Cognitive Science Learning Theories Applied1

COGNITIVE SCIENCE LEARNING THEORIES APPLIED1

Cognitive Science Learning Theories Applied to a Simulated Archaeological Dig Activity

Evan M. Silberman

New York University

Introduction

Imagine two students, John and Jane are participating in the same section of an undergraduate archaeology course. Both students are at the same point in the lecture, a section dedicated to Archaeological Dating. This is a particularly difficult concept, one that students universally struggle with. Jane’s instructor lectures on the topic with minimal supporting materials. John’s instructor uses a variety of pedagogical techniques, including a concept map of learning goals and objectives, assigned advanced readings, in-class group work, discussions, and exercises. Most notably, John’s instructor designed a multimedia simulation, which John must complete after the in-class lecture. The multimedia activity challenges John to date artifacts found in a simulated archaeological dig.

Upon the completion of the class, John and Jane compare their experiences. Jane reports struggling to understand the concepts. Her instructor gave a pop-quiz on archaeological dating. Jane reported failing the quiz. She recalls how complex the material was, and how rote her learning experience was. John, on the other hand, was graded on the simulated dig activity and received and passed the assignment. He attributes the simulation to his success and remembers being engaged in the experience.

John and Jane both have an opportunity during the summer semester to participate in an actual archaeological dig. John is confident he can transfer his practice from the simulation to a real dig and takes advantage of the summer field activity. Jane, on the other hand, remains uncertain about her knowledge of archeological dating and decides to not participate.

This is a common scenario. Two students learning the same material have different outcomes. What accounts for the John’s mastery of the subject matter? Although the two course sections are taught differently, this paper emphasizes the value of the simulated dig to promote active and deep learning. In comparing thetwo course sections taught by two different instructors using varying techniques, more standard assessments might be required. Assessments are outside the scope of this paper.

Although the above situation is hypothetical, our group designed an interactive multimedia simulation for a real-world context. Our simulation, which is the focus on this paper, is in response to Dr. Thomatos’s experience teaching archaeological dating to college students. The content is so complex, Dr. Thomatos no longer test students on it.

Our multimedia activity, a simulated archaeological dig (SAD) is an attempt to make the information comprehensible. This paper will discuss the theories of situated cognition, cognitive apprenticeship, dual-coding, cognitive load, and cognitive theory of multimedia learning, and it will explore principles akin to the theories. The paper will also discuss how the theories are applied to SAD and in some case the design of the instruction.

Theoretical Framework

General Learning Theory

Situated Cognition. One of the challenges of teaching students about archaeological dating is that the content is removed from the context. Learning archaeology in the classroom is dissimilar to digging at an excavation site, and dating items based on the dig, and lab analysis. For example, in order to apply the relative dating method, you need to know which layers of the earth an artifact is extracted from. You then date it in relationship to the items found in nearby layers. In a classroom setting, it’s difficult to replicate an excavation site.

SAD provides students a context that is a simulated archaeological dig and dating exercise. Students are presented with four artifacts to excavate. Upon excavating each item, students associate it within a strata of earth. Later in the activity, students related items to Carbon dated relics and are presented a timeline of the items in the exercise. All of which starts and ends in a virtual dig notebook, nearly identical to what’s used in the field. Therefore, SAD is a situated activity, and grounded in the theory of situated cognition.

Learning is a socially oriented process. Clancy explains: “Situated Cognition claims that every human thought is adapted to the environment, that is, situated, because what people perceive, how they conceive of their activity, and what they physically do develop together” (as cited in Driscoll, 2005, pp. 157). “The...approach is based on constructivist principles, in which a learner actively constructs an internal representation of knowledge by interacting with the material to be learned” (Plekhanova, 2005). Our in-class activities leading to the simulation, and the multimedia exercise itself, reflect situated cognition.

Legitimate peripheral participation is: “Learning viewed as situated activity has as its central defining characteristic a process [called] legitimate peripheral participation (As cited in Driscoll, 2005, pp. 165). Through LPP, learning occurs as learners deepen their participation in a community. Our classroom is the community, and in-class, group activities provide an opportunity to learn together. The simulation is a sociocultural setting where cultural context is provided for learning. It is another way for the learner to assimilate to the community of archaeology.

Students participate in their communities through cognitive apprenticeship and assessments in-situ. According to Driscoll (2005), cognitive apprenticeship is usually the relationship between a master teacher and student in which knowledge is transferred through the apprentice learning ‘on-the-job’ as guided by the apprentice. While the traditional master/apprentice relationship is difficult to reproduce in a classroom, our simulation models skills for the students required to become a master archaeologists. The assessment in-situ does not necessarily occur within SAD, despite a reflective journaling exercise. The animation is embedded within a platform that collects rich analytics. Analytics tracking can be embedded into the simulation. Therefore, the potential to gather summary statistics existsusing reporting features of systems like NYU Classes or software such as Google Analytics.According to Driscoll (2005), “McLellan (1993) recommended a three part model...the three parts provide three different kinds of assessments measures: (1) Diagnosis, (2), Summary Statistics, and (3) portfolios” (As cited in Driscoll, 2005, pp. 179). The most relative model for SAD is summary statistics such that they can help us understand how students interact with SAD.

