Search Contact Resources Library Knowledge Solutions Campus Forensic Science Bookstore Cases Home


Using Case-Based Reasoning and Cognitive Apprenticeship 
to Teach Criminal Profiling and Internet Crime Investigation

by Eoghan Casey, BS
Master of Arts Candidate in Educational Communication and Technology
New York University
Note: Eoghan Casey is a full partner of Knowledge Solutions LLC.
He can be reached for comment or consultation by contacting: 
Knowledge Solutions; 61535 S Hwy 97, #9-148; Bend, OR 97702; 
Phone 541-318-8293; Email:


Blue_Swirl52D5.gif (150 bytes) Introduction 

Blue_Swirl52D5.gif (150 bytes) Case-based reasoning

Blue_Swirl52D5.gif (150 bytes) Cognitive Apprenticeship





Blue_Swirl52D5.gif (150 bytes) Course Description: Internet Crime Investigation

Blue_Swirl52D5.gif (150 bytes) Conclusions

Blue_Swirl52D5.gif (150 bytes) References



Extensive research shows that we learn best through a combination of theory and practice [2]. Theory is inert until it is applied to a realistic situation, and practice is shallow without the influence of theory. A clear example is how modern pilots are trained. Though attending lectures and reading books is necessary, it is not sufficient training for a pilot. There is an obvious need for practical experience. This is why millions of dollars have been spent on flight simulators – to provide a safe, realistic environment to practice skills and learn from inevitable mistakes. Similarly, with Criminal Profiling and Internet Crime Investigation, students learn best by applying theory and techniques to realistic cases in an environment where they can make mistakes without the real world consequences of failure. 

The purpose of this paper is to outline a powerful approach to teaching Criminal Profiling and Internet Crime Investigation. This approach - informed by cognitive science - combines theory and practice by teaching through cases, supplying relevant theory when necessary. The two pertinent instructional strategies, Case-Based Reasoning and Cognitive Apprenticeship, are described in the beginning of this paper. The discussion of these strategies is followed by an actual example of how both strategies have been combined to implement a realistic learning environment. Finally, the main benefits of using the approaches described in this paper are summarized. 


Case-based reasoning

Using cases to teach Criminal Profiling and Internet Crime Investigation might seem like a straightforward procedure. After all, the theories and techniques that are used to investigate criminal cases were born from past cases, and are thus inextricably bound to them. However, there are some subtleties in the implementation that deserve serious consideration. It is not enough to use a single case, or to string cases together in an arbitrary manner and expect students to flourish. Having a carefully structured set of cases that clearly demonstrate how experts link and compare cases is crucial when teaching any aspect of criminal investigation. To better understand the subtlety of choosing instructional cases, let us look at two serious mistakes that investigators make: 


  1. Inappropriate case comparison – The most pronounced instances of this mistake occur when investigators cannot comprehend a heinous crime in terms of their past experiences and so they attribute the crime to less tangible influences such as the devil and "pure evil". Some investigations have become witch-hunts (or Satan worshipper hunts) as a result of this mistake, leaving the actual perpetrator free to commit more heinous acts.
  1. Failure to make appropriate case comparisons – This is so common that it has earned the name "linkage blindness" [7]. Investigators fail to notice similarities in several cases and thereby fail to link the crimes to a single perpetrator. This failure to link cases makes it significantly harder for investigators to apprehend the perpetrator thus allowing him to commit additional crimes.

In light of these recurring problems, it becomes clearer that the choice of cases in instruction is crucial. In addition to learning how to apply theory to cases, students need to learn how to compare cases and to evaluate the validity of these comparisons (i.e. critical thinking). A model of how people reason using cases, called case-based reasoning (CBR), can help us choose instructional cases that address the two problems described above. 

In the late 1970’s Roger Schank and Robert Abelson [5] initiated the research that led to the Case-Based Reasoning model. While studying under Schank, Janet Kolodner [4] refined the CBR model and used it to create a computer program that demonstrated a limited ability to reason. CBR has been further developed and is regularly used to create successful learning environments and expert systems. 

