Abstract Young, Kulikowich, & Barab (1997).
With the increasing educational use of multimedia hypertext, Internet “Web” pages, interactive CD-ROMS, and interactive videodiscs, there is a growing opportunity to exploit computer interactions for assessment. Many such environments are inherently nonlinear, open-ended and afford multiple paths to solution. Recently, we have found it useful to graphically represent transitions from screen to screen using Pathfinder (Schvaneveldt, 1990), but have found this technique excludes extensive information contained in the dribble file and fails to become part of a dynamic interaction (Barab, Fajen, Kulikowich & Young, in press). We propose the adoption of an agent-environment interaction as the unit of analysis. It is our conclusion that assessment should serve to optimize performance. Assuming the agent as detector metaphor, we recommend the interaction of a perceiving/acting agent with dynamic information-rich environment as the unit of analysis. Assessment then will produce information that will extend the detection capabilities of the agent(s) and enable them to “tune” the assessment to their changing goals and intentions. Rather than standing apart assessing static knowledge, assessment should become part of the dynamic context. We identify two candidates for defining an interative unit. The first comes from the dynamics of intentions, in which competing goals are adopted (dynamics of intentions). The second source comes from the choice points within the perceiving-acting cycle toward a single goal (intentional dynamics).
Abstract. Barab, Fajen, Kulikowich, & Young. (1996).
In response to the need for more dynamic forms of assessment, our primary aim was to evaluate the validity of Pathfinder for representing and comparing individuals’ navigation through a hypertext. Our secondary aim was to evaluate the validity of Pathfinder for generating an empirically-derived path that represents a set of individual navigational paths. Participants were divided into two groups: a Test-based group and a Problem-based group. Based on prior research, we assumed that these groups would differ with respect to navigation. Pathfinder-generated networks and the related C-measure were used to test the validity of Pathfinder for capturing differences between these groups. The results supported our hypothesis that Pathfinder would be able to distinguish between the two groups. Furthermore, it was shown that, with some modification, Pathfinder was able to generate an empirically-derived network which represented the paths of the individuals in the group from which it was derived.
Abstract. Barab, Young, & Wang (1999).
The study examined learning while using a linear text, navigational hypertext, or a generative hypertext system. In Experiment One, students were assigned one of these treatments and expected to learn the information to solve a posed problem, while students in Experiment Two learned the information to pass a reading comprehension test. Results from Experiment One revealed that the multivariate composite of self determination (SD), problem solving (PS), and reading comprehension (RC) scores explained 65% of the variance in treatments, with all three variables being significant. Students in the Navigational condition had the highest SD and PS scores, and students in the Linear condition had the highest RC scores. Results from Experiment Two revealed no significant differences among treatments for the outcome measures. However, number of generative actions explained a significant amount of the variance in PS and RC scores. Qualitative grouping of log-file data using Pathfinder illuminated patterns of user’s navigational processes.
Abstract. Young, Nastasi, & Braunhardt . (1996).
A working definition of “situated learning” and “situated cognition” initiated a 3-month project designed to replace students’ traditional math instruction with “immersion” into three complex realistic everyday mathematics problems. As the intervention unfolded, the researchers and classroom teacher engaged in a dialog that modified the initial conception of both the research and teaching efforts. Daily classroom observations structured weekly interview questions and generated hypotheses that could be investigated quantitatively and qualitatively. Conversely, feedback from quantitative tests and qualitative weekly interviews supported classroom efforts by eliciting information from students and feeding those students problem-solving strategies (primarily social) back to the classroom. As might be expected with nearly any constructivist implementation, conceptual change was evident from both the classroom teacher and educational researchers’ perspective. This may be a common attribute of action research based on constructivist and situated learning classroom interventions. Techniques and understandings developed from the intervention have continued, and a expanded working definition of situated learning has emerged.
Abstract: Young (1995).
Based on a theory of situated learning, assessment must emphasize process as much as product. Assessment must move away from a linear additive model, accepting at the outset the complex, nonlinear, and possibly chaotic nature of real learning. Assessment must adapt to and take advantage of students working with technologies that extend their perceptual and problem-solving capacities beyond what they could do alone. To illustrate these points, several assessment examples are given, including a computer-based planning assistant for a videodisc mathematics and science video, suggestions for computer-based portfolio assessment, and speculations about embedded assessment of virtual situations. In all cases, the theory of situated learning emphasized perception over memory, and the co-determined nature of learning and thinking.
