Understanding the Impact of Guiding Inquiry: The Relationship between Directive Support, Student Attributes, and Transfer of Knowledge, Attitudes, and Behaviours in Inquiry Learning
Saved in:
| Title: | Understanding the Impact of Guiding Inquiry: The Relationship between Directive Support, Student Attributes, and Transfer of Knowledge, Attitudes, and Behaviours in Inquiry Learning |
|---|---|
| Language: | English |
| Authors: | Roll, Ido (ORCID |
| Source: | Instructional Science: An International Journal of the Learning Sciences. Feb 2018 46(1):77-104. |
| Availability: | Springer. 233 Spring Street, New York, NY 10013. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-348-4505; e-mail: service-ny@springer.com; Web site: http://www.springerlink.com |
| Peer Reviewed: | Y |
| Page Count: | 28 |
| Publication Date: | 2018 |
| Document Type: | Journal Articles Reports - Research |
| Descriptors: | Inquiry, Active Learning, Discovery Learning, Physics, Electronics, Equipment, Simulation, Learning Processes, Outcomes of Education, Comparative Analysis, Guidance, Achievement Gains, Student Attitudes, Attitude Change, Instructional Effectiveness |
| DOI: | 10.1007/s11251-017-9437-x |
| ISSN: | 0020-4277 |
| Abstract: | Guiding inquiry learning has been shown to increase knowledge gains. Yet, little is known about the effect of guidance on attitudes and behaviours, its interaction with student attributes, and transfer of impact once guidance is removed. We address these gaps in the context of an interactive Physics simulation on electric circuits (https://phet.colorado.edu/en/simulation/circuit-construction-kit-dc). 49 students in the Non-Directive condition received a set of goals to focus their inquiry, in addition to implicit support built into the simulation. 48 students in the Directive condition received, in addition to these, also detailed directions and prompts. Log-file analysis found that directive support led to more formal testing and less exploration. Clustering identified two groups of learners: one with higher incoming knowledge (Higher Knowledge), the other with higher incoming perceptions of competence and control (Higher PoCC). Working with the simulation improved knowledge and attitudes across cluster groups, so that prior differences all but disappeared. With regard to guidance, adding directive support improved knowledge gains for the Higher Knowledge group, yet suppressed their attitudinal growth. The same support had no effect on knowledge gains for the Higher PoCC group, yet it boosted their attitudinal growth. A transfer activity, where directive support was no longer available, found that impact on attitudes carried forward, yet impacts on behaviours and knowledge were short-lived. Overall, the study highlights the complex interaction between guidance and student attributes. For some, supporting short-term knowledge gains may inadvertenly lead to longer term negative impact on attitudes towards inquiry. |
| Abstractor: | As Provided |
| Number of References: | 76 |
| Entry Date: | 2018 |
| Accession Number: | EJ1170537 |
| Database: | ERIC |
|
Full text is not displayed to guests.
Login for full access.
|
|
| FullText | Links: – Type: pdflink Url: https://content.ebscohost.com/cds/retrieve?content=AQICAHj0k_4E0hTGH8RJwT4gCJyBsGNe_WN95AvKlDbXJGqwxwFB_wxZbnySK9LV7fofTuS6AAAA4jCB3wYJKoZIhvcNAQcGoIHRMIHOAgEAMIHIBgkqhkiG9w0BBwEwHgYJYIZIAWUDBAEuMBEEDKwUWumJOXYdRc4bPAIBEICBmgt5KBq-IKxtQ-IO2TiuI19vPCmh3KStnC5EGlQoji4TJ3hkEOI3eBW_I1OSwusm17hUZ9hJQAduRGx-xJC5XNsSHjEAErg5t6PX3Mvq3mWENW2_AnmrLUyy9rIVJDVMF0pBSP0gMozqOYA_u6raDzL3TLb7Dg1m-T0tqks5RmC86fEvW7KKkXfN2Byr30zi9HM7LQ7e6r4Mw_w= Text: Availability: 1 Value: <anid>AN0128149542;isl01feb.18;2018Feb24.11:14;v2.2.500</anid> <title id="AN0128149542-1">Understanding the impact of guiding inquiry: the relationship between directive support, student attributes, and transfer of knowledge, attitudes, and behaviours in inquiry learning </title> <p>Guiding inquiry learning has been shown to increase knowledge gains. Yet, little is known about the effect of guidance on attitudes and behaviours, its interaction with student attributes, and transfer of impact once guidance is removed. We address these gaps in the context of an interactive Physics simulation on electric circuits (&lt;ext-link&gt;https://phet.colorado.edu/en/simulation/circuit-construction-kit-dc&lt;/ext-link&gt;). 49 students in the Non-Directive condition received a set of goals to focus their inquiry, in addition to implicit support built into the simulation. 48 students in the Directive condition received, in addition to these, also detailed directions and prompts. Log-file analysis found that directive support led to more formal testing and less exploration. Clustering identified two groups of learners: one with higher incoming knowledge (Higher Knowledge), the other with higher incoming perceptions of competence and control (Higher PoCC). Working with the simulation improved knowledge and attitudes across cluster groups, so that prior differences all but disappeared. With regard to guidance, adding directive support improved knowledge gains for the Higher Knowledge group, yet suppressed their attitudinal growth. The same support had no effect on knowledge gains for the Higher PoCC group, yet it boosted their attitudinal growth. A transfer activity, where directive support was no longer available, found that impact on attitudes carried forward, yet impacts on behaviours and knowledge were short-lived. Overall, the study highlights the complex interaction between guidance and student attributes. For some, supporting short-term knowledge gains may inadvertenly lead to longer term negative impact on attitudes towards inquiry.</p> <p>Inquiry learning, discovery-based learning; Exploratory learning environments; Interactive simulations; Assistance dilemma; Scaffolding; Transfer</p> <p>Who benefits from guiding inquiry? The interaction between directive support and learners’ transfer of knowledge, attitudes, and behaviours in inquiry learning.</p> <p>How will you look for [the definition of virtue], Socrates, when you do not know at all what it is? How will you aim to search for something you do not know at all? (Plato, Meno, 80D).</p> <p></p> <hd id="AN0128149542-2">Introduction</hd> <p>Over the last several decades, the premise of inquiry learning has been the focus of a much-heated debate. A main point of contention involves the amount of direction that is provided to students, often referred to as “the assistance dilemma” (Koedinger and Aleven [<reflink idref="bib38" id="ref1">38</reflink>] ). In Mano’s paradox of learning, quoted above, Plato identifies the challenge of learning a domain which one does not already know. Similarly, proponents of the so-called “instructionist” camp argue that lacking domain knowledge, students often fail to apply productive learning strategies in order to acquire the very knowledge that they are lacking (Kirschner et al. [<reflink idref="bib35" id="ref2">35</reflink>] ; Mayer [<reflink idref="bib47" id="ref3">47</reflink>] ). However, proponents of the so-called “constructivist” camp advocate for the value of providing space and time for learners to make sense and construct their own learning experiences within an environment. Appropriate support from this perspective should provide learners with the agency to apply scientific approaches and manage their learning (Cunningham and Duffy [<reflink idref="bib18" id="ref4">18</reflink>] ). Furthermore, in addition to learning about the domain, it is argued that students learn the value and habits of scientific inquiry (Hmelo-Silver et al. [<reflink idref="bib29" id="ref5">29</reflink>] ).</p> <p>Current research on instructional support seeks to find the right fit between type of support, learner characteristics, and desired learning outcomes within a given context (Wise and O’Neill [<reflink idref="bib71" id="ref6">71</reflink>] ; Koedinger and Aleven [<reflink idref="bib38" id="ref7">38</reflink>] ; Koedinger et al. [<reflink idref="bib39" id="ref8">39</reflink>] ). In his excellent critique on the subject, Klahr ([<reflink idref="bib36" id="ref9">36</reflink>] ) warns from the ideological nature of this debate and calls for more evidence to be presented. Specifically, Klahr challenges proponents of constructivist instruction to define a set of instructional goals, instruction and assessment (Klahr [<reflink idref="bib36" id="ref10">36</reflink>] ).</p> <p>In the present research, we accept Klahr’s challenge by evaluating the impact of a specific type of support, detailed directions, on post-secondary students’ knowledge, attitudes, and behaviours in an inquiry learning activity in the context of interactive simulations. Our study seeks to answer the following research questions, within the context of inquiry learning using interactive simulations: (i) What is the impact of directive support on students’ knowledge, attitudes and behaviours? (ii) How does directive support interact with students’ incoming knowledge and attitudes to affect learning? (iii) How does the impact of directive support on students’ knowledge and attitudes transfer to a subsequent learning activity once directive support is removed?</p> <p>We begin by reviewing the literature on guidance in inquiry learning, focusing on the three questions raised above. We then describe the interactive simulation environment which was used in the current study and our methodological choices. Finally, we outline key findings and discuss their contribution to our understanding of the impact of different kinds of support on students’ engagement and learning.</p> <hd id="AN0128149542-3">Background: guidance in scientific inquiry learning</hd> <p>Inquiry-based learning has been shown to facilitate learning of scientific principles and methods (Lazonder and Harmsen [<reflink idref="bib42" id="ref11">42</reflink>] ). It also provides students opportunities to practice scientific processes, which may be beneficial in its own right (Hmelo-Silver et al. [<reflink idref="bib29" id="ref12">29</reflink>] ; de Jong and van Joolingen [<reflink idref="bib21" id="ref13">21</reflink>] ). Yet, to achieve these potential benefits, research has provided conclusive evidence that students are in need of support (D’Angelo et al. [<reflink idref="bib19" id="ref14">19</reflink>] ; Alfieri et al. [<reflink idref="bib2" id="ref15">2</reflink>] ; Kirschner et al. [<reflink idref="bib35" id="ref16">35</reflink>] ).</p> <p>One common type of support is guidance. Lazonder and Harmsen ([<reflink idref="bib42" id="ref17">42</reflink>] ) define guidance as “any form of assistance offered before and/or during the inquiry learning process that aims to simplify, provide a view on, elicit, supplant, or prescribe the scientific reasoning skills involved” (p. 687). Here we look at a specific type of guidance, one that directs students towards prescribed effective inquiry paths by providing them with detailed instructions and task breakdown. This recipe-style guidance has several potential benefits. First, it can direct students towards effective exploration, hence increasing the efficiency of the learning process. Second, it can segment the inquiry process, making it more manageable to keep track by the teacher, the system, or the learners themselves. However, over-directive support may reduce students’ capacity for independent learning in future scenarios. In addition to learning the relevant topics, students should learn to regulate their engagement in future situations, in the absence of detailed guidance. Thus, literatures on scaffolding and self-regulated learning have identified the importance of instruction that structures support to bridge from guiding learning to fostering independence (Butler et al. [<reflink idref="bib14" id="ref18">14</reflink>] ; Roll et al. [<reflink idref="bib56" id="ref19">56</reflink>] ; Wood [<reflink idref="bib72" id="ref20">72</reflink>] ; Zohar and Barzilai [<reflink idref="bib76" id="ref21">76</reflink>] ).</p> <hd id="AN0128149542-4">Who benefits from what support?</hd> <p>Learner attributes such as prior knowledge, experiences, beliefs, and attitudes, interact with instruction in ways that influence learning (Butler and Cartier [<reflink idref="bib12" id="ref22">12</reflink>] ; Butler et al. [<reflink idref="bib13" id="ref23">13</reflink>] ). Thus, instruction and support should match learners and their contexts. Several studies have evaluated the differential effect of guidance based on student knowledge level. One intuitive patterns, as identified by the majority of the scaffolding literature, suggests that novices benefit from more directive support (Wood et al. [<reflink idref="bib73" id="ref24">73</reflink>] ). In short, this view argues for increased guidance in early stages of learning. Guidance should then fade, as learners acquire expertise. Tuovinen and Sweller ([<reflink idref="bib64" id="ref25">64</reflink>] ) explain this by suggesting that guidance acts as a substitute for missing schemas. This fits well with a Vygotskian view of learning, as novices require mediation to help them tackle learning tasks within their zone of proximal development (Vygotsky [<reflink idref="bib67" id="ref26">67</reflink>] ). Indeed, several studies found that novice learners benefit more from higher levels of guidance compared with expert learners (termed Expertise Reversal Effect; cf. Kalyuga [<reflink idref="bib31" id="ref27">31</reflink>] ).