Effects of the Adequacy of Learning Strategies in Self-Regulated Learning Settings: A Video-Based Microanalytical Lab Study

Journal of Cognitive Education and Psychology Volume 12, Number 3, 2013 Effects of the Adequacy of Learning Strategies in Self-Regulated Learning Settings: A Video-Based Microanalytical Lab Study Peter H. Ludwig University of Koblenz-Landau, Germany Claudia Finkbeiner Markus Knierim University of Kassel, Germany So far, the quality of learning strategies has been considered primarily within the framework of the “description paradigm” by investigating the relationship between the use frequency of macrostrategies and achievement. The ADEQUA study is approaching the quality of learning strategic actions in a more finely grained fashion by rating the adequacy of discrete learning strategies at the microanalytical level. Specifically, the study scrutinizes the strategies used by secondary-level students of English as a foreign language while reading an English text in a self-regulated, cooperative learning environment. The strategies they used in overcoming comprehension difficulties were identified and rated on the basis of the students’ videotaped task performance as well as a stimulated recall procedure. In regression models, the adequacy of strategic actions is of major predictive power with considerable effect sizes for students’ achievement. The hypothesis-testing approach adopted here (i.e., to assess the adequacy of every discrete strategy used by means of highly inferential ratings), appears to be promising. Keywords: learning strategies; video-based study; self-regulation; highly inferential ratings L earning strategies are regarded as an important key to successful learning, above all in self-regulated learning processes. In spite of its high plausibility (Zimmerman, 2000), empirical support for this assumption has been scarce. Most correlational field studies, at least, found little or no connections between the use of learning strategies and learning outcomes (Leutner, Leopold, & den Elzen-Rump 2007, p. 180; Pintrich & Garcia, 1994; Schwinger, Steinmayr, & Spinath, 2009; Veenman, 2005). According to Veenman (2005), learning strategies explain no more than 3% of the performance variance 374 © 2013 Springer Publishing Company http://dx.doi.org/10.1891/1945-8959.12.3.374 Adequacy of Learning Strategies 375 in questionnaire-based studies. Several reasons are given in the literature for this apparent lack of connection (e.g., Leopold, 2009, pp. 72–74; Pintrich, 2004). The two most important ones are (a) the collection of data on habitual strategy use (e.g., Dresel & Haugwitz, 2008), that is, in a context-independent and, thus, less reliable fashion (e.g., by means of prospective self-report assessment; Veenman, 2005) and (b) the distal measurement of learning outcomes through general criteria of effectiveness, such as report grades. In addition, some common inventories display a mismatch between the construct of learning strategies and their test items (e.g., Biggs, 1993; VanderStoep & Pintrich, 2003). For example, several items of the Learning and Study Strategies Inventory (LASSI; Weinstein, 1988) and the Motivated Strategies for Learning Questionnaire (MSLQ; Pintrich & De Groot, 1990; K. E. Dunn, Lo, Mulvenon, & Sutcliffe, 2012) serve at best as indirect indicators of learningrelated activities because they do not express intended actions but passive inner states, considerations, and behavior patterns. Based on the assumption that strategies are best chosen with specific domains, contexts, and learning tasks in mind (Boekaerts, 1997, p. 161), the use of strategies is increasingly being understood as a situational characteristic in line with the state concept. Thus, learning strategies are studied via retrospective assessment: Students are being observed or questioned directly after having worked on a task in order to capture any self-regulated activity “on the fly” during or shortly after its occurrence (Boekaerts & Corno, 2005). Such a proximal approach to data collection is supported by evidence from studies with a close match between observed strategies and measures of learning outcomes. Those studies generally report higher correlations (e.g., Alexander, Murphy, Woods, Duhon, & Parker, 1997; Finkbeiner, 2005; Labuhn, Bögeholz, & Hasselhorn, 2008; Schiefele, 2005). This line of research might be successfully continued by further refining the focus of observation, in particular through a microlevel analysis of the use of discrete strategies in self-regulated learning processes, as will be discussed in the following sections. Classifying learning strategies and extensively describing their use will hardly generate insights of practical relevance. Rather, the appraisal of strategies with respect to their practical usability calls for some kind of normative evaluation, because researchers’ and teachers’ interest in learning strategies revolves around the quality of strategy use and its optimization. Strategies are only relevant to performance if their use is appropriate to the learning situation. This aspect of quality is expressed in theoretical postulates, which refer to “appropriate,” “suitable,” “skillful,” or “effective” strategy use. Conceptions of adequacy are also implied in phrases that are used to describe competence in the application of learning strategies like “good strategy user” and “self-regulated learner” for students who act autonomously, efficiently, in a reflective manner, and possess specific abilities and beliefs (e.g., Wolters, 2003; Schwinger, Steinmayr, & Spinath, 2012). In second and foreign language research, this has also been discussed in the framework of the construct of the “good language learner” (e.g., Griffiths, 2008). Within the context of such theoretical-normative considerations, the question of how the adequacy of learning strategies can be determined in more detail remains unanswered. Empirical research within the framework of the “description paradigm” (Leutner et al., 2007) mainly investigates which strategies are used and how intensively they are being used. However, the utilization frequency of a strategy does not necessarily tell us anything about its usefulness with respect to a specific learning goal (Cohen, 1998). The emphasis on the quality aspect of strategies represents a paradigmatic orientation, 376 Ludwig et al. which has only been cautiously advocated up to now. Essentially, three approaches have been employed: a. Correlational studies on efficiency, which make up the bulk of learning strategy research, consider the quality aspect in so far as they look for links between strategy use and measures of learning results. That is to say, they infer in an ex post manner from noteworthy correlations that the use of specific strategy sets (e.g., surface or deeplevel strategies) results in a specific learning advantage. b. All strategy training programs including those rare studies testing training success experimentally following the “intervention paradigm” (Leutner et al., 2007, p. 180) at least implicitly presuppose theoretical concepts of strategy quality (e.g., Dignath, Büttner, & Langfeldt, 2008; Donker, De Boer, Dignath-van Ewijk, Kostons, & Van der Werf, 2012; Schunk & Ertmer, 2000). The veridicality of these a priori assumptions is globally tested (aggregated in experimental training conditions). c. The same applies to tests of students’ “strategic competence” (e.g., Schütte, Wirth, & Leutner, 2010). One of the few examples is the “Learning Strategy Knowledge Test” by Schlagmüller and Schneider (2007), which is based on experts’ a priori assessments of the adequacy of learner actions in typical classroom situations. On the basis of these approaches, four desiderata can be formulated with reference to the adequacy of strategy use: 1. Studies that follow the description paradigm and test relations between strategy use and learning outcome mostly evaluate strategies on a global level. Up to now, rather nonspecific sets of learning strategies (e.g., rehearsal, organization, elaboration, time management, planning, monitoring) have mainly been tested for adequacy. The suitability of the use of discrete strategies and techniques on a more narrowly defined level of action (e.g., the marking of important text passages) is at most considered in peripheral analyses. However, classroom teachers face student activity on the level of single strategies as represented by individual student actions. Consequently, teachers typically want and need to diagnose the strategic competence of students in order to identify improvement and scaffolding needs. In a more coarse-grained analysis, the external validity of the findings depends decisively on the “correct” grouping of single strategies into clusters or types. Alternative groupings may thus lead to different findings. 