A key assumption of the CBR model is that people understand new experiences in terms of past ones: 

"Knowing where something fits in with what we already know is a prerequisite to reasoning about it. We find it hard to suggest a solution to a problem until we understand the problem. … We use what we know about a situation to find the closest match in memory, and we use that as the basis for deriving expectations and inferences." [4] 

Even when we are reasoning deductively we are employing knowledge that has been distilled from past experiences to arrive at new information. The CBR model goes on to suggest that, to make case comparison easier, we organize cases in our minds by indexing them in two distinct ways ([4], [6]): 


  1. Type of case and how it differs from similar cases. For example, an investigator might index a case as a sadistic rapist who slaps breast and uses pliers to associate it with, but differentiate it from, other cases involving sadistic rapists. The same investigator might also index the case as a mobile offender who has committed similar crimes elsewhere. Both indexes are valid and each one is useful when comparing cases. For example, the investigator might try to recall other cases that involve a sadistic rapist, and search through literature for other sadistic rape cases. Also, the investigator might take a hint from the second index and look in surrounding geographical areas for crimes that match some part of the first index.
  1. The goals of participants in a case (e.g. acquire money) and the details surrounding those goals. For example, an investigator might index a case as sadistic rapist intended to kill victim but was interrupted. The surrounding details, adapted from Kolodner [4], might include:
  • a description of motive or intents (e.g. evil intent).
  • the plan or type of plan being used to achieve the goals (e.g. search plan; cover tracks plan; attack plan)
  • characteristics of the plan being used (e.g. easy but inappropriate; strange strategy; dirty tactics)
  • unusual conditions that influence the plan (e.g. outside opposition; difficulties along the way; outside help; normal solution will not work)
  • if and how the goals were achieved (e.g. apparent success; apparent failure; goals reached using compromise solution, goals reached after opponent quit; goals not reached because opponent successfully resisted) 

 We create an index for each aspect of a case that is distinctive or violates our expectations. Therefore, a single case can have many indexes. Also, different people will index cases very differently unless some indexing guidelines are provided. When we encounter a new case, we compare it with cases that have similar indexes, using the differences and similarities between cases to understand them. If a case does not fit into our existing understanding we can reject the case, distort the new case until it fits into our existing understanding, or create new indexes that accommodate the case. When teaching criminal investigation, the latter method is most desirable. 

To address the problem of inappropriate case comparison in criminal investigation, we can present students with cases that they cannot understand in terms of past experiences. After repeatedly failing to understand these cases using past knowledge, students will appreciate the vocabulary, theory and techniques that help them understand these unusual new cases. The purpose here is to wean students from their personal indexing systems, help them develop a new indexing system for criminal cases, and teach them the usefulness of relevant theory. 

Once students have developed several new indexes for criminal cases (e.g. sadistic rapist; motive of offender) additional cases can be introduced that have some relation to previously presented cases. A careful combination of cases can teach students to identify cases with similar behaviors, motives, MO’s (modus operandi – actions necessary to commit the crime), or signatures (actions performed by the perpetrator that were not necessary to commit the crime) [3]. 


Cognitive Apprenticeship

Teaching theory and teaching how that theory can be applied to cases is a major undertaking. The added effort and complexity of teaching through cases could explain why many instructors retain the traditional lecture format. Fortunately, a powerful instructional strategy called Cognitive Apprenticeship [2] can help us with the difficult task of case-based instruction. Cognitive Apprenticeship is founded on the assumption that people learn best in authentic environments [1] and has been used successfully to teach subjects such as writing, mathematics, multimedia production and computer programming. 

In cognitive apprenticeship the instructor first gives students an overview of what they will be learning so that they know from the outset what to expect and how things fit together. The instructor then demonstrates how to do certain tasks. As the instruction progresses, the tasks become increasingly demanding, leading the student into the “culture of expert practice”. Students try each task themselves with some coaching from the instructor. The instructor gradually reduces the coaching, ultimately leaving the students to perform the tasks on their own. Students are encouraged share understanding and to explore, reflect upon, and articulate ideas throughout the learning process. 

Ideally, the learning environment will provide students with the resources that they require to carry out tasks and become independent learners and practitioners. In cognitive apprenticeship these supportive resources are referred to as scaffolding. 

Collins, Brown, and Newman [2] provide the following framework for an ideal cognitive apprenticeship learning environment. The remainder of this paper will describe each characteristic of this ideal framework and how they relate to teaching students of criminal investigation to apply theory to cases and reason using cases. 