</p> <p>Interestingly, some evidence suggests the contrary. According to this literature, detailed guidance is effective mainly when students have relevant mental models to comprehend the given directions and learn from them. Yet, in many cases, novice learners lack the right operators to make sense of the directions, or they are only able to follow them shallowly. Guidance itself may be a victim of the same problem it comes to solve, namely, insufficient prior knowledge. Making sense of guidance is often demanding, particularly when learners lack the basic operators of the domain (Koedinger and Roll [<reflink idref="bib41" id="ref28">41</reflink>] ). For example, van Dijk et al. ([<reflink idref="bib65" id="ref29">65</reflink>] ) showed that in an interactive simulation environment, novice learners were not able to make sense of directions provided. Similarly, Brenner et al. ([<reflink idref="bib11" id="ref30">11</reflink>] ) found that detailed directions hurt learning for novices who were actively managing their learning process (Brenner et al. [<reflink idref="bib11" id="ref31">11</reflink>] ). Working on Geometry problem solving, Roll et al. ([<reflink idref="bib57" id="ref32">57</reflink>] ) found that detailed directions (in the form of hints) hurt student learning of new topics, but assisted the same students to learn the same topics once students became moderately competent. Another limitation of detailed guidance is that it encourages learners to pursue sub-goals that were defined by the activity design, and may not be the goals that they need to comprehend first. Indeed, the literature on goal specificity shows that letting students set their own goals is more effective than setting these goals externally (Miller et al. [<reflink idref="bib49" id="ref33">49</reflink>] ; Sweller et al. [<reflink idref="bib63" id="ref34">63</reflink>] ).</p> <p>Within the context of inquiry-based leaning, the impact of guidance depends on more than merely prior knowledge (Brenner et al. [<reflink idref="bib11" id="ref35">11</reflink>] ). In their study, Brenner et al. provided students with an inquiry-learning activity using a biology simulation. They clustered learners based on interactions during the learning process. Among novice learners (that is, those who enter the study with low prior knowledge), one cluster of learners specifically benefited from lower levels of guidance. Learners in this cluster were the most active in the system, even though most of their actions were unproductive. However, novice learners who were less active in the system benefited from increased levels of guidance. These results suggest that what matters to learning may be how students engage in meaning-making, rather than how accurate their initial attempts are in generating effective solutions. If this is the case, very specific directions may not be necessary to foster effective forms of learning, even for novice learners. Consistent with this interpretation, in a meta-review, Loibl et al. ([<reflink idref="bib44" id="ref36">44</reflink>] ) looked at learning from low-guidance activities as a preparation for subsequent instruction (termed Productive Failure). Their review suggests that failing due to lacking knowledge may be productive, as long as students have opportunities to make sense of their failures. Mathan and Koedinger ([<reflink idref="bib45" id="ref37">45</reflink>] ) provide one clear example of this approach. In their study, they found that learners who studied how to use a spreadsheet program benefited more from opportunities (and support) to make sense of their own mistakes, rather than when feedback was given to them upon performance of an error (Mathan and Koedinger [<reflink idref="bib45" id="ref38">45</reflink>] ).</p> <p>Notably, guidance is not the only form of support. In all of these scenarios, while guidance was limited, students were supported in a variety of alternative ways. For example, scientific inquiry environments typically offer well-designed constraints and affordances which implicitly encourage students to pursue productive paths (Wieman et al. [<reflink idref="bib68" id="ref39">68</reflink>] ). In addition, these simulations provide situational feedback, in which students’ actions are met with authentic reaction by the system (Nathan [<reflink idref="bib50" id="ref40">50</reflink>] ). Such situational feedback may provide better support than detailed guidance if it helps students to interpret the outcomes of their actions (Roll et al. [<reflink idref="bib58" id="ref41">58</reflink>] ; Mathan and Koedinger [<reflink idref="bib45" id="ref42">45</reflink>] ; Wiese and Koedinger [<reflink idref="bib69" id="ref43">69</reflink>] ). As online simulations typically offer such situational feedback, they may provide a context in which some students may benefit more from lower levels of explicit guidance.</p> <hd id="AN0128149542-5">Impact of guidance on attitudes</hd> <p>Productive attitudes towards learning are instrumental in supporting learning (Bandura et al. [<reflink idref="bib5" id="ref44">5</reflink>] ; Schunk et al. [<reflink idref="bib60" id="ref45">60</reflink>] ). One construct of interest is self-efficacy, that is, learners’ perceptions of their ability to perform a specific task in relation to a given goal (Bandura [<reflink idref="bib4" id="ref46">4</reflink>] ; Linnenbrink and Pintrich [<reflink idref="bib43" id="ref47">43</reflink>] ). Self-efficacy is situated and contextual. For example, it can describe learners’ beliefs about figuring out specific scientific principles in a specific virtual simulation. Positive self-efficacy can help students maintain motivation and persist when a task becomes challenging (Schunk et al. [<reflink idref="bib60" id="ref48">60</reflink>] ), and encourages them to apply effortful cognitive and metacognitive strategies (Linnenbrink and Pintrich [<reflink idref="bib43" id="ref49">43</reflink>] ).</p> <p>Another related construct is success attributions, where students can point to specific aspects of task and learning that make them successful (Graham and Williams [<reflink idref="bib28" id="ref50">28</reflink>] ). Students who associate success with factors within their control, such as use of particular strategies, build self-efficacy and are more likely to continue engaging in learning. Value expectancy is also important to learning, because it reflects whether students feel a task is worthwhile and interesting (Wigfield and Eccles [<reflink idref="bib70" id="ref51">70</reflink>] ). Goal orientation is another factor affecting learning (Linnenbrink and Pintrich [<reflink idref="bib43" id="ref52">43</reflink>] ): students who focus on mastering the content or the processes (mastery goals), as opposed to demonstrating competence (performance goals), tend to believe that ability can be changed and thus work to improve it (Dweck and Master [<reflink idref="bib23" id="ref53">23</reflink>] ), demonstrating more persistence.</p> <p>The impact of inquiry learning, and guidance within inquiry learning, on attitudes, is an interesting topic. Literature suggests that inquiry activities positively impact students’ attitudes towards science and inquiry (Chen and Howard [<reflink idref="bib17" id="ref54">17</reflink>] ; Nomme et al. [<reflink idref="bib51" id="ref55">51</reflink>] ). However, the impact of directive support on attitudes has been understudied (Belland et al. [<reflink idref="bib8" id="ref56">8</reflink>] ). Belenky and Nokes-Malach ([<reflink idref="bib6" id="ref57">6</reflink>] , [<reflink idref="bib7" id="ref58">7</reflink>] ) studied the effect of directive support when situational feedback is built into the design of the activity. They found that low levels of guidance improved students’ mastery goals, and subsequently, increased transfer.</p> <p>One explanation for the positive impact of low levels of directive guidance on attitudes may be agency (Zimmerman [<reflink idref="bib75" id="ref59">75</reflink>] ; Podolefsky et al. [<reflink idref="bib54" id="ref60">54</reflink>] ; Deci and Ryan [<reflink idref="bib22" id="ref61">22</reflink>] ). Agency can be defined as “the capacity to exercise control over one’s own thought processes, motivation, and action” (Bandura [<reflink idref="bib3" id="ref62">3</reflink>] , p. 1175). Fostering a sense of agency helps students take ownership of the learning process, and thus amplifies the behavioural, cognitive, and attitudinal benefits of positive self-efficacy (Butler et al. [<reflink idref="bib15" id="ref63">15</reflink>] ). Providing directive guidance may reduce students’ agency in the environment, as the learning goals and processes are externally defined. Indeed, a study by Sawyer et al. ([<reflink idref="bib59" id="ref64">59</reflink>] ) demonstrate the complex relationship between guidance, knowledge, and attitudes. They used directive guidance to manipulate student agency, and found that while a higher level of directive guidance improved domain learning outcomes, it also led students to perform more unproductive behaviours in the environment, likely because students adopted an answer-hunting approach rather than authentic inquiry.</p> <hd id="AN0128149542-6">Impact of guidance on transfer</hd> <p>Some evidence suggests that instructional practices that have better outcomes on short-term measures of learning also show better results on measures of far transfer. For example, Klahr and Nigam ([<reflink idref="bib37" id="ref65">37</reflink>] ), as well as Strand-Cary and Klahr ([<reflink idref="bib62" id="ref66">62</reflink>] ), advocate for “path independence”, arguing that it is enough to maximize short-term benefits, as robust learning follows similar patterns. Consistent with this view, several studies found that increasing guidance in inquiry activities leads to superior performance on measures of both application and transfer (Klahr and Nigam [<reflink idref="bib37" id="ref67">37</reflink>] ; Strand-Cary and Klahr [<reflink idref="bib62" id="ref68">62</reflink>] ; Matlen and Klahr [<reflink idref="bib46" id="ref69">46</reflink>] ). Holmes et al. ([<reflink idref="bib30" id="ref70">30</reflink>] ) provide another example of improved performance on transfer measures when more directive guidance was available in inquiry learning.</p> <p>However, Kapur ([<reflink idref="bib32" id="ref71">32</reflink>] ) shows that robust learning does not require immediate success. For example, letting novices explore a problem space seems to help them acquire more flexible knowledge that they can later apply more adaptively in future activities (Rittle-Johnson et al. [<reflink idref="bib55" id="ref72">55</reflink>] ). In contrast, increasing guidance by providing detailed directions during the exploratory phase may put students in an “answer hunting” mode which may be counterproductive (McDaniel and Schlager [<reflink idref="bib48" id="ref73">48</reflink>] ). Particularly notable is Strand-Cary and Klahr’s ([<reflink idref="bib62" id="ref74">62</reflink>] ) finding that the benefits associated with more directive forms of support during early learning may wear off with time. Perhaps forms of support that lead students to particular learning outcomes do not challenge their existing thought processes enough to facilitate robust knowledge and successful transfer (Kapur and Bielaczyc [<reflink idref="bib33" id="ref75">33</reflink>] ).</p> <p>Overall, this body of work suggests that, in order to evaluate the impact of guidance, it is essential to investigate not only immediate learning, but also ability to mobilize knowledge flexibly in transfer activities. Bransford and Schwartz ([<reflink idref="bib10" id="ref76">10</reflink>] ) make an especially compelling case for assessments that focus on subsequent, or future, learning. Overall, the impact of directive guidance on transfer of knowledge in inquiry learning warrants additional attention.</p> <p>A related question focuses on transfer of strategies, rather than knowledge. Detailed guidance prescribes certain paths and patterns that should be followed during the learning process. Indeed, studies show that students apply the prescribed strategies while support is in effect (de Jong and van Joolingen [<reflink idref="bib21" id="ref77">21</reflink>] ; Lazonder and Harmsen [<reflink idref="bib42" id="ref78">42</reflink>] ; Gobert et al. [<reflink idref="bib27" id="ref79">27</reflink>] ). Ideally, learners should internalize these paths and transfer them to isomorphic activities. Thus, an important question is whether students transfer the prescribed inquiry behaviours to new activities. In one example, Biswas et al. (Biswas et al. [<reflink idref="bib9" id="ref80">9</reflink>] ) found that learners were able to transfer supported behaviours once support was removed. However, their support was implicitly embedded in the learning environment and did not take the form of external guidance. Roll et al. ([<reflink idref="bib56" id="ref81">56</reflink>] ) found that in a problem-solving environment, the impact of directive guidance on help-seeking strategies transferred to new topics, but only after it was given over several months and across topics. Thus, the impact of detailed guidance on learners’ engagement in scientific inquiry needs further investigation.