Abstract Duarte, Young, & DeFranco (2000).
Experts in the area of using technology to enhance mathematics instruction have, through national standards, attempted to convey ideas concerning best practices. Research concerning Social Judgment Theory (SJT) has found many instances where expert judges, when asked, report one set of factors to be most important, but when making actual judgments, they rely on a very different weighting of factors. We developed 20 profiles of actual computer-using mathematics teachers and asked experts to rank the profiles in the context of selecting a profile for a hypothetical award. We used regression analysis to predict the award winner based on experts’ actual use of profile dimensions. Additionally, experts were asked to report weightings of dimensions they felt they used in making judgments. Actual award judgments were highly predictable from profile dimensions (adjusted R-square = .98) with teacher actions being most relevant, along with student actions and the nature of the classroom environment. But in contrast, what experts say is most important in influencing their judgments was less predictable (adjusted R-square = .25), with nature of the environment reported to be most influential, and nature of the task reported to be a negative influence. The implications for professional development are discussed.
Abstract Young, Barab, Garrett (2000).
For a moment, throw aside the concepts of memory, storage and retrieval of information, and think of students as sophisticated information detectors, instead of complex symbol manipulating information processors. Think of instructional design as the process of creating environments that afford meaningful authentic activity that tunes novice perceivers to the affordances for action of tools used by communities of mathematicians, physicists, musicians, and artists. Consider the potential to quantify scaffolding by mathematical descriptions of treatment gradients detected during automatic performance. Now consider that many students do not have goals for which these tools afford much action, so they must be led to adopt new goals (the education of intention). Descriptions of learners as detectors leads us to prefer student immersion into new environments through a well-written novel, a good video, or a virtual reality that can create interactions in which new goals emerge. Clever instructional designers can detect the affordances of these contrived environments and use them to encourage students to adopt new goals (build a bridge, design a robot, help a character solve an everyday problem). Those clever designers can also create learning environments that afford progress toward these new goals, environments that are so closely aligned with the goals that they encourage optimal performance. We have found thinking in this way to be liberating and we hope others will indulge in the enterprise with us.
Abstract Young & Barab (2000).
In this study we provide evidence that video anchors encourage students to adopt certain contrived goals over their more naturalistic goals. Well designed goals are those that increase the probability that students will detect the raison d’être of content information as they work toward the contrived goal. In those cases where problem solvers detect the raison d’être of the material, transfer should be enhanced. In experiment 1, participants worked with hypermedia in which either an anchor problem was posed, or they were preparing for a test. They were grouped based on their report of adopting one of four goals. Results showed 72% of students assigned the anchor reported their goals as “to solve a meaningful problem,” and students with such goals showed constrained navigation toward relevant information and significantly higher learning efficiency scores than students with any of the other goals. In experiment 2, results demonstrated that when an anchor encourages students to adopt goals that increase the probability that they will detect the raison d’être of the material, transfer is enhanced. Individuals assigned the anchor reconstructed significantly more information and detected qualitatively different information in a brief transfer video. Goals that enable the problem solver to detect the raison d’être of information enhance learning, focus navigation within nonlinear text, and afford transfer.
Abstract Arthurs, DeFranco & Young (1999).
Describes a study of sixth grade middle school students that examined whether tuning students’ attention to information in a mathematics problem impacted their problem-solving performance. Discusses use of the Jasper Woodbury Problem-Solving series and results of statistical analyses that showed the importance of creating mathematical problems based on students’ interests.
Abrstract of Barab, Cherkes-Julkowski, Swenson, Garrett, Shaw, & Young (1999).
Offers an explanation for how learning takes place through the individual-environment relation. The authors provide a general description and beginning vocabulary for the philosophical shift in viewpoint that comes with a nonlinear systems approach focusing on self-organization. In this view, knowing, meaning, and cognition are actualized through the dynamic between learner (self) and environment (nonself). It is argued that the ecologized, or self-organization model (relational ontology) establishes that the particles (learners), in effect, strive to order themselves once the intention has been initiated. From this perspective, instruction involves establishing the appropriate field conditions or connecting the learner into a system (set of relations) through participation in the service of an intention. The type of learning advocated cannot be handed to the learner, but develops through dynamic participation as part of a system. Central to this line of reasoning is the assertion that learner practices and meaningful relations that arise due to their functional significance as part of a dynamic system are fundamentally different from teacher or textbook descriptions, and that they are in this way far richer, more meaningful, and more functional.