2. Many studies measuring the effectiveness of strategies distally refer to a synthetic “average” of instructional settings that are subject to changes over time. This is done, for instance, through mapping the relationships between stable strategy styles and general competencies. It can be assumed, however, that individual strategies have no universal value. Therefore, it is much more important to select a strategy that is appropriate to the situation and adapt it for application. Of yet, more relevance for teaching practice will be insights that classify the usefulness of strategies with regard to specific learning situations in a microcontext while taking student-related factors into account. 3. “Although respondents report using certain learning strategies, neither the frequency of strategy use nor the total number of strategies used give any indication of whether or not the strategies are used effectively” (Leutner et al., 2007, p. 180). The implicitly Adequacy of Learning Strategies 377 introduced assumption “the more, the merrier” tends to ignore the crucial difference between quantity and quality of strategy use. To a large degree, this essential difference (. . .) has been rather neglected in previous research on learning strategies. Successful learning seems not to be (so much) a question of knowing as many learning strategies as possible, but of being familiar with a critical number of learning strategies (not necessarily many), selecting those appropriate to the situation at hand, and applying them in such a way that the specific goals of the strategies can be achieved. (Leutner et al., 2007, p. 180; cf. Schunk & Ertmer, 2000) For instance, for foreign language text comprehension, the identification of unknown terms and phrases that are central to a text is, in general, an appropriate strategy. Moreover, a suitable application of this strategy requires correct decision making about which terms are of central or peripheral importance, respectively. In analyzing the quality of strategy use, the adequacy of strategies is considered in even more detail than at the discrete strategy level. Such microanalytical studies of quality still tend to be rare. 4. Commonly, self-regulation or self-regulated learning is almost equated with using learning strategies (e.g., Perry, Hutchinson, & Thauberger, 2008; Weinstein, Acee, & Jung, 2011). This universal equation is justified in so far as every form of conscious action, including the application of learning strategies, presupposes a certain scope for decision making. Learning in schools cannot be conceived as either completely teacher-directed or as totally self-controlled by the student. Rather, most of these learning processes imply a minimum of self-regulation (Schunk & Ertmer, 2000, p. 632). From this perspective, the phrase “self-regulated learning” sounds slightly tautological. However, this basic consideration may obscure the fact that such a broad, action-theoretical concept of self-regulation does not necessarily cover instructional practices referred to as autonomous, independent, or student-oriented learning, for instance within the framework of “open,” constructivist approaches to teaching (Paris & Byrnes, 1989). Many studies on learning strategies that draw on self-regulation deal mostly with “traditional” teaching where ordinarily only moderate freedom of choice is given to students. Studies on strategic learning situated in settings that strongly emphasize student self-regulation are rare. Notably, the small scope for decision making in instructed learning is used to explain the generally weak correlations between strategy use and learning outcomes (Artelt, 2006). In conclusion, a microanalytical approach rates the choice as well as the distinct application of observed discrete strategies with respect to their adequacy. The exploration of the potential gain in knowledge via such an approach has not yet been worked upon, especially not in learning settings focusing on student autonomy. The research project ADEQUA presented here is intended to test the feasibility of this approach. RESEARCH OBJECTIVE The central interest of the ADEQUA project (“Adequacy of learning strategy use and teacher support actions”) is in gaining empirically founded insights into effectively supporting secondary-school learners during cooperative reading tasks through moderate teacher 378 Ludwig et al. intervention and the development of learning environments that are conducive to the independent reading of foreign language texts. Apart from an extensive pilot study (Finkbeiner, Ludwig, Wilden, & Knierim, 2006), ADEQUA consists of two main stages, a laboratory study and a field study. In both phases, students worked collaboratively in dyads on an English text. In the lab study, these pairs were separated from their classes to allow for the naturalistic observation on the use of learning strategies that were not influenced by teachers. In the subsequent field study, these learning scenarios were implemented in “regular” English lessons in order to observe the teachers’ support actions. This article reports on the laboratory study. The core aspect of this stage has been to “reconstruct” the students’ use of discrete strategies and the assessment of their situation-specific adequacy. In focusing on the level of actual student activity, the study aims to increase the understanding of the effects of specific discrete strategies at the microlevel. These insights may then be used to enhance the quality of learning strategy use in studentcentered learning environments. The lab study addresses two key issues: (a) Does the adequacy of learning strategy use influence the learning result? and (b) How does the relationship between “adequacy of strategy use” and learning outcome compare with that between “frequency of strategy use” and learning outcome? METHOD Design Design of the Learning Setting. The task types used were devised in accordance with the Programme for International Student Assessment (PISA) literacy concept emphasizing cognitively demanding reading comprehension. This concept was adapted for the purpose of foreign language teaching. Pairs of students were given the task of working out the content of a challenging expository or narrative text. Inspired by Vygotsky’s (1978) concept of the Zone of Proximal Development (W. E. Dunn & Lantolf, 1998), the texts were chosen in such a way that their degree of difficulty—in terms of lexical, syntactic, and cultural-pragmatic complexity—exceeded the competence level to be expected from students. Students were deliberately confronted with difficult passages (“stumbling blocks”), which they were supposed to tackle as independently as possible while cooperating with a fellow student. This process was scaffolded by the task