</p> <hd id="AN0128149542-7">Inquiry learning in virtual labs and interactive simulations</hd> <p>The review above highlights the importance of looking in a more nuanced way at the impact of directive guidance on inquiry learning. We do so in the context of interactive science simulations (de Jong [<reflink idref="bib20" id="ref82">20</reflink>] ; Wieman et al. [<reflink idref="bib68" id="ref83">68</reflink>] ). Also termed virtual labs (Yaron et al. [<reflink idref="bib74" id="ref84">74</reflink>] ) or microworlds (Gobert [<reflink idref="bib26" id="ref85">26</reflink>] ), these are open ended environments in which learners can explore different topics with the goal of understanding the underlying models that govern the behaviour of the simulations. Interactive simulations offer several advantages for learning over their real-world equivalence (in addition to low barriers in terms of budget and equipment). First, they visually represent physical elements that are otherwise obscure, such as electrons flowing through electric circuits. Second, they allow learners to change the flow of time (fast forward, slow down, or bring to a halt) and examine their experimentations carefully. Third, they allow learners to engage in experimentations that would otherwise be unsafe or impossible. Last, done right, they eliminate extraneous cognitive load and help learners focus on key aspects of the domain. The outcome is a meaningful learning experience in which learners have opportunities to engage with core concepts, set their own questions for exploration, and compare and contrast scenarios to make sense of the topic to be investigated.</p> <p>Benefits of interactive simulations for immediate learning and transfer have been identified in several studies. For example, students who learned with an interactive simulation outperformed their counterparts who learned in a physical lab using the exact same activity, even when the transfer test was done in the physical lab (Finkelstein et al. [<reflink idref="bib25" id="ref86">25</reflink>] ).</p> <p>Interactive simulations are environments in which learners are supported using carefully designed affordances for sense-making (Paul et al. [<reflink idref="bib52" id="ref87">52</reflink>] ; de Jong and van Joolingen [<reflink idref="bib21" id="ref88">21</reflink>] ). By including certain controls (and not others), learners are invited to explore certain aspects of the domain and make sense of these. Situational feedback (Nathan [<reflink idref="bib50" id="ref89">50</reflink>] ; Roll et al. [<reflink idref="bib58" id="ref90">58</reflink>] ) is provided by showing learners the outcomes of their actions (Wieman et al. [<reflink idref="bib68" id="ref91">68</reflink>] ). Overall, these affordances direct learners’ attention implicitly to core ideas in the domain, and enable succinct and iterative experimentation-analysis cycles that can inform learning. Furthermore, constraints of these environments prevent extraneous cognitive load that is introduced by more open-ended environments (such as physical labs; Finkelstein et al. [<reflink idref="bib25" id="ref92">25</reflink>] ). Thus, these environments may offer conditions for rich forms of inquiry that best support learning without extensive directive guidance (Adams et al. [<reflink idref="bib1" id="ref93">1</reflink>] ). For example, Vollmeyer et al. ([<reflink idref="bib66" id="ref94">66</reflink>] ) found that learners benefit more from answering their own questions in a Biology environment, compared to being given guidance in the form of specific sub-goals. Indeed, directive guidance was shown to reduce learners’ autonomy and exploration in virtual simulations (Chamberlain et al. [<reflink idref="bib16" id="ref95">16</reflink>] ).</p> <hd id="AN0128149542-8">Research questions</hd> <p>In the present study, we vary the presence of detailed, directive guidance and evaluate its impact on learning. All students in the study received the same focus for inquiry and worked with the same environment (with its situational feedback, affordances and constraints). The presence of detailed directions was manipulated. Students in the Non-Directive condition explored the simulation without additional guidance. Students in the Directive condition were provided with detailed guidance that defined their inquiry process by specifying sub goals, diagrams, and guiding questions, in addition to the more implicit forms of support that were common to all learners.</p> <p>As suggested by the review above, the study was designed to address the following questions: (i) What is the impact of directive support on students’ knowledge, attitudes and behaviours? (ii) How does directive support interact with students’ incoming knowledge and attitudes to affect learning? (iii) How does the impact of directive support on students’ knowledge and attitudes transfer to a subsequent learning activity once directive support is removed?</p> <hd id="AN0128149542-9">Methods</hd> <hd id="AN0128149542-10">Participants</hd> <p>One hundred students from a large Canadian university participated in the study. Half of the students who reported their gender self-identified as females and half as males. All students were near the end of their first year in the Faculty of Science. Students were recruited from two introductory Physics courses. All students were compensated for their time; students from one of the courses also received homework credit for participating in the study. All students had not learnt the study topics in their courses prior to the study. The study took place with groups of 10–20 students in a Physics lab where students used the lab computers. Assignment to conditions was done at random within each group of students.</p> <hd id="AN0128149542-11">Materials</hd> <hd id="AN0128149542-12">Instructional materials</hd> <p>Students in the study used the Circuit Construction Kit PhET simulation (<ulink href="http://phet.colorado.edu/en/simulation/circuit-construction-kit-dc">http://phet.colorado.edu/en/simulation/circuit-construction-kit-dc</ulink>). PhET is a family of over 100 free simulations in different STEM topics, developed at the University of Colorado, Boulder. PhET Simulations are used over 45,000,000 times a year. The simulation used in this study is the most popular simulation in the PhET family.</p> <p>The PhET Circuit Construction Kit simulation supports students as they explore basic properties of DC circuits. It allows students to connect wires, light bulbs, resistors, switches, and measurement instruments on a virtual test bed. Figure 1 shows examples of how learners in our study explored using the PhET CCK Simulation. The simulation provides situational feedback as it adjusts light intensity and visible speed of electrons based on circuit configuration. In other cases, when the power is too high, elements in the circuit may catch on fire. Overall, the simulation has 140 combinations of action (add, test, etc.), component (light-bulb, battery, voltmeter, etc.), and outcome (light intensity changed, voltage changed, etc.). The version of the simulation used in this study logs all student actions.The PhET D/C Circuit Construction Kit (CCK) simulation. Snapshots taken during the study</p> <p>The topic of the first activity was light-bulbs. All students worked with the same simulation and were given the following focus for their inquiry:</p> <p>Use the DC Circuit PhET to explore how voltage, current, and the brightness of light bulbs depend on: 1. The number of light bulbs in a circuit; 2. The arrangement of light bulbs in circuit. For example, a. What happens when several light bulbs are placed in the same loop? (that is, all the electrons move through the same components); b. What happens when light bulbs are placed in different loops? (that is, the electrons move through different loops); c. What happens when circuits combine (a) and (b)?</p> <p>All students also received a short written description of key components of the simulation (e.g., “To begin building circuits, drag the various circuit parts onto the blue work area”).</p> <p>During this activity, students were randomly assigned to one of two conditions. Students in the Non-Directive condition received only the support described above (i.e., focus of inquiry, and affordances, constraints, and situational feedback that are built into the simulation). The other students, in the Directive condition, also received five pages of detailed guidance. This guidance was modeled after current theories of learning, best practices using simulations (as documented on the PhET teacher resources, on the sim webpage), and with consultation with one of the course instructors. Guidance was broken to tasks and included diagrams of the circuits to build, tables of the values to measure, and compare-and-contrast and reflection prompts. Notably, this support included both domain-level directions about which circuits to build and test, as well as metacognitive-level prompts to reflect and extract patterns. Figure 2 shows the first two pages of the detailed guidance provided to students in the Directive group.Guidance given to students in the Directive condition. The materials included five such pages</p> <p>After the first activity and a short break, all students continued to Activity 2, the transfer activity, on the topic of resistors. All students were given the following focus of inquiry:</p> <p>Next we will investigate how resistors affect the behaviour of circuits, using only resistors, batteries, and wires. 1. What happens to the current and voltage when you use resistors with different resistance? 2. Investigate circuits that include multiple resistors with different resistance in a variety of arrangements; 3. Explore the properties of different combinations of resistors with the same resistance?</p> <p>Notably, the topic of Activity 2 is more conceptually challenging than Activity 1, and less familiar from everyday life. Also, when using resistors, situational feedback is less effective than with light-bulbs, as resistors do not emit light. Thus, more explicit testing with measurement tools (i.e., voltmeter or ammeter) is required.</p> <p>During this transfer activity, no directive guidance was offered. Instead, students from both conditions received support only in the form of focus of inquiry and simulation structure. This allowed us to evaluate the impact of directive guidance on transfer. Figure 3 shows the two conditions across the two activities.The study procedure</p> <hd id="AN0128149542-13">Measures</hd> <p>We collected data about students’ behaviours, knowledge, and attitudes.</p> <p>Students’ behaviours were evaluated using the system logs of the simulation. These logs capture each system event (such as a component being added or removed), together with student ID and a time stamp.