ITEM 43, 44, 45, 46, 47, 50, 58). Less obvious repe- tition-based VLS items may have been interpreted by participants as repetition strategies, such as quizzing oneself [repeatedly] (ITEM 48), labeling items with 4 SVLSS item statements listed in Tables 5-10 have been paraphrased for display convenience. Richard LaBontee 326 L2 words [to view repeatedly] (ITEM 51), or organizing words into groups or lists [to review] (ITEM 61). Items also may have been interpreted through the state- ments’ use of the word review, which potentially evokes repetition through study (ITEM 39, 43, 49, 58), and curiously does not appear in item statements that loaded onto the other factors. Issues of misrepresentation of item state- ments will be addressed in a following section regarding revision and adjust- ment of item statements. Table 5 Factor one item list “Memorization strategies” Item VLS Loading score 45 Write words down over and over. 0.678 43 Review words repeatedly over time. 0.613 46 Writing lists of words over and over. 0.597 48 Quiz self with word lists 0.579 58 Plan to review words in spaced intervals 0.574 61 Organizing words into groups/lists 0.538 63 Reflect on vocabulary learning in Swedish 0.53 44 Read words out loud over and over. 0.514 49 Review words or with others. 0.508 50 Use flashcards 0.502 57 Plan individual learning before/during/after study. 0.486 39 Mix up the order of words during review 0.463 47 Listen to recordings of words over and over 0.456 51 Labeling items (in home, workplace) 0.456 59 Plan to use free time to casually practice vocabulary 0.433 62 Paying attention to difficult Swedish vocabulary 0.429 Strategic word learning, in perhaps its most shallow but oft-practiced form, includes repetitive engagement with a word through viewing, hearing or verbaliza- tion in order to better remember it. This engagement is usually coupled with related information (e.g., a learner’s L1 translation of the L2 word). Repetition-based strat- egies have been referred to as “repetition,” “memory” or “memorization strategies” in LLS and VLS taxonomy (e.g., Fan, 2003; Gu & Johnson, 1996; Oxford, 1990; Schmitt, 1997). In Nation’s (2013) taxonomy they fit into his processing category of VLS that involve noticing, retrieving and generating vocabulary knowledge. Other items that loaded onto this factor that do not fit the theme in a more obvious way reference reflection or regulation of one’s vocabulary learn- ing (ITEM 57, 58, 59, 62, 63). One item did not score over 0.3 (ITEM 41), indicat- ing a very weak correlation, although the item cross-loaded similarly onto sev- eral other factors. These items (and other instances of item-construct fit issues found in this section) are addressed in the following section regarding cross- loading of items across factors and item pool adjustments. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 327 Table 6 Factor two item list “Depth enhancing strategies (use)” Item Loading score 13 Listen carefully for lexical information 0.643 1 Speak or write as much as possible 0.635 2 Using words in casual speech 0.622 15 Watch TV/film in Swedish/with subtitles 0.559 16 Watch TV/film without subtitles 0.533 14 Pay attention to useful/interesting words in everyday life 0.503 12 Read text to find new words 0.491 56 Use known words even though uncomfortable 0.49 6 Use words in familiar situations 0.479 3 Describe word meaning without saying the word itself 0.394 53 Motivation comes from natural interest in the language. 0.342 40 Remember words through common expressions/idioms 0.301 Factor two accounted for 6.66% of the variance and 12 items loaded onto it. The VLS that loaded onto factor two reflected strategies used for both finding new word information through attending to aural or visual target language input (ITEM 12, 13, 14, 15, 16), as well as the production of words to strengthen re- tention of vocabulary knowledge (ITEM 1, 2, 3, 6, 56). Strategies for finding and retaining new word knowledge are also seen in factor four, but lean more to- wards the use of context clues and previous knowledge for facilitation. “Word depth” as related to the factor two and four VLS lists refers to the extent of which vocabulary knowledge units (which may occur in a variety of forms – morphological, syntactic, collocational, semantic, etc.) are linked to each other in the lexicon, and within one’s cognitive network (i.e., Henriksen, 1999). VLS that reflect the expansion of word knowledge depth require complex, effortful networking of word knowledge features that can be facilitated through the production of vocabulary as a means of learning said vocabulary. Strategies of this kind have been classified as “authentic language use” and “creative ac- tivities” in the VOLSI (Stoffer, 1995), as “activation strategies” in the VLQ (Gu & Johnson, 1996), or as “social strategies” in Schmitt’s (1997) taxonomy. The factor two VLS list also contains strategies that are used to engage one’s linguistic environment in order to find and learn new vocabulary knowledge, which also contributes to extending one’s depth of vocabulary knowledge. These kinds of strategies are seen as “determination strategies” in Schmitt (1997), and “source strate- gies” in the VLSQ (Fan, 2003). The VLQ (Gu & Johnson, 1996) divides them into “diction- ary strategies” and “note-taking strategies,” as well as “using background knowledge or the wider context” and “using linguistic cues or the immediate context.” The two lowest scoring items in this list did not seem to fit the theme in an obvious way. ITEM 53 deals with an intrinsic motivation for learning words, while ITEM 40 considers the use of colloquial language for word retention. Richard LaBontee 328 Table 7 Factor three item list “Association-based strategies” Item Loading score 38 Associate word with time-related information 0.673 25 Associate word with key-words 0.589 37 Associate word with location word found 0.544 36 Associate tone or music to word 0.526 29 Connect other related TL/NL words to new/learned word 0.514 35 Connect images or pictures to word 0.506 34 Associate salient contexts (story/action/emotion) to word 0.493 31 Remember context word found in 0.45 17 Use pictures/gestures for comprehension of unknown words 0.445 32 Connect sentence context to words found in them 0.42 4 Create mnemonic phrases using word 0.406 5 Make up new sentences with word 0.397 30 Connect word to translation/opposite meaning 0.384 18 Connect word with its phonetics 0.336 Fourteen items loaded onto factor three, which accounted for 6.66% of the variance. The characterizing feature for most strategies in this group was that they operated through associating word knowledge with some kind of other infor- mation. This information involved time (ITEM 38), other words (ITEM 25, 29, 30), sound (ITEM 18, 36), imagery (ITEM 17, 25, 35), text (ITEM 4, 5, 32), location (ITEM 37), or a general sense of context (ITEM 31, 34). Building associations between words and other kinds of information is an- other form of encoding word information for retention and future recall. These as- sociations