Multimedia Specific Learning Theories

According to Plass, Homer, & Hayward (2009), there is increasing evidence that the educational efficacy of visualizations depends on how well they are designed to reflect human cognitive architecture. It is also a matter of learners cognitive ability to process and perceive the essential information presented. All this to say that how visual images are processed is complicated, and the process has the potential to increase cognitive load. The theories Plass, et al. (2009) think are most important include the dual-coding theory, cognitive load theory of multimedia learning, and cognitive load theory.

Dual-Coding Theory. Another basis of our design is Dual-Coding theory: “...Paivio and his colleagues demonstrated that people were better at remembering lists of words coded visually and verbally, rather than merely verbally...By encoding information to be learned in two modalities rather than a single modality, people have multiple retrieval cues that help them access information, thus enhancing memory” (Thompson et al., 2002). DCT is important because it demonstrates a popular model for processing verbal and visual information and associating that information with prior knowledge. The associative connections between the two representations are also important for effective instruction. According to Clark & Paivio (1992) verbal associative process are a critical component of good teaching. However, SAD plays more favorably to a modified version of DCT, the Dual-Coding Theory of Multimedia Learning. Mayer & Sims (1994) adopted the DCT for multimedia learning in which “learners construct referential connections between the mental representations of the verbal and visual information presented within a hypermedia document” (as cited in Boechler, 2006, pp. 575). They studied the effect of presenting animation and narration contiguously versus non-contiguously. The results demonstrated that learners performed better on transfer-test and problem solving when animation and narration were contiguous, also known as the contiguity effect.

We employ DCT in within our simulation where imagery is supported by text. Our animation also has contiguous narration, and our students are inexperienced, which are two factors for the contiguity effect. “Inexperienced students were better able to transfer what they had learned about a scientific system when visual and verbal explanations were presented concurrently than when visual and verbal explanations were separated (Mayer & Sims, 1994). Naturally, a concurrent presentation of visualizations and spoken text help students understand complex archaeological facts better. Through an improved understanding, they will perform well on their worked examples and graded dig report.

Figure 1. A SAD interface with pictures supported by text and contiguous narration.

Cognitive Load Theory and Cognitive Theory of Multimedia Learning. The Cognitive load theory is the most significant consideration for our SAD design. CLT is based on the idea that working memory is very limited.“The size of working memory is equal to the amount of information that can be verbally rehearsed in approximately 2 seconds...When information is attended to and enters working memory, if it is not consciously processed, it will decay in approximately 20 seconds” (Boechler, 2006). According to Baddeley (1992) as load increases on working memory, performance declines. Therefore, the goal of CLT is to reduce cognitive load. Anyway instruction decreases load is favorable, while still keeping it challenging.

CLT presents three types of loads: Intrinsic, Extraneous, and Germane. According to Mayer & Moreno (2010), intrinsic load results from the inherent complexity of the learning material, extraneous load results from mental activity that doesn’t contribute to learning, and germane load results from mental activity that does contribute to learning. Intrinsic load is within the control of the learner, while extraneous and germane load can be affected by instruction. The goal of CLT is to reduce intrinsic and extraneous load, but foster germane load.

Cognitive theory of multimedia learning addresses how learners interpret the messages of multimedia presentations. It is rooted in the multimedia principle. There are three assumptions of CTML:

●Dual Channels, according to Paivio (1986), Baddeley (1986, 1999) meaning, “humans possess separate channels for processing visual and auditory information” (as cited in Mayer, 2005 pp. 34).

●Limited capacity, according to Baddeley (1986, 1999), Chandler & Sweller (1991), meaning, “humans are limited in the amount of information that can be processed in each channel at one time” (as cited in Mayer, 2005, pp. 34).

●Active processing, according to Mayer (2001), Wittrock (1989) meaning, “meaningful learning depends on active cognitive processing during learning, including selecting relevant information for further processing, organizing selected material into a coherent mental representation, and integrating incoming material with existing knowledge” (as cited in Mayer, 2005, pp. 34).

The three assumptions are important for understanding how to present multimedia information. Presenting information using images and spoken word benefits learning because it uses the two channels available in our cognitive structure. To promote active processing, that information must be structured coherently in its presentation.

SAD focuses on reducing extraneous load, managing intrinsic load, and fostering germane load by applying several principles of CLT and CTML. The next section of this paper will review principles akin to our design and how they are implemented in SAD.

Reducing Extraneous Cognitive Load

Redundancy Principle

According to Mayer & Moreno (2010), students learn better when redundant on-screen text is eliminated from an animation using narration. In our simulation, on-screen text is excluded to the extent that our narration is unique. There is no on-screen text that is similar to the spoken text.