</p> <p>Students’ knowledge was evaluated using a pre-test prior to Activity 1 and a combined post-test on both topics at the end of Activity 2. All knowledge tests assessed conceptual knowledge and required no calculations. For example, students were asked to rank the voltage and current through elements in different circuits, such as the ones shown in Fig. 4.1 [<reflink idref="bib1" id="ref96">1</reflink>] The post-tests on Activities 1 and 2 included seven and six questions respectively. The pre-test was a subset of the post-test and included the five post-test items that did not have diagrams, so that students could not use their exploration to merely replicate the given test. Notably, all test items (pre and post) allowed for partial credit (for instance, by ranking some, but not all, of the components correctly). The combined post-test was a reliable measure of student’s knowledge, with Cronbach α = 0.75.Example knowledge test items</p> <p>Students’ attitudes were evaluated using surveys, modeled after Butler and Cartier ([<reflink idref="bib12" id="ref97">12</reflink>] ). Two constructs were evaluated throughout the study. These included success attributions and self-efficacy. Success attributions are factors that students recognize as being the determinants of success in a particular context (Graham and Williams [<reflink idref="bib28" id="ref98">28</reflink>] ). These factors reflect the way learners perceive locus of control for their success on an activity (as malleable and within their control, vs. as dependent on external factors). Success attributions in the context of simulation activities were surveyed using the following stem: ‘I think that I will succeed if…’. Self-efficacy describes students’ perceived competencies for learning or acting in context (Bandura [<reflink idref="bib4" id="ref99">4</reflink>] ). Self-efficacy was evaluated using the stem ‘I think that I can do a good job of…’. A tenth item simply asked students to rank ‘I can be successful’. Figure 5 shows all ten survey items.Perception of Competence and Control (PoCC) survey</p> <p>Because attributions and self-efficacy perceptions are highly contextualized, we expected they might shift over time through experiences with the simulation. Thus, to assess shifts in these beliefs, the same survey items were used before the first activity (at time t0), between activities (at time t1), and after the second activity (at time t2). In all cases, students were shown the relevant activity ahead of completing the survey, to ground their responses. Factor analysis found that nine out of the ten items across the two scales comprised a coherent dimension with Cronbach’s α = 0.825. Given the core meaning underlying self-efficacy and attributions, we interpreted them together as reflecting learners’ Perceptions of Competence and Control (PoCC). The only item that was left out was ‘I will succeed if the task is not too difficult’. In contrast, the nine items that comprise the PoCC dimension all reflect factors that are controllable by the learners. Students’ PoCC scores were the average of their responses to the nine items on that specific survey (t0, t1, or t2).</p> <p>To gain a more complete understanding of students’ incoming attitudes, two additional constructs were surveyed at the onset of the study, before Activity 1. These included goal orientation and task value. Two Likert items, with four levels each, asked students to report their mastery goal orientation on a scale of strongly disagree to strongly agree. These items were adapted from Elliot and Murayama ([<reflink idref="bib24" id="ref100">24</reflink>] ): ‘I am striving to understand the physics content as thoroughly as possible’ and ‘My goal is to learn as much as possible’. Students were also asked about their perceived value of working with PhET simulations, using four Likert items with four levels, on a scale of strongly disagree to strongly agree: ‘For me personally, PhET Sims are usually… boring; productive; fun; useless’.</p> <p>Lastly, to get a sense of their prior experience, students were asked three questions about their background: whether they had taken high-school or college-level physics classes; the number of simulations that they had previously worked with; and whether they had previously used the PhET CCK simulation.</p> <p>Table 1 describes the data collected in the study.</p> <hd id="AN0128149542-14">Procedure</hd> <p>Activities in the study took nearly 2 h, as seen in Fig. 3. Students first completed the pre-test (Know<subs>t0</subs>) and pre-survey (incoming attitudes and experiences, including PoCC<subs>t0</subs>, goal orientation, task value, and prior experiences), followed by 25 min of Activity 1, according to their condition. Students then completed the mid-survey (PoCC<subs>t1</subs>), followed by a short break. After the break students completed the transfer Activity 2, the subsequent survey (PoCC<subs>t2</subs>), and the post-test on both topics (Know<subs>t1</subs> and Know<subs>t2</subs>). The tests and surveys were administered online.</p> <hd id="AN0128149542-15">Analysis</hd> <hd id="AN0128149542-16">Grouping students</hd> <p>In order to examines the interaction between directive support and student attributes, we first clustered students based on incoming attitudinal and knowledge measures. This serves a dual purpose: Firstly, learners often have multidimensional profiles that combine in coherent ways. These profiles often cut across cognition and attitudes. Clustering students can uncover deeper patterns that go beyond any single dimension and that otherwise remain latent (Butler et al. [<reflink idref="bib13" id="ref101">13</reflink>] ). For example, when explaining the impact of guidance on learning, Brenner et al. ([<reflink idref="bib11" id="ref102">11</reflink>] ) found that clustering learners surfaced patterns that could not be found by controlling for prior-knowledge alone. Secondly, from a statistical perspective, evaluating the relationship between condition, learning, student attributes, and all of their 2-, 3-, and 4-way interactions is impractical.</p> <p>We clustered learners based on all four measures of incoming knowledge and attitude: PoCC, incoming knowledge level, mastery orientation, and perceived value of working with simulations. All of these measures were collected at time t0, prior to working with the simulation.</p> <p>We applied Two-Step clustering to these four factors. This advanced clustering approach consistently provides better outcomes than k-means or traditional hierarchical approaches in the social sciences (Kent et al. [<reflink idref="bib34" id="ref103">34</reflink>] ). Also in our case, Two-step clustering provided better BIC values than other clustering approaches.</p> <hd id="AN0128149542-17">Analysis of behaviours</hd> <p>Each student action in the simulation was coded as one of four types, based on the interface element that was used: Construct (e.g., adding resistors, connecting wires, splitting junctions, etc.); Test (e.g., using the testing instruments to conduct measurements); Pause (i.e., not engaging with the sim for longer than 15 s2 [<reflink idref="bib2" id="ref104">2</reflink>]), and Reset (i.e., removing all components form the testbed and starting from scratch). We counted each action individually. For example, adding four wires in a row was considered four separate actions. Such frequency analysis can surface high level patterns in the focus of students’ work. A high frequency of Construct suggests that students focus on building and revising circuits, and possibly informally evaluating these using the sim’s situational feedback. Focus on Test suggests more formal testing, as the testing instruments require deliberate use. Focus on Pauses suggests more planning and reflection, and was associated with more productive learning (cf. Perez et al. [<reflink idref="bib53" id="ref105">53</reflink>] ). Interface actions (such as zooming) and actions that did not change the structure of the circuits or the testing instruments were ignored. A MANOVA with action category as a function of Condition, Cluster, and their interaction as independent variables, was used to evaluate the impact of support on students’ actions. A similar MANVOA evaluated the transfer of the impact on behaviours during Activity 2.</p> <hd id="AN0128149542-18">Analyzing knowledge and attitudes</hd> <p>To evaluate the effect of support on knowledge and PoCC while support was in effect a 2-way MANCOVA was used with Know<subs>t1</subs> and PoCC<subs>t1</subs> as dependent measures, and Condition (Directive/Non-Directive), Cluster (as described below), and their interaction as independent variables. Notably, Cluster is a measure of group, not of individuals. Thus, the MANCOVA model also controls for PoCC<subs>t0</subs> and Know<subs>t0</subs>. When results warranted additional analysis, we followed up with separate ANCOVAs for knowledge and PoCC. When interaction terms were significant, we split the data by Cluster and evaluated the effect of Condition within each group using one-way ANCOVAs.</p> <p>A similar analysis was conducted at t2 to evaluate transfer of knowledge and attitudes once Directive support was removed.</p> <hd id="AN0128149542-19">Results</hd> <p>We begin by describing participants incoming knowledge, attitudes, and background, and how this information was used to cluster students to groups. We then describe learning while the different kinds of support were in effect (Activity 1) and in the transfer activity (Activity 2). Within t1 and t2, we first analyze students’ actions within the simulation, followed by analysis of knowledge and PoCC.</p> <hd id="AN0128149542-20">Incoming knowledge and attitudes</hd> <p>Three students received perfect scores on the pre-test and were removed from further analysis. Analysis thus included 49 and 48 students in the Non-Directive and Directive conditions respectively. There were no differences between conditions with regard to incoming knowledge or attitudes. M(SD) for Know<subs>t0</subs>: Directive: 0.47(0.17); Non-Directive: 0.47(0.18); t(<reflink idref="bib96" id="ref106">96</reflink>) = 0.09, p &gt; 0.9. M(SD) for PoCC<subs>t0</subs>: Directive: 2.76(0.47); Non-Directive: 2.85(0.47); t(<reflink idref="bib96" id="ref107">96</reflink>) = 0.99, p &gt; 0.3.</p> <p>A main goal of this study was to examine differentiated effects of guidance on students with diverse incoming attributes, namely, knowledge level, PoCC, goal orientation, and task value. To better understand the relationships between these variables, we first calculated the correlations between them. Table 2 shows that PoCC at time t0 is associated with higher perceived value towards simulations and higher mastery orientation. Other correlations, especially between knowledge and attitudinal factors, are fairly low.</p> <p>Applying two-step clustering using these four variables suggested that students could be split to two groups, with 31 and 66 students. The clustering obtained a silhouette score of 0.4, meaning that each student is a good fit within that cluster and a poor fit to the other cluster. Figure 6 shows the distributions of each variable within cluster. Students in the smaller cluster entered the activity with higher scores in domain knowledge assessment, but lower reported PoCC towards learning through simulations (see also Fig. 6). Because the strength of this group tends towards domain knowledge, we subsequently refer to this group as “Higher Knowledge”. Cohen’s d for Know<subs>t0</subs> shows that students in this cluster performed 2.1 Cohen’s d effect sizes better than students in the larger cluster. Students in the other, larger, cluster entered the activity with lower scores in prior knowledge, but reported higher PoCC towards their own capacity to learn through simulations (PoCC; Cohen’s d = 1.1). Since this group’s strength leans towards attitudes, we refer to this group as “Higher PoCC”. Students in the Higher PoCC group also reported higher mastery goal orientation and higher confidence in the sim’s ability to support their learning (perceived value), though to a lesser degree.Distribution of students on the clustering variables</p> <p>It is important to emphasize that these are characteristics of overall clusters, and not for all individuals within the clusters. To account for individual differences, the analysis also controlled for individual values, as explained above. Also, the clusters capture relationships between multiple variables. Naming them based on a single variable is made only for descriptive purposes.</p> <p>The data suggests that students’ prior experiences may have contributed to the difference in attitudes. Students in the Higher PoCC cluster reported using more sims in the past (though not this specific simulation) and being more comfortable using them. Thus, students’ prior experiences with simulations may have led them to have higher confidence in their ability to learn with them (PoCC) and trust the simulation to be a beneficial learning aid (perceived value). However, prior experiences do not explain differences in incoming knowledge, as students in both clusters reported similar experiences with high-school physics.</p> <p>Significantly for further analyses, students in both clusters were split nearly evenly across conditions, as shown in Table 3. Note that due to the small number of students in the Higher Knowledge cluster (<reflink idref="bib31" id="ref108">31</reflink>), post hoc analysis for this cluster had very limited statistical power.</p> <hd id="AN0128149542-21">Effect of directive support while present</hd> <hd id="AN0128149542-22">Effect on actions</hd> <p>Figure 7 shows the impact of different kinds of support on students’ behaviours while working through Activity 1. Impact of Support and Cluster on actions was evaluated using a 2-way MANOVA with frequency of Build, Pause, Reset, and Test as DV, and Condition, Cluster, and their interaction as IV. The MANOVA showed a significant main effect for Condition: F(<reflink idref="bib4" id="ref109">4</reflink>,<reflink idref="bib89" id="ref110">89</reflink>) = 9.11, p &lt; 0.0001, η<sups>2</sups> = 0.29. Cluster and its interaction with Condition were not significant, p &gt; 0.7 for either. Individual ANOVAs showed that students in both conditions were equally likely (and mainly, unlikely) to pause and reset. Regarding the more common actions, students in the Directive condition were much more likely to Test: F(<reflink idref="bib1" id="ref111">1</reflink>,<reflink idref="bib92" id="ref112">92</reflink>) = 27.6, p &lt; 0.0005, η<sups>2</sups> = 0.23, while students in the Non-Directive condition were much more likely to build, F(<reflink idref="bib1" id="ref113">1</reflink>,<reflink idref="bib92" id="ref114">92</reflink>) = 8.8, p = .004, η<sups>2</sups> = .088.Students actions in the simulation during Activity 1</p> <hd id="AN0128149542-23">Student knowledge</hd> <p>Both post-tests (Know<subs>t1</subs> and Know<subs>t2</subs>) were much more challenging than the pre-test (Know<subs>t0</subs>), and thus raw scores cannot be compared across tests. Instead, analysis of identical items between Know<subs>t0</subs> and the combined Know<subs>t1,t2</subs> showed significant learning across conditions, from 0.47 (0.17) to 0.62 (0.23), t(<reflink idref="bib96" id="ref115">96</reflink>) = .61 p = &lt;0.0005, Cohen’s d = 0.75. This result corresponds to a large effect size.