are intended to connect word knowledge to varied memory nodes, hopefully resulting in faster, more regular and more primed activation. Associations might involve linking new words to related words, concepts, grammaticality, vis- ual/audio information, notes, kinesthetic actions, or tactile sense impressions such as musicality or tone (Gu, 2003; Oxford, 1990). Association strategies have been represented by other VLS taxonomy as “creating mental linkages through visual- auditory means, physical action and organizing words” by Stoffer (1995), “memory strategies for encoding” by Gu and Johnson (1996), and “association strategies,” “grouping strategies” and “analysis strategies” by Fan (2003). These kinds of VLS, like memorization strategies in factor one, fall into Nation’s (2013) category of VLS for processing information through noticing, retrieving and generating. Although all items in this group could be fitted into the interpreted theme, some statements seemed to be pointing towards other strategy types, though may have been misinterpreted by participants. For example, ITEM 17 represents a strategy for finding new word knowledge, but may have been construed by participants as a strategy for associating picture/gesture information to known information for better retention. This class of ambiguity is discussed in the later section on item pool adjustments. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 329 Table 8 Factor four item list “Depth enhancing strategies (sources)” Item Loading score 7 Note common/important words for later look-up 0.566 8 Use dictionary to look-up meaning 0.563 65 Reflect on error/accuracy of word use 0.495 11 Guess word meaning/pronunciation/spelling based on previous knowledge 0.485 9 Look-up words found in everyday life. 0.481 54 Motivation from other interests than the language. 0.478 42 Write down vocabulary notes 0.47 24 Use cognates 0.425 10 Ask others about word knowledge 0.389 33 Use previous exposure/knowledge of words 0.384 28 Connect Swedish words to NL words 0.359 60 Use technology to assist learning 0.32 Twelve items loaded onto factor four, which accounted for 5.80% of the vari- ance. This item list seems to represent a kind of extension of the depth enhancing strategy types found in factor two, but with a focus on such strategies as those rep- resented as “linguistic cues,” “the immediate context,” “dictionary strategies,” and “note-taking strategies” as seen in Gu and Johnson’s (1996) VLQ. The lion’s share of VLS that loaded onto factor four represented strategies performed to establish new vocabulary knowledge through the use of note-taking (ITEM 7, 42), look-up strate- gies (ITEM 8, 9, 10), using previous knowledge (ITEM 11, 24, 28, 33). Table 9 Factor five item list “Self-regulation and reflection strategies” Item Loading score 66 Reflect on stress/anxiety from vocabulary use 0.62 71 Reflect on socio-cultural impact of word knowledge 0.57 72 Reflect on feeling from knowing words 0.518 69 Reflect on comparing TL to other known languages 0.509 67 Reflect on using VLS and their impact on learning 0.502 52 Stare at words to remember visually 0.49 74 Reflect on the sound of TL when learning words 0.463 64 Reflect on learning style/strengths/weaknesses of learning 0.433 68 Reflect on importance of word knowledge on language learning 0.423 70 Think about direction of translation for word knowledge 0.421 55 Skip difficult/unknown words 0.37 Items that were not obviously connected to the central factor theme in- cluded reflecting on errors (ITEM 65), using extrinsic motivation (ITEM 54), and the use of technology to assist in learning (ITEM 60). It is possible that partici- pants may have perceived error correction and use of technology as continued means of finding new word knowledge, which would place those strategies more meaningfully within the central theme of the item group. However, this Richard LaBontee 330 remains conjecture as the wording of the VLS item statements may have been misleading or overly vague for the participant audience, resulting in misrepre- sentation. Misleading and vague item statements in the item pool will be ad- dressed in the following section. Eleven items loaded onto factor five, which explained 5.75% of the variance. The majority feature that VLS in this item group possessed was the use of reflection on and regulation of their vocabulary learning experience. These reflections were characterized by affective considerations and self-regulation (ITEM 55, 66, 72), con- sidering the impacts of word learning (ITEM 67, 68, 71), considering personal learning style (ITEM 64), and considering the nature of the target language (ITEM 69, 70, 74). Meta-strategies are strategies used to regulate one’s learning with regards to motivation, emotions, planning, strategy choice, and how to use those strategies. These kinds of strategies have been represented in other VLS taxonomy as “self-mo- tivation” and “overcoming anxiety” strategies by Stoffer (1995), “metacognitive strat- egies” by Schmitt (1997), “management strategies” in Fan (2003), and “selective at- tention” and “self-initiation” strategies in Gu and Johnson (1996). Gu and Johnson also included a section in their VLQ on “beliefs about vocabulary learning” that per- haps most closely resembles the reflective statements that loaded onto factor five. Nation’s (2013) taxonomy includes an entire classification of strategies used for plan- ning what vocabulary knowledge to learn, how to learn it, and when to learn it. An item that did not obviously fit into the central item group theme was “staring at words to help remember them visually” (ITEM 52), a statement that seems to reflect a more rehearsal-based approach to word learning. However, the extent to which this statement reflects an actual strategy is suspect due to its perceived lack of a clear goal besides retention of a word’s orthographic form, which has been covered elsewhere more explicitly (ITEM 19, factor six). Issues related to the core strategic concepts and redundancy of item statements will be covered in the following revisions section. Table 10 Factor six item list “Lexical information strategies” Item Loading score 26 Organize words by word-type 0.727 20 Pay attention to how words change with grammar 0.699 21 Pay attention to en/ett word distinction 0.573 22 Pay attention to words inside compound words 0.562 27 Organize words into semantically similar groups 0.538 23 Organize words using language rules 0.537 73 Pay attention to morphology of words 0.531 19 Pay attention to orthography 0.451 Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 331 The final item grouping, factor six, accounted for 5.30% of the variance with eight items loading onto it. The items in this group, while including VLS that rep- resented association-making techniques (ITEM 23, 26, 27) as well as analytical techniques (ITEM 20, 21, 22, 73, 19), all shared the core characteristic of being concerned with the strategic use of lexical information to better understand and retain related vocabulary knowledge. The lexical information occurring in