3.1.1 Attentional selectivity and control

Our experience of the world seems to us complete, accurate, and stable. This is largely an illusion. Our attentional processes filter out all but a tiny fraction of all the information available in our sensory registers. A particularly striking demonstration of this principle became widely known after a seminal research paper by Simons and Chabris (1999) on what has come to be known as “inattentional blindness.” These researchers asked their participants to watch videos showing members of two basketball teams, each passing a ball to other members of their teams. The participants were instructed to carefully count the number of passes within one of the teams. Midway through these videos, an unexpected event occurred. A tall woman with an umbrella, or a shorter woman dressed in a gorilla suit, entered the field and walked out again. Across various conditions, only about half of the participants noticed this unexpected event, despite its close spatial proximity to, even at times visually overlapping with, the basketball players.

The core phenomenon here is this: we tend not to notice—and thus we are unable to process—anything that our attention is not fully focused on. This was established already during the nineteen-fifties and -sixties, in a number of different experimental paradigms (Dosher & Lu, 2010; Folk, 2010), investigating both visual and auditory attention.

Attentional selectivity is universal and all-pervasive. We tend not to notice it, however, and thus we are generally unaware of how it affects us. In a teaching and learning context, attentional selectivity will matter. If the learner’s attention is not properly focused on what is important to the learning itself, little or no processing will occur, and little or no learning will be the end result (Sweller, Ayres, & Kalyuga, 2011).

Given this, it seems important to know something about how our attentional focus is controlled and directed. Two basic principles are at work in attentional control: bottom-up processes and top-down processes. Attentional focus is partly controlled from the bottom up by events in our environment. Certain stimuli will serve to pull attentional focus away from whatever it is currently directed at. A loud, unexpected noise in a quiet room, for instance, will simply demand attention. In such circumstances, a shift of attentional focus—an orienting response—is almost reflexive. But less extreme deviations from background stimuli will tend to attract attentional focus too. Someone entering or exiting a room or an auditorium, even discretely, will rarely be completely ignored. A tiny flashing light on your smart phone, or a small notification pop-up in the corner of your laptop screen, will tend to tug at the beam of your attentional spotlight, sometimes successfully persuading you to abandon your current goals.

The implications of this in a teaching and learning context are relatively banal, but important none the less. If we surround ourselves with stimuli irrelevant to our current learning tasks, our attentional focus will tend to jump around, allowing irrelevant information to occupy limited working memory capacity. This in turn reduces the amount of germane, learning-focused processing, and will serve to undermine our efforts to learn.

But if our attentional control is so fickle and easily derailed by random stimuli bottom up, why then do the participants in Simon and Chabris’ experiments not notice the umbrella lady or the gorilla? As stimuli go, they are both unexpected, and certainly out of place. There are two related explanations. First, Simon and Chabris cleverly manipulated the unexpected events to visually blend in to the background stimuli. The gorilla suit, for instance, is black, and the unattended basketball team members wear black t-shirts. More interestingly, their participants’ attention is strongly influenced by top down control factors. They have clear instructions, a very clear goal (count the passes in the white team), and most of them are probably highly motivated (“I’m so gonna get this right!”).

Again, deriving implications for teaching and learning is relatively straightforward. Clear, specific goals and instructions help focus attention from the top down, allowing us to ignore irrelevant stimuli, even when they are otherwise quite noticeable. This again prevents our limited working memory from becoming overburdened, leaving that capacity for processing that leads to learning.

Clear goals may have several other advantages, too. For instance, not all distractions are external stimuli; quite often, we allow ourselves to become distracted by internal stimuli. They may be bodily sensations or emotions, or they may be thoughts concerning goals related to other aspects of our lives not directly related to what we are trying to learn.

In fact, there is a substantial research base in education science supporting communication of clear goals and success criteria for learning tasks. John Hattie of Visible Learning fame (Hattie, 2009), when summarizing the conclusions of his mega synthesis, lists providing “clear learning intentions and criteria for success” (Hattie, 2011, p. 134) as one of three main claims for higher education (for more on lessons to be learned from the Visible learning synthesis, see Section 3.4.1).

3.1.2 Limits on sustained, focused attention

Besides the selectivity and the tendency to be derailed by irrelevant stimuli, our attentional apparatus suffers from another limitation. Maintaining focused attention over time is hard mental work, and it almost seems as if our attention tires after having been directed at the same stimuli for a while. Whatever may be the root cause, we do suffer occasional attentional lapses over time. In ergonomics, this phenomenon is known as the “vigilance decrement” (e.g., Smit, Eling, & Coenen, 2004).

While controlled studies of this phenomenon in realistic teaching and learning situations are surprisingly few and far between (see Wilson & Korn, 2007), there is some evidence that attention lapses during lectures. An early observational study found inattention at the very start of a lecture, followed by a period of relatively high attentiveness, and then lapses of attention tending to occur after about 10–18 minutes into the lecture. Also, lapses of attention became more frequent further into the session (Johnstone & Percival, 1976). A more recent study found lapses of attention arising after just a few (2–3 and 5–6) minutes, and replicated a pattern of shorter intervals between lapses as time goes by (Bunce, Flens, & Neiles, 2010), although in this study, lecture segments were relatively short (10–12 minutes).

