Adv in Health Sci Educ DOI 10.1007/s10459-014-9573-x
Manipulation of cognitive load variables and impact on auscultation test performance Ruth Chen • Lawrence Grierson • Geoffrey Norman
Received: 20 January 2014 / Accepted: 20 November 2014 Springer Science+Business Media Dordrecht 2014
Abstract Health profession educators have identified auscultation skill as a learning need for health professional students. This article explores the application of cognitive load theory (CLT) to designing cardiac and respiratory auscultation skill instruction for senior-level undergraduate nursing students. Three experiments assessed student auscultation performance following instructional manipulations of the three primary components of cognitive load: intrinsic, extraneous, and germane load. Study 1 evaluated the impact of intrinsic cognitive load by varying the number of diagnoses learned in one instruction session; Study 2 evaluated the impact of extraneous cognitive load by providing students with single or multiple examples of diagnoses during instruction; and Study 3 evaluated the impact of germane cognitive load by employing mixed or blocked sequences of diagnostic examples to students. Each of the three studies presents results that support CLT as explaining the influence of different types of cognitive processing on auscultation skill acquisition. We conclude with a discussion regarding CLT’s usefulness as a framework for education and education research in the health professions. Keywords
Cognitive load theory Auscultation skill Instructional design Near transfer
Background In preparation for future clinical practice, students in health professional programs acquire a range of skills during their training and education. Cardiac and respiratory auscultation
R. Chen (&) Faculty of Health Sciences, School of Nursing, McMaster University, 1280 Main Street West, Hamilton, ON HSC 2J34H, Canada e-mail: [email protected] L. Grierson Department of Family Medicine, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada G. Norman Department of Clinical Epidemiology and Biostatistics, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada
123
R. Chen et al.
skill development forms part of the core competencies of many health professions, including students in undergraduate nursing education programs. However, health professional students have demonstrated weakness with auscultation skills in both the cardiac and respiratory systems, and educators agree that this requires greater focus in health professions curricula (Holmboe 2004; Mangione and Nieman 1997, 1999; Mangione 2001). It is our position that cognitive psychology will prove useful in guiding the development of interventions that will improve student auscultation skills. We have identified cognitive load theory (CLT) as particularly relevant. Cognitive load theory relies heavily on the construct of a human working memory resource: a cognitive space responsible for the active integration of sensory-perceptual information and long term memory (Mayer 2002) for the purposes of active computation, decision making, and performance control (Baddeley 1992; Schnotz and Ku¨rschner 2007). Importantly, working memory is understood to be constrained and limited with respect to the amount of information—or cognitive load—it can process at any one time (Van Merrienboer and Ayres 2005). The critical implication is that if the cognitive load exceeds the capacity limits of working memory, then performance of tasks dependent on working memory will suffer. With respect to skill learning, this implies that instruction that exceeds the capacities of a learner’s working memory will not be retained (Sweller 1994). Thus, consideration of the cognitive load imposed on a learner during instruction can help educators and curricula developers understand why some interventions facilitate learner success (Mousavi et al. 1995; Paas et al. 2003) whereas others lead to ineffective learning (Doolittle et al. 2005). Cognitive load theory describes three types of cognitive load: intrinsic, extraneous, and germane (Van Merrie¨nboer and Sweller 2005). Intrinsic cognitive load (ICL) speaks to the information processed specifically as part of the learning task and is influenced by factors such as the learner’s level of expertise and the inherent difficulty of the learning goal. Extraneous cognitive load describes the features of instruction that are neither relevant nor necessary for the particular learning goal or task and that impose an unnecessary burden on working memory (Sweller et al. 1998). Germane cognitive load also contributes additional cognitive burden to working memory processes, but this load is considered useful as it serves to facilitate a learner’s understanding and automation of the information received during instruction into long-term memory (Paas and Van Merrie¨nboer 1994). Taken together, CLT explains that student learning can be optimized through instruction that maintains intrinsic cognitive load within the limits of a learner’s working memory, decreases extraneous load, and facilitates the type of processing associated with germane cognitive load. Importantly, the amount of working memory capacity occupied by intrinsic load is not a fixed quantity. Rather, it changes for any given task as a function of learner expertise. Specifically, as a learner progresses along the continuum of expertise, the integration of information into long-term memory is accomplished by way of schema construction and automation, chunking the invariant features of a task so that the total processing impact on working memory is reduced and more capacity is freed up for additional processing (Artino 2008; Sweller 1988). The dependence of ICL on level of expertise highlights the importance in understanding the characteristics of the learner when determining whether an instructional design will contribute to extraneous cognitive load or foster germane cognitive load (Leahy and Sweller 2005). Accordingly, the increased attention that CLT is receiving as a useful framework in the health professions education literature (Leppink et al. 2013; Mayer 2010; Pociask et al. 2013; Van Merrie¨nboer and Sweller 2010) provides
123
Cognitive load auscultation skill
researchers an opportunity to evaluate how cognitive load variables interact to facilitate learning within students at various levels of expertise (Van Merrie¨nboer et al. 2006). The present studies reflect a first effort in accomplishing this. In particular, this work investigates the effects of manipulating cognitive load variables on the auscultation performance of novice learners. The tests used in these studies assess performance using several measures. While overall diagnostic success is the primary outcome measure to assess learning of cardiac and respiratory sounds, measurements also include both retention and transfer performances. For the following studies, near transfer refers to a participant’s ability to identify a different example of a diagnostic sound that was not previously heard. For example, if a participant heard an example of a ‘‘wheeze’’ during instruction, they would demonstrate near transfer if they correctly identified a different example of a ‘‘wheeze’’ during the auscultation test. Additionally, while we tested both cardiac and respiratory systems in the auscultation tests, the primary objective was not to report on the differences in performance between the two body systems. We wished to evaluate the cardiac and respiratory systems to assess whether the outcomes and trends in performance would be similar or consistent between two systems of auscultation instruction. The goal of these studies is to understand how, or if, instructional design modifications impact learner performance and skill maintenance for auscultation learning.
Study 1: Modification of intrinsic cognitive load (ICL) and impact on auscultation test performance When designing instructional content for learners, educators should consider the optimal ICL to achieve desired learning goals. One approach to modifying ICL is to reduce the number of components of a learning task that a learner must retain simultaneously in working memory (Bannert 2002; Sweller 2010). Educators can then add complexity to the learning task at later stages of learning. However, ecological validity is a concern: as complexity is decreased for a learning task and the instruction sessions cover less and less content, there will be a point where the content of each session is unrealistically minimal and the number of sessions required to cover all the desired content is infeasible. Determining how much content to cover in an instructional session is difficult with complex tasks like auscultation skills. Educators must balance the effects of reducing the scope and complexity of the instructional session with the need to cover all the necessary learning material. Reducing the number of diagnoses per instructional session would likely improve performance. However, would doubling the number of diagnoses learned in one instructional session necessarily result in significantly poorer performance? The purpose of this first study was to evaluate performance outcomes by manipulating the intrinsic cognitive load (ICL) of the learning goal, as measured by the number of diagnoses presented in one instructional session. Furthermore, the goal was to evaluate whether there were additional benefits of minimizing ICL that resulted in enhanced performance outcomes for near transfer and on delayed testing. Research question Is there a significant difference in cardiac and respiratory auscultation test performance when intrinsic cognitive load is modified by instruction on either three or six diagnoses?
123
R. Chen et al.
Methods Participants Senior level undergraduate nursing students (n = 44) were recruited for this study. Participants were provided with both Cardiac (n = 21) and Respiratory (n = 23) sounds instruction. Participants were randomized to receive instruction on either three (‘‘3 Group’’) or six (‘‘6 Group’’) diagnoses. Respiratory instruction occurred first. Participants were given an auscultation test immediately following instruction and then a Respiratory retest 1 week later. Following the Respiratory retest, participants then received Cardiac instruction and a post-instruction Cardiac auscultation test; we were not able to conduct a Cardiac retest 1 week later due to logistical issues with student schedules. Study design (see Fig. 1) Six diagnoses were used for the Respiratory and Cardiac sounds interventions. A group of clinical experts was consulted to determine which three diagnoses from each body system were most difficult to learn. Based on the feedback of the clinical experts, three Cardiac and three Respiratory diagnoses were identified as most difficult of the six. The 3 group received instruction on these three diagnoses. The 6 group learned all six diagnoses. The rationale for selecting the three most difficult Cardiac or Respiratory diagnoses was to minimize superior performance by the 3 group participants simply because they had easier diagnoses to learn. Time for learning each diagnosis was standardized across both the 3 and 6 instruction groups. For each diagnosis learned, each participant was provided with 2 min of auditory instruction for each sound (i.e. listening to the cardiac or respiratory sound) followed by a 10 s rest period before proceeding to the next diagnosis. During the 2-min instruction period, participants were exposed to three different examples of the diagnosis. These diagnosis examples were recorded audio (i.e. mp3) files from a database of cardiac and respiratory sounds, obtained from the Internet through non-proprietary sources. We wanted examples that would not require an educator to purchase any particular vendor’s materials. Therefore, a research assistant went online and looked for audio examples of the sounds we wished to use for instruction and testing. These sounds were then recorded and formatted for our instruction. Brief written descriptions accompanied the auditory examples of all sounds provided; the written descriptions were identical for the 3 and 6 groups. Auscultation test Following the instruction period, participants were tested with four different audio examples of each of the diagnoses learned. Two examples were taken from the sounds presented during the instruction period and evaluated participant recall (OLD sounds); two examples were not previously heard and evaluated participant near transfer (NEW sounds). Therefore, the participants in the 3 group were tested on twelve (12) total sounds and the participants in the 6 group were tested on twenty-four (24) sounds. All audio examples were shuffled and a Test CD was created with the audio files. Auscultation test performance scores were calculated based on percentage of total sounds identified correctly (0–100 %). The post-instruction test and retest content was identical.