Temporal Contiguity Principle

According to the temporal contiguity principle: “If meaningful learning depends on holding corresponding words and pictures in working memory at the same time, then successive presentation of narration and animation can easily overload the learner’s cognitive system” (Mayer & Moreno, 2010). Therefore, presenting narration and animation simultaneously reduces extraneous cognitive load. Each segment of SAD presents animation and narration contiguously.

Reducing Intrinsic Cognitive Load

Segmenting Principle

According to the segmenting principle: “People learn more deeply when a multimedia message is presented in user-paced segments rather than as a continuous unit” (Sweller, 2005). Chunking or segmenting an animation to give the learner control of the presentation reflects the segmenting principle. Each step of the activities embedded in SAD are separate. The user can control the progress of SAD by choosing when to continue to the previous or next section. Likewise, narration can be turned on or off, or paused so the student controls the portion of audio and its pacing.

Modality Principle

According to Mayer & Moreno (2010), the modality principle is important when presenting challenging material in a multimedia presentation. According to the modality principle, presenting animation and narration is preferred over animations and on-screen text. In other words, using spoken word to explain animations helps share the load in our different processing channels. SAD adheres to this principle by providing its explanations through spoken-word, thereby eliminating the need for written descriptions of visual concepts in most cases.

Other principles

Multimedia Principle

According to Fletcher & Tobias (2005) the multimedia principle states: “...people learn better from words and pictures than from words alone...that people learn more or more deeply when appropriate pictures are added to text” (as cited in Mayer, 2005, pp. 117). This principle is a foundation of SAD because the simulation largely uses pictures and text, especially on third screen of the virtual dig notebook. The pictures and text correspond and together represent artifacts and explanatory information about the items. They provide context for the relationship between an item and its history, particularly the date.

Worked-Out Example Principle

According to Renkel (2005), a worked-out example generally consists of problem formulation, solution steps, and the final solution. A rule is first introduced, the worked-out example is presented, and one or more problems to be solved are provided.

Worked-out examples are an effective tool for complex subjects like archaeology, especially for archaeological dating. Both the simulation and the course design implement worked-out examples to help scaffold learning. For example, students are asked to complete a partial equation for carbon dating given some of its variables. In another activity, students are given a blank harris matrix and information from a logbook and are asked to complete the matrix.

Worked-out examples are especially effective in our simulation because our learners are low-prior knowledge students. As their expertise increases, the effectiveness of worked-out examples, especially fully or partially completed models will be less effective (i.e. expertise reversal effect).

Over the course of the class, and in combination with the simulation, we adhere to these guidelines. Students are learning these concepts for the first time, and we think problem-solving through worked-out examples will deepen their understanding of the concepts, and sharpen their skills.

Active Processing and Reflection Principle

A significant aspect of our simulation is the end of the activity in which we ask students to write a dig report. We provide an open-ended forum for them to reflect on their simulated dig. Journaling, in this scenario, is meant to promote active cognitive engagement. It uses the active processing principle of CTML as a basis, because “Students may fail to learn unless instruction includes methods aimed at engaging the learner…” (Moreno & Mayer, 2010). It is important to note, that reflection is only accessible after the student correctly completes the preceding activities. As noted by Moreno & Mayer (2010), reflection against correct information deepens learning.

Feedback Principle

According to the feedback principle: “Novice students learn better when presented with explanatory feedback during learning” (Moreno & Mayer, 2010). During the dating segments of SAD, students are prompted with explanatory feedback to aid learning. Like the reflection principle, students will learn better if given corrective feedback plus explanation. This helps them create mental models of the newly acquired information.

Personalization Principle

According to Moreno & Mayer (2010), a personalized environment aids active and meaningful learning. In a related study, Mayer (2003) discovered that social cues in multimedia messages ignite social conversation schema in learners. This results in conversational like behavior and enacts rules as if the person was in a social conversation. This is particularly true when the voice of the spoken text is a standard-accented human Therefore, deep cognitive processing is possible when narration is conversational and human.

The narration in our simulation will be a familiar and welcoming voice. It will address students as if they have a personal relationship using terms like “you”, “we”, and “our”. By scripting the spoken words to be conversational, the simulation will feel more personal.

Pretraining Principle

According to the pretraining principle: “people learning more deeply from a multimedia messages when they know the names and characteristics of the main concepts” (Mayer, 2005). SAD is completed after the in-class lecture on archeological dating. The names and characteristics of the main concepts are provided through the lecture, group work, reading, and other activities leading up to the simulation. Therefore, the students have the prior knowledge necessary for deeper learning because pretraining, according to Mayer (2005), also helps reduce intrinsic cognitive load.

Guided Activity Principle

According to the guided activity principle: “instruction that allows students to interact by dialoguing and manipulating the learning materials is more likely to lead to meaningful learning then instruction that does not…” (Moreno & Mayer, 2010). SAD is an interactive simulation, and students must manipulate objects to complete the partial worked-out examples. The interaction is indented to provide students with deeper learning. For example, students must select a trench, then an artifact, and ultimately place the selected relic in the appropriate location on two different graphics to represent its age. Each step of the process is guided through a narration. Together, the spoken text and control of objects promotes active learning.