VLS items here include representations of the form (ITEM 22, 73, 19), meaning (ITEM 27) and use (ITEM 26, 20, 21, 23) dimensions of word knowledge. The wording of the item statements clustered onto this factor may have had an influence on loading scores due to their nearly uniform use of “I pay attention to” or, “I organize” when describing each represented VLS. This group- ing could result in misinterpretation of items as being related to learners who are simply “attentive to lexical information” when word learning, rather than actually representing methods of strategic learning. Steps to combat such issues are discussed in the revisions section. Table 11 Cronbach alpha per factor group Construct Cronbach alpha Total items 1. Memorization strategies 0.856 17 2. Depth enhancing strategies (Use) 0.795 12 3. Context- and association-based strategies 0.832 14 4. Depth enhancing strategies (Sources) 0.743 12 5. Self-regulation and reflection strategies 0.753 11 6. Lexical information strategies 0.812 8 A Cronbach’s alpha measurement was obtained for each factor group in order to evaluate the internal consistency of item lists (see Table 11). Alpha co- efficients between 0.743 and 0.856 for all factor groupings suggest an adequate reliability of item stability per construct (Nunnally & Bernstein, 1994), though may also have been influenced by the relatively high number of items per group. 4.2. Item deletion, retention and revision The EFA performed using data collected by the SVLSS 1.2 was intended to guide interpretation of the underlying theoretical VLS categories represented by the items included on the instrument. The six-factor solution rotated factor pattern values, however, highlight issues with the item pool for the SVLSS in terms of possible re- dundancy of items and conceptual accuracy related to strategic learning. Further- more, a careful review of the six-factor VLS model in comparison to other VLS tax- onomy (LaBontee, in press) indicates a need for careful consideration of what kinds of VLS information the SVLSS stands to collect, and what kinds of VLS information it Richard LaBontee 332 does not. Concerns regarding readability and accessibility are also raised and ad- dressed in order to facilitate further revisions aimed at improving the instrument. The remainder of this paper will present adjustments made to the item pool of the SVLSS 1.2 in relation to item-construct fit, accuracy of strategic concept intended for representation, taxonomical revisions, and overall readability and accessibility. 4.2.1. EFA results as a means of examining the item pool It should be noted at the outset that although EFA has commonly been used to help facilitate questionnaire validation and design, it remains only an interpretive tool that does not offer a concrete formula for instrument refinement (Osborne, 2014). That said, although the exploratory nature of EFA may result in misleading and non-generalizable conclusions (i.e., Petrić & Czárl, 2003), it can also be used as a tool to help to point towards potential hazards that may exist for nascent instrument designs. As such, EFA results were revisited in order to help guide at- tention towards potentially problematic items in the SVLSS. The purpose of this examination was to operationalize the deletion or revision of inappropriate items in the item pool by seeking out problematic patterns suggested by EFA results. Low and diffuse scores for individual items were examined with regard to their fit within the chosen VLS construct mapped onto factor groupings. Score ratings (high, moderate, low) were applied to the rotated 6-factor pattern matrix values in order to establish thresholds used to examine issues with cross-loaded items. A survey of all scores resulted in labels of low if they were between 0.2- 0.299, medium between 0.3-0.399, and high over 0.4, and shown in Table 12 below. Scores below 0.2 were removed for visual clarity. Three types of cross-loading arose from an examination of scores: items that scored high on one factor but also cross-loaded onto other factors with moderate or low scores, items that cross-loaded with similarly moderate scores across several factors, and items that cross-loaded with similarly low scores across factors. These cross-loaded items were then subject to a review that con- sidered where the diffuse loading may have originated from. Looking primarily at the VLS text statements representing each item, three patterns of issues emerged related to item-construct fit, over-specificity of VLS representation, or conversely, overly vague representation of VLS. First, several items that returned low or cross-loaded scores seemed to all share a feature that their representative VLS did not “fit” into a single one of the six factor groups but could ostensibly represent aspects of several groups. For ex- ample, ITEM41, “I try to memorize words however I can” reflects use of memoriza- tion strategy for word learning, but the vague wording of “however I can” is worded in a way that is highly susceptible to a variety of interpretations by participants. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 333 Similarly, ITEM59, “I plan to use my free time to casually practice Swedish vocabu- lary” and ITEM42, “I write down vocabulary notes” are perhaps worded in ways that are far too open to interpretation to be reliable representations of VLS groups. Table 12 Rotated factor pattern for 6-factor solution (cross-loading scores only) Si ng le -f ac to rh ig h- sc or e (w ith m ed an d lo w ) 1 2 3 4 5 6 ITEM04 0.318*** * 0.406*** ITEM14 -0.078*** 0.503*** 0.065*** 0.382*** 0.266*** ITEM15 -0.1*** 0.559*** 0.334*** ITEM21 0.227*** 0.323*** -0.085*** 0.077*** 0.018*** 0.573*** ITEM27 0.139*** -0.073*** 0.348*** 0.026*** 0.094*** 0.538*** ITEM28 -0.152*** -0.019*** 0.255*** 0.359*** -0.024*** 0.085*** ITEM32 0.017*** 0.317*** 0.42*** -0.02*** 0.339*** 0.289*** ITEM43 0.613*** 0.061*** -0.062*** 0.316*** -0.047*** 0.108*** ITEM49 0.508*** -0.162*** 0.343*** -0.104*** 0.043*** 0.06*** ITEM59 0.433*** 0.322*** -0.095*** 0.251*** 0.207*** 0.033*** ITEM60 0.15*** 0.077*** 0.168*** 0.32*** 0.092*** 0.099*** ITEM61 0.538*** -0.07*** 0.3*** 0.068*** 0.011*** 0.112*** ITEM64 0.273*** 0.358*** 0.01*** 0.235*** 0.433*** 0.005*** ITEM65 0.065*** 0.223*** 0.019*** 0.495*** 0.381*** 0.056*** ITEM67 0.302*** 0.159*** 0.12*** 0.132*** 0.502*** 0.084*** ITEM68 0.245*** 0.195*** -0.09*** 0.331*** 0.423*** -0.071*** ITEM69 -0.01*** 0.041*** 0.089*** 0.425*** 0.509*** -0.039*** ITEM73 0.147*** 0.192*** -0.14*** 0.143*** 0.362*** 0.531*** M ul tip le fa ct or hi gh /m ed (w ith lo w ) sc or e ITEM03 0.146*** 0.394*** 0.354*** -0.162*** -0.038*** -0.134*** ITEM05 0.165*** 0.374*** 0.397*** 0.071*** -0.073*** 0.096*** ITEM10 -0.193*** 0.218*** 0.243*** 0.389*** 0.09*** 0.084*** ITEM18 0.201*** 0.032*** 0.336*** 0.296*** 0.01*** 0.157*** ITEM30 0.108*** 0.042*** 0.384*** 0.006*** 0.372*** 0.232*** ITEM33 -0.082*** 0.1*** 0.32*** 0.384*** 0.135*** 0.29*** ITEM40 -0.017*** 0.301*** 0.242*** 0.238*** 0.065*** 0.139*** ITEM42 0.451*** -0.058*** -0.009*** 0.47*** 