</p> <p>Figure 8 shows the effect of directive guidance on students’ knowledge and attitudes in the first activity. MANCOVA of Know<subs>t1</subs> and POCC<subs>t1</subs> as DV, and Cluster, Condition, and their interaction as IV, controlling for Know<subs>t0</subs> and PoCC<subs>t0</subs>, was significant, warranting individual ANOVAs: F(<reflink idref="bib2" id="ref116">2</reflink>, 90) = 3.96, p = .023, η<subs>p</subs><sups>2</sups> = .081.Knowledge and attitudes following Activity 1</p> <p>ANCOVA with Know<subs>t1</subs> as a DV found a marginally-significant Condition X Cluster interaction: F(<reflink idref="bib1" id="ref117">1</reflink>, 91) = 3.56, p = .054, η<sups>2</sups> = .040. To interpret the interaction, we evaluated the impact of Condition for each cluster separately. Students in the Higher Knowledge cluster benefited significantly more from Directive support, compared with their peers in the Non-Directive condition: F(<reflink idref="bib1" id="ref118">1</reflink>,<reflink idref="bib27" id="ref119">27</reflink>) = 4.5, p = .04, η<sups>2</sups> = 0.14, which corresponds to a large effect size. Condition had no effect on knowledge in the Higher PoCC cluster, p = 0.3. Instead, these students, who came to the study with low prior knowledge, seemed to benefit equally in both Directive and Non-Directive conditions. In both conditions, performance on Know<subs>t1</subs> showed a reduced effect for Cluster, Cohen’s d = 0.40 (down from 2.1 at Know<subs>t0</subs>).</p> <hd id="AN0128149542-24">Student attitudes</hd> <p>As seen in Fig. 8, PoCC scores of students in both Conditions show a modest increase. ANCOVA with PoCC- as a DV found a significant Condition by Cluster interaction: F(<reflink idref="bib1" id="ref120">1</reflink>,<reflink idref="bib91" id="ref121">91</reflink>) = 5.2, p = 0.025, η<sups>2</sups> = 0.054. For the Higher PoCC cluster, the Directive support boosted their attitudes towards inquiry learning with simulations: F(<reflink idref="bib1" id="ref122">1</reflink>,<reflink idref="bib65" id="ref123">65</reflink>) = 3.9, p = 0.05, η<sups>2</sups> = .06, which corresponds to a medium effect size. Within the Higher Knowledge cluster, descriptively, students in the Directive condition had a smaller growth in their PoCC scores. However, this does not reach significance, possibly due to small sample size: F(<reflink idref="bib1" id="ref124">1</reflink>,<reflink idref="bib27" id="ref125">27</reflink>) = 2.1, p = 0.15, η<sups>2</sups> = .07.</p> <hd id="AN0128149542-25">Effect of support on transfer activity</hd> <p>In this section, we analyze students’ behaviours, PoCC and knowledge during the second, transfer activity, once all students worked with a Non-Directive version of the activity. We repeat the same analyses as was done above.</p> <hd id="AN0128149542-26">Effect on behaviours</hd> <p>To evaluate the effect of guidance on students’ behaviours in the transfer activity, we calculated a 2-way MANOVA with frequencies of Built, Test, Reset, and Pause as DV, and Condition, Cluster, and their interaction as IV. As seen in Fig. 9, the MANOVA showed no statistical effect of Condition, p &gt; 0.5.Students’ actions in the simulation during Activity 2</p> <hd id="AN0128149542-27">Overall effect</hd> <p>Figure 10 shows PoCC<subs>t2</subs> and Know<subs>t2</subs> as a function of Condition and Cluster. A 2-way MANCOVA with Know<subs>t2</subs> and PoCC<subs>t2</subs> as DV, Cluster, Condition, and their interaction as DV, controlling for Know<subs>t0</subs> and PoCC<subs>t0</subs> found only a marginally significant interaction F(<reflink idref="bib2" id="ref126">2</reflink>,<reflink idref="bib90" id="ref127">90</reflink>) = 2.4, p &lt; 0.1, η<sups>2</sups> = 0.5. To allow for comparisons across activities, we follow up with individual ANCOVAs, even though the interaction term was only marginally significant. No main effects approached significance.Knowledge and attitudes following Activity 2</p> <hd id="AN0128149542-28">Student knowledge</hd> <p>At the end of the transfer task (t2), there was no statistically significant effect of Condition or Cluster on students’ knowledge, nor of their interaction, controlling for incoming knowledge. Comparing both clusters, Cohen’s d for Know<subs>t2</subs> = 0.65, down from 2.1 at Know<subs>t0</subs>.</p> <hd id="AN0128149542-29">Student attitudes</hd> <p>The transfer activity showed a persistent pattern in relation to learners’ PoCC. Students in the Higher PoCC cluster kept their high PoCC scores, and students in the Higher Knowledge cluster kept improving their scores. Descriptively, the effect of Condition seems minor. That said, there was a significant interaction between Condition and Cluster F(<reflink idref="bib1" id="ref128">1</reflink>,<reflink idref="bib91" id="ref129">91</reflink>) = 4.0, p &lt; 0.05, η<sups>2</sups> = 0.04. Students in the Higher PoCC cluster had higher PoCC scores in the Directive condition: F(<reflink idref="bib1" id="ref130">1</reflink>,<reflink idref="bib62" id="ref131">62</reflink>) = 4.4, p = 0.04, η<sups>2</sups> = 0.07, which corresponds to a medium effect size. The PoCC of students in the Higher Knowledge cluster remained lower in the Directive condition, though, possibly due to a small sample size, this effect is not statistically significant: F(<reflink idref="bib1" id="ref132">1</reflink>,<reflink idref="bib27" id="ref133">27</reflink>) = 1.3, p = 0.27, η<sups>2</sups> = 0.5. This finding suggests that having directive forms of support had differentiated effect on students’ perceptions about their ability to learn in simulations, even after support had faded. Students in the Higher PoCC cluster developed slightly higher perceptions of competence and control following the Directive support, compared with their peers in the Non-Directive condition. In contrast, students in the Higher Knowledge cluster who received Directive support benefited least in terms of building a sense of competence and control (or self-efficacy). These lower levels of PoCC persisted into the transfer activity, even when all students received the same level of support (and lack of directive guidance).</p> <hd id="AN0128149542-30">Discussion</hd> <p>Table 4 summarizes the results as described above.</p> <p>Results summary</p> <ct id="AN0128149542-31"></ct> <p> <ephtml> &lt;table frame="hsides" rules="groups"&gt;&lt;tr&gt;&lt;th align="left"&gt;Effect of condition on&amp;#8230;&lt;/th&gt;&lt;th align="left"&gt;Students in higher PoCC cluster&lt;/th&gt;&lt;th align="left"&gt;Students in higher knowledge cluster&lt;/th&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td align="left"&gt;Actions&lt;/td&gt;&lt;td align="left" colspan="2"&gt;Students in the Directive condition tested more and built less while support was in effect (t1). This effect did not hold once Directive support was removed (t2)&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td align="left"&gt;Knowledge&lt;/td&gt;&lt;td align="left"&gt;Not much impact of Directive support. Students in this cluster increased their knowledge to a level comparable to that of the Non-Directive students in the Higher Knowledge cluster, regardless of condition (t1 and t2)&lt;/td&gt;&lt;td align="left"&gt;Students in this cluster learned more when Directive support was given (t1). However, the effect did not hold once Directive support was no longer available (t2)&lt;/td&gt;&lt;/tr&gt;&lt;tr&gt;&lt;td align="left"&gt;Perceptions of Competence and Control&lt;/td&gt;&lt;td align="left"&gt;Prior experiences were related to higher incoming attitudes (t0), which remained high. Directive support in Activity 1 (t1) was related to higher PoCC in both activities (t1 and t2)&lt;/td&gt;&lt;td align="left"&gt;Students improved their PoCC with more experiences (t1 and t2). PoCC levels approached those of their peers in the Non-Directive condition in the Higher PoCC cluster (t2). Growth in PoCC was smaller&amp;#8212;for students in the Directive condition (t1 and t2)&lt;/td&gt;&lt;/tr&gt;&lt;/table&gt; </ephtml> </p> <hd id="AN0128149542-32">Effect of directive guidance on behaviours</hd> <p>Students who received Directive support, from both clusters, were more likely to test deliberately and less likely to build while support was present. Testing is often considered productive, as collecting relevant data can support sense making and learning. Reducing the amount of building suggests a more restricted exploration. Thus, it seems that the detailed guidance given to students in the Directive condition directed them towards prescribed and more formal strategies, possibly reducing their agency and self-driven exploration.</p> <p>More interesting, however, is the impact that Directive support did not have. There was no impact for type of support on two high-level factors: overall number of actions and frequency of pauses. These are perhaps the two most important aspects of behaviours in this sim. Overall number of actions is a proxy of activity level. Having no impact from detailed guidance is surprising. Pauses are often a marker of metacognitive processes, either reflecting backwards or planning ahead. It has previously been shown that pauses are a significant predictor of learning (Shih et al. [<reflink idref="bib61" id="ref134">61</reflink>] ), also with the current simulation (Perez et al. [<reflink idref="bib53" id="ref135">53</reflink>] ). Thus, from a self-regulatory perspective, the impact of support was very limited.</p> <p>Furthermore, as soon as support was removed, there were no differences in frequencies of any actions between the two conditions. Whether productive or not, the impact of support was short lived. The two activities required only near transfer: they both took place in the same setting, with minutes between them, using the same simulation. The activities themselves were nearly isomorphic, as far as the given focus of inquiry. Students in the Directive conditions could have readily applied the directive guidance to Activity 2, as the same circuits, tests, and prompts that were prescribed for Activity 1 would have been beneficial also in Activity 2. Thus, the fact that there was no impact for condition on students’ behaviours in Activity 2 is not obvious. It seems that students in the Directive condition interpreted the support as being very specific to the initial context, Activity 1, and did not see its relevance to the transfer activity. Earlier we posed one of the greater challenges of support—to prepare students for independent learning. In the context of the current simulation, the guidance provided failed to achieve that.</p> <hd id="AN0128149542-33">Effect of guidance on knowledge</hd> <p>The impact of type of support on learning was more complex. Echoing earlier studies on inquiry learning, directive guidance improved learning for some students while it was available. Notably, this was the case only for students in the Higher Knowledge cluster. Students in the Higher PoCC cluster were not affected by the presence of directive guidance. Furthermore, as was the case with students’ behaviours, there were no long-term benefits for guidance learning, as measured in the transfer activity. This is a failure to transfer on two levels: first, students in the Directive group within the Higher Knowledge cluster failed to transfer their greater knowledge acquired during Activity 1. Second, students in both Directive conditions failed to transfer the ways of learning from the support available to them in Activity 1. Thus, overall, directive guidance was useful to learning—but only for some students (a third of our sample), and only while it was in place.