A more interesting result from the study by Bunce et al. is that the frequency of attentional lapses decreased markedly during session segments using student-centered pedagogies facilitated by clickers (a type of student response system—see Section 4.3.3) compared to lecture-based segments. More interesting still, lapses of attention were less frequent in lecture segments following the clicker segments compared to lecture segments leading up to clicker segments (Bunce et al., 2010). It seems then that attentional lapses are more likely to be reported during lecture segments than during segments of student activity, and that bouts of meaningful student activity can somehow refresh students’ ability to maintain attentional focus immediately afterwards.

Similar results have been obtained in studies of mind wandering during online, video-recorded lectures. Two studies estimated mind wandering to occur as often as from around 35 to around 52 percent of the time, and both studies found this proportion to be higher towards the end of the lectures (Risko, Anderson, Sarwal, Engelhardt, & Kingstone, 2012; Szpunar, Khan, & Schacter, 2013). Notably, in the study by Risko et al. (2012), mind wandering was associated with reduced note taking and poorer retention. Echoing the results from live lectures by Bunce et al. (2010), Szpunar, Khan et al. (2013) found markedly less mind wandering when online, video-recorded lectures were interspersed with test questions throughout.

It seems safe to conclude, then, that teaching by lecturing is likely to induce attentional lapses and mind wandering, both during live teaching and in online, recorded teaching. Lapses are more likely the longer the lecturing segment lasts. On the bright side, it also seems that we can partially avert lapses by mixing in student-centered, learning-focused activity, such as clicker quizzes or test questions.

Why is this important to IL teachers, or to teachers in general? In their review of the research on attention lapses and mind wandering, Szpunar, Moulton, and Schacter (2013) argue that “mind wandering is particularly relevant to education, because learning depends critically on attention in ways that other activities do not” (p. 5). Recall that earlier (in Section 3.1) we described attention as the gatekeeper to our working memory; as the processes that determine what information is selected for further processing. To the extent that our attention is directed at something that is irrelevant to our learning goals, be it external stimuli or the encounters of our own mind wanderings, we are prevented from engaging in the processing that generates learning. Let us now consider working memory, the very component of our cognitive architecture where that processing actually happens.

3.1.3 Working hard with working memory

Working memory, sometimes called short-term memory, refers to the processes that allow us to hold on to and manipulate information we are currently aware of and working on. When following a conversation or a presentation, or when reading, our working memories retain what was just previously said or read and relates it to incoming information, allowing us to group together words and concepts to form larger, meaningful chunks. When solving a problem, we use our working memory to hold information about the current problem state, our next possible solution step and its consequences, and to relate these to the new possibilities they open up and to the intended solution.

Our working memory is severely limited in both capacity (the amount of information it can hold) and duration. Cognitive psychology pioneer George Miller (1956) famously estimated short-term memory capacity at around 7 pieces of information. This is reasonably accurate for some tasks, such as simply holding on to uncomplicated information (e.g., digits), but it is an overestimate for more complex information and for tasks that require relating bits of information to each other. Modern consensus gauges the central capacity limit of human working memory at about three to five pieces of information (Cowan, 2010).

To get a feel for these limitations, consider a common working memory capacity measure, the digit span task. In this task, the test administrator reads a sequence of digits, and the task of the person being tested is to repeat the same digits back in the correct sequence immediately afterwards. The test starts with 2 or 3 digits, increasing stepwise until the person being tested starts to fail, at which point an estimate of individual capacity is established.

Below this paragraph is a sequence of digits. Read it once, then close the book or cover your reading tablet, and try to repeat the correct numbers in the correct sequence. Go on, try it!

$280619842617$

That was a sequence of 12 digits. Even under ideal circumstances, this far exceeds the normal limits of our working memory. Unless you cheated (or cleverly spotted a pattern—more on that soon), you would not have been able to retain all those digits and repeat them correctly after closing your book cover.

The mental strain involved in trying to retain too much information is almost like a physical sensation, and a vaguely unpleasant one at that. When reading the sequence of digits from left to right, our central storage limit is reached around the fourth to seventh digit. Reading subsequent digits, we feel the first few ones start to slip away from us. It is as if each new digit rudely shoves an old one out to make room for itself, despite our best efforts to hold on to them all.

As teachers, we are in constant danger of overburdening our students’ working memories in a similar fashion. And because IL teachers are often allowed only one or two sessions with each student group, that danger looms even larger for us. As we saw in Chapter 2, Information Literacy: The What and How, we often feel tempted, or compelled, to squeeze a lot of information into each session. There is so much students need to know, after all. The nature of working memory implies we should make an effort to resist that urge and focus our energies only on the most central concepts and skills.

While we argue in this book in favor of using student-centered, collaborative, activity based teaching (see Chapter 6, Teaching It All), there are still occasions when a teacher needs to provide information or instructions. If we try to serve up too much, or if we serve it up too quickly, students are likely to be unable to hold on to it long enough to process it properly, thus reducing their chances of transferring it to more permanent long-term storage.