0.024*** 0.045*** Multiple factor (low only) score ITEM41 0.269*** 0.186 0.155 0.248*** 0.134*** 0.227*** *** High ** Medium * Low Note. Scores > 0.2 filtered from table. A second problem was observed in overly contextually-niche VLS items that may have been too specific to load onto a specific construct, instead cross-loading onto several factors, or not finding adequate representation within the factor solu- tion extracted. For example, ITEM03, “I practice vocabulary by describing the mean- ing of words in Swedish without saying them,” ITEM04, “I create new mnemonic phrases to help me remember Swedish words,” and ITEM15, “I watch TV or film [in Swedish or with Swedish subtitles] to try and find or practice Swedish words” may Richard LaBontee 334 not have been strategies that are readily comprehensible or may have been misin- terpreted by participants due to non-exposure to such specific practices. The third problem was characterized by high-scoring items that seemed to fit the factor group they loaded onto (ITEM05, ITEM10, ITEM14, ITEM32, ITEM40, ITEM61), but suffered from vague or imprecise wording (e.g., ITEM05, “I make up my own sentences to help me remember specific words”), or repre- sented multiple or clustered iterations of VLS in a single statement (e.g., ITEM10, “I ask others (teachers, friends, natives) about Swedish word knowledge”). These and the above issues with the ambiguity, overly-contextual nature, or in- terpretability of item-statement wording have potentially contributed to the col- lection of “unclean data,” or responses that do not necessarily link up with the constructs thought to be represented by the instrument (Osbourne, 2014). This potential for misrepresentation likely contributed to low scoring and cross-load- ing issues found in the rotated component matrix. The above issues with the items listed were detected through the guidance of EFA results, but will be con- sulted during revisions made to the entirety of the SVLSS 1.2 item pool. Table 13 Examples of unclear VLS representation in item statements Item Original text Complication Revised text 51 I label items (i.e., in my home, workplace) in order to review them often over time. Multiple VLS in single item state- ment. I label items (furniture, utensils, etc.) to help me remember the Swedish words for them. 19 [I] try to remember how to spell words. Does not reflect strategic behav- ior. I try to learn spelling of words let- ter by letter. Adjustments to statements revised to clarify the VLS concept represented by the item (ITEM 43, 46, 48, 39, 47, 51, 1, 2, 15, 16, 6, 17, 32, 11, 9, 55, 26, 22) reconsidered the core strategic element of the intended VLS, then revised the text according to that element. For example, ITEM32, “I remember the sentence context for words I found in those sentences” became, “I connect words to the sentence, phrase or story I find them in” (see more examples in Table 13). As the strategy in question sought to determine if participants encode word infor- mation using associated textual context, the overly vague wording was replaced with a concrete statement exemplifying the core VLS. Additionally, it was observed that some items loaded onto factors poten- tially due to the influence of similarly worded item statements. For example, items in factor six (lexical information strategies) nearly all included wording formula- tions using “I pay attention to” or “I organize.” The item statements included in this factor group were differentiated from their original uniformity in wording choices as a means of avoiding item-factor clustering due to statement wording rather than the VLS intended to be represented by the item. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 335 There were a significant number of statements that were problematic with regard to their underlying conceptual representativeness of strategic be- havior. These complications were identified as: items that did not actually rep- resent strategic behavior, but rather, seemed to represent reflections on individ- ual learning beliefs or individual learning style(s); items that were overly vague and non-representative of any clear VLS group(s); and items that were redun- dantly representative of other VLS items already included in the item pool. Ex- amples of complicating items are shown in Table 14 below. Table 14 Examples of other issues with item statements Item Text Complication Construct NOT STRATEGIC - BELIEF 53 My motivation for Swedish word learning comes from my natural interest in the language. Motivational belief. Non-strategic. 2 68 I reflect on the importance of learning Swedish vo- cabulary in terms of my overall language learning. Reflective practice. Non-strategic. 5 NOT STRATEGIC – L. STYLE 60 I use technology as a means of learning. Modality of study. Non-strategic in itself. 4 VAGUE 41 I memorize words however I can. Does not represent any specific VLS. 1 REDUNDANT 58 I plan to review [specific amounts of] vocabulary over specific time periods. Reflects same core VLS as #43. 1 43 I will review words or my own notes repeatedly over time. 1 The above complications led to the deletion of 28 items from the SVLSS 1.2 instrument. Nearly the entire item group of “self-regulation and reflection strategies” was deleted as, after consideration of what strategic behavior is, these items did not, in fact, reflect planned, conscious, goal-oriented strategies. Rather, they repre- sented reflective practices related to vocabulary learning, which are important for self-regulation of learning, but non-strategic as stated. Deleted items were classified as reflecting preferred learning styles and self-regulative, but not strategic, behavior (ITEM 60, 63, 64, 65, 66, 67, 68, 69, 71, 72), items reflecting learning beliefs and motivations (ITEM 53, 54, 56, 74), and items that were too vague to reflect any con- crete VLS concept (ITEM 3, 13, 14, 18, 41, 52, 57, 62, 70, 73). Items reflecting already included VLS concepts (ITEM 40, 42, 58, 61) were also removed. Two of these problematic items were retained and revised for use (ITEM 19, 20). For example, ITEM19, “I try to remember how to spell words” is not inherently strategic – the statement does not indicate purpose or goal-orienta- tion, it simply asks participants to rate how salient a word’s orthographic form is to them. This item was revised as, “I try to learn words spelling letter by letter” in order to reflect a strategic plan-of-action for retaining vocabulary knowledge. Richard LaBontee 336 4.2.2. Readability and accessibility The original text5 used in the SVLSS 1.2 was measured for readability using the Flesch Reading Ease (FRE) test. The FRE test uses a formula for determining read- ability of a text with mathematical comparisons between sentence length and total amounts of words, and between total syllable count and the total amount of words (Flesch, 1948). The FRE measurement for the SVLSS 1.2 was 65, con- sidered to reflect a reading difficulty of “plain English” easily understood by age 13-to-15-year-old native English speakers. Following EFA and resulting item list adjustments, item statements were scrutinized with regard to their readability for a wide audience of English-user- proficiencies. Accessibility was revised through adjusting statements to reflect simple, concise language that avoids the use of jargon wherever possible (see Table 15 for examples). Table 15 Examples of item statement adjustments for readability Item Original text Complication Revised text 25 I make use of pictures (in text) or gestures (in speech) to help me un- derstand Swedish vocabulary that I find in that context. Overly complicated, unclear use of understand, use of difficult words. When I see pictures nearby or re- lated to Swedish text, I use them to help me guess the meaning of Swedish words there. 