types developed for the purposes of the study. More specifically, the tasks provide guidelines to help students structure their reading process. However, the tasks neither point students at “solutions” for presupposed comprehension difficulties (as many textbooks still do by providing a glossary) nor do they tell students explicitly how to tackle any stumbling blocks. That is, the tasks leave considerable room for self-regulatory processes such as choosing adequate strategies, monitoring one’s comprehension, adjusting one’s task completion procedure against situation- and ability-specific constraints, and managing one’s affect (e.g., by coping with frustration). For more details on the task types developed and implemented in the ADEQUA study, see Finkbeiner, Knierim, Ludwig, and Wilden (2008) and Finkbeiner, Knierim, Smasal, and Ludwig (2012). The cooperative task design requires verbal exchange, negotation, and an agreement about a joint approach among the students. Cooperative learning (e.g., Slavin, 1995) serves here as a diagnostic tool to ensure the ongoing observability of students’ cognitive, metacognitive, Adequacy of Learning Strategies Pre-test/-survey Introductory lesson 379 Task session Follow-up survey, achievement test Interview, stimulated recall • approx. 30 min/dyad • approx. 10 min/dyad • approx. 30 min/dyad • 1 class period • 3 class periods • recorded on video • no video recording • no video recording • recorded on video • recorded on video • N  164 • N  164 • N  164 • N  352 • N  352 (82 learner dyads) (82 learner dyads) (82 learner dyads) FIGURE 1. Stages of data collection. and socioaffective strategic actions. This design feature resembles a think-aloud procedure, although it largely avoids the artificiality and risk of reactivity of the measurement frequently associated with thinking aloud (Webb, Campbell, Schwartz, & Sechrest, 1981). Usually, self-regulated learning in the classroom is staged as a student activity that is initiated and accompanied by teaching activities. The absence of the teacher in the lab study is not meant to argue for a change of this concept. Teacher support has been dispensed with only to observe “typical” student behavior, minimizing current external influences. Research Design. Data collection took place in three main stages: (a) the pretest survey, (b) the video recording of student pairs while working on tasks, and (c) a follow-up (see Figure 1). Pretest/Presurvey and Introductory Lesson. Prior to working on the task, all students of the participating classes took three class periods to fill in a comprehensive written survey and test to determine dispositional characteristics. During an introductory lesson, students were acquainted with the complex format of the task by watching a video demonstrating the task procedure and by trying the respective task format afterward. During the pilot phase, this standardized video presentation of a behavioral model had proven to be superior to oral instructions in getting students to adhere to the intended task procedure, in particular with regards to communicating aloud with their partners. Moreover, in the introductory lesson, the subsequent data collection procedure was simulated through the presence of investigators and video cameras in an effort to enable the students to become accustomed to both and, thus, to desensitize them to the research setting. Video-Recorded Work on Tasks. A few weeks (M 5 5.3) later, the individually audio- and video-recorded task sessions were administered by two researchers. Each student pair worked on the task for about 30 min. Follow-up. Immediately after the task session, the students were tested for their learning outcomes and retrospectively questioned on various situational characteristics of the learning by means of a questionnaire. This was followed by a video-recorded interview that included a stimulated recall procedure: The students were shown clips from the recorded task session that contained, from the researchers’ vantage point, ambiguous sequences of student actions. These clips had been preselected during the task session by the researchers. The students were asked to explain and comment on their actions. In this way, it was possible to make initially ambiguous learning activities “visible” post hoc. Measures The following overview is limited to those data collection instruments deployed whose results will be reported in this article. For details on the psychometric properties of instruments developed by other authors, the references given can be consulted. 380 Ludwig et al. In the presurvey/pretest, habitual characteristics of students were determined by means of written tests and students’ self-assessment reports. Three measures were used to assess the students’ ability: Grades in the last report for the subjects English, German, and mathematics as indicators of the degree of prior subject knowledge; English language proficiency measured with the standardized Oxford Quick Placement Test (Geranpayeh, 2003), and general verbal cognitive ability determined with the verbal section of the intelligence test “Kognitiver Fähigkeitstest” [Test of Cognitive Ability] (Heller & Perleth, 2000). The students’ self-reported traits include habitual interest in reading English texts measured with a shortened version of an inventory by Finkbeiner (2005; Cronbach’s a 5 .925), dispositional frequency of using strategies in the reading of English texts as determined with a shortened version of a standardized inventory by Finkbeiner, and the self-concept of ability in the subject English assessed with a scale by Helmke (1992) and adapted by Finkbeiner. During the follow-up, the learning result was measured, which represents the objective yield of dealing with the learning task. Therefore, the students took an achievement test consisting of 11 items to determine their degree of text comprehension after completion of the task. The task-sensitive multiple-choice test was developed by the authors based on concepts of van Dijk and Kintsch (1983). The students were asked to reply to questions that could be answered correctly if the content of the text was understood. An example of one test item referring to a text on tornado storm chasers is “Which one of the following items does not provide protection against an upcoming tornado?” with the answer choices “mobile home— walk-in closet—basement—bathroom” given. As expected, the learning result scores correlate significantly with English proficiency (r 5 .58; p , .001) and with verbal intelligence (r 5 .59; p , .001). Both coefficients can be viewed as measures of criterion validity. The internal consistency of the test, as an indicator of its reliability, is moderately sufficient (Cronbach’s a 5 .66). Besides this test of the learning result, the learning gain as another achievement measure was determined. This second variable was introduced because the term learning result for the test score does not correspond exactly with the current performance in understanding the text because the test score is not only influenced by the student’s current learning process but also by his or her prior knowledge and mastery of the English language as well as general cognitive ability. Hence, to obtain a rough indicator of the learning gain, the unstandardized residuals of the observed test scores and those outcome values predicted via regression (R2 5 .40) from English language proficiency and verbal intelligence were calculated. Positive residuals characterize situational “overachievers” (i.e., students who have achieved better test scores than could be expected from their language proficiency and their verbal abilities). Negative values point out situational “underachievers” whose performance on the task at hand is weaker than expected. This regression procedure