3.1.4 Two modes of processing in working memory

Before considering the role of long-term memory in teaching and learning, we should briefly consider another feature of working memory.

During the 1970s, scientists studying working memory had participants perform more than one demanding task requiring its use (see, e.g., Baddeley, 2012, for a historical account). Given our limited capacity, this of course, is usually very hard to do. Researchers noticed, however, that dual task performance was reasonably good if the two tasks required different types of processing. Performing a phonological task (e.g., recalling digit sequences of varying length) together with a visuospatial task (judging the relative spatial location of two letters) led to only a modest performance reduction. On the other hand, performing two tasks relying on the same type of processing (e.g., attempting to recall digit sequences of varying length while repeating a simple word aloud—both dependent on phonological processing), caused more severe performance reduction. This led to the proposal that working memory is a multi-component system capable of somewhat independent processing of verbal/phonological information and visual/spatial information, respectively.

This means that an overburdened working memory, easy enough to inflict on students in the first place, is even more likely to occur if we present information such that different sources require processing of the same type. For instance, if we present verbal information visually while simultaneously presenting verbal information auditorily, students will need to process both using the verbal/phonological subcomponent of working memory. Interference is very likely, resulting in suboptimal processing of both the visual and auditory verbal information. This is what happens when presenters use slides with plenty of text while talking at the same time. Students will attempt to process both messages using the same working memory subsystem (the phonological), but they are likely to fail.

This empirically derived theoretical distinction between phonological and visuospatial subcomponents of working memory forms part of the basis for developments in the science of multimedia instruction; the promotion of understanding and learning using both pictures and words (Mayer, 2009). Generally, multimedia instruction more effectively promotes learning (both retention and problem solving transfer) than instruction relying exclusively on verbal or pictorial mediation. One reason is that utilizing both subsystems of our working memory leads to a de facto capacity increase, allowing more resources to be devoted to integration and active sense making. Another is that building coherent mental models of to-be-learned material is easier when it can rely on both verbal and spatial mental representations. However, the effectiveness of multimedia instruction is dependent on certain boundary conditions. Importantly, verbal and visuospatial information need to be integrated, i.e., presented together; spatially if words are printed, and temporally if they are narrated.

Being aware of this important property of our working memories can help us carefully consider how to coordinate visual–spatial and verbal information in expository teaching and when designing our own educational materials and resources. This in turn helps us avoid overburdening our students’ limited processing capacities, preventing confusion and promoting understanding.

3.1.5 Prior knowledge and long-term learning

There are, though, other, possibly even more powerful ways, in which we can indirectly strengthen our inherently limited working memories. Consider our 12-digit sequence from Section 3.1.3 again. While central working memory capacity is only about 4 items, there are usually features present in any given situation that will affect how much we are able to hold on to or process (Cowan, 2010). When asked to repeat a string of numbers, for instance, we may be able to rehearse some of them before our capacity limit is reached, or, more relevant to teaching and learning, we may be able to process the information by grouping or organizing it; by imposing on it some sort of structure. In fact, in any group of people challenged to repeat a 12-digit sequence, always a handful is actually quite successful. When asked how they managed to retain all the digits, their answer is invariably along the lines of “I grouped them into three sets of four” or “I thought I saw a day–month–year pattern in the first 8 digits.”

And in fact, there are patterns in the particular 12-digit sequence we tried to remember above. The first four digits are indeed in a day–month format, and June 28th is an important date in western history (the signing of the Versailles treaty in 1919). The next four digits are both a year and the title of a famous novel. The last four digits is the world record for running 10,000 m on a track. If we had hinted at what sort of patterns were hiding in that number sequence, would you have been able to remember it more easily? You most certainly would.

Why do hints like that help us? Note that recognizing or imposing patterns on new information depends on our long-term store of knowledge. If we had not learned how to keep track of time using calendars and clocks, and if we had not learned what sort of expressions of time would fit a given type of time keeping, these concepts would not be available to us as organizational devices, and we would not be able to hold on to as many digits. So, our little 12-digit experiment not only demonstrates the limits of working memory, it also exemplifies another very important general principle. Prior knowledge exerts strong influences on what and how much we are able to retain, process, and learn (see, e.g., Ambrose, Bridges, DiPietro, Lovett, & Morman, 2010; Hattie & Yates, 2014, for other expositions of this idea).

In fact, one of the most striking contrasts between novices and experts is the remarkable differences in the amount of information they are able to temporarily retain and process. This was first studied in chess players (Chase & Simon, 1973). Experienced players are capable of remembering board positions much more accurately than novices. This difference between experts and novices is much smaller, however, if board positions are random rather than meaningful (in the sense that they could have occurred during a game of chess). Hence, it is not that experts have larger working memory capacities per se. What they do have is a well-established and rich store of domain relevant long-term memories—in the case of chess, of strategically meaningful positions. This allows them to recognize, interpret, label, and chunk or subsume together into larger units what is actually a lot of complex information. In other words, our working memories can utilize existing long-term memories to make sense of, hold on to, and allow much more efficient processing of the information we are attending to. This strong dependence of efficient information processing on the availability of relevant long-term memories has led experts on expertise to label the phenomenon a “long-term working memory” (Ericsson & Kintsch, 1995).

