Anchored Instruction and Anchored Assessment: An Ecological Approach to Measuring Situated Learning.

Anchored instruction and anchored assessment are described and illustrated through a mathematics problem from the Jasper problem solving series developed at Vanderbilt University in Nashville (Tennessee). Anchored instruction is instruction situated in a context complex enough to provide meaning and reasons for why information is useful. Problems anchored in a complex context require anchored assessment, assessment that is a seamless, to the extent possible, part of the instruction process. A prototype assessment approach, the Jasper Planning Assistant (JPA), is described. Transfer from a single mathematical problem solving activity to reading comprehension of passages with analogous content, and the absence of transfer across content domains demonstrated in a study of 121 middle school students, is described. It is speculated that cross-domain transfer will require anchored instruction that provides a generator set of situations across which students could detect invariants that specify when higher order thinking skills would be useful. Assessment techniques for anchored instruction and situated learning must adapt to accommodate the non-linear topological dynamics that are seen when complex realistic problem solving is described as a perception-action cycle. Eight figures illustrate :Ihe discussion. Three appendixes provide sample Jasper verbal protocol and analysis, and two samples of JPA output. Thirty-eight references are included. (SLD) *********************************************************************** * Reproductions supplied by EDRS are the best that can be made from the original document. *********************************************************************** U.S. DEPARTMENT OF EDUCATION Office of Educational Research and Improvement EDUCATIONAL RESOURCES INFORMATION CENTER (ERIC) document has been reproduced as received from the person or organitatic,n originating a C' Minor changes have been made to improve reproduction quality Points of mew or opinions stated in this docu rnent do not necessarily represent official OERI position or policy "PERMISSION TO REPRODUCE THIS MATERIAL HAS BEEN GRANTED BY P-1/046-4. F You,* TO THE EDUCATIONAL RESOURCES INFORMATION CENTER (ERIC)." Anchored Instruction and Anchored Assessment: An Ecological Approach to Measuring Situated Learning Michael F. Young and Jonna M. Kulikowich Educational Psychology Department The University of Connecticut Box U-4 249 Glenbrook Rd. Storrs CT 06269-2004 USA E-Mail, Bitnet: myoung@uconnvm Internet: myoung@uconnvm.uconn.edu Applelink: young.Mike Am. Online: Mike Young Paper presented at AERA annual meeting, session 28.30, "Grounding mathematical problem solving in meaningful contexts: Research implications and outcomes", Wed. April 22 1992, San Francisco, CA. Acknowledgment: This work has been supported in part by a grant from the UConn Research Foundation. Significant contributions to the ideas expressed in this paper were made by discussions with John Rickards and Scott Brown, by collaborative writing with Mike McNeese, and by the ideas of Bob Shaw, Michael Turvey, and the members of the Cognition and Technology Group at Vanderbilt. Many thanks also go to Marianne Cavenaugh and the students at Glastonbury's Gideon Wells Middle School who worked so hard to solve the Jasper problem, and Lynette Braunhardt and the students at Mansfield Middle School whose problem solving activities laid the groundwork for figuring out our data. BEcT Crj7f'tL.1-C e,13r1 wrl 411 w JPA and Anchored Assessment 2 Anchored Instruction and Anchoreci Assessment: An Ecological Approach to Measuring Situated Learning The benefits of teaching in a complex realistic context have been suggested by many sources from Dewey (1938) to the recent discussions of situated learning and situated cognition. Context provides meaning, enriches perception, and affords development of complex problem solving and higher level thinking skills. if this is so, then we must develop means to creatively assess the effects of anchored instruction, benefits to mathematical thinking and beyond, across subject domains and across transfer situations. This paper discusses the Jasper Planning Assistant as an assessment of higher level mathematical thinking. In addition, we describe transfer from a single mathematical problem solving activity to reading comprehension of passages with analogous content, and the absence of transfer across content domains. We speculate that crossdomain transfer will require anchored instruction that provides a "generator set" of situations across which students can detect invariants that specify when higher order thinking skills, such as planning, would be useful. We conclude that assessment techniques for situated learning and anchored instruction must adapt to accommodate the nonlinear topological dynamics that are inherent when complex realistic problem solving is described as a perception-action cycle. The