50 ... use flashcards/index cards (or similar study tools) to help me re- view words. Jargon Use. I write words on one side of a card, and the meaning on the other side to help me review words. Forty-three item statements were adjusted in order to address concerns with the wording of text. Most changes to item statements that were revised for language simplicity (ITEM 45, 44, 49, 12, 38, 37, 36, 29, 35, 34, 31, 4, 5, 24, 10, 33, 21) were minor adjustments intended to direct participants more towards the particular VLS in question. For example, ITEM29, “I try to connect other re- lated Swedish (or native language) words to the word being learned” became “I connect other related Swedish words to the one I am learning.” “Trying” is inher- ent in all VLS use due to their being effortful by definition, but using the word acts to weakens the statement (i.e., trying does not necessarily reflect doing), and connecting TL words to L1/additional language translations is a somewhat natural function of language learning that does not need to be included in the item statement, risking further confusion of the VLS being represented. Item statements revised for use of jargon (ITEM 50, 25, 7, 8, 28, 27) generally were overly verbose, or used inappropriately academic terminology for participants. 5 All item statements, demographic survey questions, introductory instructions, and open- ended questions. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 337 For example, ITEM27, “I try to put Swedish words into groups that have similar meaning or themes to help me remember them” became “I group words into cat- egories (e.g., animals, utensils).” Providing a simple example better explains the core VLS concept of the item than using a moderately complex text explanation. These adjustments to item statements, in combination with deletion of items according to EFA findings and VLS taxonomy concerns, as well as addition of items from other VLS taxonomy (LaBontee, in press) resulted in an FRE score of 72.5 for the SVLSS 2.0, a marked improvement from the 65 score received the SVLSS 1.2. This score indicates a level of readability that should be easily under- stood by 13- to 14-year-old native speakers of English. Although the question- naire is used by English L2 users, they are adults who use English at the univer- sity level in Sweden, suggesting that the evaluated level of readability for the SVLSS should be accessible to its target audience. 5. Discussion The main thrusts of this study have centered on the importance of a transparent instrumentation process, as well as the development of a context-specific VLS taxonomy from the ground-up. Transparency, here, is conceived of as the act of reporting on the steps taken to create, evaluate, revise and utilize a research instrument, all while keeping in mind the intended use of said instrument. As past survey-driven research into learners’ VLS use has suffered from a lack of clear reporting on instrumentation practices, the findings from that research can be somewhat difficult to interpret in connection with how they were ob- tained. The findings emergent from the analyses performed on the SVLSS 1.2 indicate that even with carefully planned steps regarding design, piloting and revision, discrepancies and issues can still arise during instrument development. As data sets are restricted by the scope of the instruments and practices used to collect them, first obtaining a transparent illustration of what conceptual structures underlie those instruments and practices is paramount to be able to interpret principal data sets in valid and reliable ways. Although other surveys used in research may have undergone similar development regimens, the lack of clear reporting on these processes obscures the reasons for why certain items or concepts are included (or excluded), and how certain findings are arrived at through the type of data that is actually collected. However, SVLSS development has not gone without its own limitations. The SVLSS 1.2 item pool was generated through a data collection and analysis process that used interviews and learning tasks to elicit and observe strategy use for vocabulary learning in Swedish as a L2. Representativeness of the SVLSS 1.2 item pool should thus be considered restrained to data gathered during that Richard LaBontee 338 initial investigation, and to feedback collected from pilot participants who con- tributed to the revision of the SVLSS version 1.0 and 1.1. A benefit of this pro- cess, however, lies in the instrument being contextually situated in and by its target audience. This supports one of the main purposes of the instrument, as it is intended to represent strategic learning specific to the adult, Swedish L2 learning context (e.g., ITEM 21, “Pay attention to en/ett word distinction”)6. The instrumentation process reported on here allows for a transparent understand- ing of how this instrument has been contextually situated, and how it has ar- rived at the underlying taxonomies that govern further development. The next step in the revision and adjustment process for the SVLSS will focus on a comparison of other VLS taxonomy and how the SVLSS item list compares to them in terms of VLS representativeness. This will allow for the illustration of pos- sible gaps (or inappropriateness of items) in the SVLSS instrument, and help situ- ate what kind of representativeness it holds for VLS constructs as compared to other instruments. Such a comparison can better demonstrate what kinds of VLS information the SVLSS aims to (and does not aim to) collect from participants. In LaBontee (in press), the six-category VLS model that emerged from in- terpreted EFA results here is compared to VLS taxonomy appearing in Stoffer’s (1995) VOLSI, Gu and Johnson’s (1996) VLQ, Schmitt’s (1997) VLS questionnaire, and Nation’s (2013) VLS taxonomy. This comparison results in the six-category VLS model being altered to better integrate elements from the other VLS taxon- omy reviewed, centered on theoretical scaffolding from Nation’s VLS taxonomy. The resulting model encompasses a four-category model that classifies VLS as: strategies for improving word knowledge (using rehearsal and encoding), pro- ductive activation strategies, strategies for establishing new word knowledge (using sources and contexts), and self-regulative strategies. As a means of supporting this updated VLS taxonomy with regards to con- struct representation, the revised item pool from the SVLSS version 1.2 is ad- justed further through the addition of items found in other VLS taxonomy. A visual representation of the changes made from the six-category model (SVLSS 1.2) to the four-category model (SVLSS 2.0) is provided in Table 16. For an in- depth discussion concerning the adoption of the four-category VLS taxonomy and the changes made to the SVLSS item pool related to this shift, see La- Bontee’s (in press) VLS list meta-analysis. 