123
Cognitive load auscultation skill
Randomization (n = 23) Respiratory “3” Instruction (n = 11)
Respiratory “6” Instruction (n = 12)
Sounds Used for Instruction and for Test/Retest Respiratory “3”
Respiratory TEST (post-Instruction) Respiratory RETEST (1 week later)
Cardiac “3” Instruction (n = 10)
Cardiac “6” Instruction (n = 11)
Cardiac TEST (post-Instruction)
Crackles Rhonchi Stridor Respiratory “6” Above, plus Bronchial breath sounds Vesicular breath sounds Wheeze
Cardiac “3” Aortic Insufficiency Atrial Septal Defect Mitral Valve Prolapse Cardiac “6” Above, plus Aortic Stenosis Normal heart sounds Ventricular Septal Defect
Fig. 1 Study design: 3 versus 6 instruction
Ethics The joint Research Ethics Board of McMaster University Faculty of Health Sciences, Hamilton Health Sciences, and St. Joseph’s Healthcare Hamilton approved all studies. All participants provided written and informed consent to participate. Statistical analysis IBM SPSS Statistics version 21.0 was used for the following data analyses. In the analysis of the post-instruction auscultation test, a 2 9 2 9 2 mixed ANOVA was conducted, with system (Cardiac, Respiratory) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (3 group, 6 group) treated as a between subject factor. In the analysis of the Respiratory system test and retest, a 2 9 2 9 2 mixed ANOVA was conducted, with test (post-instruction, retest 1 week later) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (3 group, 6 group) treated as a between
123
R. Chen et al.
subject factor. Alpha level was set at 0.05. Performance outcomes on only the three diagnoses common both to the 3 group and the 6 group were used in the analyses to ensure comparability between groups. Performance of the 6 group on the additional diagnoses was not analyzed; this allowed direct comparison of the effect of increasing intrinsic load on auscultation test performance. Results Overall means (SD) for auscultation test performance of the 3 and 6 groups for both analyses are found in Table 1. Post-instruction test comparison of 3 and 6 groups, for Cardiac and Respiratory sounds As predicted, there was a significant difference in auscultation test performance between the 3 and 6 instruction groups (see Table 1). The 3 group outperformed the 6 group, F(1,21) = 18.06, p \ 0.001 across both cardiac and respiratory groups [Interaction F(1,19) = 0.047, p = 0.83]. There was no interaction between OLD/NEW sounds and the 3 or 6 instruction group F(1,19) = 0.33, p = 0.58. Test–retest auscultation test comparison of 3 and 6 groups for Respiratory sounds The main effect of group demonstrated a significant difference between the Respiratory 3 and 6 instruction groups’ performance across both test administrations with the 3 group achieving higher scores than the 6 group F(1,19) = 26.35, p \ 0.001. There was a significant decline in test performance between Test and Retest F(1,19) = 31.02, p \ 0.001, and a significant three-way interaction between 3–6 group, Old–New sounds, and Test– Retest performance F(1,19) = 23.74, p \ 0.001. Data were analyzed further to understand the three-way interaction; OLD sounds and NEW sounds were analyzed separately. Analysis of the OLD Respiratory sounds showed a significant difference between 3 and 6 groups F(1,19) = 12.01, p = 0.002. There was a significant decline in scores from Test to Retest F(1,19) = 52.54, p \ 0.001, and a significant interaction between 3–6 group and Test–Retest performance with F(1,19) = 34.73, p \ 0.001, indicating that the performance of the 6 group declined more than the 3 group. Analysis of NEW Respiratory sounds showed a significant advantage for the 3 group across both tests, F(1,19) = 19.34, p \ 0001, but there was not a significant decline Table 1 Post-instruction Cardiac and Respiratory auscultation test scores, and respiratory retest scores between instruction groups Cardiac subscore Mean (SD)
Respiratory subscore Mean (SD)
Test
Test
OLD
NEW
‘‘3’’ Group instruction 75.76 (21.3) 81.6 (14.6)
123
NEW
71.21 (18.5) 73.3 (21.2)
‘‘6’’ Group instruction 58.33 (22.3) 61.3 (20.2)
OLD
Retest
83.3 (15.2) 81.6 (14.6)
NEW
59.17 (19.7) 58.3 (14.5)
55.83 (19.1) 53.4 (13.5)
OLD
78.3 (24.20)
60.0 (20.5)
34.17 (19.4) 30.0 (13.2)
33.3 (13.6)
38.6 (13.9)
Cognitive load auscultation skill
between Test and Retest F(1,19) = 0.66, p = 0.43, and no 3–6 group and Test–Retest interaction F(1,19) = 0.29, p = 0.60. The three-way interaction, therefore, was carried predominantly by the OLD Respiratory sounds performance on Retest by the 6 group, which dramatically fell from Test to Retest. Summary of results For the three diagnoses common to both the 3 group and the 6 group, the 3 group consistently demonstrated superior auscultation test performance on the post-instruction test. When evaluating Respiratory auscultation test performance on retest, the 3 group continued to demonstrate superior performance than the 6 group. Of note, whereas the 6 group performed significantly worse in recalling OLD sounds on retest, there was no significant decline in test performance on NEW sounds between test and retest. Discussion This study explored the effect of adjusting ICL for learning cardiac and respiratory auscultation sounds in novice learners. As expected, those students with fewer diagnoses to learn demonstrated better performance on auscultation test performance than students with twice as many diagnoses to learn. Increasing intrinsic load by simply increasing the quantity of material to be learned yielded the expected outcomes. However, an interesting finding was that, when looking at performance on NEW sounds between Test and Retest, there was no significant attenuation in near transfer performance for both the 3 group and the 6 group. These results suggest that, even though the 6 group learned twice as many diagnoses, the near transfer performance between test and retest was comparable for both instructional groups. For cardiac and respiratory auscultation skill development, it is also necessary to address other cognitive load components—in particular, strategies that decrease extraneous load and facilitate germane load—to promote learning and transfer while intrinsic load remains constant. Manipulating extraneous and germane cognitive load variables will be the focus of the next two studies.
Study 2: Adding extraneous cognitive load (ECL) or fostering germane cognitive load (GCL) through multiple representations? Once an instructional goal has been determined, CLT also supports removing aspects of the instruction that are unnecessary for learning, factors that contribute extraneous cognitive load. Decreasing extraneous load is less important for learning tasks that do not place a high intrinsic load on a learner’s working memory because there are additional working memory resources available to help the learner process both the necessary intrinsic load and the unhelpful extraneous load (Van Merrie¨nboer and Sweller 2005). However, if the intrinsic load of a learning task increases, then it is important to decrease extraneous load so that the total cognitive burden does not exceed the learner’s working memory capacity. The goal in instructional design is to use the remaining cognitive resources available in working memory by decreasing extraneous load and by facilitating germane load for the particular instructional goal. Just as there are no standard metrics to quantify the intrinsic cognitive load of an instructional task without having to consider the expertise level of the learner, so too, what
123
R. Chen et al.
constitutes germane load for one level of learner may be unnecessary, extraneous cognitive load for another (McKeough et al. 1995; Paas and Van Merrie¨nboer 1994; Quilici and Mayer 1996, 2002; Scheiter et al. 2004). One such example in the literature is the use of multiple examples or representations in instructional design. Providing multiple examples during instruction helps learners distinguish relevant from irrelevant features (Van Merrie¨nboer and Sweller 2005) and facilitates transfer to new contexts (McKeough et al. 1995; Reed and Bolstad 1991). The use of multiple examples, or representations, for instruction has also been found to be beneficial in health professions education with cardiac physiology instruction (Norman et al. 2007; Norman 2009). However, educators must exercise caution when using multiple examples in instruction because requiring learners to process multiple representations may increase extraneous cognitive load and may, therefore, be detrimental to learning (Sankey and Nooriafshar 2005). In Study 1, all learners were provided with three examples of each diagnosis, whether they were randomized to the 3 group or the 6 group. This study explores whether providing only a single example to this group of novice learners might result in decreased performance, especially on near transfer, thereby suggesting that multiple examples were adding germane load and were therefore beneficial for auscultation test performance. Conversely, if learners provided with only a single example of each diagnosis showed superior performance in comparison to the group provided with multiple examples, these results would suggest that the multiple examples contributed extraneous cognitive load on learners. Research question How does providing multiple examples, or representations, of cardiac and respiratory diagnoses impact on auscultation test performance when compared with providing only single examples of diagnoses? Methods Participants Senior level undergraduate nursing students (n = 32) were recruited for this study in June– July 2011. Participants were randomized to receive instruction with a single example (Single group) or with multiple examples (Multiple group) of the Cardiac and Respiratory diagnoses. Study design (see Fig. 2) Eight diagnoses were used for the Single and Multiple sounds interventions. The audio examples provided to the Single group consisted of recorded audio (i.e. mp3) files from the pediatric Laerdal human patient simulators. These audio files were compiled into an instruction CD for the Single group. The audio examples provided to the Multiple group consisted of three audio examples of each diagnosis. One example was taken from Laerdal simulators, identical to what the Single group heard. The two additional examples were taken from a database of cardiac and respiratory sounds. Brief written
123
Cognitive load auscultation skill
Randomization (n = 32)
Single Instruction Group (n = 16)
Multiple Instruction Group (n = 16)
Sounds Used for Instruction and for Test/Retest Cardiac
Auscultation TEST (post-Instruction) Auscultation RETEST (1 week later)
Normal heart sounds Aortic Insufficiency Aortic Stenosis Mitral Valve Prolapse Respiratory Bronchial breath sounds Crackles Stridor Wheeze
Fig. 2 Study design: single versus multiple instruction
descriptions accompanied the auditory examples of all sounds provided; the written descriptions were identical for the Single and Multiple groups. Time for learning each diagnosis was standardized across both the Single and Multiple instruction groups. For each diagnosis learned, each participant was provided with 2 min of auditory instruction (i.e. listening to the cardiac or respiratory sound) followed by a 10 s rest period before proceeding to the next diagnosis. During the 2-min instruction period, participants were exposed to either one example of the diagnosis (Single group) or three different examples of the diagnosis (Multiple group). Auscultation test Following the instruction period, participants were tested with four different examples of each of the diagnoses learned. For the Single group, the example heard during instruction was included to test for recall (OLD sound) and three additional examples which were not previously heard tested for near transfer (NEW sound). For the Multiple group, two examples were taken from the sounds presented during the instruction period and evaluated participant recall (OLD sounds); two examples were not previously heard and evaluated participant near transfer (NEW sounds). All audio examples used for the auscultation test were shuffled and a Test CD was created with the audio files. Auscultation test performance scores were calculated based on percentage of total sounds identified correctly (0–100 %). Participants completed both the post-instruction test and a retest of the same
123
R. Chen et al.
sounds 1 week following instruction. As with Study 1, the retest content was identical to the post-instruction test content. Statistical analyses IBM SPSS Statistics version 21.0 was used for the following data analyses. A 2 9 2 9 2 mixed ANOVA was conducted, with system (Cardiac, Respiratory) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (Single, Multiple) treated as a between subject factor. Alpha level was 0.05. Results Overall means (SD) Auscultation Test performance scores for the Single and Multiple instruction groups are reported in Table 2. There was no significant difference between Single and Multiple group scores F(1,28) = 0.000, p = 0.99. Furthermore, there was no significant decline in scores from Test to Retest F(1,28) = 0.13, p = 0.73; there was also no interaction between the Single or Multiple instruction group and Test/Retest scores. No further interactions were found between instruction group with Cardiac or Respiratory performance as well as with recall (OLD) or near transfer (NEW) performance. Discussion These results suggest that for novice nursing students, multiple representations of cardiac and respiratory auscultation diagnoses may either contribute to extraneous cognitive load or may have no additional instructional benefit over providing students with single examples of the diagnoses. Extraneous load involves all the irrelevant features in the instructional design, but also includes aspects of the instruction that increase/exceed working memory capacity without facilitating schema construction and automation (Sweller 1994). Therefore, for the novice learners we used in these studies, being exposed to multiple examples of a cardiac diagnosis, for example, did not introduce ‘‘irrelevant features’’ per se, but rather, the multiple examples may have introduced greater complexity that the students’ working memories were unable to process. It is important to note that students were not provided with any additional instruction around the diagnoses presented. Therefore, there was no instruction regarding how these
Table 2 Auscultation test scores, cardiac and respiratory test and retest subscores between instruction groups Total score Mean (SD)
Cardiac subscore Mean (SD)
Respiratory subscore Mean (SD)
Test
Retest
Test
Retest
Test
Retest
Single instruction group
62.55 (10.8)
60.82 (14.9)
61.77 (14.16)
62.50 (21.2)
62.61 (13.2)
59.14 (14.2)
Multiple instruction group
59.81 (23.4)
63.33 (20.4)
59.14 (29.6)
60.58 (26.4)
60.48 (22.8)
66.09 (17.8)
123
Cognitive load auscultation skill
examples highlighted the central features of each cardiac or respiratory diagnosis. These results underscore, perhaps, the need for further explanatory information to accompany the auditory and basic written instructions when providing multiple examples during instruction. The non-significant results, however, could also be due to a type-II error as a result of an insufficient sample size. Power calculations were performed which demonstrated that for a sample size of 15 participants per instruction group, a difference of 15 % points in test performance could be detected with 80 % power. Therefore, the study results must be interpreted with caution. While there is sufficient justification in the literature around the use of multiple representations with instruction, the findings were not replicated in this learning context and with this level of learner. The next study investigates another approach that has been discussed in the literature as a strategy to facilitate germane load: interleaving examples during instruction.