</p> <p>Lack of effect of directive guidance on learners in the Higher PoCC cluster is somewhat surprising. Why was learning in this group not affected by guidance, even though their actions were? This cluster includes the majority of learners in our study. One explanation may be that students in this cluster lacked sufficient knowledge to make sense of the prescribed guidance, as overall, their prior knowledge was lower. In a similar scenario, van Dijk et al. ([<reflink idref="bib65" id="ref136">65</reflink>] ) found that only high-prior knowledge students were successful at making use of available support. Also, Brenner et al. ([<reflink idref="bib11" id="ref137">11</reflink>] ) found that while some novices benefit from more Directive support, other novices benefited from less guidance. These results appear to be in striking contrast to the expertise reversal effect, which suggests that learners of low knowledge level are in need of greater support (Kalyuga [<reflink idref="bib31" id="ref138">31</reflink>] ). Cognitive Load theory may explain this discrepancy. The Directive support was not simple. All students needed to learn how to construct circuits, use the testing instrument, and learn from their experimentations. Students in the Directive condition also had to decode diagrams, interpret questions, and most importantly, reflect on their measurements, identify trends, and explain variations. Coordinating and executing these multiple tasks demand a high level of cognitive load. While all students were able to follow the motions (as suggested by the higher rate of Testing events), it may be that only learners with more cognitive resources or higher domain knowledge, which was typical of the Higher Knowledge cluster, were positioned to benefit from directive support.</p> <p>An alternative explanation relies on students’ prior experiences. Students in the cluster that benefited from detailed directions were also the students with fewer experiences with simulations, and lower expectations about the value of these simulations. It may be that these students could benefit from more “handholding” in their earlier steps with simulations, until they become comfortable (and confident) in their learning in these environments. Students in the Higher PoCC cluster, on the other hand, were more proficient in using simulations, and thus were not in need for such detailed support.</p> <p>While students in the Higher PoCC cluster did not benefit from directive guidance, the implicit support in the form of focus of inquiry and simulation affordances was highly effective. Even though the two clusters differed in their knowledge level coming into the study, this difference was not significant by the end of the study. That is, Higher PoCC learners from both conditions were able to catch up to the knowledge level of their Higher Knowledge peers.</p> <p>Several factors may have contributed to the effectiveness of implicit support for learners in the Higher PoCC cluster. Students in this cluster were armed with two important elements: they had more prior experiences with simulations, and they had positive attitudes towards inquiry learning and towards their ability to engage in inquiry: They attributed success to their own actions; they had high self-efficacy; and they perceived simulations to be effective. Thus, it seems that these students were prepared to learn from the activity, and make use of the implicit support that was available to them.</p> <p>At times, instructors discount informal ways of learning. For example, when giving students math equations or story problems of the very same equations, teachers erroneously think that students perform best on equations, while students are better at applying informal strategies to word problems (Koedinger and Nathan [<reflink idref="bib40" id="ref139">40</reflink>] ). A similar situation may apply here. It may be that for this group of learners and for this activity, informal learning using the simulation’s constraints and affordances is more beneficial (or at least as beneficial) as following more formal logic.</p> <p>Looking at the overall shifts from Know<subs>t0</subs> to Know<subs>t2</subs>, students who learned the most were those in the Higher PoCC cluster, regardless of condition. This result exemplifies the value of coming in with the right attitudes for learning. Students who came in with positive attitudes towards inquiry were able to learn the most, regardless of given support, and overcome lower incoming levels of knowledge. An alternative explanation is that students in the Higher PoCC cluster had more room for growth, as their prior knowledge level was lower.</p> <hd id="AN0128149542-34">Effect on attitudes</hd> <p>Overall, attitudes of all learners shifted upwards. Student attributes at time t0 showed positive relationship between attitudes and prior experiences. This pattern continued throughout the study—the more experiences students gained, the higher their PoCC became. This relationship also explains the discrepancy between students in the two clusters. Learners who started with fewer prior experiences, in the Higher Knowledge cluster, increased their attitudes as they progressed through the activity, and narrowed the gap. It may be that this study gave students in this cluster the missing experiences to build up their confidence and perceptions towards working with simulations.</p> <p>While overall attitudes of all students shifted upwards, differences that were still visible after Activity 2 suggest that type of support has a lasting impact on students’ attitudes. For students in the Directive condition, attitudes within the Higher PoCC cluster increased, while attitudes within the Higher Knowledge cluster were the lowest. That is, students whose knowledge benefited from directive guidance were also the ones whose attitudes were hurt by it. Giving students detailed guidance helps their learning but may reduce their sense of agency (as shown by Sawyer et al. [<reflink idref="bib59" id="ref140">59</reflink>] ), and accordingly, perceptions of competence and control. While impact on knowledge was short-lived, impact on attitudes remained also once support was removed.</p> <p>The study presented here has several limitations. Its main limitation is the small sample size. Thus, statistical power is small and some findings are more suggestive than conclusive. A second related limitation is that this was a single study, in the context of a single simulation, using a single population. It is of interest to evaluate how these results extend to other simulations, science topics, forms of support, and populations. A third significant limitation is the lack of qualitative data from the study. Thus, the results lack in-depth description of how students in the clusters differed, how support affected students’ inquiry, and how students understood it to affect their perceptions of competence and control. Last, for both types of measures (knowledge and attitudes), gaps at the end of the study are smaller than incoming gaps. Our current measures cannot reveal whether this is due to a low ceiling effect.</p> <hd id="AN0128149542-35">Summary</hd> <p>We evaluated the effect of directive guidance on learning with the PhET CCK simulation. Our first research question focused on the effect of directive guidance on behaviours, knowledge, and attitudes. Analysis of students’ behaviours in the simulation shows that Directive support guided learners towards more formal testing and less exploration. However, significant behaviours such as number of actions and frequency of pauses were not affected by the guidance. With regard to learning, when focusing on short-term knowledge gains (as is commonly done), directive guidance helped some students to learn more. However, the big picture is much more nuanced, as explained below.</p> <p>Our second research question asked about interaction between guidance and student incoming knowledge and attitudes. We found two clusters of students. Students in a group that was characterized by high prior knowledge and lower attitudes benefited from directive guidance. However, directive guidance also slowed down attitudinal growth for these students. Students in the second cluster, that was characterized with higher attitudes and more prior experiences, learned the most and were much less sensitive to the presence (or lack) of directive support. The detailed guidance did not affect their knowledge gains, yet it improved their perceptions of competence and control. As a reminder, all of our analyses controlled for incoming knowledge (Know<subs>t0</subs>) and attitudes (PoCC<subs>t0</subs>). Thus, variations in incoming knowledge or attitudes, in isolation, were controlled for statistically and do not explain the differentiated effect of guidance. Instead, the impact of cluster suggests that these groups of learners have deeper traits that transcend isolated variables. These are a combination of knowledge, attitudes, experiences, and likely additional untested factors.</p> <p>Our third research question asked about transfer of the impact to a subsequent activity, once directive guidance is no longer available. We evaluated the same factors in a minimal transfer setting: same simulation, same session, isomorphic activities. Impact on students’ actions and knowledge all but disappeared in the transfer activity. Yet, the mixed impact on attitudes transferred, suggesting that that impact was more substantial.</p> <p>Overall, this study demonstrates the positive impact that guidance has on inquiry behaviours, and in some cases, learning gains. However, it also reveals the short-lived nature of these gains. Furthermore, results show that adding guidance may come at the expense of attitudes towards inquiry, and that the suppressed attitudes may persist even once directive support is removed.</p> <p>As shown above, no one type of support was effective for all learners. When people need support, they are often provided with more structure. Results of this study suggest otherwise: support for learners who enter a learning situation with low knowledge but high attitudes for learning should instead focus on sense making. These results also show the importance of positive attitudes towards learning, and that experiences with inquiry learning helps students improve these attitudes. Thus, there may be value in having students explore concepts with carefully designed implicit support that invites inquiry, prior to providing them with high directive guidance.</p> <hd id="AN0128149542-36">References</hd> <hd id="AN0128149542-37">Citations</hd> <p>1 Adams, W. K., Paulson, A., &amp; Wieman, C. E. (2008). What levels of guidance promote engaged exploration with interactive simulations? In, AIP conference proceedings (Vol. 1064, p. 59).</p> <ulist> <item>2 Alfieri L, Brooks PJ, Aldrich NJ, Tenenbaum HR, Does discovery-based instruction enhance learning?, Journal of Educational Psychology, 2011</item> <item>3 Bandura A, Human agency in social cognitive theory, American Psychologist, 1989, 44, 9, 1175, 10.1037/0003-066X.44.9.1175</item> <item>4 Bandura A, Self-efficacy: The exercise of control, 1997, New York, Macmillan</item> <item>5 Bandura A, Barbaranelli C, Caprara GV, Pastorelli C, Multifaceted impact of self-efficacy beliefs on academic functioning, Child Development, 1996, 67, 1206, 1222, 10.2307/1131888</item> <item>6 Belenky DM, Nokes-Malach TJ, Motivation and transfer: The role of mastery-approach goals in preparation for future learning, Journal of the Learning Sciences, 2012</item> <item>7 Belenky DM, Nokes-Malach TJ, Mastery-approach goals and knowledge transfer: An investigation into the effects of task structure and framing instructions, Learning and Individual Differences, 2013, 25, 21, 34, 10.1016/j.lindif.2013.02.004</item> <item>8 Belland BR, Kim C, Hannafin MJ, A framework for designing scaffolds that improve motivation and cognition, Educational Psychologist, 2013, 48, 4, 243, 270, 10.1080/00461520.2013.838920</item> <item>9 Biswas, G., Roscoe, R., Jeong, H., &amp; Sucler, B. (2009). Promoting self-regulated learning skills in agent-based learning environments. In, Proceedings of the 17th international conference on computers in education.</item> <item>10 Bransford JD, Schwartz DL, Rethinking transfer: A simple proposal with multiple implications, Review of Research in Education, 2001, 24, 3, 61, 100</item> <item>11 Brenner DG, Matlen BJ, Timms MJ, Gochyyev P, Grillo-Hill A, Luttgen K, Varfolomeeva M, Modeling student learning behavior patterns in an online science inquiry environment, Technology, Knowledge and Learning, 2017, 22, 405, 425, 10.1007/s10758-017-9325-0</item> <item>12 Butler, D. L., &amp; Cartier, S. C. (2004). Assessing self-regulated learning in two different activities: Reading to learn and conducting research. Paper presented at the 2004 annual meetings of the Canadian Society for Studies in Education, Winnipeg, MB.