6 Swedish nouns are gendered into common (en) words and neuter (ett) words, commonly learned with their associated article: e.g., en katt (a cat), ett äpple (an apple). Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 339 Figure 2 Side-by-side comparison of VLS taxonomy for SVLSS 1.0 and 2.0 The updated SVLSS 2.0 is being used and evaluated in a study investigating patterns and possible profiles of adult, Swedish L2 learners’ VLS use related to individual differences between learner groups in LaBontee (2018). 6. Summary This study has focused on providing a transparent report on the methods used and decision-making that has facilitated initial interpretations of a VLS questionnaire that was developed through data-driven item pool generation, and thus without a priori assumption of its underlying construct make-up. This method of instrumen- tation was chosen explicitly, as to avoid simply adapting an instrument that was cre- ated for use with another target audience than the one intended. Rather, this method was employed with the intention of creating a questionnaire that is driven by data collected from the target audience (adult, beginner Swedish L2 learners), built in a ground-up fashion, and then ready for distribution to the same population. However, as mentioned earlier, this resulted in the initial item pool being repre- sentative of only the data collected to create it, not necessarily representative of a wider range of VLS available to those learners. As a response to this, this study sought to first better understand what the item pool did represent, and then once thus informed, explore what else the item pool could represent. Understanding the SVLSS 1.0 # Items SVLSS 2.0 # Items Influences TOTAL 72 TOTAL 69 Nation's VLS Taxonomy Pr oc es se s: N ot ic in g, R et ri ev in g, G en er at in g So ur ce s: Fi nd in g in fo rm at io n ab ou tw or ds Pl an ni ng : W ha tt o fo cu s on an d w he n Sk ill in us e: E nr ic hi ng K no w le dg e VLQ, Nation: Rehearsal & Encoding Strategy Distinction VLQ: 4 items Schmitt: Establishing new word knowledge distinction Schmitt: 1 item VLQ: 3 items VOLSI: 2 items Nation: Generating/Skill in use VOLSI: 2 items VOLSI: 4 items VLQ: 2 items VLQ, Nation: Rehearsal & Encoding distinction VLQ: 3 items Schmitt: 3 items VOLSI: 1 item 8 15 8 14 12 9 24 8 13 7 Depth Increasing Strategies (via Sources) STRATEGIES FOR ESTABLISHING NEW WORD KNOWLEDGE (Sources) STRATEGIES FOR ESTABLISHING NEW WORD KNOWLEDGE (Contexts) Self-regulation & reflection STRATEGIC SELF- REGULATION 12 11 Depth Increasing Strategies (via Use) PRODUCTIVE ACTIVATION Memorization Strategies STRATEGIES FOR IMRPOVING WORD KNOWLEDGE (Rehearsal) (Lexical information strategies) STRATEGIES FOR IMPROVING WORD KNOWLEDGE (Encoding)(Context & Association Based strategies) Richard LaBontee 340 item pool representation was operationalized through EFA and scaffolded through comparisons to other VLS taxonomy. The EFA performed on collected data from the SVLSS resulted in the pro- posal of a six-category VLS taxonomy used to explain the constructs represented by the SVLSS item pool. This taxonomy was interpreted and described through the use of comparisons to VLS classifications found in other VLS lists and vocab- ulary knowledge taxonomy. These comparisons also acted as a means of exam- ining representativeness of items included (and not included) in the SVLSS. Alongside these comparisons, the SVLSS item pool was reviewed using EFA find- ings as a guide that suggested possible issues with item statement wording as related to redundancy and representation of VLS concepts. Revisions to item statements as well as inclusion (or exclusion) of items in the item pool were made on the basis of the core strategic behavior represented (or not repre- sented) by statements in relation to the six-category VLS model interpreted from EFA results. Other revisions to the item pool were made in response to issues regarding statement clarity and readability of item statements included in the SVLSS, resulting in an overall improvement to readability measures for the next iteration of the instrument, the SVLSS 2.0. Of course, continued development of the SVLSS will never produce a “perfect instrument.” Rather, revision is performed with the intention of new iterations of the instrument that reflect an increasingly clearer construct structure represented by the items that are included on it. Also, the instrument will hopefully become more accessible to a wider range of English users that may engage with it. Further revali- dation procedures for the SVLSS must be performed reiteratively with a variety of samples in order to develop the instrument further. Finally, establishing suggested practices for the SVLSS’s use by researchers and instructors should also be ad- dressed in order to extend accessibility and application of the instrument. 7. Conclusion The findings observed in this study suggest that the item list for the SVLSS 1.2 has shown some level of representation regarding several key conceptual areas connected to different types of VLS that appear elsewhere in L2 vocabulary ac- quisition research. Comparing instruments created to operate in similar fields provides a measure of reliability in establishing which strategic learning behav- iors exist and can be catalogued, but also provides insight into which strategies are more (or less) relevant to a specific context. This comparison occurs using other taxonomy as a lens for iterative evaluation, leading to a synthesis of con- cepts that establish a new taxonomy. Reporting on this connection-making is an important step in establishing a transparent instrumentation process, as it allows Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 341 the reader to see why an instrument was developed, for what purpose, and to more clearly understand what kinds of data it might actually collect. The organization and arrangement of VLS concepts into a taxonomy is done so for the benefit of future research as well as learners who engage in the classification of learning strategies. Future VLS research in the Swedish L2 learn- ing context will be able to draw from this and future versions of the taxonomy proffered here, hopefully expediting expansion into the area. The clarity of VLS categorizations can lead to smoother strategy instruction, improving learners’ vocabulary acquisition and reflective learning processes. Richard LaBontee 342 References Allestam, C. (2007). Ordinlärningsstrategier i svenska som andraspråk, engelska