  • Domain knowledge 
  • Heuristic strategies 
  • Control strategies 
  • Learning strategies


  • Modeling 
  • Coaching 
  • Scaffolding and fading 
  • Articulation 
  • Reflection 
  • Exploration


  • Increasing complexity 
  • Increasing diversity 
  • Global before local skills


  • Situated learning 
  • Culture of expert practice 
  • Intrinsic motivation 
  • Exploiting cooperation 
  • Exploiting competition



Domain knowledge is the conceptual, factual and procedural knowledge that we normally associate with a subject. This is the knowledge that is usually found in books and covered in traditional lectures. While working on cases, students will have to use this domain knowledge to assist them in an investigation. It is the responsibility of the instructor to present this knowledge in a clear and accessible manner and sensitize students to when this knowledge is applicable. 

Heuristic strategies are the rules of thumb that expert investigators use to accomplish a task. For example, whenever possible, an investigator will look through all of the available evidence before forming an opinion or focusing on specifics in the case. An expert Criminal Profiler will examine crime scenes first, victims second, and possible offenders last. A more difficult heuristic strategy to teach is how to identify pieces of information (clues) that are likely to lead to more information. An instructor can teach heuristic strategies by example and can make their rationale and usefulness clear whenever the opportunity arises. 

Experts use control strategies to choose applicable approaches during an investigation. These control strategies involve self-monitoring, self-diagnosis, and self-correction. An expert will identify the need for correction, will determine what needs to be corrected, and will choose a method of correction. Experts use control strategies to choose appropriate domain knowledge and heuristic strategies, as well as to decide when to change strategies. For example, when expert investigators get stuck on one part of a case, they will move to a different part of the case for a while. Again, an instructor can teach students these control strategies by example, and by making self-monitoring and evaluation an inherent part of the learning process. Having students compare their approaches with the approaches of experts and of other students will help them develop a refined understanding of where they are going astray and how they can correct their mistakes. 

Learning strategies are used to acquire additional domain knowledge, heuristic strategies, and control strategies. The instructor can encourage these learning strategies by teaching students to use additional sources. Also, the instructor can help students evaluate the work of other investigators’, by identifying flaws and picking out useful techniques. 



Modeling is one of the most important aspects of cognitive apprenticeship and should be given a great deal of attention. This is because “observation plays a surprisingly key role” [2] in learning. By observing the instructor perform a specific task, the students get a better idea of what they are supposed to do and how they are supposed to do it. It is tempting to make checklists rather than demonstrate how to do something. However, checklists create inflexible learners who have difficulty progressing beyond or operating without the checklists. 

It takes some trial and error before students can perform the tasks that they have seen modeled. The instructor can coach students through this trial and error process by showing them where they went wrong and how to correct the error. For example, after showing students how to read and interpret an autopsy report or an email header, the instructor can ask students to interpret a new autopsy report or email header. The instructor and other students can then examine and learn from each other's interpretations. If necessary, the instructor can model the process again for students who are having difficulty. Over time, this coaching should be reduced and ultimately stopped. This reduction is called fading. 

Any additional support (other than modeling and coaching) that helps students develop necessary skills is called scaffolding. For instance, to give students the satisfaction of making sense of a case, the instructor can carry out tasks that are too advanced for students and can then allow the students to put the prefabricated pieces together. Additional support can be provided by organizing case materials in a logical way, making them easy to access and search, and by giving students questions to ask themselves as they look through the materials. 

Creating a forum where students can offer each other support and share their understanding is another form of scaffolding. Such a forum can also be used to help students articulate and reflect upon what they have learned. The more students practice these high level thought processes, the better they will understand what they are learning. Shared articulation and reflection usually magnifies the benefits of these processes. 

Encouraging and facilitating exploration is key to making students more independent. Giving students an interesting assignment with only generally formulated goals gives students the latitude to explore and thus extend their understanding of a subject. Exploration can also help students gain confidence in their ability to learn on their own. 



We have already considered how to sequence cases when teaching criminal investigation. It remains to discuss how to sequence the vocabulary, theory and techniques required to understand these cases. The three main recommendations in cognitive apprenticeship are increasing complexity, increasing diversity and global before local skills. 