approximates a pre–post measurement since the development of a real pretest, which would have to be taken before the working phase without knowing the text, was not possible because of the text-specific nature of the comprehension test. In brief, the learning test result is a measure of the degree of text comprehension (absolute achievement result) whereas the learning gain is an assessment of the degree of improvement of knowledge through working on the text (relative learning result in terms of a “net yield”). Evaluation of the Video Recordings The data sources that represent the learning processes during the task session (i.e., video recordings and follow-up interviews) were used to reconstruct students’ learning activities Adequacy of Learning Strategies 381 from the chains of actions displayed by the students. In a first step, these chains of actions were segmented into events, during each of which the students addressed a specific comprehension problem, such as an unknown word or idiom, an unfamiliar syntactic structure, and difficulties in understanding relations between propositions or a new concept. With respect to the identification of strategies, the analytical unit was the individual student. Where both students of a dyad were involved in the use of a strategy, this was coded for both students. In a second round of coding, each strategy that had been identified was classified by specifically trained expert raters according to a strategy taxonomy: Each observed instance of strategic behavior was allocated to 1 of 69 types of learning strategy. This classification was developed and documented in a comprehensive coding manual. The coding scheme is based on O’Malley and Chamot’s (1990) well-established taxonomy of foreign language learning strategies, which has been inductively differentiated based on the analysis of a part of the videotaped material. The taxonomy distinguishes between cognitive, metacognitive, and socioaffective strategies. When readers apply cognitive strategies, for example, they summarize the text, infer meanings of unknown words from context or based on their prior knowledge of the foreign language, look up words in a dictionary, or connect the text to their world knowledge (i.e., elaboration). Readers applying metacognitive strategies plan, monitor, and evaluate their reading comprehension processes. Finally, socioaffective strategies include, for example, selftalk, asking for clarification, giving feedback, and cooperation in general. Finally, in a third coding step, each strategic action was judged in a dichotomous scheme as being either adequate or inadequate. Since the bundling of discrete microstrategies into macrostrategies, as is typical of previous studies, was considered suboptimal (see earlier text), the strategies’ adequacy was assessed at the level of the smallest analytical unit: In a time-consuming rating procedure, each learning activity previously identified as an instance of strategy use was judged separately with respect to its adequacy, independent of its prior allocation to a strategy type. For example, if a dictionary was used on three occasions to determine the meaning of an unknown lexical item (representing a “resourcing” strategy type), each of these instances was judged separately. A similar approach to the determination of the adequacy of learning strategic actions can be found in Schütte et al. (2010) and Schlagmüller and Schneider (2007), who dealt however with a smaller range of strategies. Following Oxford (2003), the use of a strategy was judged to be adequate if it had the potential to contribute to a complete or partial solution of the respective problem in understanding the text. In this highly inferential judgment, the quality of (a) choice and (b) the concrete implementation of the strategy were taken into account (cf. Ludwig, 2009, 2012). The judgment of adequacy was based on thorough scrutiny (a) of the comprehension problem which the students intend to resolve with the learning strategy, (b) of the microcontext in which the students make their strategic moves, and (c) of the competencies needed for the implementation of the strategy (Leopold, 2009, pp. 81–85). Some examples may serve to clarify the definition. For instance, when a student reduces the unknown word “safety” to a known root (“safe”) by applying the strategy “L2 lexical/ morphological transfer,” this is considered an adequate strategic action. Similarly, inferring the gender of a person’s first name “Yolanda” in the text from knowing that in several languages the ending “a” indicates a female first name (“academic elaboration” strategy) would be strategically adequate. In contrast, it would be strategically inadequate to attempt to infer the meaning of an unknown word from the context by guessing or reasoning (“inferencing” or “between-parts elaboration” strategy) if the phrase or passage in question 382 Ludwig et al. contains further unknown words. Also, trying to decode the meaning of “close calls” by looking up “close” and “calls” in the dictionary (resourcing strategy) would be considered inadequate since the student fails to recognize that close calls is an idiom, the meaning of which cannot be derived from the sum of its parts. As a final example, it would be inadequate for a student to assume that the verb “(to) last” is related in meaning to the adjective “last” (L2 lexical/morphological transfer strategy) while ignoring contextual cues that preclude such an assumption. The rating process was preceded by a training of the coders, including joint analysis of action sequences, anchor examples, discussion of complex learning sequences as well as independent ratings followed by intercoder comparisons. This elaborate training was absolutely necessary because of the highly inferential nature of some of the decisions. The video recording of each student pair was coded independently by two coders using the software program Transana (Woods & Fassnacht, 2007). Discrepancies between these codings were then resolved through discussion, involving a third rater if necessary (discursive validation). If consensus could not be reached, the code in question was removed from the data set and any further analysis. The adequacy codings of a student’s discrete learning strategies were aggregated into a total value which represents the degree of adequacy of his or her strategic actions. This index is the percentage of adequate strategies with regard to all registered strategies of a student. Participants In the pretest/presurvey, 15 complete 9th-grade classes with 352 students took part. The students (age M 5 14.76; SD 5 0.72) were drawn from 11 German schools at all three secondarylevel tracks. From each class of the total sample, 10–12 students were selected for the videotaped task session (including follow-up) and grouped in pairs. Selection criteria were students’ willingness, parents’ consent to participate, and social preferences about the dyad partner. Moreover, the sample was to include students from a wide range of English foreign language proficiency levels and to represent both genders equally. After excluding video recordings that could not be used for technical reasons, the data from 74 student pairs (148 students) were available for analysis. The recordings of task sessions and follow-up interviews have a length of 60–90 min per dyad, amounting to approximately 110 hr of video material. RESULTS Quality of the Adequacy Ratings The distribution of learning strategy adequacy indices is skewed to the right (M 5 93.70%; min 5 64.70%; max 5 100.00%). This means that most students employed strategies in an extremely situationally appropriate way. Compared with the theoretical range of the scale (0.00%–100.00%), the values vary only slightly (SD 5 6.87%). Yet the analyses described further