Study 3: Fostering germane cognitive load (GCL) through an interleaving (or, mixed format) instructional approach The advantages of interleaving or mixed instruction, where practice examples from multiple categories are mixed together, has been well documented in the cognitive psychology and medical education literature, and research on this approach spans several decades (Hatala et al. 2003; Richland et al. 2004). In the field of auscultation skill development, interleaving would involve including examples from two or more diagnoses during instruction. This has also been referred to as mixed instruction. Instruction with all examples from each diagnosis presented in succession before moving on to the next diagnosis would be an example of blocked instruction. Interleaving approaches in education were first documented over two decades ago for motor skills acquisition (Shea and Morgan 1979). Further studies in motor learning followed and the benefits of the interleaving approach, alternatively termed ‘‘mixed practice’’ or ‘‘practice involving contextual interference’’, have been demonstrated for a variety of motor skills instruction, from children throwing beanbags (Carson and Wiegand 1979), to baseball players’ batting skills (Hall et al. 1994), to basketball players’ basket-shooting skills (Landin and Hebert 1997). In all the above examples from the motor learning literature, interleaving instruction demonstrated benefits. Blocked instruction may yield better outcomes during the practice session over mixed instruction (Bjork 1994, 1999; Rohrer and Taylor 2007; Taylor and Rohrer 2010) and learners also perceive blocked instruction to be more effective for their learning than mixed instruction (Birnbaum et al. 2012; Kornell and Bjork 2008; Kornell et al. 2010; Simon and Bjork 2001). However with passage of time, the benefits of mixed instruction are revealed. Therefore, even as learners believed that the blocked instruction format was more effective for their own learning and the mixed instruction group perceiving the instructional format to be less effective, their performance results demonstrated the opposite outcomes. In cognitive psychology, the interleaving instructional approach has been termed a ‘‘desirable difficulty’’ in learning (Bjork 1994), where cognitive load is increased for the learner, but in a manner that facilitates rather than hinders learning. Inducing these ‘‘desirable difficulties’’ in learning facilitate GCL even as instruction is more taxing to working memory when compared with a blocked instruction approach (Rohrer and Pashler 2010). The benefits of the interleaving approach have been found with instruction around differentiating painting styles (Kang and Pashler 2012; Kornell and Bjork 2008), learning bird and butterfly species (Birnbaum et al. 2012), understanding astronomy concepts
123
R. Chen et al.
(Richland et al. 2005), and applying mathematical concepts (Mayfield and Chase 2002; Rohrer and Taylor 2007; Taylor and Rohrer 2010). In health professions education, interleaving, or mixed practice during instruction, has been helpful in the instruction of psychiatric diagnoses (Zulkiply et al. 2012) and in electrocardiogram interpretation (Hatala et al. 2003). While the advantage of interleaving or mixing has been documented with motor and cognitive skills acquisition, questions remain whether this approach would also be beneficial for learners receiving auditory instruction for cardiac and respiratory auscultation skill development. The theoretical rationale for the benefits of interleaving for cognitive goals is that this approach encourages learners to make contrasts between the examples of the learning goals provided. The following study explored whether the interleaving approach for auscultation skill development would similarly allow novice health professional learners to make auditory contrasts between examples to enhance learning of cardiac and respiratory diagnoses. Research question Does providing instruction of cardiac and respiratory diagnoses in either a mixed or blocked format have an impact on auscultation test performance? Methods Participants Senior level undergraduate nursing students (n = 22) were recruited for this study in May– June 2012. Participants were randomized to receive instruction with diagnoses examples presented in an interleaved format (Mixed group) or with diagnoses presented together in succession (Blocked group) of the Cardiac and Respiratory diagnoses. Study design (see Fig. 3) Eight total diagnoses were used for the Mixed and Blocked group interventions. For the Blocked instruction group, three examples of each diagnosis were provided in succession. All three examples of one diagnosis were played before moving to next diagnosis (i.e. Diagnosis 1, examples A, B, C; Diagnosis 2, examples A, B, C; etc.). For the Mixed instruction group, three examples of each diagnosis were provided. However, the diagnoses were presented in pairs and the examples of each diagnosis were interleaved (i.e. Diagnosis 1, example A; Diagnosis 2, example A; Diagnosis 1, example B, C; Diagnosis 2, example B, C; Diagnosis 3, example A; Diagnosis 4, example A; Diagnosis 3, example B, C; Diagnosis 4, example B, C, etc.). The audio examples were audio files taken from Laerdal simulators and from a database of cardiac and respiratory sounds. All examples for each diagnosis were the same for the Mixed and Blocked instruction group; only the order in which the examples were presented was changed. Brief written descriptions accompanied the auditory examples of all sounds provided; the written descriptions were identical for the Mixed and Blocked instruction groups. Time for learning each diagnosis was standardized across both the Mixed and Blocked instruction groups. For each diagnosis learned, each participant was provided with 2 min of total auditory instruction (i.e. listening to the cardiac or respiratory sound).
123
Cognitive load auscultation skill
Randomization (n = 22)
Blocked Instruction Group (n = 11)
Mixed Instruction Group (n = 11)
Sounds Used for Instruction and for Test/Retest Cardiac
Auscultation TEST (post-Instruction) Auscultation RETEST (1 week later)
Normal heart sounds Aortic Insufficiency Aortic Stenosis Mitral Valve Prolapse Respiratory Bronchial breath sounds Crackles Stridor Wheeze
Fig. 3 Study design: mixed versus blocked instruction
Auscultation test Following the instruction period, participants were tested with four different examples of each of the diagnoses learned. Two Auscultation Test sounds for each diagnosis were taken from the examples presented during the instruction period and evaluated participant recall (OLD sounds); two examples were not previously heard and evaluated participant near transfer (NEW sounds). All audio examples used for the auscultation test were shuffled and a Test CD was created with the audio files. Auscultation test performance scores were calculated based on percentage of total sounds identified correctly (0–100 %). Participants completed both the post-instruction test and a retest of the same sounds 1 week following instruction; test and retest content were identical. Statistical analyses IBM SPSS Statistics version 21.0 was used for the following data analyses. A 2 9 2 9 2 mixed ANOVA was conducted, with system (Cardiac, Respiratory) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (Mixed, Blocked) treated as a between subject factor. Alpha level was 0.05. Results Overall means (SD) Auscultation Test performance scores for the Mixed and Blocked instruction groups are reported in Table 3.
123
R. Chen et al.
There was a significant difference in auscultation test performance between the Mixed and Blocked instruction groups, with the Mixed instruction group consistently superior to the Blocked instruction group for Total score and for the Cardiac and Respiratory subscores F(1, 20) = 15.95, p = 0.001. While all scores decreased on Retest, the decrease in scores was not significant F(1, 20) = 3.67, p = 0.07. There was no interaction between Mixed or Blocked instruction and Test and Retest F(1, 20) = 0.01, p = 0.93. A significant finding was that there was an interaction between Mixed or Blocked instruction and recall (OLD) and near transfer (NEW) performance F(1,20) = 7.85, p = 0.01; see Fig. 4. The Blocked instruction group showed substantially poorer transfer to new examples than the Mixed group. Discussion Results from this study for cardiac and respiratory auscultation test performance are consistent with the literature addressing the beneficial learning effects of mixed instruction. Furthermore, the mixed approach not only improved overall test performance, but the interaction between OLD and NEW sounds also suggests that it may facilitate near Table 3 Auscultation test scores, cardiac and respiratory test and retest subscores between instruction groups Total score Mean (SD)
Cardiac subscore Mean (SD)
Respiratory subscore Mean (SD)
Test
Retest
Test
Retest
Test
Retest
Blocked instruction group
49.91 (14.2)
49.18 (14.7)
44.27 (15.5)
39.58 (14.8)
55.56 (19.4)
58.79 (19.1)
Mixed instruction group
71.06 (10.7)
64.17 (8.9)
64.38 (12.5)
54.38 (9.7)
77.75 (12.2)
73.96 (12.5)
Fig. 4 Interaction between Mixed and Blocked instruction with auscultation test performance on recall (OLD) and Near Transfer (NEW)—Total Scores
123
Cognitive load auscultation skill
transfer. If the goal of education is to help learners not simply remember information or knowledge that has been presented, but to apply the learning to new examples or contexts, the results from this study are encouraging and suggest that interleaving facilitates near transfer. These findings support previous research of instruction on painting styles differentiation (Kang and Pashler 2012). This study suggests that for novice learners of cardiac and respiratory diagnoses, mixed instruction through interleaving contributed to germane load and was beneficial for learning. Comparing this study with Study 2’s exploration of Single vs. Multiple examples, the findings suggest that simply providing multiple examples in blocked format, without further guidance or elaboration, may contribute to extraneous load whereas mixed instruction in this study promoted germane cognitive processing by allowing learners to make active comparisons between the examples, even when learners were not specifically instructed to do so. Education researchers have speculated that the interleaving approach encourages learners to differentiate and discriminate between examples (Kang and Pashler 2012; Taylor and Rohrer 2010). The simple act of juxtaposing examples from different categories (e.g. wheezes vs. crackles) within a learning domain (e.g. respiratory diagnoses) encourages learners to recognize differences and enhances inductive learning (Birnbaum et al. 2012). The discriminative contrast hypothesis postulates that interleaving promotes inductive learning whereas temporal spacing does not (Kang and Pashler 2012; Kornell and Bjork 2008). Throughout this study, the definition of interleaving was intentionally specific regarding alternated examples within a shared learning domain. The findings in this study have implications for auscultation skill development for health professional students. Mixed, or interleaved, instruction may introduce increased cognitive load on a learner’s working memory, but this increased cognitive load was found to contribute to germane cognitive processing and facilitated near transfer performance. Blocked instruction may be perceived by learners to be more effective, but this perception is not reflected in better learning outcomes (Bjork and Bjork 2011). It is likely that the positive perceptions around blocked instruction are due to the decreased cognitive load that learners experience during instruction. Therefore, learners perceive that this format is more effective for learning simply because it is less taxing on working memory. Blocked instruction may help learners process similarities within categories whereas interleaving may facilitate processing between categories (Birnbaum et al. 2012; Bjork and Bjork 2011). According to Bjork (1994), the ‘‘conditions that introduce difficulties for the learner—and appear to slow the rate of the learning—can enhance long-term retention and transfer … [and are considered] ‘desirable difficulties’’’. Indeed, increasing germane load through mixed instruction does result in an increased working memory load. The auscultation test performance results overall and for near transfer suggests that this approach facilitates inductive learning, even if these beneficial effects are not be perceived by the learners during instruction.