Increasing complexity involves gradually putting more responsibility on the students, forcing them to perform more of the tasks required in an investigation and making those tasks more challenging. For instance, after giving students the satisfaction of putting prefabricated pieces of a case together, the instructor can present a new case that requires the students to do some of the work that was previously done for them. 

Increasing diversity involves choosing cases that require a wider and wider variety of strategies and skills. The main purpose of increasing diversity is to teach students that what they are learning is applicable in a variety of different situations. Essentially, this is an effort to help students generalize what they are learning so that the knowledge is not permanently locked within the cases that they learned from. Also, increasing the variety of strategies and skills required in a case can develop students' control strategies. 

A good example of presenting global skills before local skills can be found in tailoring apprenticeship. Apprentices learn “to put together a garment using precut pieces before they learn to draw and cut out the pieces themselves” [2]. The purpose of this is to demonstrate the ultimate purpose of a process before presenting students with all of the details. Once the students understand the ultimate purpose of a process they can develop the necessary skills to achieve this purpose. In criminal investigation, the same effect can be achieved by giving students evidence that has already been interpreted before asking them to interpret evidence themselves. 



By choosing cases to teach criminal investigation, we are situating the learning process in an realistic context. In cognitive apprenticeship, the subsequent learning that takes place is called situated learning. Collins, Brown, and Newman summarize several general purposes served by situated learning. “First, students come to understand the purposes or uses of the knowledge they are learning. Second, they learn by actively using knowledge rather than passively receiving it. Third, they learn the different conditions under which their knowledge can be applied.” [2] 

One of the main goals of cognitive apprenticeship is to bring students into the culture of expert practice. This is achieved through situated learning, modeling and comparing student and expert methods. An additional step can be taken when teaching Criminal Profiling and Internet Crime Investigation by encouraging students to get involved in the expert discussions that occur in related mailing lists and discussion groups on the Internet. 

Intrinsic motivation is an important aspect of creating a learning environment. If students do not see the purpose of what they are learning and do not find it interesting, they will not be motivated to learn. When teaching criminal investigation using cases, an effort must be made to present cases in a way that encourages students to become involved. If a case is described in minute detail, students might absorb some of the details, but they will not necessarily be engaged by it or even learn from it. On the other hand, if a case is presented in a puzzle or problem solving format where students must put the pieces together before they can understand the case, they are more likely to be engaged and they are more likely to be motivated to learn. 

On balance, collaboration will enhance learning. Unless students are encouraged to share and compare what they are learning, learning will occur more slowly because every student will have to make every mistake in order to learn from it. However, collaboration has some pitfalls. For example, asking students to work as a group on a project often hurts students by placing added load on stronger students and sheltering weaker students. One way to avoid this pitfall is to have an open forum that enables the instructor to identify and support weaker students. Weaker students will call attention to themselves by making serious blunders or by remaining silent. 

Cognitive apprenticeship draws attention to competition in a learning environment but is not very helpful with its resolution. The reality is that some students are reluctant to share their work because they fear that they will seem less intelligent than other students. No matter how engaging and supportive a learning environment is, it still takes some luck and inventiveness for constructive competition to take place. 


Course Description: Internet Crime Investigation

An example of how these strategies can be integrated to teach Internet Crime Investigation is given here in an effort to clarify how case-based reasoning and cognitive apprenticeship can be used to teach criminal investigation. 

The course is fully situated by teaching it over the Internet using instructional cases. Supporting domain knowledge is given on a set of Web pages that are referenced when applicable. A Web-based class forum is used to communicate and share information. Email plays a lesser role, enabling the students and instructor to communicate issues that require special attention. The course begins by giving students an overview of crime on the Internet. Several simple cases are presented to give students a better idea of what is involved in an investigation and how experts investigate. 

The students are then presented with their first case. They are told that a complaint has been made against an individual and they are asked to examine some information that has already been gathered to determine if he has done anything illegal. Students are asked to submit a report to the class forum. The students are then informed that the subject of this case is the founder of MIRSO (Male Intergenerational Relationship Support Organization) and are instructed to search the Web for additional information. Students are asked to submit another report based on the new information that they have acquired. Each report should include a few of the student’s indexes of the case (as described in section 2). The purpose here is to have the students articulate their understanding of the case so that they can look back later to see how their understanding changes as they learn. Also, students can learn from each other’s indexes, these indexes facilitate case comparisons later in the course, and the instructor can identify any severe misunderstandings by looking at the indexes. 