in the following text show that an explanatory power is to be ascribed to the adequacy judgments in spite of their low variance. Within the framework of the data evaluation reported here, the resources were not available for a complete documentation of the parallel coding to test for intercoder agreement at the microlevel. (This kind of agreement is usually labeled “interrater reliability,” although the Adequacy of Learning Strategies 383 designation “interrater objectivity” would in fact be more appropriate.) Prior to the discursive validation procedure, the agreements of the initial raw codings of the two rater teams were estimated to be between 70% and 80%. Presently, the video data is being used to test further approaches to assessing strategic adequacy that are more differentiated and also incorporate alternative criteria. For these reanalyses, a subsample of 26 students was coded by two independent coders. In identifying the learning activities indicating strategy use, they achieved an agreement of 86% (units coded: n 5 1,165 discrete strategies) after the training. In assigning these individual strategies to specific strategy types, the agreement amounted to 89.2% (in accordance with Cohen’s kappa coefficient of .887; p , .001). For the ratings of different measures of adequate learning strategy use, the kappa values ranged between .73 and .86 (each with p , .001). These parameters, together with the following analyses of validity, indicate that this highly inferential variable has been measured with sufficient objectivity and validity. Furthermore, the characteristic “adequate learning strategy use” proves to be relatively independent compared with the traditional frequency measures of learning strategy research: The scales of the inventory on habitual strategy use are not significantly correlated with adequacy. The frequencies of only 12 of the 66 types of strategies, as identified effectively in situational use, correlate significantly with the adequacy index, in the main marginally (.14 # r # .25; .002 , p , .04). Effects of Strategy Frequency On average, students used 41.3 recorded learning strategies per task session (SD 5 18.7; min 5 12; max 5 108), 38 (approximately 92%) of which were adequate. If the traditional core variable of learning strategy research, namely frequency of use, was to play an important role for the learning outcomes achieved, we would have expected differences in frequencies between students with high and low learning gains. To test this assumption, the frequencies of the strategies used by two extreme student groups were compared, namely between the quartiles with the lowest (distinctive underachievers) and the highest learning gain (marked overachievers). Only 4 of the 66 recorded strategy types show significant differences in frequency of use for these student groups (.001 , p , .041; unpaired t tests). Generally speaking, this seems to support the argument that differences in learning success cannot be attributed to different frequencies of strategy use. Moreover, it emphasizes the necessity to take account of the quality of strategy use. Effects of Strategic Adequacy The more adequately students act, the better their learning results should be. The testing of this hypothesis simultaneously serves to check the predictive validity of adequacy ratings for the external criterion “learning result.” Relationship Between Adequate Strategy Use and Measures of Learning Success. The students’ adequacy indices correlate moderately and highly significantly with the learning results (r 5 .384; p , .001; N 5 148): A more strategically adequate way of working on the task thus goes hand in hand with improved text comprehension. The magnitude of this correlation is mitigated when the influences of English foreign language proficiency and verbal intelligence are partialled out but it still remains highly significant (r 5 .26; p 5 .002). That is, the covariance between adequate strategy use and achieved reading comprehension is partly caused by the influence of subject-related competence and intelligence but cannot be exhaustively explained by these factors. 384 Ludwig et al. Comparison With “Traditional” Coefficients of Strategy Use. When compared to variables traditionally used to represent learning strategy use, the construct adequacy of strategy use proves to be more criterion valid. Frequencies of habitual strategy use as well as frequencies of observed actual strategy use correlate less, on average, with measures of learning success than strategic adequacy: • The scales of the inventory measuring students’ “frequencies of dispositional strategy use” do not correlate significantly with (relative) learning gain. Four of six subscales of the frequency instrument still correlate significantly with (absolute) learning results albeit weaker than the adequacy variable (.16 # r # .28; .001 , p , .035). • The frequencies of only 2 of all 66 strategy types observed correlate significantly with learning gain (r 5 .19; p 5 .026 and r 5 .38; p , .001, respectively). The frequencies of 10 of the 66 types display significant relations with the absolute learning result (.16 # r # .34; rM 5 .21; .001 , p # .049). Relative Significance of Adequate Strategy Use. Several regression models have been calculated. They examine the question of whether strategic adequacy contributes independently to explaining variance in the learning results if competing influences of further dispositional factors are controlled for. The best fit is achieved in a model that explains approximately 45% of the variance observed (F[5, 131] 5 21.13; R2 5 .446; p , .001): In addition to adequacy of strategy use, explanatory predictors are language proficiency (b 5 .23; p 5 .014), verbal cognitive ability (b 5 .34; p , .001), habitual interest in reading English texts (b 5 .08; p 5 .22), and the classwise standardized grades in mathematics (b 5 .18; p 5 .006). In this model, the explanatory potential of strategic adequacy with a beta weight of b 5 .21 (p 5 .003) is exceeded only by language proficiency and intelligence. The additional inclusion of the domain-specific self-concept of ability and the report grade in English do not yield further explanation of variance to the model. Their explanatory power seems to be largely covered by English language proficiency, which is plausible: Language proficiency was measured in absolute terms (standardized over all school tracks) as was the dependent variable learning result, whereas the awarding of grades is oriented toward the social reference frame of the school track or the individual class, respectively. A similar reference group effect applies to self-concept. Therefore, the poor predictive power of German report grades as an additional independent variable is hardly surprising, even though the subjects of German and English can be situated within the same domain. The grades in both languages are correlated with r 5 .52 (p , .001; N 5 335). However, it is worth noting that the grades in mathematics, which as a subject does not appear close to English initially, contribute 4% to the explanation of variance. Testing student dyad data for interdependence does not seem necessary. If variables of the learning partner (adequacy, language proficiency, intelligence) are added to the model, they do not increase the total amount of variance explained. The intended influence of the partner through the negotiation of strategic actions seems to be sufficiently represented in the adequacy index of the “protagonist.” The adequacy indices of learning partners correlate at r 5 .49 (p , .001). Effect Sizes. In order to assess the practical significance of adequate strategy use for the learning results, student groups with different degrees of adequate strategy use were compared. For this purpose, the