benefits of teaching in realistic contexts have been touted for a long time (e.g., Dewey, 1938; Whitehead, 1929). For example it has been suggested that situated and experiential learning in everyday settings both provide meaning to our current activities (e.g., cognitive apprenticeships, Collins, Brown, & Newman, 1989) and impel or give meaning to what happens next (Lave, 1988); that is, the environment influences both perception and action. At a basic level, nearly all agree that situations are a part of learning, whether in the form of episodic memories, merely as part of the backdrop for learning, or more fundamentally integrated with learning as described by the ecological psychology of James Gibson (1979). The ideas of 'situated learning" take the view that situations are inevitably an integral part of what is learned: all learning is situated and, therefore, learning should be done in authentic (see Note 1) settings (see for example, Brown, Collins, & Duguid, 1989). Recently, an even stronger assertion about situations has been put forth: that not only learning, but all thinking is, in fact, situated. This is the psychology of situated cognition (Clancey & Roschelle, in press; Greeno, 1989). These acknowledgments for the importance of situations in learning and thinking compel researchers to consider more closely the advantages and limitations of teaching through situations (anchored instruction) and to develop means to assess "situated learning" (anchored assessment). We view complex situated problem solving from the perspective of ecological psychology. From this perspective we acknowledge the primacy of the interaction between the skills and abilities brought to the situation by the problem solver (effectivities) and the affordances for action provided by the problem environment or problem space -a symmetry of acausal interactions (Shaw, Turvey & Mace, 1982). This relationship is captured in the perceiving-acting cycle that temporally unfolds through the problem-solving process. It would not be meaningful to characterize the problem solving of an individual apart from the context in which that problem solving occurs. A situated cognitive analysis of thinking must describe both the abilities that each person brings to the table as well as all the relevant attributes (affordances) of the JP:\ and Anchored Assessment 3 environment including dimensions of the problem and problem space that afford certain actions. In contrast to Skinner's (1987) impoverished conception of environmental stimuli, an ecological approach acknowledges the richness and complexity of the information available in the environment and its co-determinant role in thinking. The "situatedness" of knowledge is consistent with much of what psychology has learned in several areas. Psycho linguists have strongly acknowledged the importance of context, citing simple examples such as "indexicals" (I, you, here, there) that only have specific meaning in a particular situation (Bruner, 1986; Miller & Gildea, 1987). Sociologists have also acknowledged that context (particularly culture) arises from and gives meaning to social interactions (Coulter, 1989; Saxe, 1991). Using an ecological perspective, we have found it useful to apply this analysis to mathematical and scientific thinking, specifically problem solving. That is, when viewed from this perspective, knowledge, thinking and problem solving are not properties of individuals, but rather, they "live" in the interaction between the capabilities of problem solvers and attributes of the problem (specifically the problem and solution spaces). When perception is emphasized over memory, it is the information picked up from the environment that is perceived and acted upon, and which must become part of assessment, not simply the actions or results of problem solving. Kugler, Shaw, Vicente, and Kinsella-Shaw (1991) described goal-directed activity, such as problem solving, as an interaction of attractor sets, specifically the attractor sets supplied by a complex realistic context (affordances) and the attractor processes (effectivities) by which we achieve the goal-states set up by our problem solving intentions. Their analysis suggests that it is the information available from a situation that guides and constrains problem solving: "The behavior of inanimate systems is lawfully determined by a force field, whereas, the behavior of animate systems is lawfully specified by an information field (p. 408)." In their analysis of selforganization and intentional systems (such as people), they give mathematic substance to Gibson's (1979) principle of "organism-environment mutuality." Applying these ideas to anchored instruction suggests that problem solving is an interaction between the problem solving skills of the individual problem solver and the activities and manipulations that a particular problem affords. In their terms the interaction between agent and environment (problem solver and problem) is inherently nonlinear: Fields that have hidde