For three weeks, students are asked to search different parts of the Internet for more information on the suspect. Search techniques are modeled using simpler or more general cases (e.g. searching for obscene materials). Relevant domain knowledge is provided on easily accessible Web pages. For every new search assignment related to the suspect, students are asked to submit a report specifying what they found, when and how they found it. Students are especially encouraged to describe any difficulties that they had. Again, each report includes some of the author’s indexes for the case. Students will find that the vocabulary, tools and techniques they learn during the course will improve their understanding of the case and will change the indexes they use for the case. 

The instructor addresses specific problems, providing coaching when necessary. The instructor also discusses heuristic and control strategies and asks probing questions that give students a hint about what clues to look for. As the main case evolves, students follow the suspect into different parts of the Internet, gradually uncovering his criminal behavior. By this time, students will be competing to find more incriminating evidence and will be learning from each other as they read each other’s reports and messages. 

After three weeks, students are asked to review the case and summarize the main things that they learned. At this time, another major case is presented that requires all of the previously accumulated skills and tools as well as some new ones. Again subskills are modeled and developed using simpler, more general cases. The instructor encourages case comparisons throughout the course by having students compare motive, MO, signature and other characteristics of each case. These comparisons will be fairly simple to begin with, especially among the smaller cases, but comparisons will become more involved as the major cases develop. 

In this course, the Internet is used as scaffolding. The instructor helps students explore and become comfortable with the Internet and students learn to use its many resources to support their learning. Later in the course, students are encouraged to become involved in the culture of expert practice by participating in on-line discussion forums such as Ideally, after students finish the course they will continue to participate in and learn from the culture of expert practice. 



Expert investigators have much more to offer students than a book or lecture series. The instructional models and methods advocated in this paper can help instructors create a rich learning environment that brings students closer to the fields they are studying than a traditional lecture course would. Students invariably learn more from this engaging and supportive learning experience because they have the opportunity to create a personal understanding as they apply what they learn in authentic ways. Any personal understanding that students create as they learn will be more meaningful and useful than abstracted concepts and facts. Having students articulate, share and reflect upon what they have learned encourages them to think more deeply about a topic. Finally, by providing students with scaffolding, an instructor enables them to become more independent and continue learning after instruction has ceased. 




  1. Brown, J. S., Collins, A., and Duguid, P. (1989). Situated Cognition and the Culture of Learning. Educational Researcher (18) 1:32-42.  

  2. Collins, A., J. S. Brown, and S. E. Newman (1989). Cognitive Apprenticeship: Teaching the Crafts of Reading Writing and Mathematics, In L. B. Resnick (Ed.), Knowing, learning, and instruction: Essays in honor of Robert Glaser, ed. L. Resnick (pp. 140-185). Hillsdale, NJ: Lawrence Erlbaum Associates.  

  3. Douglas, J. E., Burgess, A. W., Burgess, A. G., Ressler, R. K. (1992), Crime Classification Manual: A Standard System for Investigating and Classifying Violent Crimes. New York, NY: Lexington Books.  

  4. Kolodner, J. (1994). From Natural Language Understanding to Case-Based Reasoning and Beyond: A Perspective on the Cognitive Model That Ties It All Together. In R. C. Schank (Ed.), Beliefs, Reasoning and Decision Making (pp. 55-110). Hillsdale, NJ: Lawrence Erlbaum Associates.  

  5. Schank, R. C., Abelson, R. (1977) Scripts, Plans, Goals and Understanding. Hillsdale, NJ: Lawrence Erlbaum Associates.  

  6. Schank, R. C. (1990). Tell me a Story: Narrative and Intelligence. Evanston, IL: Northwestern University Press. 

  7. Egger, S. A. (1990), Serial Murder: An Elusive Phenomenon. London: Praeger. 


This site is designed and produced by Knowledge Solutions LLC
 1996 - 2001 Knowledge Solutions LLC; All rights reserved.
Last update: 2/19