continuously metric learning strategy adequacy index was reduced to a discrete ordinal scale (post hoc blocking) by assigning the students to four groups Adequacy of Learning Strategies 385 TABLE 1. Learning Results by Level of Adequate Strategy Use Learning Result Level of Adequate Strategy Use 1 Low level of adequate strategy use 2 Moderate–low level of adequate strategy use 3 Moderate–high level of adequate strategy use 4 High level of adequate strategy use M N SD Min Max 53.91 65.87 64.97 70.98 37 36 34 41 17.2 14.6 19.3 18.1 21.43 35.71 9.09 28.57 81.82 92.86 100.00 100.00 Note. Higher means represent better results on the reading comprehension test. (quartiles) of similar size with homogeneous levels of strategic adequacy. In a one-way analysis of variance design, these four groups, distinguished by their levels of adequacy, differ significantly in their learning results (F[3, 144] 5 6.52; p , .001; h2 5 .12; see Table 1). The group that employs strategies most adequately outperforms their fellow students with the least adequate strategic actions by almost one standard deviation (Cohen’s d 5 0.96). Subgroup comparisons using the Scheffé test show significant differences between the extreme groups (p , .001) and between the first and second quartile (d 5 0.75; p 5 .038). The difference between the first and third quartile is only “marginally significant” (d 5 0.60; p 5 .072). Even when adjusted for the influence of language proficiency and verbal intelligence as covariates, between-group differences remain statistically significant, while retaining somewhat lower effect sizes (d 5 .60 between extreme groups). DISCUSSION The prevalent approach in learning strategy research puts emphasis on identifying macrolearning strategies and drawing conclusions about the quality of strategy use from relationships between frequencies of use and learning outcome. In this type of research, the a priori development of concepts of strategy quality to be tested is not strictly required. In contrast, a hypothesis-testing approach was chosen for the ADEQUA laboratory study. The quality of discrete strategic acts was directly assessed on the molecular level. These separate assessments represent virtual mini-hypotheses about quality, subsequently tested on the student level. In a learning setting that encourages and requires self-regulation, student dyads were given the task to independently reconstruct the meaning of a text in English as a foreign language. Based on the video recording of the task session, the learning strategies they used were assessed with respect to their adequacy. Overall, these assessments have proven valid. Adequate strategy use has turned out to be an important predictor for the variance of learning success, a fact that supports the claim that the measurement of this highly inferential variable is criterion valid. The effect sizes are considerable. Thus, further analyses exploring the use of discrete strategies in a differentiated manner (Knierim, 2013) and examining students’ trait and state characteristics as predictors of adequate strategy use seem promising. Such predictors can provide teachers with hints about which students are probably in special need of their support while employing strategies. The moderate bivariate correlations between adequate use of strategies and measures of learning success (r # .38) are remarkable insofar as the theoretical range of the adequacy scale is not even remotely covered by the observed values. 386 Ludwig et al. Contrasted with classical learning strategy research, which focuses on frequencies of strategy use, this approach to assessing the quality of the strategies used can stand its ground: Within the data set of this study, frequency measures correlate less with learning success than strategic adequacy does. In addition, our findings pertaining to the strategic adequacy construct fare very well against frequency measures used in other studies as well. Specifically, the coefficients of the correlations between adequacy and learning success are high compared to (a) most studies using distal forms of measurement (e.g., Artelt, 2006; Dent, Cooper, & Koenka, 2012; Pintrich, 1989; Pintrich & Garcia, 1994; Pokay & Blumenfeld, 1990) as well as to (b) the bulk of studies adopting a proximal approach to measuring frequencies of strategy use (e.g., Schiefele, 2005). The findings presented here, together with other studies that consider the qualitative aspect of (a more limited range of) learning strategies, yield a consistent picture: Comparatively strong associations between the quality of strategy use and learning success can be demonstrated. Simpson, Olejnik, Tam, and Supattathum (1994) report a correlation of r 5 .76 between the quality of verbal elaboration strategies and reading comprehension. Hall, Bailey, and Tillman (1997) found a correlation of r 5 .61 between the quality of the execution of a visualization strategy (drawing pictures) and learning success. In a study carried out by Leopold, den Elzen-Rump, and Leutner (2006), proximal measurements of the quality use of strategies and of the learning outcome upon reading a scientific text reach coefficients up to r 5 .59. In the German data set of PISA 2000, dispositional knowledge about the adequate use of learning strategies for reading proved to be a good indicator of reading competence (r 5 .55; Artelt, 2006). Schütte et al. (2010, 2011) calculated a coefficient of r 5 .39 between the adequacy value of two selected strategies used while reading native-language expository texts and the knowledge gain achieved in the process. In their study, the two adequacy measurements of learning strategy use turned out to be the best predictors of learning success. In analogy, Finkbeiner (2005) reports on a strong connection between the quality of strategy use and the level of text deep processing (r 5 .52) in a linear structural relations (LISREL) analysis. The current state of research in the field indicates that the quality of using strategies is of substantial importance for effective learning and that a high quality is not per se associated with the application of certain strategies. In establishing further determining factors for understanding the students’ reading comprehension achievement, we discovered an anomaly that could not be resolved with the available data: The math grades proved to be a powerful predictor. However, the learning results counterintuitively correlate positively with classwise z-standardized math grades (r 5 .18; p 5 .02). That is, poorer grades in mathematics (here, low performance is indicated by high numbers) go together with better results. This is all the more astonishing since the grades in English, German, and mathematics correlate positively with one another to a highly significant degree (Cronbach’s homogeneity index a 5 .64; intraclass correlation [ICC] 5 .373; p , .001). This means that the reciprocal relationship between math grade and learning result cannot be explained by means of the popular myth that high competencies in languages and STEM domains exclude each other. Limitations and Outlook. To the authors’ knowledge, this study has undertaken for the first time a comprehensive microanalytical assessment of the adequacy of all discrete strategic actions taken by students in a given self-regulated learning environment. It has therefore not been possible to build on previous experience or well-established category systems in Adequacy of Learning Strategies 387 assessing the situational adequacy of strategy use. Although this approach has proved useful, its implementation in this study probably bears potential for refinement. The adequacy was coded only dichotomously (adequate vs. inadequate). For the initial testing of this highly inferential approach, such a binary nominal scale seemed