This phenomenon—being able to virtually expand working memory capacity by finding patterns or ways to organize new information—has important implications for teaching. We will return to them in Chapter 6, Teaching It All, but let us first expand upon them a little bit here.

We mentioned above that piling on too much information in too little time is likely to cause working memory overflow. The digit span example provides us with another reason to remind ourselves to slow down. In order to find meaningful patterns in new information, we need time (and capacity) to try to fit it to concepts from our long-term memory. A careful consideration of how we segment our teaching sessions, making sure we occasionally assign some time for thought, will allow our students the breathing space they need to relate our message to prior knowledge.

Apart from portioning and pacing information and instructions when teaching, what else can we do to ease the pattern-finding processing of students when introducing them to new ideas and procedures? We can provide pointers as to what sort of organizational conceptual devices might be appropriate. This core idea was recognized early and is the basis of the concept of “advance organizers” (Ausubel, 1960). Providing advance organizers involves indicating concepts that can subsume new information. Usually, this means activating relevant prior knowledge, thus increasing the amount of information that can be held and manipulated in working memory, leading to increased probability of long-term retention. For instance, when introducing Boolean operators, then using familiar concepts from basic set theory may help some learners make sense of their role in database searching. For medically trained learners versed in the concepts of sensitivity and specificity as they relate to screening, conceptualizing a systematic search as a mass screening of a population of literature in order to identify the possibly “afflicted” individuals (i.e., the potentially relevant sources) can be a helpful metaphor. But advance organizers need not be metaphors to be effective. A simple classification scheme or story-like structure for a sequence of events can provide all the help learners need to chunk unfamiliar material.

This fascinating power of established long-term memories to aid working memory in the processing of new information we attend to highlights the reciprocal nature of the interplay of comprehension and knowledge, of understanding and memory. In the next section, we will see that understanding something is an important precondition for learning, and that trying to understand something for one self is probably a better approach to learning something for the long run than a rote learning approach geared toward reproduction. But while this is true, it is also the case that knowing something is an important precondition for understanding. We all make sense of new information only by applying facts, concepts, and schemas already stored in our long-term memory. This is important to keep in mind lest we, in our eagerness to promote understanding and higher order thinking, forget the value of actually remembering stuff. Noel Entwistle, one of the foremost advocates of teaching for understanding, reminds us of this: “In other words, memorizing often plays a supportive role in building up initial understanding, but also later on, ensuring that understanding is firmly lodged in the memory” (Entwistle, 2009, p. 32).

3.2.1 Consequences of learning approaches

Looking at the portrayal of the two learning approaches above, most IL teachers will immediately feel an affinity to the second one—the deep approach. After all, adeptly digging for evidence, critically appraising it, and relating it to the core ideas of an argument is arguably at the very heart of information literate activity.

In their original studies, Marton et al. noted two important differences in the kind and quality of learning resulting from the two approaches (Marton & Säljö, 1976a, 1976b). First and foremost, there was a correspondence between the approach students adopted and the content they typically remembered. That is, the two approaches led to retention of different information. Students adopting a deep level learning approach did indeed tend to remember more of the authors’ intended meaning, producing “conclusion oriented” answers, supported by relevant detail. Their summaries were more precise and displayed a fundamental understanding. Surface level studiers on the other hand tended to produce answers that were low on information, sometimes just restating the probe question, and occasionally providing rather exact recall of parts of the material they had studied.

Marton and Säljö also noted, however, that the two learning approaches seemed to be associated with different time courses of learning. Surface level studiers tended to remember a number of propositions from the studied material immediately after study, but had trouble remembering much at all on a delayed test 45 days later. Long-term retention (i.e., durable learning) was much better for propositions expressing the more fundamental content of the learning material. Interestingly, deep level learners also remembered long term more of the sort of detail that surface learners were able to remember only for the short term.

It seems, then, that students adopting a deep level approach to a learning task will be more likely to learn what are in fact the most important aspects of the materials they are working with, and also more of the supporting detail. Focusing on the main ideas or conclusions provides a frame that gives meaning to details, making them easier to remember over time. It is almost as if, in adopting a deep learning approach, students generate their own advance organizers or subsuming concepts (see Section 3.1.5), and can reap the benefits this entails for working memory processing and long-term memory encoding.

Going beyond the original findings of a different (and arguably better) quality of learning resulting from a deep level approach, several studies have examined the association between learning approaches and academic achievement as measured by exam results and grade point averages. This research seems to confirm what one would expect from the above descriptions of the learning approaches. Quite a number of studies have found a link such that a deeper approach is associated with better performance and/or that reliance on a surface approach is associated with poorer performance (e.g., Dennehy, 2014; Diseth, 2007; May, Chung, Elliott, & Fisher, 2012; Reid, Duvall, & Evans, 2007; Salamonson et al., 2013).