Conclusion The overarching objective for this set of studies was to evaluate auscultation learning outcomes by manipulating intrinsic, extraneous, and germane cognitive load during instruction. Near transfer performance and auscultation performance following 1-week delayed testing were also explored. Results from Study 1 suggest that, when comparing instruction with three or six diagnoses, there was no significant attenuation in near transfer
123
R. Chen et al.
between test and retest for participants who were provided with twice as many diagnoses to learn relative to the comparison group. While this was a simple example of intrinsic cognitive load manipulation, the results from this first study served to inform the subsequent instructional interventions by identifying a reasonable number of cardiac and respiratory diagnoses to present to learners during a single instruction session. Study 2 points to the role of learner expertise in instructional intervention effectiveness. Providing multiple examples of diagnoses to more experienced learners may have yielded improved auscultation performance. However, as demonstrated in this study, providing multiple examples of diagnoses to novice learners may be ineffective due to the extraneous cognitive load burden on working memory. Again, the results of Study 2 must be interpreted with caution. Finally, the findings in Study 3 of interleaving instruction support the benefits of incorporating a mixed instruction approach to develop auscultation skill in health professional students. Mixed instruction was most effective in facilitating near transfer of auscultation learning. While further studies are needed to explore the role of learner expertise in auscultation skill development and to understand more clearly the relationship between learner expertise and the effectiveness of the instructional approaches explored here, these three studies offered an authentic exploration of cognitive load theory in the design and evaluation of auscultation skills instruction for novice nursing students. Conflict of interest The authors have no financial disclosures or conflicts of interest to declare.
References Artino, A. R, Jr. (2008). Cognitive load theory and the role of learner experience: An abbreviated review for educational practitioners. Aace Journal, 16(4), 425–439. Baddeley, A. (1992). Working memory. Science, 255(5044), 556–559. doi:10.1126/science.1736359. Bannert, M. (2002). Managing cognitive load-recent trends in cognitive load theory. Learning and Instruction, 12(1), 139–146. Birnbaum, M. S., Kornell, N., Bjork, E. L., & Bjork, R. A. (2012). Why interleaving enhances inductive learning: The roles of discrimination and retrieval. Memory & Cognition, (41)3, 392–402. Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. Retrieved from http://psycnet.apa.org.libaccess.lib.mcmaster.ca/psycinfo/1994-97967-009. Bjork, R. A. (1999). F 5 assessing our own competence: Heuristics and illusions. Retrieved from http:// bjorklab.psych.ucla.edu/pubs/Bjork_1999.pdf. Bjork, E. L., & Bjork, R. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning (pp. 56–64). New York, NY: Worth Publishers. Carson, L. M., & Wiegand, R. L. (1979). Motor schema formation and retention in young children: A test of Schmidt’s schema theory. Journal of Motor Behavior. Retrieved from http://psycnet.apa.org.libaccess. lib.mcmaster.ca/?fa=main.doiLanding&uid=1981-10140-001. Doolittle, P. E., McNeill, A. L., Terry, K. P., & Scheer, S. B. (2005). Multimedia, cognitive load and pedagogy. In Interactive Multimedia in Education and Training (pp. 184–212). Retrieved from http:// www.igi-global.com.libaccess.lib.mcmaster.ca/chapter/instructional-design-concepts-methodologiestools/51901. Hall, K. G., Domingues, D. A., & Cavazos, R. (1994). Contextual interference effects with skilled baseball players. Perceptual and Motor Skills, 78(3), 835–841. Hatala, R. M., Brooks, L. R., & Norman, G. R. (2003). Practice makes perfect: the critical role of mixed practice in the acquisition of ECG interpretation skills. Advances in Health Sciences Education, 8(1), 17–26. Holmboe, E. S. (2004). Faculty and the observation of trainees’ clinical skills: problems and opportunities. Academic Medicine, 79(1), 16. Kang, S. H. K., & Pashler, H. (2012). Learning painting styles: Spacing is advantageous when it promotes discriminative contrast. Applied Cognitive Psychology, 26(1), 97–103. doi:10.1002/acp.1801. Kornell, N., & Bjork, R. A. (2008). Optimising self-regulated study: The benefits—and costs—of dropping flashcards. Memory, 16(2), 125–136.
123
Cognitive load auscultation skill Kornell, N., Castel, A. D., Eich, T. S., & Bjork, R. A. (2010). Spacing as the friend of both memory and induction in young and older adults. Psychology and Aging, 25(2), 498. Landin, D., & Hebert, E. P. (1997). A comparison of three practice schedules along the contextual interference continuum. Research Quarterly for Exercise and Sport, 68(4), 357–361. Leahy, W., & Sweller, J. (2005). Interactions among the imagination, expertise reversal, and element interactivity effects. Journal of Experimental Psychology: Applied, 11(4), 266. Leppink, J., Paas, F., Vleuten, C. P. M., Gog, T., & Merrie¨nboer, J. J. G. (2013). Development of an instrument for measuring different types of cognitive load. Behavior Research Methods, 45(4), 1058–1072. doi:10.3758/s13428-013-0334-1. Mangione, S. (2001). Cardiac auscultatory skills of physicians-in-training: A comparison of three Englishspeaking countries. The American Journal of Medicine, 110(3), 210. Mangione, S., & Nieman, L. Z. (1997). Cardiac auscultatory skills of internal medicine and family practice trainees. JAMA: The Journal of the American Medical Association, 278(9), 717–722. Mangione, S., & Nieman, L. Z. (1999). Pulmonary auscultatory skills during training in internal medicine and family practice. American Journal of Respiratory and Critical Care Medicine, 159(4), 1119–1124. Mayer, R. E. (2002). Cognitive theory and the design of multimedia instruction: An example of the two-way street between cognition and instruction. New Directions for Teaching and Learning, 2002(89), 55–71. Mayer, R. E. (2010). Applying the science of learning to medical education. Medical Education, 44(6), 543–549. doi:10.1111/j.1365-2923.2010.03624.x. Mayfield, K. H., & Chase, P. N. (2002). The effects of cumulative practice on mathematics problem solving. Journal of Applied Behavior Analysis, 35(2), 105–123. McKeough, A., Lupart, J. L., & Marini, A. (1995). Teaching for transfer: Fostering generalization in learning. Routledge. Retrieved from http://books.google.ca.libaccess.lib.mcmaster.ca/books?hl=en&lr= &id=jnsr4C9UBgcC&oi=fnd&pg=PR7&dq=teaching?for?transfer?McKeough&ots=PkzfRyWuPR&sig= pATnqGItYapgZ0It05d13jZKhWc. Mousavi, S. Y., Low, R., & Sweller, J. (1995). Reducing cognitive load by mixing auditory and visual presentation modes. Journal of Educational Psychology, 87(2), 319–334. Norman, G. (2009). Teaching basic science to optimize transfer. Medical Teacher, 31(9), 807–811. doi:10. 1080/01421590903049814. Norman, G., Dore, K., Krebs, J., & Neville, A. J. (2007). The power of the plural: Effect of conceptual analogies on successful transfer. Academic Medicine, 82(10), S16–S18. Paas, F., Tuovinen, J. E., Tabbers, H., & Van Gerven, P. W. (2003). Cognitive load measurement as a means to advance cognitive load theory. Educational Psychologist, 38(1), 63–71. Paas, F. G. W. C., & Van Merrie¨nboer, J. J. G. (1994). Variability of worked examples and transfer of geometrical problem-solving skills: A cognitive-load approach. Journal of Educational Psychology, 86(1), 122. Pociask, F. D., DiZazzo-Miller, R., & Samuel, P. S. (2013). Reducing cognitive load while teaching complex instruction to occupational therapy students. The American Journal of Occupational Therapy, 67(5), e92–e99. Quilici, J. L., & Mayer, R. E. (1996). Role of examples in how students learn to categorize statistics word problems. Journal of Educational Psychology, 88(1), 144. Quilici, J. L., & Mayer, R. E. (2002). Teaching students to recognize structural similarities between statistics word problems. Applied Cognitive Psychology, 16(3), 325–342. Reed, S. K., & Bolstad, C. A. (1991). Use of examples and procedures in problem solving. Journal of Experimental Psychology. Learning, Memory, and Cognition, 17(4), 753–766. Richland, L. E., Bjork, R. A., Finley, J. R., & Linn, M. C. (2005). Linking cognitive science to education: Generation and interleaving effects. In Proceedings of the twenty-seventh annual conference of the cognitive science society. Mahwah, NJ: Erlbaum. Retrieved from http://learninglab.uchicago.edu. libaccess.lib.mcmaster.ca/Publications_files/5-CogsciIddeas2005.pdf Richland, L. E., Finley, J. R., & Bjork, R. A. (2004). Differentiating the contextual interference effect from the spacing effect. In Proceedings of the twenty-sixth annual conference of the cognitive science society (p. 1624). Retrieved from http://www.cogsci.northwestern.edu/cogsci2004/ma/ma239.pdf Rohrer, D., & Pashler, H. (2010). Recent research on human learning challenges conventional instructional strategies. Educational Researcher, 39(5), 406–412. doi:10.3102/0013189X10374770. Rohrer, D., & Taylor, K. (2007). The shuffling of mathematics problems improves learning. Instructional Science, 35(6), 481–498. Sankey, M., & Nooriafshar, M. (2005). Multiple representations in multimedia and e-learning materials: An issue of literacy. Retrieved from http://eprints.usq.edu.au/211 Scheiter, K., Gerjets, P., & Schuh, J. (2004). The impact of example comparisons on schema acquisition: do learners really need multiple examples? In Proceedings of the 6th international conference on
123
R. Chen et al. Learning sciences (pp. 457–464). Retrieved from http://dl.acm.org.libaccess.lib.mcmaster.ca/citation. cfm?id=1149182 Schnotz, W., & Ku¨rschner, C. (2007). A reconsideration of cognitive load theory. Educational Psychology Review, 19(4), 469–508. doi:10.1007/s10648-007-9053-4. Shea, J. B., & Morgan, R. L. (1979). Contextual interference effects on the acquisition, retention, and transfer of a motor skill. Journal of Experimental Psychology: Human Learning and Memory, 5(2), 179–187. Simon, D. A., & Bjork, R. A. (2001). Metacognition in motor learning. Journal of Experimental Psychology. Learning, Memory, and Cognition, 27(4), 907–912. Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design. Learning and Instruction, 4(4), 295–312. Sweller, J. (2010). Element Interactivity and Intrinsic, Extraneous, and Germane Cognitive Load. Educational Psychology Review, 22(2), 123–138. doi:10.1007/s10648-010-9128-5. Sweller, J., Van Merrienboer, J. J., & Paas, F. G. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10(3), 251–296. Taylor, K., & Rohrer, D. (2010). The effects of interleaved practice. Applied Cognitive Psychology, 24(6), 837–848. Van Merrienboer, J. J., & Ayres, P. (2005). Research on cognitive load theory and its design implications for e-learning. Educational Technology Research and Development, 53(3), 5–13. Van Merrie¨nboer, J. J., Kester, L., & Paas, F. (2006). Teaching complex rather than simple tasks: Balancing intrinsic and germane load to enhance transfer of learning. Applied Cognitive Psychology, 20(3), 343–352. Van Merrie¨nboer, J. J., & Sweller, J. (2005). Cognitive load theory and complex learning: Recent developments and future directions. Educational Psychology Review, 17(2), 147–177. Van Merrie¨nboer, J. J., & Sweller, J. (2010). Cognitive load theory in health professional education: design principles and strategies. Medical Education, 44(1), 85–93. Zulkiply, N., McLean, J., Burt, J. S., & Bath, D. (2012). Spacing and induction: Application to exemplars presented as auditory and visual text. Learning and Instruction, 22(3), 215–221. doi:10.1016/j. learninstruc.2011.11.002.