to be sufficient. In future research, it would be worthwhile to explore whether a more finely grained coding of the adequacy of strategy use leads to even more valid assessments, for instance, by considering multiple dimensions such as the adequacy for the overall learning situation, the individual student’s abilities, or the time available for the given task. In the adequacy assessment, the finest possible analysis unit has been selected, namely each instance of strategy use during an event representing one specific comprehension problem. In doing so, we wanted to avoid the risk of losing sight of necessary differentiations (e.g., by assessing adequate strategy use for each discrete instance of strategy use instead of assessing the adequacy of macrostrategies or bundles of discrete strategies used). This process of assessing adequate strategy use is very time-consuming. Subsequent studies should explore whether the applied assessment criteria can be made more explicit and consolidated into heuristics that allow for more time-economical coding as well as transfer to other learning scenarios and domains. Furthermore, options for aggregating discrete strategies into superordinate categories, without sacrificing insights that might be relevant for classroom practice, should be explored on the basis of the (further refined) criteria for assessing adequate strategy use. Considerable efforts have been made to identify as many instances of strategy use as possible, while recognizing the limits to this endeavor, for example, because of students’ limited capacity for self-reflection. The approach adopted here, however, assumes that the strategies identified are representative of all learning strategies applied, including those that might have been overlooked, at least with regard to their degree of adequacy. Indeed, the prognostic validity of the adequacy assessments reported in this article indicates that the degree of adequate strategy use can be fruitfully exploited to diagnose a student’s “learning strategic competence” as part of his or her “self-regulatory competence.” The adequacy ratings were based on the definition given previously, which leaves room for implicit decision making on part of the raters. In this respect, the rating procedure is similar to the way teachers would try to diagnose their students’ comprehension problems and use of strategies. In other words, the rating process was not based on an explicit and elaborate theory of the adequacy of learning strategies. The quality of these assessments is thus dependent on the diagnostic competence of the raters (Cooksey, 1996). The detailed and systematic explication of the raters’ subjective theories of adequate strategy use was beyond the scope of this project. In this point, ADEQUA resembles other studies that also build on expert judgment of the appropriateness of the strategies used; typically, though, these studies focus on a more limited number of strategies than ADEQUA does (e.g., Schlagmüller & Schneider, 2007; Schütte et al., 2010). A theory of such complexity (initially domain- and task-specific) has still to be developed. Future research may start to develop this theory by following the microanalytical approach adopted in this study and by documenting the raters’ decision making processes in assessing the adequacy of individual instances of strategy use. For example, our raters often observed that students did not try to verify or falsify their hypotheses about the meaning of an unfamiliar word in a systematic way by putting their supposition into the context and testing it for meaningfulness. Based on such documentation, first steps can be undertaken toward an a posteriori theory of “good strategic action.” In this vein, additional qualitative analyses of the ADEQUA data corpus might yield some initial insights. 388 Ludwig et al. REFERENCES Alexander, P. A., Murphy, P., Woods, B., Duhon, K., & Parker, D. (1997). College instruction and concomitant changes in student’s knowledge, interest, and strategy use. Contemporary Educational Psychology, 22, 125–146. Artelt, C. (2006). Lernstrategien in der Schule [Learning strategies in school]. In H. Mandl & H. F. Friedrich (Eds.), Handbuch Lernstrategien (pp. 337–351). Göttingen, Germany: Hogrefe. Biggs, J. B. (1993). What do inventories of students’ learning processes really measure? A theoretical review and clarification. British Journal of Educational Psychology, 63, 3–19. Boekaerts, M. (1997). Self-regulated learning: A new concept embraced by researchers, policy makers, educators, teachers, and students. Learning and Instruction, 7(2), 161–186. Boekaerts, M., & Corno, L. (2005). Self-regulation in the classroom: A perspective on assessment and intervention. Applied Psychology: An International Review, 54, 199–231. Cohen, A. D. (1998). Strategies in learning and using a second language. London, United Kingdom: Longman. Cooksey, R. W. (1996). Judgment analysis. San Diego, CA: Academic Press. Dent, A., Cooper, H. M., & Koenka, A. C. (2012, April). A synthesis of research on the relation between study skills and academic performance. Paper presented at the 93rd Annual Meeting of the American Educational Research Association, Vancouver, BC, Canada. Dignath, C., Büttner, G., & Langfeldt, H. P. (2008). How can primary school students acquire selfregulated learning most efficiently? A meta-analysis on interventions that aim at fostering selfregulation. Educational Research Review, 3, 101–129. Donker, A. S., De Boer, H., Dignath-van Ewijk, C. C., Kostons, D. D. N. M., & van der Werf, M. P. C. (2012, April). Effective self-regulation strategies: A meta-analysis. Paper presented at the 93rd Annual Meeting of the American Educational Research Association, Vancouver, BC, Canada. Dresel, M., & Haugwitz, M. (2008). A computer-based approach to fostering motivation and selfregulated learning. The Journal of Experimental Education, 77(1), 3–18. Dunn, K. E., Lo, W., Mulvenon, S. W., & Sutcliffe, R. (2012). Revisiting the Motivated Strategies for Learning Questionnaire. Educational and Psychological Measurement, 72(2), 312–331. Dunn, W. E., & Lantolf, J. P. (1998). Vygotsky’s Zone of Proximal Development and Krashen’s i11: Incommensurable constructs, incommensurable theories. Language Learning, 48(3), 411–442. Finkbeiner, C. (2005). Interessen und Strategien beim fremdsprachlichen Lesen [Interests and strategies in foreign language reading]. Tübingen, Germany: Narr Finkbeiner, C., Knierim, M., Ludwig, P. H., & Wilden, E. (2008). Textbasierte kooperative Lernumgebungen im Englischunterricht—Das ADEQUA-Projekt [Text-based cooperative learning environments in the EFL classroom—The ADEQUA project]. In Kasseler Forschergruppe Empirische Bildungsforschung (Ed.), Lernumgebungen auf dem Prüfstand (pp. 81–99). Kassel, Germany: Kassel University Press. Finkbeiner, C., Knierim, M., Smasal, M., & Ludwig, P. H. (2012). Self-regulated cooperative EFL reading tasks: Students’ strategy use and teachers’ support. Language Awareness, 21(1–2), 57–84. Finkbeiner, C., Ludwig, P. H., Wilden, E., & Knierim, M. (2006). ADEQUA—Bericht über ein DFGForschungsprojekt zur Förderung von Lernstrategien im Englischunterricht [ADEQUA—Report about a DFG-funded research project on fostering learning strategies in the English as a foreign language classroom]. Zeitschrift für Fremdsprachenforschung, 17(2), 257–274. Geranpayeh, A. (2003). A quick review of the English Quick Placement Test. UCLES Research Notes, 12, 8–10. Griffiths, C. (2008). Lessons from Good Language Learners. Cambridge, United Kingdom: Cambridge University Press. Hall, V. C., Bailey, J., & Tillman, C. (1997). Can student-generated illustrations be worth ten thousand words? Journal of Educational Psychology, 89, 667–681. Adequacy of Learning Strategies 389 Heller, K. A., & Perleth, C. (2000). Kognitiver