Some of the results from these studies are particularly noteworthy. In the study by Reid et al. (2007), the authors went to great lengths to plan and implement a medical education course designed to encourage a deep approach and prevent a surface approach. They formulated learning objectives with this in mind, and they carefully aligned teaching methods, assessment, and objectives in order to achieve this. (For more on the importance of alignment, see Section 6.2.1.) Measuring both learning approaches and achievement on three separate assessment types in several cohorts worth of students, they were able to test a considerable number of correlations. As expected, a deep approach more often correlated positively with achievement, while a surface approach tended to correlate negatively. However, while Reid and colleagues had expected to find this pattern for in-course essay assignments, specifically designed to draw out higher level thinking, and perhaps not so much for the MCQ (multiple-choice questions) exams, which are often considered tests of mere factual recall, they found exactly the opposite. That is, a deep approach was associated with better performance on the MCQ exam, but not on the essay exam. This is a reminder that we should be very careful not to assume that MCQs cannot measure advanced thinking, or that essay questions always do. In other words, there is no straightforwardly predictable correspondence between the kind of assessment used and the kind of learning approach that pays off. Perhaps then, using a variety of assessment methods is just as important as using a variety of teaching methods (see Section 3.4.1 on variety of teaching methods).

Another particularly interesting finding is from the study by May et al. (2012). These researchers found what is arguably one of the clearest associations between learning approaches and learning outcomes, as measured by an exam. The most remarkable bit is that they measured learning with a high stakes, multi-component clinical performance examination (again we are dealing with medical students), measuring complex knowledge applied in realistic settings, yielding scores on dimensions such as overall patient satisfaction, physical examination, information sharing, and physician–patient interaction. Most studies on learning approaches typically assess performance on more conventional verbal, usually written, examinations or learning measures. This study by May et al. allows us to be optimistic about the “real life” value of learning resulting from a deep approach to studying.

We have seen that the way in which we approach a learning task is likely to influence the amount, quality, and durability of our learning. Taking the deep level approach—an approach that seems well aligned with the ultimate goals of information literacy—tends to yield the better results, also reflected in exam results and grade point averages.

Still, a caveat is in order here. While most studies find an association between learning approaches and academic performance in the expected directions, others do so only for some of the correlations tested (Davidson, 2002; Reid et al., 2007), and in general, correlations tend to be small to moderate, rather than strong.

Why is this so? Why are the associations between learning approaches and measures of academic achievement not even stronger and clearer? After all, it seems to make perfect sense, perhaps especially to us IL teachers, that a deep approach should lead to better learning. Disregarding less interesting explanations that may apply to any and all failed attempts to relate grades and exam scores to anything at all (e.g., sampling error and the less than perfect reliability of grading and scoring schemes), one possible explanation suggests itself: a lack of alignment between stated goals, teaching methods, and assessments of student achievement.

Even if we have the very best intentions to support student understanding, and even if some students are inclined toward a deep level approach, deep learning may become invisible if assessments are mostly geared toward reproducing factual, descriptive knowledge. This in turn may quickly lead students to conclude that maintaining a deep approach, supported by IL, is just not paying off over the long run. For IL teachers, who may not be directly involved in designing assessments, this may become a particular instructional challenge. If we cannot show students where and how IL and deep learning become visible to the assessors of student performance, and how it is reflected in assessment scores or grades, we may not be able to motivate students to become information literate, even if we otherwise design our IL teaching to the highest standards. This suggests that working with teaching staff in the disciplines and departments to influence assessment design should be a priority.

Nevertheless, despite this caution, it still seems that a student’s approach to learning tasks and materials is likely to influence what she learns, the quality and durability of her learning, and her performance on assessments designed to measure that learning. Given this, we should turn to the question of whether and how we can influence student learning approaches.

3.2.2 Determinants of learning approaches

At the beginning of this chapter we briefly characterized our own view of teaching and learning as broadly constructivist, with a conception of teaching as an activity supportive of students’ self-directed learning, and a conception of learning as conceptual change brought about by active sense making. While these ideas hopefully make sense from a number of perspectives, they are not based solely on intuition or the teachings of great thinkers of the past. Nor were they introduced simply as an organizational aid or conceptual crutch to help you make sense of what we are trying to argue in this book. Most importantly, they were introduced because we believe that the conceptions we hold about teaching and learning really matter.

Research attempting to measure teachers’ conceptions or approaches to teaching seems to confirm this belief. Simplifying somewhat, conceptions of teaching can be construed as anchored in the extremes of a dimension from a teacher-centered focus on transmitting information on the one end, to a student-centered focus on conceptual change on the other (see, e.g., Postareff & Lindblom-Ylänne, 2008; Trigwell, Prosser, & Waterhouse, 1999). In a teacher-centered approach to teaching, the emphasis is on what the teacher does to convey facts and skills to students. Students’ prior knowledge is considered relatively unimportant and their role in the teaching-learning process need not be active. In a student-centered approach, on the other hand, the emphasis is on what the student does to learn, and on how the teacher can help them change their conceptions. Students’ prior knowledge is considered important, and in order for change to occur, students need to actively make sense of what is to-be-learned.