123
Manipulation of cognitive load variables and impact on auscultation test performance Ruth Chen • Lawrence Grierson • Geoffrey Norman
Received: 20 January 2014 / Accepted: 20 November 2014 Springer Science+Business Media Dordrecht 2014
Abstract Health profession educators have identified auscultation skill as a learning need for health professional students. This article explores the application of cognitive load theory (CLT) to designing cardiac and respiratory auscultation skill instruction for senior-level undergraduate nursing students. Three experiments assessed student auscultation performance following instructional manipulations of the three primary components of cognitive load: intrinsic, extraneous, and germane load. Study 1 evaluated the impact of intrinsic cognitive load by varying the number of diagnoses learned in one instruction session; Study 2 evaluated the impact of extraneous cognitive load by providing students with single or multiple examples of diagnoses during instruction; and Study 3 evaluated the impact of germane cognitive load by employing mixed or blocked sequences of diagnostic examples to students. Each of the three studies presents results that support CLT as explaining the influence of different types of cognitive processing on auscultation skill acquisition. We conclude with a discussion regarding CLT’s usefulness as a framework for education and education research in the health professions. Keywords
Cognitive load theory Auscultation skill Instructional design Near transfer
Background In preparation for future clinical practice, students in health professional programs acquire a range of skills during their training and education. Cardiac and respiratory auscultation
R. Chen (&) Faculty of Health Sciences, School of Nursing, McMaster University, 1280 Main Street West, Hamilton, ON HSC 2J34H, Canada e-mail: [email protected] L. Grierson Department of Family Medicine, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada G. Norman Department of Clinical Epidemiology and Biostatistics, Faculty of Health Sciences, McMaster University, Hamilton, ON, Canada
123
R. Chen et al.
skill development forms part of the core competencies of many health professions, including students in undergraduate nursing education programs. However, health professional students have demonstrated weakness with auscultation skills in both the cardiac and respiratory systems, and educators agree that this requires greater focus in health professions curricula (Holmboe 2004; Mangione and Nieman 1997, 1999; Mangione 2001). It is our position that cognitive psychology will prove useful in guiding the development of interventions that will improve student auscultation skills. We have identified cognitive load theory (CLT) as particularly relevant. Cognitive load theory relies heavily on the construct of a human working memory resource: a cognitive space responsible for the active integration of sensory-perceptual information and long term memory (Mayer 2002) for the purposes of active computation, decision making, and performance control (Baddeley 1992; Schnotz and Ku¨rschner 2007). Importantly, working memory is understood to be constrained and limited with respect to the amount of information—or cognitive load—it can process at any one time (Van Merrienboer and Ayres 2005). The critical implication is that if the cognitive load exceeds the capacity limits of working memory, then performance of tasks dependent on working memory will suffer. With respect to skill learning, this implies that instruction that exceeds the capacities of a learner’s working memory will not be retained (Sweller 1994). Thus, consideration of the cognitive load imposed on a learner during instruction can help educators and curricula developers understand why some interventions facilitate learner success (Mousavi et al. 1995; Paas et al. 2003) whereas others lead to ineffective learning (Doolittle et al. 2005). Cognitive load theory describes three types of cognitive load: intrinsic, extraneous, and germane (Van Merrie¨nboer and Sweller 2005). Intrinsic cognitive load (ICL) speaks to the information processed specifically as part of the learning task and is influenced by factors such as the learner’s level of expertise and the inherent difficulty of the learning goal. Extraneous cognitive load describes the features of instruction that are neither relevant nor necessary for the particular learning goal or task and that impose an unnecessary burden on working memory (Sweller et al. 1998). Germane cognitive load also contributes additional cognitive burden to working memory processes, but this load is considered useful as it serves to facilitate a learner’s understanding and automation of the information received during instruction into long-term memory (Paas and Van Merrie¨nboer 1994). Taken together, CLT explains that student learning can be optimized through instruction that maintains intrinsic cognitive load within the limits of a learner’s working memory, decreases extraneous load, and facilitates the type of processing associated with germane cognitive load. Importantly, the amount of working memory capacity occupied by intrinsic load is not a fixed quantity. Rather, it changes for any given task as a function of learner expertise. Specifically, as a learner progresses along the continuum of expertise, the integration of information into long-term memory is accomplished by way of schema construction and automation, chunking the invariant features of a task so that the total processing impact on working memory is reduced and more capacity is freed up for additional processing (Artino 2008; Sweller 1988). The dependence of ICL on level of expertise highlights the importance in understanding the characteristics of the learner when determining whether an instructional design will contribute to extraneous cognitive load or foster germane cognitive load (Leahy and Sweller 2005). Accordingly, the increased attention that CLT is receiving as a useful framework in the health professions education literature (Leppink et al. 2013; Mayer 2010; Pociask et al. 2013; Van Merrie¨nboer and Sweller 2010) provides
123
Cognitive load auscultation skill
researchers an opportunity to evaluate how cognitive load variables interact to facilitate learning within students at various levels of expertise (Van Merrie¨nboer et al. 2006). The present studies reflect a first effort in accomplishing this. In particular, this work investigates the effects of manipulating cognitive load variables on the auscultation performance of novice learners. The tests used in these studies assess performance using several measures. While overall diagnostic success is the primary outcome measure to assess learning of cardiac and respiratory sounds, measurements also include both retention and transfer performances. For the following studies, near transfer refers to a participant’s ability to identify a different example of a diagnostic sound that was not previously heard. For example, if a participant heard an example of a ‘‘wheeze’’ during instruction, they would demonstrate near transfer if they correctly identified a different example of a ‘‘wheeze’’ during the auscultation test. Additionally, while we tested both cardiac and respiratory systems in the auscultation tests, the primary objective was not to report on the differences in performance between the two body systems. We wished to evaluate the cardiac and respiratory systems to assess whether the outcomes and trends in performance would be similar or consistent between two systems of auscultation instruction. The goal of these studies is to understand how, or if, instructional design modifications impact learner performance and skill maintenance for auscultation learning.
Study 1: Modification of intrinsic cognitive load (ICL) and impact on auscultation test performance When designing instructional content for learners, educators should consider the optimal ICL to achieve desired learning goals. One approach to modifying ICL is to reduce the number of components of a learning task that a learner must retain simultaneously in working memory (Bannert 2002; Sweller 2010). Educators can then add complexity to the learning task at later stages of learning. However, ecological validity is a concern: as complexity is decreased for a learning task and the instruction sessions cover less and less content, there will be a point where the content of each session is unrealistically minimal and the number of sessions required to cover all the desired content is infeasible. Determining how much content to cover in an instructional session is difficult with complex tasks like auscultation skills. Educators must balance the effects of reducing the scope and complexity of the instructional session with the need to cover all the necessary learning material. Reducing the number of diagnoses per instructional session would likely improve performance. However, would doubling the number of diagnoses learned in one instructional session necessarily result in significantly poorer performance? The purpose of this first study was to evaluate performance outcomes by manipulating the intrinsic cognitive load (ICL) of the learning goal, as measured by the number of diagnoses presented in one instructional session. Furthermore, the goal was to evaluate whether there were additional benefits of minimizing ICL that resulted in enhanced performance outcomes for near transfer and on delayed testing. Research question Is there a significant difference in cardiac and respiratory auscultation test performance when intrinsic cognitive load is modified by instruction on either three or six diagnoses?
123
R. Chen et al.
Methods Participants Senior level undergraduate nursing students (n = 44) were recruited for this study. Participants were provided with both Cardiac (n = 21) and Respiratory (n = 23) sounds instruction. Participants were randomized to receive instruction on either three (‘‘3 Group’’) or six (‘‘6 Group’’) diagnoses. Respiratory instruction occurred first. Participants were given an auscultation test immediately following instruction and then a Respiratory retest 1 week later. Following the Respiratory retest, participants then received Cardiac instruction and a post-instruction Cardiac auscultation test; we were not able to conduct a Cardiac retest 1 week later due to logistical issues with student schedules. Study design (see Fig. 1) Six diagnoses were used for the Respiratory and Cardiac sounds interventions. A group of clinical experts was consulted to determine which three diagnoses from each body system were most difficult to learn. Based on the feedback of the clinical experts, three Cardiac and three Respiratory diagnoses were identified as most difficult of the six. The 3 group received instruction on these three diagnoses. The 6 group learned all six diagnoses. The rationale for selecting the three most difficult Cardiac or Respiratory diagnoses was to minimize superior performance by the 3 group participants simply because they had easier diagnoses to learn. Time for learning each diagnosis was standardized across both the 3 and 6 instruction groups. For each diagnosis learned, each participant was provided with 2 min of auditory instruction for each sound (i.e. listening to the cardiac or respiratory sound) followed by a 10 s rest period before proceeding to the next diagnosis. During the 2-min instruction period, participants were exposed to three different examples of the diagnosis. These diagnosis examples were recorded audio (i.e. mp3) files from a database of cardiac and respiratory sounds, obtained from the Internet through non-proprietary sources. We wanted examples that would not require an educator to purchase any particular vendor’s materials. Therefore, a research assistant went online and looked for audio examples of the sounds we wished to use for instruction and testing. These sounds were then recorded and formatted for our instruction. Brief written descriptions accompanied the auditory examples of all sounds provided; the written descriptions were identical for the 3 and 6 groups. Auscultation test Following the instruction period, participants were tested with four different audio examples of each of the diagnoses learned. Two examples were taken from the sounds presented during the instruction period and evaluated participant recall (OLD sounds); two examples were not previously heard and evaluated participant near transfer (NEW sounds). Therefore, the participants in the 3 group were tested on twelve (12) total sounds and the participants in the 6 group were tested on twenty-four (24) sounds. All audio examples were shuffled and a Test CD was created with the audio files. Auscultation test performance scores were calculated based on percentage of total sounds identified correctly (0–100 %). The post-instruction test and retest content was identical.
123
Cognitive load auscultation skill
Randomization (n = 23) Respiratory “3” Instruction (n = 11)
Respiratory “6” Instruction (n = 12)
Sounds Used for Instruction and for Test/Retest Respiratory “3”
Respiratory TEST (post-Instruction) Respiratory RETEST (1 week later)
Cardiac “3” Instruction (n = 10)
Cardiac “6” Instruction (n = 11)
Cardiac TEST (post-Instruction)
Crackles Rhonchi Stridor Respiratory “6” Above, plus Bronchial breath sounds Vesicular breath sounds Wheeze
Cardiac “3” Aortic Insufficiency Atrial Septal Defect Mitral Valve Prolapse Cardiac “6” Above, plus Aortic Stenosis Normal heart sounds Ventricular Septal Defect
Fig. 1 Study design: 3 versus 6 instruction
Ethics The joint Research Ethics Board of McMaster University Faculty of Health Sciences, Hamilton Health Sciences, and St. Joseph’s Healthcare Hamilton approved all studies. All participants provided written and informed consent to participate. Statistical analysis IBM SPSS Statistics version 21.0 was used for the following data analyses. In the analysis of the post-instruction auscultation test, a 2 9 2 9 2 mixed ANOVA was conducted, with system (Cardiac, Respiratory) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (3 group, 6 group) treated as a between subject factor. In the analysis of the Respiratory system test and retest, a 2 9 2 9 2 mixed ANOVA was conducted, with test (post-instruction, retest 1 week later) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (3 group, 6 group) treated as a between