Fähigkeitstest für 4. bis 12. Klassen, Revision [Cognitive ability test for grades 4 through 12, revised]. Göttingen, Germany: Beltz. Helmke, A. (1992). Selbstvertrauen und schulische Leistungen [Self-confidence and school performance]. Göttingen, Germany: Hogrefe. Knierim, M. (2013). Strategien beim aufgabenorientierten Lesen in der Fremdsprache Englisch. Eine empirische Studie zur effektiven Sequenzierung von Lernerstrategien [Strategies in task-based EFL reading. An empirical study on the effective sequencing of learner strategies]. Forthcoming doctoral dissertation, University of Kassel, Germany. Labuhn, A., Bögeholz, S., & Hasselhorn, M. (2008). Lernförderung durch Anregung der Selbstregulation im naturwissenschaftlichen Unterricht [Facilitating learning by stimulating self-regulation in the science classroom]. Zeitschrift für Pädagogische Psychologie, 22(1), 13–24. Leopold, C. (2009). Lernstrategien und Textverstehen [Learning strategies and text comprehension]. Münster, Germany: Waxmann. Leopold, C., den Elzen-Rump, V., & Leutner, D. (2006). Selbstreguliertes Lernen aus Sachtexten [Selfregulated learning from expository texts]. In M. Prenzel & L. Allolio-Näcke (Eds.), Untersuchungen zur Bildungsqualität von Schule (pp. 268–288). Münster, Germany: Waxmann. Leutner, D., Leopold, C., & den Elzen-Rump, V. (2007). Self-regulated learning with a text-highlighting strategy: A training experiment. Zeitschrift für Psychologie, 215(3), 174–182. Ludwig, P. H. (2009). “Anything goes—Quality stays.” Quality standards for qualitative educational research in the context of justification. EducatiON-line. Retrieved from http://www.leeds.ac.uk/ educol/documents/180695.pdf Ludwig, P. H. (2012). Thesen zur Debatte um Gütestandards in der qualitativen Bildungsforschung [Thesis statements on debating quality standards in qualitative educational research]. In M. GläserZikuda, T. Seidel, C. Rohlfs, A. Gröschner, & S. Ziegelbauer (Eds.), Mixed Methods in der empirischen Bildungsforschung (pp. 79–89). Münster, Germany: Waxmann. O’Malley, J. M., & Chamot, A. U. (1990). Learning strategies in second language acquisition. Cambridge, United Kingdom: Cambridge University Press. Oxford, R. L. (2003). Language learning styles and strategies. International Review of Applied Linguistics, 41, 271–278. Paris, S. G., & Byrnes, J. (1989). The constructivist approach to self-regulation and learning in the classroom. In B. J. Zimmerman & D. H. Schunk (Eds.), Self-regulated learning and academic achievement (pp. 169–200). New York, NY: Springer Publishing. Perry, N. E., Hutchinson, L., & Thauberger, C. (2008). Talking about teaching self-regulated learning. International Journal of Educational Research, 47, 98–108. Pintrich, P. R. (1989). The dynamic interplay of student motivation and cognition in the college classroom. In M. Maehr & C. Ames (Eds.), Advances in motivation and achievement: Motivation enhancing environments (pp. 117–160). Greenwich, CT: JAI Press. Pintrich, P. R. (2004). A conceptual framework for assessing motivation and self-regulated learning in college students. Educational Psychology Review, 16, 385–407. Pintrich, P. R., & De Groot, E. V. (1990). Motivational and self-regulated learning components of classroom academic performance. Journal of Educational Psychology, 82(1), 33–40. Pintrich, P. R., & Garcia, T. (1994). Self-regulated learning in college students. In P. R. Pintrich, D. R. Brown, & C. E. Weinstein (Eds.), Student motivation, cognition, and learning (pp. 113–133). Hillsdale, NJ: Erlbaum. Pokay, P., & Blumenfeld, P. C. (1990). Predicting achievement early and late in the semester. Journal of Educational Psychology, 82(1), 41–50. Schiefele, U. (2005). Prüfungsnahe Erfassung von Lernstrategien und deren Vorhersagekraft für nachfolgende Lernleistungen [Measuring learning strategies in proximity to an exam situation and their predictive power for subsequent learning performance]. In C. Artelt & B. Moschner (Eds.), Lernstrategien und Metakognition (pp. 13–41). Münster, Germany: Waxmann. 390 Ludwig et al. Schlagmüller, M., & Schneider, W. (2007). Würzburger Lesestrategie-Wissenstest für die Klassen 7-12 [Wuerzburg test of reading strategy knowledge for grades 7–12]. Göttingen, Germany: HogrefeTestzentrale. Schunk, D. H., & Ertmer, P. (2000). Self-regulation and academic learning. In M. Boekaerts, P. R. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 631–649). San Diego, CA: Academic Press. Schütte, M., Wirth, J., & Leutner, D. (2010). Selbstregulationskompetenz beim Lernen aus Sachtexten— Entwicklung und Evaluation eines Kompetenzstrukturmodells [Self-regulatory competence in learning from texts—Development and evaluation of a competence structure model]. Zeitschrift für Pädagogik, 56, 249–257. Schütte, M., Wirth, J., & Leutner, D. (2011, August/September). The influence of self-regulation competencies on learning with expository texts. Paper presented at the EARLI 2011 conference, University of Exeter, United Kingdom. Schwinger, M., Steinmayr, R., & Spinath, B. (2009). How do motivational regulation strategies affect achievement: Mediated by effort management and moderated by intelligence. Learning and Individual Differences, 19, 621–627. Schwinger, M., Steinmayr, R., & Spinath, B. (2012). Not all roads lead to Rome—Comparing different types of motivational regulation profiles. Learning and Individual Differences, 22, 269–279. Simpson, M., Olejnik, S., Tam, A., & Supattathum, S. (1994). Elaborative verbal rehearsals and college students’ cognitive performance. Journal of Educational Psychology, 86, 267–278. Slavin, R. E. (1995). Cooperative learning. Boston, MA: Allyn and Bacon. van Dijk, T. A., & Kintsch, W. (1983). Strategies of discourse comprehension. San Diego, CA: Academic Press. VanderStoep, S., & Pintrich, P. R. (2003). Learning to learn. Upper Saddle River, NJ: Prentice Hall. Veenman, M. V. (2005). The assessment of metacognitive skills. In C. Artelt & B. Moschner (Eds.), Lernstrategien und Metakognition (pp. 77–99). Münster, Germany: Waxmann. Vygotsky, L. S. (1978). Mind in society. Cambridge, MA: Harvard University Press. Webb, E., Campbell, D., Schwartz, R., & Sechrest, L. (1981). Unobstrusive measures. Chicago, IL: Rand McNally. Weinstein, C. (1988). Assessment and training of student learning strategies. In R. R. Schmeck (Ed.), Learning strategies and learning styles (pp. 291–316). New York, NY: Plenum Press. Weinstein, C., Acee, T. W., & Jung, J. (2011). Self-regulation and learning strategies. New Directions for Teaching and Learning, 126, 45–53. Wolters, C. A. (2003). Regulation of motivation. Educational Psychologist, 38(4), 189–205. Woods, D., & Fassnacht, C. (2007). Transana 2.20 [Computer software]. Madison, WI: The Board of Regents of the University of Wisconsin System. Zimmerman, B. (2000). Attaining self-regulation. In M. Boekaerts, P. Pintrich, & M. Zeidner (Eds.), Handbook of self-regulation (pp. 13–39). San Diego, CA: Academic Press. Acknowledgments. ADEQUA was funded by grants from the German Research Foundation (DFG; grant no. FI 684/13-1 and -2. At various stages, Eva Wilden, Marc Smasal, and Sylvia Fehling were members of the ADEQUA research group. The project was magnificently supported by undergraduate, graduate, and doctoral students: Anne Berger, Björn Bierschenk, Marie-Louis Freyberg, Jennifer Friedrich, Sebastian Groll, Kristina Joppich, Ruth Kamin, Regina Kesting, Sven-Nils Körner, Marcus Kourdji, Nadine Merkator, Agnes Olson, Jennifer Rehling, Peer-Rouven Reich, Delia Uhlenbrock. Correspondence regarding this article should be directed to Peter H. Ludwig, Institut für Bildung im Kindes-und Jugendalter, Fachbereich 5: Erziehungswissenschaften, Universität Koblenz-Landau, Campus Landau, August-Croissant-Str. 5, D-76829 Landau, Germany. E-mail: [email protected]