From our review of the human cognitive architecture, we already have every reason to believe that students’ prior knowledge, and whether and how it can influence their processing, is important for their learning. Holding this idea, along with the other components of a student-centered approach to teaching, may be advantageous for our students. Findings from some very interesting studies indicate that the conceptions of, or approaches to, teaching that teachers hold are indeed in turn related to the learning approaches students tend to adopt.

Gow and Kember (1993) and Kember and Gow (1994) identified two orientations to teaching, roughly parallel to the teacher-centered and student-centered conceptions outlined above. They called them the knowledge transmission orientation and the learning facilitation orientation. They then measured correlations between orientations to teaching aggregated on a departmental level and the approaches to learning generally adopted by students in these various departments. Notably, they found evidence of change in student approaches to learning over the course of their student careers, such that in departments where a knowledge transmission orientation dominated, students’ use of a deep level learning approach declined substantially over time.

Trigwell et al. (1999) found similar results at the level of individual teachers after surveying students and teachers in 48 first year higher education classes. They found that in classes where teachers adopted a teacher-centered, information-transmission approach to teaching, students reported adopting more surface approaches to learning. In classes where teachers were less teacher-centered in their approach, students tended to adopt less surface level and more deep level approaches.

This is evidence that conceptions of teaching matter. Hence, there is every reason for us IL teachers to seriously consider our own ideas about teaching and learning, and what our role as IL teachers is and should be. In the words of educational researcher John Hattie: “It is less what teachers do in their teaching, but more how they think about their role. It is their mind frames, or ways of thinking about teaching and learning, that are most critical” (Hattie, 2015, p. 81).

But teachers’ orientations are not the only influences on student approaches to learning. The research literature on learning approaches is rife with studies investigating a number of different factors and how they relate to learning approaches (see Baeten, Kyndt, Struyven, & Dochy, 2010, for a review). Most of these factors originate in a student perspective, and most have to do with the learning environment or, perhaps more accurately, with the learning environment as the student perceives it.

We all have some trouble accurately gauging the quality of our own learning. One example of this is the well-known Dunning–Kruger effect. This is a general tendency to overestimate our own competence, and it reveals itself in academic learning situations, as well as on other arenas (see, e.g., Dunning, Johnson, Ehrlinger, & Kruger, 2002; Kruger & Dunning, 1999). It is particularly pronounced when actual competence is low, hence we are more likely to overlook gaps in our knowledge when being exposed to unfamiliar ideas in a learning situation. This phenomenon helps explain why students often choose study strategies that provide cognitive ease, but are perhaps not the most effective (see Section 4.2). It may also explain why students sometimes voice resistance to student-centered teaching geared toward active learning.

Given this, it may seem paradoxical that students’ perceptions of their learning environment are related to their learning approach, and hence to the quality of their learning. However, there need be no contradiction here. While students’ may misjudge their own mastery, they are more likely to adopt a deep learning approach if they judge the quality of their teachers, and the teaching they do, to be high. Quality teaching in this context is, in the words of Paul Ramsden (2003), “teaching which is perceived to combine certain human qualities with explanatory skills…” (p. 73–74). Some of these aspects are teaching that involves providing feedback, being sensitive to student difficulties, displaying an interest in what students have to say, and enabling them to see the relevance of the subject matter.

Two particular aspects of perceived learning environment deserve special mention: (1) workload and (2) assessment. We have already mentioned assessment as a possible explanation for the sometime invisibility of the advantages of a deep level approach (see Section 3.2.1), and we have suggested that if assessments are perceived to require mostly simple reproduction of factual knowledge, then this may put students off a deep learning approach. However, studies that have looked for changes in student approaches to learning as a consequence of the adoption of assessment methods considered innovative and learner-centered, have disappointingly found shifts toward more surface level approaches (see Baeten et al., 2010, p. 247). Again, this highlights the complex relationships between assessment and learning approaches; there is no simple match of assessment type and the quality of learning. Perhaps the best we can strive for is assessment of IL concepts and skills that are congruent with the high-level goals of IL, with our own conceptions of teaching and learning, and that match the variety of teaching methods we employ (cf. Section 3.4.1).

Perceived workload has turned up repeatedly in research on the determinants of learning approaches. Results are generally quite clear. To the extent that students feel demands are excessive, they tend to adopt more surface approaches and less deep approaches to learning. Perhaps the mechanism is a need to find short cuts in order to cope (Kember, 2004). A detail of particular interest to IL practitioners is the finding that perceived information overload in an online course was associated with lower scores on deep approach and higher scores on surface approach measures (Svirko & Mellanby, 2008). This result seems to provide further support for the conclusions we arrived at in Section 3.1.3, i.e., that we should be careful to resist the natural urge of the IL teacher to pour piles of information into our IL sessions.

3.2.3 Encouraging deeper learning

Based on what we now know about approaches to learning and teaching, it may be possible to suggest one or two implications for IL teaching practice. First, our own ideas about what learning is and what teaching should be, do seem to influence our students’ approaches to learning. Hence, continually working to develop and refine our own conceptions of teaching and learning is likely to be a sound investment. Having said this, we should add that a certain type of conception, one that perhaps inclines us toward adopting teaching methods that allow and stimulate students to think deeply about what they are trying to learn, seems superior. Second, striving for and demonstrating quality in our teaching is likely to support a deep learning approach in our students. This striving probably involves keeping an eye on the alignment between learning goals, teaching approaches, and assessments. We will return to these themes in Chapter 6, Teaching It All.