123
R. Chen et al.
subject factor. Alpha level was set at 0.05. Performance outcomes on only the three diagnoses common both to the 3 group and the 6 group were used in the analyses to ensure comparability between groups. Performance of the 6 group on the additional diagnoses was not analyzed; this allowed direct comparison of the effect of increasing intrinsic load on auscultation test performance. Results Overall means (SD) for auscultation test performance of the 3 and 6 groups for both analyses are found in Table 1. Post-instruction test comparison of 3 and 6 groups, for Cardiac and Respiratory sounds As predicted, there was a significant difference in auscultation test performance between the 3 and 6 instruction groups (see Table 1). The 3 group outperformed the 6 group, F(1,21) = 18.06, p \ 0.001 across both cardiac and respiratory groups [Interaction F(1,19) = 0.047, p = 0.83]. There was no interaction between OLD/NEW sounds and the 3 or 6 instruction group F(1,19) = 0.33, p = 0.58. Test–retest auscultation test comparison of 3 and 6 groups for Respiratory sounds The main effect of group demonstrated a significant difference between the Respiratory 3 and 6 instruction groups’ performance across both test administrations with the 3 group achieving higher scores than the 6 group F(1,19) = 26.35, p \ 0.001. There was a significant decline in test performance between Test and Retest F(1,19) = 31.02, p \ 0.001, and a significant three-way interaction between 3–6 group, Old–New sounds, and Test– Retest performance F(1,19) = 23.74, p \ 0.001. Data were analyzed further to understand the three-way interaction; OLD sounds and NEW sounds were analyzed separately. Analysis of the OLD Respiratory sounds showed a significant difference between 3 and 6 groups F(1,19) = 12.01, p = 0.002. There was a significant decline in scores from Test to Retest F(1,19) = 52.54, p \ 0.001, and a significant interaction between 3–6 group and Test–Retest performance with F(1,19) = 34.73, p \ 0.001, indicating that the performance of the 6 group declined more than the 3 group. Analysis of NEW Respiratory sounds showed a significant advantage for the 3 group across both tests, F(1,19) = 19.34, p \ 0001, but there was not a significant decline Table 1 Post-instruction Cardiac and Respiratory auscultation test scores, and respiratory retest scores between instruction groups Cardiac subscore Mean (SD)
Respiratory subscore Mean (SD)
Test
Test
OLD
NEW
‘‘3’’ Group instruction 75.76 (21.3) 81.6 (14.6)
123
NEW
71.21 (18.5) 73.3 (21.2)
‘‘6’’ Group instruction 58.33 (22.3) 61.3 (20.2)
OLD
Retest
83.3 (15.2) 81.6 (14.6)
NEW
59.17 (19.7) 58.3 (14.5)
55.83 (19.1) 53.4 (13.5)
OLD
78.3 (24.20)
60.0 (20.5)
34.17 (19.4) 30.0 (13.2)
33.3 (13.6)
38.6 (13.9)
Cognitive load auscultation skill
between Test and Retest F(1,19) = 0.66, p = 0.43, and no 3–6 group and Test–Retest interaction F(1,19) = 0.29, p = 0.60. The three-way interaction, therefore, was carried predominantly by the OLD Respiratory sounds performance on Retest by the 6 group, which dramatically fell from Test to Retest. Summary of results For the three diagnoses common to both the 3 group and the 6 group, the 3 group consistently demonstrated superior auscultation test performance on the post-instruction test. When evaluating Respiratory auscultation test performance on retest, the 3 group continued to demonstrate superior performance than the 6 group. Of note, whereas the 6 group performed significantly worse in recalling OLD sounds on retest, there was no significant decline in test performance on NEW sounds between test and retest. Discussion This study explored the effect of adjusting ICL for learning cardiac and respiratory auscultation sounds in novice learners. As expected, those students with fewer diagnoses to learn demonstrated better performance on auscultation test performance than students with twice as many diagnoses to learn. Increasing intrinsic load by simply increasing the quantity of material to be learned yielded the expected outcomes. However, an interesting finding was that, when looking at performance on NEW sounds between Test and Retest, there was no significant attenuation in near transfer performance for both the 3 group and the 6 group. These results suggest that, even though the 6 group learned twice as many diagnoses, the near transfer performance between test and retest was comparable for both instructional groups. For cardiac and respiratory auscultation skill development, it is also necessary to address other cognitive load components—in particular, strategies that decrease extraneous load and facilitate germane load—to promote learning and transfer while intrinsic load remains constant. Manipulating extraneous and germane cognitive load variables will be the focus of the next two studies.
Study 2: Adding extraneous cognitive load (ECL) or fostering germane cognitive load (GCL) through multiple representations? Once an instructional goal has been determined, CLT also supports removing aspects of the instruction that are unnecessary for learning, factors that contribute extraneous cognitive load. Decreasing extraneous load is less important for learning tasks that do not place a high intrinsic load on a learner’s working memory because there are additional working memory resources available to help the learner process both the necessary intrinsic load and the unhelpful extraneous load (Van Merrie¨nboer and Sweller 2005). However, if the intrinsic load of a learning task increases, then it is important to decrease extraneous load so that the total cognitive burden does not exceed the learner’s working memory capacity. The goal in instructional design is to use the remaining cognitive resources available in working memory by decreasing extraneous load and by facilitating germane load for the particular instructional goal. Just as there are no standard metrics to quantify the intrinsic cognitive load of an instructional task without having to consider the expertise level of the learner, so too, what
123
R. Chen et al.
constitutes germane load for one level of learner may be unnecessary, extraneous cognitive load for another (McKeough et al. 1995; Paas and Van Merrie¨nboer 1994; Quilici and Mayer 1996, 2002; Scheiter et al. 2004). One such example in the literature is the use of multiple examples or representations in instructional design. Providing multiple examples during instruction helps learners distinguish relevant from irrelevant features (Van Merrie¨nboer and Sweller 2005) and facilitates transfer to new contexts (McKeough et al. 1995; Reed and Bolstad 1991). The use of multiple examples, or representations, for instruction has also been found to be beneficial in health professions education with cardiac physiology instruction (Norman et al. 2007; Norman 2009). However, educators must exercise caution when using multiple examples in instruction because requiring learners to process multiple representations may increase extraneous cognitive load and may, therefore, be detrimental to learning (Sankey and Nooriafshar 2005). In Study 1, all learners were provided with three examples of each diagnosis, whether they were randomized to the 3 group or the 6 group. This study explores whether providing only a single example to this group of novice learners might result in decreased performance, especially on near transfer, thereby suggesting that multiple examples were adding germane load and were therefore beneficial for auscultation test performance. Conversely, if learners provided with only a single example of each diagnosis showed superior performance in comparison to the group provided with multiple examples, these results would suggest that the multiple examples contributed extraneous cognitive load on learners. Research question How does providing multiple examples, or representations, of cardiac and respiratory diagnoses impact on auscultation test performance when compared with providing only single examples of diagnoses? Methods Participants Senior level undergraduate nursing students (n = 32) were recruited for this study in June– July 2011. Participants were randomized to receive instruction with a single example (Single group) or with multiple examples (Multiple group) of the Cardiac and Respiratory diagnoses. Study design (see Fig. 2) Eight diagnoses were used for the Single and Multiple sounds interventions. The audio examples provided to the Single group consisted of recorded audio (i.e. mp3) files from the pediatric Laerdal human patient simulators. These audio files were compiled into an instruction CD for the Single group. The audio examples provided to the Multiple group consisted of three audio examples of each diagnosis. One example was taken from Laerdal simulators, identical to what the Single group heard. The two additional examples were taken from a database of cardiac and respiratory sounds. Brief written
123
Cognitive load auscultation skill
Randomization (n = 32)
Single Instruction Group (n = 16)
Multiple Instruction Group (n = 16)
Sounds Used for Instruction and for Test/Retest Cardiac
Auscultation TEST (post-Instruction) Auscultation RETEST (1 week later)
Normal heart sounds Aortic Insufficiency Aortic Stenosis Mitral Valve Prolapse Respiratory Bronchial breath sounds Crackles Stridor Wheeze
Fig. 2 Study design: single versus multiple instruction
descriptions accompanied the auditory examples of all sounds provided; the written descriptions were identical for the Single and Multiple groups. Time for learning each diagnosis was standardized across both the Single and Multiple instruction groups. For each diagnosis learned, each participant was provided with 2 min of auditory instruction (i.e. listening to the cardiac or respiratory sound) followed by a 10 s rest period before proceeding to the next diagnosis. During the 2-min instruction period, participants were exposed to either one example of the diagnosis (Single group) or three different examples of the diagnosis (Multiple group). Auscultation test Following the instruction period, participants were tested with four different examples of each of the diagnoses learned. For the Single group, the example heard during instruction was included to test for recall (OLD sound) and three additional examples which were not previously heard tested for near transfer (NEW sound). For the Multiple group, two examples were taken from the sounds presented during the instruction period and evaluated participant recall (OLD sounds); two examples were not previously heard and evaluated participant near transfer (NEW sounds). All audio examples used for the auscultation test were shuffled and a Test CD was created with the audio files. Auscultation test performance scores were calculated based on percentage of total sounds identified correctly (0–100 %). Participants completed both the post-instruction test and a retest of the same
123
R. Chen et al.
sounds 1 week following instruction. As with Study 1, the retest content was identical to the post-instruction test content. Statistical analyses IBM SPSS Statistics version 21.0 was used for the following data analyses. A 2 9 2 9 2 mixed ANOVA was conducted, with system (Cardiac, Respiratory) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (Single, Multiple) treated as a between subject factor. Alpha level was 0.05. Results Overall means (SD) Auscultation Test performance scores for the Single and Multiple instruction groups are reported in Table 2. There was no significant difference between Single and Multiple group scores F(1,28) = 0.000, p = 0.99. Furthermore, there was no significant decline in scores from Test to Retest F(1,28) = 0.13, p = 0.73; there was also no interaction between the Single or Multiple instruction group and Test/Retest scores. No further interactions were found between instruction group with Cardiac or Respiratory performance as well as with recall (OLD) or near transfer (NEW) performance. Discussion These results suggest that for novice nursing students, multiple representations of cardiac and respiratory auscultation diagnoses may either contribute to extraneous cognitive load or may have no additional instructional benefit over providing students with single examples of the diagnoses. Extraneous load involves all the irrelevant features in the instructional design, but also includes aspects of the instruction that increase/exceed working memory capacity without facilitating schema construction and automation (Sweller 1994). Therefore, for the novice learners we used in these studies, being exposed to multiple examples of a cardiac diagnosis, for example, did not introduce ‘‘irrelevant features’’ per se, but rather, the multiple examples may have introduced greater complexity that the students’ working memories were unable to process. It is important to note that students were not provided with any additional instruction around the diagnoses presented. Therefore, there was no instruction regarding how these
Table 2 Auscultation test scores, cardiac and respiratory test and retest subscores between instruction groups Total score Mean (SD)
Cardiac subscore Mean (SD)
Respiratory subscore Mean (SD)
Test
Retest
Test
Retest
Test
Retest
Single instruction group
62.55 (10.8)
60.82 (14.9)
61.77 (14.16)
62.50 (21.2)
62.61 (13.2)
59.14 (14.2)
Multiple instruction group
59.81 (23.4)
63.33 (20.4)
59.14 (29.6)
60.58 (26.4)
60.48 (22.8)
66.09 (17.8)
123
Cognitive load auscultation skill
examples highlighted the central features of each cardiac or respiratory diagnosis. These results underscore, perhaps, the need for further explanatory information to accompany the auditory and basic written instructions when providing multiple examples during instruction. The non-significant results, however, could also be due to a type-II error as a result of an insufficient sample size. Power calculations were performed which demonstrated that for a sample size of 15 participants per instruction group, a difference of 15 % points in test performance could be detected with 80 % power. Therefore, the study results must be interpreted with caution. While there is sufficient justification in the literature around the use of multiple representations with instruction, the findings were not replicated in this learning context and with this level of learner. The next study investigates another approach that has been discussed in the literature as a strategy to facilitate germane load: interleaving examples during instruction.