och moderna språk (Unpublished doctoral dissertation). Gothenburg: Gö- teborgs Universitet. Cerny, C. A., & Kaiser, H.F. (1977). A study of a measure of sampling adequacy for factor-analytic correlation matrices. Multivariate Behavioral Research, 12(1), 43-47. Dörnyei, Z., & Csizér, K. (2012). How to design and analyze surveys in second lan- guage. In A. Mackey & S. Gass (Eds.), Research methods in second language acquisition: A practical guide (pp. 74-94). Oxford, UK: Wiley-Blackwell. Fan, M. (2003). Frequency of use, perceived usefulness, and actual usefulness of second language vocabulary strategies: A study of Hong Kong learners. Modern Language Journal, 87(2), 222-241. Flesch, R. (1948). A new readability yardstick. Journal of Applied Psychology, 32, 221-233. Granberg, N. (2001). The dynamics of second language learning: A longitudinal and qualitative study of an adult’s learning of Swedish. Department of Con- temporary Literature and Scandinavian Studies. Umeå: Umeå University. Gu, Y. (2003). Vocabulary learning in a second language: Person, task, context and strategies. TESL-EJ, 7(2). Retrieved from http://www.tesl-ej.org/word press/issues/volume7/ej26/ej26a4/ Gu, Y. (2013). Vocabulary learning strategies. In C. A. Chapelle (Ed.), The encyclopedia of applied linguistics. Blackwell Publishing. Retrieved from http://doi.org/ 10.1002/9781405198431.wbeal1329 Gu, Y. (2018). Validation of an online questionnaire of vocabulary learning strat- egies for ESL learners. Studies in Second Language Learning and Teaching, 8(2), 325-350. Gu, Y., & Johnson, R. (1996). Vocabulary learning strategies and language learn- ing outcomes. Language Learning, 46(4), 643-679. Henriksen, B. (1999). Three dimensions of vocabulary development. Studies in Second Language Acquisition, 21, 303-317. LaBontee, R. (2016). Investigating reported vocabulary learning strategy use in Swedish second language learning: From interviews to questionnaires. Jour- nal of Foreign Language Teaching and Applied Linguistics, 3(3), 131-140. LaBontee, R. (in press). Comparing vocabulary learning strategy lists: Design, context and content. Situating Strategy Use 2 Conference Proceedings. Ko- motini, Greece: Democritus University of Thrace. LaBontee, R. (2018). Likert-scale questionnaire analysis of reported vocabulary learning strategy use in the Swedish second language learning context. Manuscript submitted for publication. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 343 Magnusson, C., & Öggesjö, M. (2013). Ordinlärningsstrategier En kvalitativ studie i engelska och svenska som andraspråk (Unpublished doctoral dis- sertation). Gothenburg: Göteborgs Universitet. Malmberg, P. (Ed.) (2000). I huvudet på en elev: Projektet STRIMS: Strategier vid inlärning av moderna språk. Stockholm: Bonnier. Nation, I. S. P. (2013). Learning vocabulary in another Language. Cambridge: Cambridge University Press. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York, NY: McGraw-Hill, Inc. O’Malley, J. M., & Chamot, A. U. (1990). Learning strategies in second language acquisition. Cambridge: Cambridge University Press. Osborne, J. W. (2014). Best practices in exploratory factor analysis. Scotts Valley, CA: CreateSpace Independent Publishing. Oxford, R., & Scarcella, R. C. (1994). Second language vocabulary learning among adults: State of the art in vocabulary instruction. System, 22(2), 231-243. Oxford, R. L. (1990). Language learning strategies: What every teacher should know. Boston, MA: Heinle & Heinle. Oxford, R. L. (2011). Teaching and researching language learning strategies. Har- low: Pearson Education Limited. Oxford, R. L. (2017). Teaching and researching language learning strategies: Self-regulation in context (2nd ed.). Abingdon, England: Routledge. Park, G.-P. (2011). The validation process of the SILL: A confirmatory factor anal- ysis. English Language Teaching, 4(4),21-27. Petrić, B., & Czárl, B. (2003). Validating a writing strategy questionnaire. System, 31, 187-215. Rubin, J. (1975). What the “good language learner” can teach us. TESOL Quar- terly, 9(1), 41-51. Sandh, C. (2013). Strategier för ordinlärning bland vuxenstuderande i svenska som andraspråk (Unpublished doctoral dissertation). Gothenburg: Göte- borgs Universitet. Schmitt, N. (1997). Vocabulary learning strategies. In N. Schmitt & M. McCarthy (Eds.), Vocabulary: Description, acquisition and pedagogy (pp. 199-227). Cambridge: Cambridge University Press. Schmitt, N. (2010). Researching vocabulary: A vocabulary research manual. UK: Palgrave Press. Stoffer, I. (1995). University foreign language students’ choice of vocabulary learning strategies as related to individual difference variables (Un- published doctoral dissertation). Alabama: University of Alabama. Richard LaBontee 344 Tornberg, U., Öman, M., Bergström, I., Håkanson, U. (2000). I huvudet på en elev: Projektet STRIMS; strategier vid inlärning av moderna språk. Sanoma Ut- bildning: Stockholm. Tseng, W. -T., Dörnyei, Z., & Schmitt, N. (2006). A new approach to assessing strategic learning: The case of self-regulation in vocabulary acquisition. Applied Linguistics, 27(1), 78-102. Wareborn, M. (2004). Minst 100 nya ord i veckan ska läras in: Vem har motivation och uthållighet för det? Andraspråksinlärares strategianvändning i ordinlärn- ing (Unpublished doctoral dissertation). Gothenburg: Göteborgs Universitet. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 345 APPENDIX A Swedish Vocabulary Learning Strategy Survey (Version 1.2) Pr od uc ti on Ba se d (6 it em s) ITEM1 I speak or write in Swedish as much as I can to practice vocabulary. ITEM2 I use vocabulary words in casual speech to help me remember them better. ITEM3 I practice vocabulary by describing the meaning of words in Swedish without saying the word out loud. ITEM4 I create new mnemonic phrases [sentence used to assist memory] to help me remember vocabulary. ITEM5 I make up my own sentences to help me remember specific words. ITEM6 I will use words over and over in similar situations to help me remember them. In ve st ig at io n Ba se d (1 1 it em s) ITEM7 I take note of common/important words I don't know so I can review them later. ITEM8 I look up word meaning or details by using some kind of dictionary. ITEM9 I look up Swedish words that I find in everyday life. ITEM10 I ask others (teachers, friends, natives) about Swedish word knowledge. ITEM11 I try to guess word meaning, pronunciation, or spelling based on what I already know about the word or where I find it. ITEM12 I try to read Swedish texts to find new vocabulary. ITEM13 I try to listen carefully for Swedish vocabulary information. ITEM14 I try to pay attention to useful or interesting Swedish language that I find in everyday life. ITEM15 I watch TV or film [in Swedish, or with Swedish subtitles] to try and find or practice vocabu- lary. ITEM16 I watch Swedish TV or film specifically without subtitles on to try and learn vocabulary from context. ITEM17 I make use of pictures (in text) or gestures (in speech) to help me understand Swedish vo- cabulary that I find in that context. Le xi