3.3.1 Motivational patterns and mindsets

In research on motivation for learning, we commonly distinguish between two different motivational patterns, sometimes called a learning goal orientation and a performance goal orientation (see, e.g., Dweck, 1986). Students with a learning goal orientation engage in studying behavior and learning tasks with the primary aim of attaining mastery and increasing their competence. These students tend to hold a view of ability or intelligence as something malleable, something that can be tended and grown; what motivational scientist Carol Dweck describes as a “growth mindset.” This motivational pattern is usually adaptive. It is associated with choosing challenging tasks, with tolerance or even enjoyment of effort, and with persistence in the face of difficulties.

To students with a performance goal orientation, on the other hand, increasing mastery and competence is less important than gaining favorable, and avoiding unfavorable, evaluations of their competence. They tend to hold an entity view of intelligence or ability, i.e., seeing intelligence as something fixed. These students can apply themselves vigorously to tasks if they believe they have high ability and are likely to succeed. They may, however, experience the need to exert effort as threatening, perhaps because they see it as a sign of low ability. Hence, this motivational pattern is often maladaptive. It is associated with avoiding challenges and with more easily giving up in the face of difficulties.

The implications of this research seem to be that we should, if at all possible, try to teach in a manner that supports a growth mindset and a learning goal orientation. What might this involve in an IL teaching context? First, we should probably strive to espouse a conception of IL competencies as something that can be acquired through learning and practice, and that mastery of IL skills is not dependent on any inherent personal qualities. Such a conception seems self-evident to most IL teachers, but may not be so to the student who tends to operate under a performance goal orientation.

An interesting twist here arises from the fact that IL teachers frequently observe that many of the students they teach are overconfident. Several studies confirm this observation (e.g., Gross & Latham, 2012; Nierenberg & Fjeldbu, 2015). They may consider themselves expert searchers, be confident they can avoid plagiarism, and perceive their skills at critical source evaluation to be more than good enough, all while having large gaps in their actual knowledge and skills. These students are exhibiting symptoms of the Dunning–Kruger effect mentioned earlier. Hence, it is often an IL teacher’s unhappy predicament to crank them down a notch or two. This is risky. If a student’s first experience of IL instruction is one of defeat and frustration, she may attempt to ignore her knowledge gaps and avoid the effort required to seal them, especially if she tends to adopt a performance goal orientation.

Still, providing opportunities for students to confront their own, sometimes failed, attempts to master what they are trying to learn is probably necessary. And sometimes students need to come to the realization that they are not as competent at something as they imagined themselves to be. When setting up these situations, there are, besides conveying the conviction that mastery of IL skills is not dependent on any inherent personal ability, a couple of things we can do to discourage an entity view of IL mastery, and to encourage a growth mindset. First, avoid praising students’ intelligence. While successes should be celebrated, explicit judgements of being «clever» or «smart» or similar, are likely to encourage an entity view of competencies and a performance goal orientation. Second, we should look for opportunities to help students attribute their successes and failures to effort, or a lack thereof. For instance, when demonstrating in students an error of judgement with regard to what constitutes plagiarism, this moment of defeat can turn into an opportunity to learn, not only what constitutes plagiarism, but that a targeted effort to practice good judgement is necessary to build competence and that it can lead to success. Providing an opportunity to apply what has been learned from an erroneous response to one exercise to another, similar instance, can make those small moments of defeat pay off as subsequent success, while at the same time allowing us to point to the preceding effort and frustration as the immediate cause of successful mastery.

Another theoretical framework from the science of motivation that meshes well with most IL teachers’ views of IL and its role in academic life is self-determination theory. In brief, this theory holds that personal growth and well-being are promoted to the extent that satisfaction of three basic psychological needs—autonomy, competence, and relatedness—is supported by the environment. We will explore these ideas further in Chapter 5, Toward Academic Integrity and Critical Thinking.

3.4.1 Lessons from the Visible Learning synthesis

One of the most impressive efforts to summarize research on what works best in teaching and learning has been led by the New Zeeland educational researcher John Hattie. His approach has been to attempt to synthesize findings from meta-analyses (currently more than 1200; Hattie, 2015) of studies on the influences on student achievement. From this exceedingly rich and complex material, we think there are important lessons to be learned.

At the highest level of abstraction, the main lesson from Hattie’s (2009, 2011, 2015) synthesis is that the key to effective teaching is making student learning visible. Hattie (2011) identifies three themes he believes should form the premises on which to develop teaching that promotes visible learning.