Study 3: Fostering germane cognitive load (GCL) through an interleaving (or, mixed format) instructional approach The advantages of interleaving or mixed instruction, where practice examples from multiple categories are mixed together, has been well documented in the cognitive psychology and medical education literature, and research on this approach spans several decades (Hatala et al. 2003; Richland et al. 2004). In the field of auscultation skill development, interleaving would involve including examples from two or more diagnoses during instruction. This has also been referred to as mixed instruction. Instruction with all examples from each diagnosis presented in succession before moving on to the next diagnosis would be an example of blocked instruction. Interleaving approaches in education were first documented over two decades ago for motor skills acquisition (Shea and Morgan 1979). Further studies in motor learning followed and the benefits of the interleaving approach, alternatively termed ‘‘mixed practice’’ or ‘‘practice involving contextual interference’’, have been demonstrated for a variety of motor skills instruction, from children throwing beanbags (Carson and Wiegand 1979), to baseball players’ batting skills (Hall et al. 1994), to basketball players’ basket-shooting skills (Landin and Hebert 1997). In all the above examples from the motor learning literature, interleaving instruction demonstrated benefits. Blocked instruction may yield better outcomes during the practice session over mixed instruction (Bjork 1994, 1999; Rohrer and Taylor 2007; Taylor and Rohrer 2010) and learners also perceive blocked instruction to be more effective for their learning than mixed instruction (Birnbaum et al. 2012; Kornell and Bjork 2008; Kornell et al. 2010; Simon and Bjork 2001). However with passage of time, the benefits of mixed instruction are revealed. Therefore, even as learners believed that the blocked instruction format was more effective for their own learning and the mixed instruction group perceiving the instructional format to be less effective, their performance results demonstrated the opposite outcomes. In cognitive psychology, the interleaving instructional approach has been termed a ‘‘desirable difficulty’’ in learning (Bjork 1994), where cognitive load is increased for the learner, but in a manner that facilitates rather than hinders learning. Inducing these ‘‘desirable difficulties’’ in learning facilitate GCL even as instruction is more taxing to working memory when compared with a blocked instruction approach (Rohrer and Pashler 2010). The benefits of the interleaving approach have been found with instruction around differentiating painting styles (Kang and Pashler 2012; Kornell and Bjork 2008), learning bird and butterfly species (Birnbaum et al. 2012), understanding astronomy concepts
123
R. Chen et al.
(Richland et al. 2005), and applying mathematical concepts (Mayfield and Chase 2002; Rohrer and Taylor 2007; Taylor and Rohrer 2010). In health professions education, interleaving, or mixed practice during instruction, has been helpful in the instruction of psychiatric diagnoses (Zulkiply et al. 2012) and in electrocardiogram interpretation (Hatala et al. 2003). While the advantage of interleaving or mixing has been documented with motor and cognitive skills acquisition, questions remain whether this approach would also be beneficial for learners receiving auditory instruction for cardiac and respiratory auscultation skill development. The theoretical rationale for the benefits of interleaving for cognitive goals is that this approach encourages learners to make contrasts between the examples of the learning goals provided. The following study explored whether the interleaving approach for auscultation skill development would similarly allow novice health professional learners to make auditory contrasts between examples to enhance learning of cardiac and respiratory diagnoses. Research question Does providing instruction of cardiac and respiratory diagnoses in either a mixed or blocked format have an impact on auscultation test performance? Methods Participants Senior level undergraduate nursing students (n = 22) were recruited for this study in May– June 2012. Participants were randomized to receive instruction with diagnoses examples presented in an interleaved format (Mixed group) or with diagnoses presented together in succession (Blocked group) of the Cardiac and Respiratory diagnoses. Study design (see Fig. 3) Eight total diagnoses were used for the Mixed and Blocked group interventions. For the Blocked instruction group, three examples of each diagnosis were provided in succession. All three examples of one diagnosis were played before moving to next diagnosis (i.e. Diagnosis 1, examples A, B, C; Diagnosis 2, examples A, B, C; etc.). For the Mixed instruction group, three examples of each diagnosis were provided. However, the diagnoses were presented in pairs and the examples of each diagnosis were interleaved (i.e. Diagnosis 1, example A; Diagnosis 2, example A; Diagnosis 1, example B, C; Diagnosis 2, example B, C; Diagnosis 3, example A; Diagnosis 4, example A; Diagnosis 3, example B, C; Diagnosis 4, example B, C, etc.). The audio examples were audio files taken from Laerdal simulators and from a database of cardiac and respiratory sounds. All examples for each diagnosis were the same for the Mixed and Blocked instruction group; only the order in which the examples were presented was changed. Brief written descriptions accompanied the auditory examples of all sounds provided; the written descriptions were identical for the Mixed and Blocked instruction groups. Time for learning each diagnosis was standardized across both the Mixed and Blocked instruction groups. For each diagnosis learned, each participant was provided with 2 min of total auditory instruction (i.e. listening to the cardiac or respiratory sound).
123
Cognitive load auscultation skill
Randomization (n = 22)
Blocked Instruction Group (n = 11)
Mixed Instruction Group (n = 11)
Sounds Used for Instruction and for Test/Retest Cardiac
Auscultation TEST (post-Instruction) Auscultation RETEST (1 week later)
Normal heart sounds Aortic Insufficiency Aortic Stenosis Mitral Valve Prolapse Respiratory Bronchial breath sounds Crackles Stridor Wheeze
Fig. 3 Study design: mixed versus blocked instruction
Auscultation test Following the instruction period, participants were tested with four different examples of each of the diagnoses learned. Two Auscultation Test sounds for each diagnosis were taken from the examples presented during the instruction period and evaluated participant recall (OLD sounds); two examples were not previously heard and evaluated participant near transfer (NEW sounds). All audio examples used for the auscultation test were shuffled and a Test CD was created with the audio files. Auscultation test performance scores were calculated based on percentage of total sounds identified correctly (0–100 %). Participants completed both the post-instruction test and a retest of the same sounds 1 week following instruction; test and retest content were identical. Statistical analyses IBM SPSS Statistics version 21.0 was used for the following data analyses. A 2 9 2 9 2 mixed ANOVA was conducted, with system (Cardiac, Respiratory) and sounds (OLD, NEW) treated as repeated measures factors and instruction group (Mixed, Blocked) treated as a between subject factor. Alpha level was 0.05. Results Overall means (SD) Auscultation Test performance scores for the Mixed and Blocked instruction groups are reported in Table 3.
123
R. Chen et al.
There was a significant difference in auscultation test performance between the Mixed and Blocked instruction groups, with the Mixed instruction group consistently superior to the Blocked instruction group for Total score and for the Cardiac and Respiratory subscores F(1, 20) = 15.95, p = 0.001. While all scores decreased on Retest, the decrease in scores was not significant F(1, 20) = 3.67, p = 0.07. There was no interaction between Mixed or Blocked instruction and Test and Retest F(1, 20) = 0.01, p = 0.93. A significant finding was that there was an interaction between Mixed or Blocked instruction and recall (OLD) and near transfer (NEW) performance F(1,20) = 7.85, p = 0.01; see Fig. 4. The Blocked instruction group showed substantially poorer transfer to new examples than the Mixed group. Discussion Results from this study for cardiac and respiratory auscultation test performance are consistent with the literature addressing the beneficial learning effects of mixed instruction. Furthermore, the mixed approach not only improved overall test performance, but the interaction between OLD and NEW sounds also suggests that it may facilitate near Table 3 Auscultation test scores, cardiac and respiratory test and retest subscores between instruction groups Total score Mean (SD)
Cardiac subscore Mean (SD)
Respiratory subscore Mean (SD)
Test
Retest
Test
Retest
Test
Retest
Blocked instruction group
49.91 (14.2)
49.18 (14.7)
44.27 (15.5)
39.58 (14.8)
55.56 (19.4)
58.79 (19.1)
Mixed instruction group
71.06 (10.7)
64.17 (8.9)
64.38 (12.5)
54.38 (9.7)
77.75 (12.2)
73.96 (12.5)
Fig. 4 Interaction between Mixed and Blocked instruction with auscultation test performance on recall (OLD) and Near Transfer (NEW)—Total Scores
123
Cognitive load auscultation skill
transfer. If the goal of education is to help learners not simply remember information or knowledge that has been presented, but to apply the learning to new examples or contexts, the results from this study are encouraging and suggest that interleaving facilitates near transfer. These findings support previous research of instruction on painting styles differentiation (Kang and Pashler 2012). This study suggests that for novice learners of cardiac and respiratory diagnoses, mixed instruction through interleaving contributed to germane load and was beneficial for learning. Comparing this study with Study 2’s exploration of Single vs. Multiple examples, the findings suggest that simply providing multiple examples in blocked format, without further guidance or elaboration, may contribute to extraneous load whereas mixed instruction in this study promoted germane cognitive processing by allowing learners to make active comparisons between the examples, even when learners were not specifically instructed to do so. Education researchers have speculated that the interleaving approach encourages learners to differentiate and discriminate between examples (Kang and Pashler 2012; Taylor and Rohrer 2010). The simple act of juxtaposing examples from different categories (e.g. wheezes vs. crackles) within a learning domain (e.g. respiratory diagnoses) encourages learners to recognize differences and enhances inductive learning (Birnbaum et al. 2012). The discriminative contrast hypothesis postulates that interleaving promotes inductive learning whereas temporal spacing does not (Kang and Pashler 2012; Kornell and Bjork 2008). Throughout this study, the definition of interleaving was intentionally specific regarding alternated examples within a shared learning domain. The findings in this study have implications for auscultation skill development for health professional students. Mixed, or interleaved, instruction may introduce increased cognitive load on a learner’s working memory, but this increased cognitive load was found to contribute to germane cognitive processing and facilitated near transfer performance. Blocked instruction may be perceived by learners to be more effective, but this perception is not reflected in better learning outcomes (Bjork and Bjork 2011). It is likely that the positive perceptions around blocked instruction are due to the decreased cognitive load that learners experience during instruction. Therefore, learners perceive that this format is more effective for learning simply because it is less taxing on working memory. Blocked instruction may help learners process similarities within categories whereas interleaving may facilitate processing between categories (Birnbaum et al. 2012; Bjork and Bjork 2011). According to Bjork (1994), the ‘‘conditions that introduce difficulties for the learner—and appear to slow the rate of the learning—can enhance long-term retention and transfer … [and are considered] ‘desirable difficulties’’’. Indeed, increasing germane load through mixed instruction does result in an increased working memory load. The auscultation test performance results overall and for near transfer suggests that this approach facilitates inductive learning, even if these beneficial effects are not be perceived by the learners during instruction.