ca l-A ss oc ia ti ve (1 3 it em s) WHEN LEARNING SWEDISH VOCABULARY, I… ITEM18 … try to remember words by the way they sound. ITEM19 … try to remember how to spell words. ITEM20 … try to remember how words change with grammar in Swedish. ITEM21 … try to remember if a word is an ‘ett ord’ [ett word] or ‘en ord’ [en word]. ITEM22 … try to remember compound words (words made up of 2 or more words) by paying atten- tion to the different words inside them. ITEM23 … try to remember words by making language rules to connect them. ITEM24 … use cognates (similar words across languages) to help remember words. ITEM25 … try to use key-words (words used to help us recall other words) to help learn or remember words. ITEM26 … try to organize words [or groups of words] by word types to help me remember them. ITEM27 … try to put Swedish words into groups that have similar meanings or themes to help me remember them. ITEM28 … try to connect Swedish words to words I already know in my native language. ITEM29 … try to connect other related Swedish (or native language) words to the word being learned. ITEM30 … try to connect Swedish words to their opposite meaning or translation to help me remem- ber. Co nt ex t- A ss oc ia ti ve (1 0 it em s) WHEN LEARNING OR REMEMBERING SWEDISH VOCABULARY, I… ITEM31 ... remember the context I find words in to help me remember them. ITEM32 ... remember the sentence context for words I found in those sentences. ITEM33 ... remember previous times that I saw or read a word to help me remember it better. ITEM34 ... use interesting contexts (e.g., a story, action, emotion) to help me remember words. ITEM35 ... connect images or pictures in my mind with words to help me remember them. ITEM36 ... connect tone or music to words to help me remember them. ITEM37 ... remember the location I saw certain words to help me remember them. ITEM38 ... use time-related information to help me remember words. ITEM39 ... mix up the order of words I am learning to help train myself learn them in a different way. ITEM40 ... will remember words through common expressions or idioms that they are found in. Richard LaBontee 346 M em or y Ba se d (1 2 it em s) WHEN LEARNING OR REMEMBERING SWEDISH VOCABULARY, I… ITEM41 ... memorize words however I can. ITEM42 ... write down vocabulary notes (when I study alone, in class, etc.). ITEM43 ... will review words or my own notes repeatedly over time. ITEM44 ... read words out loud [or quietly to myself] over and over. ITEM45 ... write individual words down over and over. ITEM46 ... write down lists of grouped words over and over. ITEM47 ... listen to recordings of vocabulary words I want to remember over and over. ITEM48 ... use lists of words that I write down to give myself quizzes. ITEM49 ... review words in study groups or with others. ITEM50 ... use flashcards/index cards (or similar study tools) to help me review words. ITEM51 ... label items (i.e., in my home, workplace) in order to review them often over time. ITEM52 ... stare at words to help me remember them visually. M ot iv at io n an d Pl an ni ng of Le ar n- in g (1 0 it em s) ITEM53 My motivation for Swedish word learning comes from my natural interest in the language. ITEM54 My motivation for Swedish word learning comes from other reasons than my interest in the language. ITEM55 I skip difficult/unknown words that I come across in order to read/listen for overall meaning. ITEM56 I am willing to use Swedish vocabulary even if I might make mistakes or it makes me feel uncomfortable. ITEM57 I plan my own learning before, during, or after I study. ITEM58 I plan to review [specific amounts of] vocabulary over specific time periods. ITEM59 I plan to use my free time to casually practice Swedish vocabulary. ITEM60 I use technology as a means of learning. ITEM61 I put words into different kinds of groups or lists so I can keep myself organized while I learn. ITEM62 I pay attention to Swedish vocabulary that is especially difficult for me to remember. A w ar en es s of Le ar ni ng Si tu at io n (1 2 it em s) ITEM63 I reflect on vocabulary learning in Swedish in general. ITEM64 I reflect on my personal learning style, strengths, and/or weaknesses when studying Swedish vocabulary. ITEM65 I reflect on errors I make and my accuracy when using my Swedish vocabulary. ITEM66 I reflect on my stress and anxiety from using Swedish vocabulary. ITEM67 I reflect on how using different vocabulary learning strategies affects my learning of Swedish vocabulary. ITEM68 I reflect on the importance of learning Swedish vocabulary in terms of my overall language learning. ITEM69 I reflect on comparing Swedish to other languages I know or speak when learning Swedish vocabulary. ITEM70 I think about my direction of translation when studying Swedish words. ITEM71 I reflect on the socio-cultural impact of my learning Swedish vocabulary. ITEM72 I reflect on the feeling I get from Swedish words that I find. ITEM73 I notice grammar associated with how Swedish words change in different situations to help me learn them better. ITEM74 I reflect on the nature of the sound of Swedish language when learning and recalling words. Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 347 APPENDIX B Initial total variance explained for SVLSS 1.2 Factor Total % of Variance Cumulative % 1 12.655 17.102 17.102 2 4.804 6.492 23.593 3 3.381 4.57 28.163 4 2.596 3.508 31.671 5 2.576 3.481 35.152 6 2.462 3.327 38.479 7 1.984 2.681 41.16 8 1.883 2.545 43.705 9 1.816 2.454 46.16 10 1.669 2.255 48.415 11 1.568 2.119 50.534 12 1.518 2.051 52.586 13 1.483 2.004 54.59 14 1.447 1.956 56.545 15 1.411 1.906 58.452 16 1.298 1.754 60.205 17 1.27 1.716 61.922 18 1.233 1.666 63.588 19 1.188 1.605 65.193 20 1.118 1.511 66.704 21 1.069 1.444 68.148 22 1.004 1.356 69.504 Note. Factors 23-74, below Eigenvalue 1 suppressed. Richard LaBontee 348 APPENDIX C Rotated factor pattern for 6-factor solution Factor 1 2 3 4 5 6 ITEM1 0.635 ITEM2 0.622 ITEM3 0.394 0.354 ITEM4 0.318 0.406 ITEM5 0.374 0.397 ITEM6 0.479 ITEM7 0.355 0.566 ITEM8 0.563 ITEM9 0.481 ITEM10 0.389 ITEM11 0.485 ITEM12 0.491 ITEM13 0.643 ITEM14 0.503 0.382 ITEM15 0.559 0.334 ITEM16 0.533 ITEM17 0.445 ITEM18 0.336 ITEM19 0.451 ITEM20 0.699 ITEM21 0.323 0.573 ITEM22 0.562 ITEM23 0.537 ITEM24 0.425 ITEM25 0.589 ITEM26 0.727 ITEM27 0.348 0.538 ITEM28 0.359 ITEM29 0.514 ITEM30 0.384 0.372 ITEM31 0.45 ITEM32 0.317 0.42 0.339 ITEM33 0.32 0.384 ITEM34 0.493 ITEM35 0.506 ITEM36 0.526 ITEM37 0.544 ITEM38 0.673 ITEM39 0.463 ITEM40 0.301 ITEM41 ITEM42 0.451 0.47 ITEM43 0.613 0.316 ITEM44 0.514 ITEM45 0.678 Questionnaire instrumentation for strategic vocabulary learning in the Swedish as a second language. . . 349 ITEM46 0.597 ITEM47 0.456 ITEM48 0.579 ITEM49 0.508 0.343 ITEM50 0.502 ITEM51 0.456 ITEM52 0.49 ITEM53 0.342 ITEM54 0.478 ITEM55 0.37 ITEM56 0.49 ITEM57 0.486 ITEM58 0.574 ITEM59 0.433 0.322 ITEM60 0.32 ITEM61 0.538 ITEM62 0.429 ITEM63 0.53 ITEM64 0.358 0.433 ITEM65 0.495 0.381 ITEM66 0.62 ITEM67 0.302 0.502 ITEM68 0.331 0.423 ITEM69 0.425 0.509 ITEM70 0.421 ITEM71 0.57 ITEM72 0.518 ITEM73 0.362 0.531 ITEM74 0.463