First, teachers should communicate clear learning intentions and criteria for success. In other words, teachers should clearly describe what students are supposed to learn, and what it looks like when this learning has been attained. Earlier in this chapter (see Section 3.1.1), we saw that clear, specific learning goals make perfect sense from an information processing perspective. They help focus attention from the top down, allowing us to ignore irrelevant stimuli, both external and internal, leaving precious working memory capacity for processing that leads to learning. Importantly, clear learning goals form the basis on which both students and teachers can evaluate the effectiveness of the teaching, and they should guide the teacher’s choice of instructional content and technique. Last, but not least, the learning goals should guide the assessment of student learning. Learning goals should be challenging (Hattie, 2011), and provide a proper balance between deep and surface learning (Hattie, 2015). In Section 6.2.1, we take a closer look at the specifics of formulating learning outcomes and providing alignment between outcomes, teaching methods, and assessment.

Clear learning intentions and success criteria comprise a necessary condition for the second of the three themes: seeking feedback about the effectiveness of our teaching and providing feedback to students about the effectiveness of their learning. The primary function of feedback is to decrease the distance between students’ current mastery and the stated learning intentions (the learning outcome statements) of the session or course. Feedback interventions, while evincing some of the largest effects sizes in the academic achievement literature, are also characterized by the largest variability. It seems, then, that providing feedback is easy to get wrong, but very effective when done right. Providing feedback in the form of reward, punishment, or praise is not likely to be effective. Rather, effective feedback ensues when both students and teacher look for answers to what Hattie (2011) calls the major feedback questions: Where am I going? How am I doing? Where to next? We will return to the topic of evaluation of teaching effectiveness and the provision of feedback from assessments in Section 6.4.

The third foundational theme for visible learning is using a variety of teaching methods that emphasize student perspectives, and that support their development as self-regulated learners. This involves teaching in ways that help students obtain control over the cognitive processes involved in learning, i.e., that help them develop active, metacognitive learning strategies. This visible learning theme matches one of the major premises of this book. That the scope of IL teaching in higher education should include learning strategies. Given its centrality to our message, we devote an entire chapter (see Chapter 4, Learning Strategies) to learning strategies in an IL perspective.

3.4.2 Active and collaborative learning improves student achievement

In Section 3.2.2, we saw that when teachers conceive of learning as a process of conceptual change that requires students to be active, to interact and discuss, then students are more likely to report using a deep level approach to learning (Trigwell et al., 1999). While we may not know exactly the mechanisms involved, one likely causal path is via the use of teaching methods that foster active learning. Active learning can be loosely defined as any instructional method “… that engages students in the learning process (…)” and that “requires students to do meaningful learning activities and think about what they are doing” (Prince, 2004, s. 223). Probably, teachers with this conception of learning are more likely to adopt interactive teaching styles.

And indeed, one of the clearest and most important findings to emerge from research on teaching and learning over the last few decades is that more and better learning is attained when the learner is somehow actively engaged in the learning process itself (see, e.g., Freeman et al., 2014; Prince, 2004). The meta-analysis by Freeman et al. summarized the results of more than 150 studies comparing some form of active learning instructional technique to lecture-based teaching control conditions. The results are compelling, with substantial average gains in examination scores and grades, as well as large reductions in failure rates (students in lecturing classes were 1.5 times more likely to fail than students in active learning classes). From a cognitive perspective, these results should not be surprising. We know that the limitations of both attention and working memory create problems when students try to focus on and absorb information in a passive, receptive mode, such as they may experience in a traditional lecture. And we know the benefits of taking the time to relate to-be-learned information to established prior knowledge.

One particular variety of active learning, in which students work together in groups, sometimes called collaborative learning, is worth special mention. A meta-analysis of 168 studies comparing the effects of cooperative learning (a type of collaborative learning) to those of individual or competitive learning on academic achievement in university or adult students found substantial average gains in favor of collaborative learning teaching methods (Johnson, Johnson, & Smith, 2014). The studies were drawn from a number of different disciplines, and used a variety of achievement measures, capturing both lower level and higher level cognitive processes. Importantly, Johnson et al. analyses indicate that cooperative learning positively influences other valuable educational objectives besides those of academic achievement. One such is the quality of interpersonal relationships, measured as interpersonal attraction, group loyalty, social cohesiveness, and trust. These authors also argue convincingly that it is through collaborative learning that students develop into members of an academic community.

Several other systematic reviews strengthen our confidence that structured learning activities in an inherently social setting (e.g., collaborative, cooperative, small-group, and team-based learning) usually contribute positively to student learning and achievement (Burgess, McGregor, & Mellis, 2014; Pai, Sears, & Maeda, 2015; Tomcho & Foels, 2012).

We argue in this book that the scope of IL teaching should widen to include learning strategies and values, and to place more emphasis on IL as an integral part of learning how to learn. Adopting collaborative learning may provide one of the best avenues of approach to achieve that goal, by very elegantly and implicitly supporting the process without the need to preach the morality of academia in a teacher-tells-and-students-absorb style. The active thinking required and allowed in well-structured cooperative learning groups may also supply an environment that naturally plays to the strengths of the human psyche, while minimizing the impact of our cognitive limitations (Kirschner, Paas, & Kirschner, 2009). We will return to the practicalities of implementing active and collaborative learning in IL teaching in Chapter 6, Teaching It All.