Conclusion The overarching objective for this set of studies was to evaluate auscultation learning outcomes by manipulating intrinsic, extraneous, and germane cognitive load during instruction. Near transfer performance and auscultation performance following 1-week delayed testing were also explored. Results from Study 1 suggest that, when comparing instruction with three or six diagnoses, there was no significant attenuation in near transfer
123
R. Chen et al.
between test and retest for participants who were provided with twice as many diagnoses to learn relative to the comparison group. While this was a simple example of intrinsic cognitive load manipulation, the results from this first study served to inform the subsequent instructional interventions by identifying a reasonable number of cardiac and respiratory diagnoses to present to learners during a single instruction session. Study 2 points to the role of learner expertise in instructional intervention effectiveness. Providing multiple examples of diagnoses to more experienced learners may have yielded improved auscultation performance. However, as demonstrated in this study, providing multiple examples of diagnoses to novice learners may be ineffective due to the extraneous cognitive load burden on working memory. Again, the results of Study 2 must be interpreted with caution. Finally, the findings in Study 3 of interleaving instruction support the benefits of incorporating a mixed instruction approach to develop auscultation skill in health professional students. Mixed instruction was most effective in facilitating near transfer of auscultation learning. While further studies are needed to explore the role of learner expertise in auscultation skill development and to understand more clearly the relationship between learner expertise and the effectiveness of the instructional approaches explored here, these three studies offered an authentic exploration of cognitive load theory in the design and evaluation of auscultation skills instruction for novice nursing students. Conflict of interest The authors have no financial disclosures or conflicts of interest to declare.
References Artino, A. R, Jr. (2008). Cognitive load theory and the role of learner experience: An abbreviated review for educational practitioners. Aace Journal, 16(4), 425–439. Baddeley, A. (1992). Working memory. Science, 255(5044), 556–559. doi:10.1126/science.1736359. Bannert, M. (2002). Managing cognitive load-recent trends in cognitive load theory. Learning and Instruction, 12(1), 139–146. Birnbaum, M. S., Kornell, N., Bjork, E. L., & Bjork, R. A. (2012). Why interleaving enhances inductive learning: The roles of discrimination and retrieval. Memory & Cognition, (41)3, 392–402. Bjork, R. A. (1994). Memory and metamemory considerations in the training of human beings. Retrieved from http://psycnet.apa.org.libaccess.lib.mcmaster.ca/psycinfo/1994-97967-009. Bjork, R. A. (1999). F 5 assessing our own competence: Heuristics and illusions. Retrieved from http:// bjorklab.psych.ucla.edu/pubs/Bjork_1999.pdf. Bjork, E. L., & Bjork, R. (2011). Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning (pp. 56–64). New York, NY: Worth Publishers. Carson, L. M., & Wiegand, R. L. (1979). Motor schema formation and retention in young children: A test of Schmidt’s schema theory. Journal of Motor Behavior. Retrieved from http://psycnet.apa.org.libaccess. lib.mcmaster.ca/?fa=main.doiLanding&uid=1981-10140-001. Doolittle, P. E., McNeill, A. L., Terry, K. P., & Scheer, S. B. (2005). Multimedia, cognitive load and pedagogy. In Interactive Multimedia in Education and Training (pp. 184–212). Retrieved from http:// www.igi-global.com.libaccess.lib.mcmaster.ca/chapter/instructional-design-concepts-methodologiestools/51901. Hall, K. G., Domingues, D. A., & Cavazos, R. (1994). Contextual interference effects with skilled baseball players. Perceptual and Motor Skills, 78(3), 835–841. Hatala, R. M., Brooks, L. R., & Norman, G. R. (2003). Practice makes perfect: the critical role of mixed practice in the acquisition of ECG interpretation skills. Advances in Health Sciences Education, 8(1), 17–26. Holmboe, E. S. (2004). Faculty and the observation of trainees’ clinical skills: problems and opportunities. Academic Medicine, 79(1), 16. Kang, S. H. K., & Pashler, H. (2012). Learning painting styles: Spacing is advantageous when it promotes discriminative contrast. Applied Cognitive Psychology, 26(1), 97–103. doi:10.1002/acp.1801. Kornell, N., & Bjork, R. A. (2008). Optimising self-regulated study: The benefits—and costs—of dropping flashcards. Memory, 16(2), 125–136.
123
Cognitive load auscultation skill Kornell, N., Castel, A. D., Eich, T. S., & Bjork, R. A. (2010). Spacing as the friend of both memory and induction in young and older adults. Psychology and Aging, 25(2), 498. Landin, D., & Hebert, E. P. (1997). A comparison of three practice schedules along the contextual interference continuum. Research Quarterly for Exercise and Sport, 68(4), 357–361. Leahy, W., & Sweller, J. (2005). Interactions among the imagination, expertise reversal, and element interactivity effects. Journal of Experimental Psychology: Applied, 11(4), 266. Leppink, J., Paas, F., Vleuten, C. P. M., Gog, T., & Merrie¨nboer, J. J. G. (2013). Development of an instrument for measuring different types of cognitive load. Behavior Research Methods, 45(4), 1058–1072. doi:10.3758/s13428-013-0334-1. Mangione, S. (2001). Cardiac auscultatory skills of physicians-in-training: A comparison of three Englishspeaking countries. The American Journal of Medicine, 110(3), 210. Mangione, S., & Nieman, L. Z. (1997). Cardiac auscultatory skills of internal medicine and family practice trainees. JAMA: The Journal of the American Medical Association, 278(9), 717–722. Mangione, S., & Nieman, L. Z. (1999). Pulmonary auscultatory skills during training in internal medicine and family practice. American Journal of Respiratory and Critical Care Medicine, 159(4), 1119–1124. Mayer, R. E. (2002). Cognitive theory and the design of multimedia instruction: An example of the two-way street between cognition and instruction. New Directions for Teaching and Learning, 2002(89), 55–71. Mayer, R. E. (2010). Applying the science of learning to medical education. Medical Education, 44(6), 543–549. doi:10.1111/j.1365-2923.2010.03624.x. Mayfield, K. H., & Chase, P. N. (2002). The effects of cumulative practice on mathematics problem solving. Journal of Applied Behavior Analysis, 35(2), 105–123. McKeough, A., Lupart, J. L., & Marini, A. (1995). Teaching for transfer: Fostering generalization in learning. Routledge. Retrieved from http://books.google.ca.libaccess.lib.mcmaster.ca/books?hl=en&lr= &id=jnsr4C9UBgcC&oi=fnd&pg=PR7&dq=teaching?for?transfer?McKeough&ots=PkzfRyWuPR&sig= pATnqGItYapgZ0It05d13jZKhWc. Mousavi, S. Y., Low, R., & Sweller, J. (1995). Reducing cognitive load by mixing auditory and visual presentation modes. Journal of Educational Psychology, 87(2), 319–334. Norman, G. (2009). Teaching basic science to optimize transfer. Medical Teacher, 31(9), 807–811. doi:10. 1080/01421590903049814. Norman, G., Dore, K., Krebs, J., & Neville, A. J. (2007). The power of the plural: Effect of conceptual analogies on successful transfer. Academic Medicine, 82(10), S16–S18. Paas, F., Tuovinen, J. E., Tabbers, H., & Van Gerven, P. W. (2003). Cognitive load measurement as a means to advance cognitive load theory. Educational Psychologist, 38(1), 63–71. Paas, F. G. W. C., & Van Merrie¨nboer, J. J. G. (1994). Variability of worked examples and transfer of geometrical problem-solving skills: A cognitive-load approach. Journal of Educational Psychology, 86(1), 122. Pociask, F. D., DiZazzo-Miller, R., & Samuel, P. S. (2013). Reducing cognitive load while teaching complex instruction to occupational therapy students. The American Journal of Occupational Therapy, 67(5), e92–e99. Quilici, J. L., & Mayer, R. E. (1996). Role of examples in how students learn to categorize statistics word problems. Journal of Educational Psychology, 88(1), 144. Quilici, J. L., & Mayer, R. E. (2002). Teaching students to recognize structural similarities between statistics word problems. Applied Cognitive Psychology, 16(3), 325–342. Reed, S. K., & Bolstad, C. A. (1991). Use of examples and procedures in problem solving. Journal of Experimental Psychology. Learning, Memory, and Cognition, 17(4), 753–766. Richland, L. E., Bjork, R. A., Finley, J. R., & Linn, M. C. (2005). Linking cognitive science to education: Generation and interleaving effects. In Proceedings of the twenty-seventh annual conference of the cognitive science society. Mahwah, NJ: Erlbaum. Retrieved from http://learninglab.uchicago.edu. libaccess.lib.mcmaster.ca/Publications_files/5-CogsciIddeas2005.pdf Richland, L. E., Finley, J. R., & Bjork, R. A. (2004). Differentiating the contextual interference effect from the spacing effect. In Proceedings of the twenty-sixth annual conference of the cognitive science society (p. 1624). Retrieved from http://www.cogsci.northwestern.edu/cogsci2004/ma/ma239.pdf Rohrer, D., & Pashler, H. (2010). Recent research on human learning challenges conventional instructional strategies. Educational Researcher, 39(5), 406–412. doi:10.3102/0013189X10374770. Rohrer, D., & Taylor, K. (2007). The shuffling of mathematics problems improves learning. Instructional Science, 35(6), 481–498. Sankey, M., & Nooriafshar, M. (2005). Multiple representations in multimedia and e-learning materials: An issue of literacy. Retrieved from http://eprints.usq.edu.au/211 Scheiter, K., Gerjets, P., & Schuh, J. (2004). The impact of example comparisons on schema acquisition: do learners really need multiple examples? In Proceedings of the 6th international conference on
123
R. Chen et al. Learning sciences (pp. 457–464). Retrieved from http://dl.acm.org.libaccess.lib.mcmaster.ca/citation. cfm?id=1149182 Schnotz, W., & Ku¨rschner, C. (2007). A reconsideration of cognitive load theory. Educational Psychology Review, 19(4), 469–508. doi:10.1007/s10648-007-9053-4. Shea, J. B., & Morgan, R. L. (1979). Contextual interference effects on the acquisition, retention, and transfer of a motor skill. Journal of Experimental Psychology: Human Learning and Memory, 5(2), 179–187. Simon, D. A., & Bjork, R. A. (2001). Metacognition in motor learning. Journal of Experimental Psychology. Learning, Memory, and Cognition, 27(4), 907–912. Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design. Learning and Instruction, 4(4), 295–312. Sweller, J. (2010). Element Interactivity and Intrinsic, Extraneous, and Germane Cognitive Load. Educational Psychology Review, 22(2), 123–138. doi:10.1007/s10648-010-9128-5. Sweller, J., Van Merrienboer, J. J., & Paas, F. G. (1998). Cognitive architecture and instructional design. Educational Psychology Review, 10(3), 251–296. Taylor, K., & Rohrer, D. (2010). The effects of interleaved practice. Applied Cognitive Psychology, 24(6), 837–848. Van Merrienboer, J. J., & Ayres, P. (2005). Research on cognitive load theory and its design implications for e-learning. Educational Technology Research and Development, 53(3), 5–13. Van Merrie¨nboer, J. J., Kester, L., & Paas, F. (2006). Teaching complex rather than simple tasks: Balancing intrinsic and germane load to enhance transfer of learning. Applied Cognitive Psychology, 20(3), 343–352. Van Merrie¨nboer, J. J., & Sweller, J. (2005). Cognitive load theory and complex learning: Recent developments and future directions. Educational Psychology Review, 17(2), 147–177. Van Merrie¨nboer, J. J., & Sweller, J. (2010). Cognitive load theory in health professional education: design principles and strategies. Medical Education, 44(1), 85–93. Zulkiply, N., McLean, J., Burt, J. S., & Bath, D. (2012). Spacing and induction: Application to exemplars presented as auditory and visual text. Learning and Instruction, 22(3), 215–221. doi:10.1016/j. learninstruc.2011.11.002.
123