</item> <item>13 Butler DL, Cartier SC, Schnellert L, Gagnon F, Giammarino M, Secondary students’ self regulated engagement in reading: Researching self-regulation as situated in context, Psychological Test and Assessment Modeling, 2011, 53, 1, 73, 105</item> <item>14 Butler DL, Schnellert L, Cartier SC, Layers of self- and co-regulation: Teachers’ co-regulating learning and practice to foster students’ self-regulated learning through reading, Education Research International, 2013</item> <item>15 Butler DL, Schnellert L, MacNeil K, Collaborative inquiry and distributed agency in educational change: A case study of a multi-level community of inquiry, Journal of Educational Change, 2015, 16, 1, 1, 26, 10.1007/s10833-014-9227-z</item> <item>16 Chamberlain JM, Lancaster K, Parson R, Perkins KK, How guidance affects student engagement with an interactive simulation, Chemistry Education Research and Practice, 2014, 15, 4, 628, 638, 10.1039/C4RP00009A</item> <item>17 Chen C-H, Howard BC, Effect of live simulation on middle school students’ attitudes and learning toward science, Educational Technology &amp; Society, 2010, 13, 1, 133, 139</item> <item>18 Cunningham D, Duffy T, Constructivism: implications for the design and delivery of instruction, Handbook of Research for Educational Communications and Technology, 1996, 51, 170, 198</item> <item>19 D’Angelo C, Rutstein D, Harris C, Bernard R, Borokhovski E, Haertel G, Simulations for STEM learning: Systematic review and meta-analysis, 2014, Menlo Park, SRI International</item> <item>20 Jong T, Technological advances in inquiry learning, Science, 2006, 312, 5773, 532, 533, 10.1126/science.1127750</item> <item>21 Jong T, Joolingen WR, Scientific discovery learning with computer simulations of conceptual domains, Review of Educational Research, 1998, 68, 179, 201, 10.3102/00346543068002179</item> <item>22 Deci EL, Ryan RM, The” what” and” why” of goal pursuits: Human needs and the self-determination of behavior, Psychological Inquiry, 2000, 11, 4, 227, 268, 10.1207/S15327965PLI1104_01</item> <item>23 Dweck CS, Master A, Wentzel KR, Wigfield A, Self-theories and motivation, Handbook of motivation at school, 2009, New York, Routledge, 123, 140</item> <item>24 Elliot AJ, Murayama K, On the measurement of achievement goals: Critique, illustration, and application, Journal of Educational Psychology, 2008, 100, 3, 613, 628, 10.1037/0022-0663.100.3.613</item> <item>25 Finkelstein ND, Adams WK, Keller CJ, Kohl PB, Perkins KK, Podolefsky NS, Reid S, When learning about the real world is better done virtually: A study of substituting computer simulations for laboratory equipment, Physical Review Special Topics-Physics Education Research, 2005, 1, 1, 10103, 10.1103/PhysRevSTPER.1.010103</item> <item>26 Gobert, J. (2015). Microworlds. In, Encyclopedia of Science Education (pp. 638–639).</item> <item>27 Gobert JD, Sao Pedro M, Raziuddin J, Baker RS, From log files to assessment metrics: Measuring students’ science inquiry skills using educational data mining, Journal of the Learning Sciences, 2013, 22, 4, 521, 563, 10.1080/10508406.2013.837391</item> <item>28 Graham S, Williams C, Wentzel K, Wigfield A, An attributional approach to motivation in school, Handbook of motivation at school, 2009, Hillsdale, NJ, Lawrence Erlbaum, 11, 34</item> <item>29 Hmelo-Silver CE, Golan Duncan R, Chinn CA, Scaffolding and achievement in problem-based and inquiry learning: A response to Kirschner, Sweller, and Clark (2006), Educational Psychologist, 2007, 42, 2, 99, 107, 10.1080/00461520701263368</item> <item>30 Holmes NG, Day J, Park AHK, Bonn DA, Roll I, Making the failure more productive: Scaffolding the invention process to improve inquiry behaviors and outcomes in invention activities, Instructional Science, 2014, 42, 4, 523, 538, 10.1007/s11251-013-9300-7</item> <item>31 Kalyuga S, Expertise reversal effect and its implications for learner-tailored instruction, Educational Psychology Review, 2007, 19, 4, 509, 539, 10.1007/s10648-007-9054-3</item> <item>32 Kapur M, Examining productive failure, productive success, unproductive failure, and unproductive success in learning, Educational Psychologist, 2016, 51, 1, 11, 10.1080/00461520.2016.1155457</item> <item>33 Kapur M, Bielaczyc K, Designing for productive failure, Journal of the Learning Sciences, 2012, 21, 1, 45, 83, 10.1080/10508406.2011.591717</item> <item>34 Kent P, Jensen RK, Kongsted A, A comparison of three clustering methods for finding subgroups in MRI, SMS or clinical data: SPSS TwoStep Cluster analysis, Latent Gold and SNOB, BMC Medical Research Methodology, 2014, 14, 1, 113, 10.1186/1471-2288-14-113</item> <item>35 Kirschner PA, Sweller J, Clark RE, Why minimal guidance during instruction does not work: An analysis of the failure of constructivist, discovery, problem-based, experiential, and inquiry-based teaching, Educational psychologist, 2006, 41, 2, 75, 86, 10.1207/s15326985ep4102_1</item> <item>36 Klahr, D. (2010). Coming up for air: But is it oxygen or phlogiston? A response to Taber’s review of constructivist instruction: Success or failure?, Education Review, 13(<reflink idref="bib13" id="ref141">13</reflink>). Retrieved 2015, March 15 from, http://www.edrev.info/essays/v13n13.pdf.</item> <item>37 Klahr D, Nigam M, The equivalence of learning paths in early science instruction—Effects of direct instruction and discovery learning, Psychological Science, 2004, 15, 10, 661, 667, 10.1111/j.0956-7976.2004.00737.x</item> <item>38 Koedinger KR, Aleven V, Exploring the assistance dilemma in experiments with cognitive tutors, Educational Psychology Review, 2007, 19, 3, 239, 264, 10.1007/s10648-007-9049-0</item> <item>39 Koedinger KR, Corbett AT, Perfetti C, The Knowledge-Learning-Instruction framework: Bridging the science-practice chasm to enhance robust student learning, Cognitive Science, 2012, 36, 5, 757, 798, 10.1111/j.1551-6709.2012.01245.x</item> <item>40 Koedinger KR, Nathan MJ, The real story behind story problems: Effects of representations on quantitative reasoning, The Journal of the Learning Sciences, 2004, 13, 2, 129, 164, 10.1207/s15327809jls1302_1</item> <item>41 Koedinger KR, Roll I, Holyoak KJ, Morrison RG, Learning to think: Cognitive mechanisms of knowledge transfer, The oxford handbook of thinking and reasoning, 2012, New York, Oxford University Press, 789, 806</item> <item>42 Lazonder AW, Harmsen R, Meta-Analysis of inquiry-based learning: Effects of guidance, Review of Educational Research, 2016, 86, 3, 681, 718, 10.3102/0034654315627366</item> <item>43 Linnenbrink EA, Pintrich PR, The role of self-efficacy beliefs in student engagement and learning in the classroom, Reading &amp; Writing Quarterly, 2003, 19, 2, 119, 137, 10.1080/10573560308223</item> <item>44 Loibl K, Roll I, Rummel N, Towards a theory of when and how problem solving followed by instruction supports learning, Educational Psychology Review, 2016</item> <item>45 Mathan SA, Koedinger KR, Fostering the intelligent novice: Learning from errors with metacognitive tutoring, Educational Psychologist, 2005, 40, 4, 257, 265, 10.1207/s15326985ep4004_7</item> <item>46 Matlen BJ, Klahr D, Sequential effects of high and low instructional guidance on children’s acquisition of experimentation skills: Is it all in the timing?, Instructional Science, 2013, 41, 3, 621, 634, 10.1007/s11251-012-9248-z</item> <item>47 Mayer RE, Should there be a three-strikes rule against pure discovery learning, American Psychologist, 2004, 59, 1, 14, 19, 10.1037/0003-066X.59.1.14</item> <item>48 McDaniel MA, Schlager MS, Discovery learning and transfer of problem-solving skills, Cognition and Instruction, 1990, 7, 2, 129, 159, 10.1207/s1532690xci0702_3</item> <item>49 Miller CS, Lehman JF, Koedinger KR, Goals and learning in microworlds, Cognitive Science: A Multidisciplinary Journal, 1999, 23, 3, 305, 336, 10.1207/s15516709cog2303_2</item> <item>50 Nathan MJ, Knowledge and situational feedback in a learning environment for algebra story problem solving, Interactive Learning Environments, 1998, 5, 1, 135, 159, 10.1080/1049482980050110</item> <item>51 Nomme, K., Butler, D. L., Pollock, C., Nakonechny, J., &amp; Cartier, S. C. (2012). Understanding post-secondary students’ perceptions about learning in science from within an inquiry-oriented biology laboratory. Paper presented at the annual meeting of the American Education Research Association (AERA).</item> <item>52 Paul, A., Podolefsky, N., &amp; Perkins, K. (2012). Guiding without feeling guided: Implicit scaffolding through interactive simulation design. In, Proceedings of the Physics Education Research Conference (, PERC).</item> <item>53 Perez, S., Massey-Allard, J., Butler, D., Ives, J., Bonn, D., Yee, N., &amp; Roll, I. (2017). Identifying productive inquiry in virtual labs using sequence mining. In, Proceedings of the International Conference on Artificial Intelligence in Education (pp. 287–298). Cham: Springer.</item> <item>54 Podolefsky NS, Rehn D, Perkins KK, Affordances of play for student agency and student-centered pedagogy, AIP Conference Proceedings, 2013, 1513, 1, 306, 309, 10.1063/1.4789713</item> <item>55 Rittle-Johnson B, Star JR, Durkin K, Developing procedural flexibility: Are novices prepared to learn from comparing procedures?, British Journal of Educational Psychology, 2012, 82, Pt 3, 436, 455, 10.1111/j.2044-8279.2011.02037.x</item> <item>56 Roll I, Aleven V, McLaren BM, Koedinger KR, Improving students’ help-seeking skills using metacognitive feedback in an intelligent tutoring system, Learning and Instruction, 2011, 21, 267, 280, 10.1016/j.learninstruc.2010.07.004</item> <item>57 Roll I, Baker RSJD, Aleven V, Koedinger KR, On the benefits of seeking (and avoiding) help in online problem solving environment, Journal of the Learning Sciences, 2014, 23, 4, 537, 560, 10.1080/10508406.2014.883977</item> <item>58 Roll, I., Wiese, E. S., Long, Y., Aleven, V., &amp; Koedinger, K. R. (2014b). Tutoring self-and co-regulation with intelligent tutoring systems to help students acquire better learning skills. In, Design recommendations for intelligent tutoring systems (p. 169).</item> <item>59 Sawyer, R., Smith, A., Rowe, J., Azevedo, R., &amp; Lester, J. (2017). Is more agency better? The impact of student agency on game-based learning. In, Proceedings of International conference on artificial intelligence in education (pp. 335–346). Retrieved from Google Scholar.</item> <item>60 Schunk DH, Pintrich PR, Meece JL, Motivation in education: Theory, research, and applications, 2008, Upper Saddle River, NJ, Pearson/Merrill Prentice Hall</item> <item>61 Shih B, Koedinger KR, Scheines R, Shih B, Koedinger KR, Scheines R, A response time model for bottom-out hints as worked examples, Handbook of educational data mining, 2011, Boca Raton, CRC Press, 201, 212</item> <item>62 Strand-Cary M, Klahr D, Developing elementary science skills: Instructional effectiveness and path independence, Cognitive Development, 2008</item> <item>63 Sweller J, Mawer RF, Ward MR, Development of expertise in mathematical problem solving, Journal of Experimental Psychology: General, 1988, 112, 4, 639, 661, 10.1037/0096-3445.112.4.639</item> <item>64 Tuovinen JE, Sweller J, A comparison of cognitive load associated with discovery learning and worked examples, Journal of Educational Psychology, 1999, 91, 2, 334, 10.1037/0022-0663.91.2.334</item> <item>65 Dijk AM, Eysink TH, Jong T, Ability-related differences in performance of an inquiry task: The added value of prompts, Learning and Individual Differences, 2016, 47, 145, 155, 10.1016/j.lindif.2016.01.008</item> <item>66 Vollmeyer R, Burns BD, Holyoak KJ, The impact of goal specificity on strategy use and the acquisition of problem structure, Cognitive Science, 1996, 20, 1, 75, 100, 10.1207/s15516709cog2001_3</item> <item>67 Vygotsky L, Mind in society, 1978, Cambridge, MA, Harvard University Press</item> <item>68 Wieman CE, Adams WK, Perkins KK, PhET: Simulations that enhance learning, Science, 2008, 322, 5902, 682, 683, 10.1126/science.1161948</item> <item>69 Wiese ES, Koedinger KR, Designing grounded feedback: Criteria for using linked representations to support learning of abstract symbols, International Journal of Artificial Intelligence in Education, 2017, 27, 3, 448, 474, 10.1007/s40593-016-0133-9</item> <item>70 Wigfield A, Eccles JS, Expectancy-value theory of achievement motivation, Contemporary Educational Psychology, 2000, 25, 1, 68, 81, 10.1006/ceps.1999.1015</item> <item>71 Wise AF, O’Neill K, Duffy T, Tobias S, Beyond more versus less: A reframing of the debate on instructional guidance, Constructivist instruction: Success of failure?, 2009, New York, Routledge/Taylor &amp; Francis Group</item> <item>72 Wood D, Scaffolding, contingent tutoring, and computer-supported learning, International Journal of Artificial Intelligence in Education, 2001, 12, 3, 280, 293</item> <item>73 Wood D, Bruner JS, Ross G, The role of tutoring in problem solving, Journal of Child Psychology and Psychiatry, 1976, 17, 2, 89, 100, 10.1111/j.1469-7610.1976.tb00381.x</item> <item>74 Yaron D, Karabinos M, Lange D, Greeno JG, Leinhardt G, The chemcollective—Virtual labs for introductory chemistry courses, Science, 2010, 328, 5978, 584, 585, 10.1126/science.1182435</item> <item>75 Zimmerman BJ, Self-regulation involves more than metacognition: A social cognitive perspective, Educational Psychologist, 2010, 30, 4, 217, 221, 10.1207/s15326985ep3004_8</item> <item>76 Zohar A, Barzilai S, A review of research on metacognition in science education: Current and future directions, Studies in Science Education, 2013, 49, 2, 121, 169, 10.1080/03057267.2013.847261</item> <ref id="AN0128149542-38"> <title>Footnotes</title> <blist> <bibl id="bib1" idref="ref93" type="bt">1</bibl> <bibtext>Due to an error, several students from both conditions were missing several test-items from the post-test on the first topic, Knowt1. Thus, we first converted actual scores to z scores based on the available items, and then analyzed these. </bibtext> </blist> <blist> <bibl id="bib2" idref="ref15" type="bt">2</bibl> <bibtext>The threshold of 15 s was derived from data. It denotes the start of the long tail of the distribution of time between subsequent actions. Changing this parameter has no impact on the analysis presented below. </bibtext> </blist> </ref> </ulist> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <p>PHOTO (COLOR)</p> <aug> <p>By Ido Roll; Deborah Butler; Nikki Yee; Ashley Welsh; Sarah Perez; Adriana Briseno; Katherine Perkins and Doug Bonn</p> </aug> <nolink nlid="nl1" bibid="bib38" firstref="ref1"></nolink> <nolink nlid="nl2" bibid="bib35" firstref="ref2"></nolink> <nolink nlid="nl3" bibid="bib47" firstref="ref3"></nolink> <nolink nlid="nl4" bibid="bib18" firstref="ref4"></nolink> <nolink nlid="nl5" bibid="bib29" firstref="ref5"></nolink> <nolink nlid="nl6" bibid="bib71" firstref="ref6"></nolink> <nolink nlid="nl7" bibid="bib39" firstref="ref8"></nolink> <nolink nlid="nl8" bibid="bib36" firstref="ref9"></nolink> <nolink nlid="nl9" bibid="bib42" firstref="ref11"></nolink> <nolink nlid="nl10" bibid="bib21" firstref="ref13"></nolink> <nolink nlid="nl11" bibid="bib19" firstref="ref14"></nolink> <nolink nlid="nl12" bibid="bib14" firstref="ref18"></nolink> <nolink nlid="nl13" bibid="bib56" firstref="ref19"></nolink> <nolink nlid="nl14" bibid="bib72" firstref="ref20"></nolink> <nolink nlid="nl15" bibid="bib76" firstref="ref21"></nolink> <nolink nlid="nl16" bibid="bib12" firstref="ref22"></nolink> <nolink nlid="nl17" bibid="bib13" firstref="ref23"></nolink> <nolink nlid="nl18" bibid="bib73" firstref="ref24"></nolink> <nolink nlid="nl19" bibid="bib64" firstref="ref25"></nolink> <nolink nlid="nl20" bibid="bib67" firstref="ref26"></nolink> <nolink nlid="nl21" bibid="bib31" firstref="ref27"></nolink> <nolink nlid="nl22" bibid="bib41" firstref="ref28"></nolink> <nolink nlid="nl23" bibid="bib65" firstref="ref29"></nolink> <nolink nlid="nl24" bibid="bib11" firstref="ref30"></nolink> <nolink nlid="nl25" bibid="bib57" firstref="ref32"></nolink> <nolink nlid="nl26" bibid="bib49" firstref="ref33"></nolink> <nolink nlid="nl27" bibid="bib63" firstref="ref34"></nolink> <nolink nlid="nl28" bibid="bib44" firstref="ref36"></nolink> <nolink nlid="nl29" bibid="bib45" firstref="ref37"></nolink> <nolink nlid="nl30" bibid="bib68" firstref="ref39"></nolink> <nolink nlid="nl31" bibid="bib50" firstref="ref40"></nolink> <nolink nlid="nl32" bibid="bib58" firstref="ref41"></nolink> <nolink nlid="nl33" bibid="bib69" firstref="ref43"></nolink> <nolink nlid="nl34" bibid="bib5" firstref="ref44"></nolink> <nolink nlid="nl35" bibid="bib60" firstref="ref45"></nolink> <nolink nlid="nl36" bibid="bib4" firstref="ref46"></nolink> <nolink nlid="nl37" bibid="bib43" firstref="ref47"></nolink> <nolink nlid="nl38" bibid="bib28" firstref="ref50"></nolink> <nolink nlid="nl39" bibid="bib70" firstref="ref51"></nolink> <nolink nlid="nl40" bibid="bib23" firstref="ref53"></nolink> <nolink nlid="nl41" bibid="bib17" firstref="ref54"></nolink> <nolink nlid="nl42" bibid="bib51" firstref="ref55"></nolink> <nolink nlid="nl43" bibid="bib8" firstref="ref56"></nolink> <nolink nlid="nl44" bibid="bib6" firstref="ref57"></nolink> <nolink nlid="nl45" bibid="bib7" firstref="ref58"></nolink> <nolink nlid="nl46" bibid="bib75" firstref="ref59"></nolink> <nolink nlid="nl47" bibid="bib54" firstref="ref60"></nolink> <nolink nlid="nl48" bibid="bib22" firstref="ref61"></nolink> <nolink nlid="nl49" bibid="bib3" firstref="ref62"></nolink> <nolink nlid="nl50" bibid="bib15" firstref="ref63"></nolink> <nolink nlid="nl51" bibid="bib59" firstref="ref64"></nolink> <nolink nlid="nl52" bibid="bib37" firstref="ref65"></nolink> <nolink nlid="nl53" bibid="bib62" firstref="ref66"></nolink> <nolink nlid="nl54" bibid="bib46" firstref="ref69"></nolink> <nolink nlid="nl55" bibid="bib30" firstref="ref70"></nolink> <nolink nlid="nl56" bibid="bib32" firstref="ref71"></nolink> <nolink nlid="nl57" bibid="bib55" firstref="ref72"></nolink> <nolink nlid="nl58" bibid="bib48" firstref="ref73"></nolink> <nolink nlid="nl59" bibid="bib33" firstref="ref75"></nolink> <nolink nlid="nl60" bibid="bib10" firstref="ref76"></nolink> <nolink nlid="nl61" bibid="bib27" firstref="ref79"></nolink> <nolink nlid="nl62" bibid="bib9" firstref="ref80"></nolink> <nolink nlid="nl63" bibid="bib20" firstref="ref82"></nolink> <nolink nlid="nl64" bibid="bib74" firstref="ref84"></nolink> <nolink nlid="nl65" bibid="bib26" firstref="ref85"></nolink> <nolink nlid="nl66" bibid="bib25" firstref="ref86"></nolink> <nolink nlid="nl67" bibid="bib52" firstref="ref87"></nolink> <nolink nlid="nl68" bibid="bib66" firstref="ref94"></nolink> <nolink nlid="nl69" bibid="bib16" firstref="ref95"></nolink> <nolink nlid="nl70" bibid="bib24" firstref="ref100"></nolink> <nolink nlid="nl71" bibid="bib34" firstref="ref103"></nolink> <nolink nlid="nl72" bibid="bib53" firstref="ref105"></nolink> <nolink nlid="nl73" bibid="bib96" firstref="ref106"></nolink> <nolink nlid="nl74" bibid="bib89" firstref="ref110"></nolink> <nolink nlid="nl75" bibid="bib92" firstref="ref112"></nolink> <nolink nlid="nl76" bibid="bib91" firstref="ref121"></nolink> <nolink nlid="nl77" bibid="bib90" firstref="ref127"></nolink> <nolink nlid="nl78" bibid="bib61" firstref="ref134"></nolink> <nolink nlid="nl79" bibid="bib40" firstref="ref139"></nolink> |
|---|---|
| Header | DbId: eric DbLabel: ERIC An: EJ1170537 AccessLevel: 3 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Understanding the Impact of Guiding Inquiry: The Relationship between Directive Support, Student Attributes, and Transfer of Knowledge, Attitudes, and Behaviours in Inquiry Learning – Name: Language Label: Language Group: Lang Data: English – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Roll%2C+Ido%22">Roll, Ido</searchLink> (ORCID <externalLink term="http://orcid.org/0000-0001-7295-9059">0000-0001-7295-9059</externalLink>)<br /><searchLink fieldCode="AR" term="%22Butler%2C+Deborah%22">Butler, Deborah</searchLink><br /><searchLink fieldCode="AR" term="%22Yee%2C+Nikki%22">Yee, Nikki</searchLink><br /><searchLink fieldCode="AR" term="%22Welsh%2C+Ashley%22">Welsh, Ashley</searchLink><br /><searchLink fieldCode="AR" term="%22Perez%2C+Sarah%22">Perez, Sarah</searchLink><br /><searchLink fieldCode="AR" term="%22Briseno%2C+Adriana%22">Briseno, Adriana</searchLink><br /><searchLink fieldCode="AR" term="%22Perkins%2C+Katherine%22">Perkins, Katherine</searchLink><br /><searchLink fieldCode="AR" term="%22Bonn%2C+Doug%22">Bonn, Doug</searchLink> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="SO" term="%22Instructional+Science%3A+An+International+Journal+of+the+Learning+Sciences%22"><i>Instructional Science: An International Journal of the Learning Sciences</i></searchLink>. Feb 2018 46(1):77-104. – Name: Avail Label: Availability Group: Avail Data: Springer. 233 Spring Street, New York, NY 10013. Tel: 800-777-4643; Tel: 212-460-1500; Fax: 212-348-4505; e-mail: service-ny@springer.com; Web site: http://www.springerlink.com – Name: PeerReviewed Label: Peer Reviewed Group: SrcInfo Data: Y – Name: Pages Label: Page Count Group: Src Data: 28 – Name: DatePubCY Label: Publication Date Group: Date Data: 2018 – Name: TypeDocument Label: Document Type Group: TypDoc Data: Journal Articles<br />Reports - Research – Name: Subject Label: Descriptors Group: Su Data: <searchLink fieldCode="DE" term="%22Inquiry%22">Inquiry</searchLink><br /><searchLink fieldCode="DE" term="%22Active+Learning%22">Active Learning</searchLink><br /><searchLink fieldCode="DE" term="%22Discovery+Learning%22">Discovery Learning</searchLink><br /><searchLink fieldCode="DE" term="%22Physics%22">Physics</searchLink><br /><searchLink fieldCode="DE" term="%22Electronics%22">Electronics</searchLink><br /><searchLink fieldCode="DE" term="%22Equipment%22">Equipment</searchLink><br /><searchLink fieldCode="DE" term="%22Simulation%22">Simulation</searchLink><br /><searchLink fieldCode="DE" term="%22Learning+Processes%22">Learning Processes</searchLink><br /><searchLink fieldCode="DE" term="%22Outcomes+of+Education%22">Outcomes of Education</searchLink><br /><searchLink fieldCode="DE" term="%22Comparative+Analysis%22">Comparative Analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Guidance%22">Guidance</searchLink><br /><searchLink fieldCode="DE" term="%22Achievement+Gains%22">Achievement Gains</searchLink><br /><searchLink fieldCode="DE" term="%22Student+Attitudes%22">Student Attitudes</searchLink><br /><searchLink fieldCode="DE" term="%22Attitude+Change%22">Attitude Change</searchLink><br /><searchLink fieldCode="DE" term="%22Instructional+Effectiveness%22">Instructional Effectiveness</searchLink> – Name: DOI Label: DOI Group: ID Data: 10.1007/s11251-017-9437-x – Name: ISSN Label: ISSN Group: ISSN Data: 0020-4277 – Name: Abstract Label: Abstract Group: Ab Data: Guiding inquiry learning has been shown to increase knowledge gains. Yet, little is known about the effect of guidance on attitudes and behaviours, its interaction with student attributes, and transfer of impact once guidance is removed. We address these gaps in the context of an interactive Physics simulation on electric circuits (https://phet.colorado.edu/en/simulation/circuit-construction-kit-dc). 49 students in the Non-Directive condition received a set of goals to focus their inquiry, in addition to implicit support built into the simulation. 48 students in the Directive condition received, in addition to these, also detailed directions and prompts. Log-file analysis found that directive support led to more formal testing and less exploration. Clustering identified two groups of learners: one with higher incoming knowledge (Higher Knowledge), the other with higher incoming perceptions of competence and control (Higher PoCC). Working with the simulation improved knowledge and attitudes across cluster groups, so that prior differences all but disappeared. With regard to guidance, adding directive support improved knowledge gains for the Higher Knowledge group, yet suppressed their attitudinal growth. The same support had no effect on knowledge gains for the Higher PoCC group, yet it boosted their attitudinal growth. A transfer activity, where directive support was no longer available, found that impact on attitudes carried forward, yet impacts on behaviours and knowledge were short-lived. Overall, the study highlights the complex interaction between guidance and student attributes. For some, supporting short-term knowledge gains may inadvertenly lead to longer term negative impact on attitudes towards inquiry. – Name: AbstractInfo Label: Abstractor Group: Ab Data: As Provided – Name: Ref Label: Number of References Group: RefInfo Data: 76 – Name: DateEntry Label: Entry Date Group: Date Data: 2018 – Name: AN Label: Accession Number Group: ID Data: EJ1170537 |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=eric&AN=EJ1170537 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.1007/s11251-017-9437-x Languages: – Text: English PhysicalDescription: Pagination: PageCount: 28 StartPage: 77 Subjects: – SubjectFull: Inquiry Type: general – SubjectFull: Active Learning Type: general – SubjectFull: Discovery Learning Type: general – SubjectFull: Physics Type: general – SubjectFull: Electronics Type: general – SubjectFull: Equipment Type: general – SubjectFull: Simulation Type: general – SubjectFull: Learning Processes Type: general – SubjectFull: Outcomes of Education Type: general – SubjectFull: Comparative Analysis Type: general – SubjectFull: Guidance Type: general – SubjectFull: Achievement Gains Type: general – SubjectFull: Student Attitudes Type: general – SubjectFull: Attitude Change Type: general – SubjectFull: Instructional Effectiveness Type: general Titles: – TitleFull: Understanding the Impact of Guiding Inquiry: The Relationship between Directive Support, Student Attributes, and Transfer of Knowledge, Attitudes, and Behaviours in Inquiry Learning Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Roll, Ido – PersonEntity: Name: NameFull: Butler, Deborah – PersonEntity: Name: NameFull: Yee, Nikki – PersonEntity: Name: NameFull: Welsh, Ashley – PersonEntity: Name: NameFull: Perez, Sarah – PersonEntity: Name: NameFull: Briseno, Adriana – PersonEntity: Name: NameFull: Perkins, Katherine – PersonEntity: Name: NameFull: Bonn, Doug IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 02 Type: published Y: 2018 Identifiers: – Type: issn-print Value: 0020-4277 Numbering: – Type: volume Value: 46 – Type: issue Value: 1 Titles: – TitleFull: Instructional Science: An International Journal of the Learning Sciences Type: main |
| ResultId | 1 |
