Sunday, January 11, 2009

Teaching Students to Find the Needle in a Haystack

Introduction
All leading portals predominantly display internet search tools and these can even be installed on to the toolbars in browsers. All the users have to do is to type in the word and click on the search button. The Oxford English Dictionary and Merriam Webster Collegiate Dictionary now recognize “google” as a verb, meaning “to use the Google search engine to obtain information about (as a person) on the World Wide Web.”


While many of us may walk way with an impression that it is fairly easy to search for information on the internet and this is not something that needs to be formally taught as a part of the school curriculum, Muthukumar, a curriculum designer who focuses on integration of ICT, differs. Given the expansive nature of the web and difficulty in ascertaining the accuracy and reliability of the information available, he advises that information search on the internet is not something that should be treated as commonplace. In a study he undertook as a part of his recent Ph.D. thesis, this was the focus.
Muthukumar asserts that it is not correct to assume that students know how to maximize the information potential of the internet. To be able to efficiently and effectively search for required information on the internet, the student must know how to use the internet search engines, the advanced search options they provide and apply Boolean logic rules. The information search process is also impacted by external factors such as students’ domain knowledge and knowledge about computer systems, such as file formats and country markers.
In the past few years, problem-solving learning activities have becoming increasingly prevalent in educational institutions. Internet-based information searching skills definitely play a critical role in enabling students to seek and apply relevant information to solve given problems.
Working with a hypothesis that there is a need to methodically train students in information search skills using the internet, in 2005 he conducted a study. The study looked at how students engaged in problem solving can be trained in the application of different information searching skills. The study was conducted at a unique institution in Singapore that relies on a problem based learning methodology across all subjects and levels- a perfect setting to measure the impact of internet information searching techniques.

“Information searching skills are non-trivial and complex to be left to be developed intuitively or autonomously by students on their own.”- Kumar

Research Design
This study used both quantitative and qualitative methodologies. The data sources were:
Open ended surveys, students’ documentation, audio recordings of classroom interactions, screen capture of students’ online navigation patterns and reflection journal entries. The design is detailed and complex to be described here but in summary used both well-structured and ill-structured problems and measurements were done before and after intervention. A set of 6 well and ill structured were problems were chosen for the study.

Well-structured problems are bounded by rules, with concepts organized in a predictive/ sequential manner and have singular, convergent solutions. Ill structured problems are open-ended, ambiguous, containing some unknown elements and have divergent/ conflicting solutions. An example of an ill-designed problem:
Space is very limited in Singapore and every square meter is precious.
In order to maximize space, design a 100 storey skyscraper that does
not use lifts, staircases, or escalators

Ill structured problems are relevant because unlike in a traditional educational environment, problems faced in our work environment are likely to be poorly structured, open-ended and multi-disciplinary.

Muthukumar developed and tested a framework that can be used to train students for information search skills in the context of problem solving. The framework enables the educator to provide scaffolding through a four stage model for ill-structured problems. For well-structured problems, stages 1 and 2 are adequate.

Stage 1: Students learn about the use of internet search tools such as search engines, meta-search engines (that search across multiple search engines) and directories.

Stage 2: Students learn to use information search strategies: using multiple search engines, verifying information through cross referencing, searching specific file formats, exact phrases, Boolean operators and key words

Stage 3: Students learn to craft essential questions- questions that invoke curiosity, force evaluation between choices and warrant exploration. Students then develop foundation or subsidiary questions that help them achieve search objectives conveyed by the essential question.

Illustration: Essential question: “Should the wetland areas in the USA be preserved?”
Possible foundation questions: “What is a wetland?”, “What are the reasons for saving wetlands?”, “Why are wetlands being destroyed?”, “Who is destroying wetlands?”, “How many acres of wetlands exist in the USA?”, “At what rate are wetlands being destroyed?” and “What are the best methods for saving wetlands?”

Stage 4: Students learn to develop and organize search keywords, construct a key word concept map.


Quantitative metrics
The students’ navigational pathways in searching for information and solving problem tasks were described in terms of two metrics: Path compactness and Stratum (Rivlin, Botafogo, Shneiderman, 1994). These are mathematically computed indicators of the linearity and connectedness of network-based structures defining users’ online navigational visitations. A high compactness value (1) indicates a navigational style involving easy movement from each node to other nodes within the network structure due to accessing a large number of cross referencing links. A low compactness (0) means little or no access of cross referencing links. Stratum indicates if there is natural/ linear order in which the nodes are visited. The value is highest (1) when the user navigates in a linear sequential fashion and lowest (0) when navigation is iterative or cyclical.

Study limitations
There are limitations to the study, in terms of small sample size (25 students, working in groups of 5) and lack of a control group for comparison. The study had to be conducted within the limits of pre-determined curriculum, limiting the choice of problems that could be administered to the students.

Research findings
Phase I
Preliminary investigation (before intervention) revealed that a majority of the students acknowledged the internet as a vital information provider but only about half the students made a plan when initiating an internet search and only 12% of them made an effort to formulate the keywords for the subject matter being researched. About 60% of the students were either not aware of differences between the different search engines or felt that there were none in terms of their functional attributes. On the positive side, most students were conscious about the authenticity and accuracy of information available and drawbacks of information overload.
When problems are well structured, information needs are limited and awareness of internet search tools and basic searching techniques are adequate. The navigational pattern in internet information search was found to be sequential and deep after some initial trial and error. However, when they are problems are ill-structured, information needs become complex, cross disciplinary and extensive. In such situations, the students’ navigational patterns in internet information search were complex, more iterative and diffused. Students would also redefine search terms frequently.

Phase II- Post intervention
Students appreciated the interventionist training program and felt that it improved their information seeking abilities, made them more confident about seeking and using the internet information, especially in problem based learning environments. Crafting foundation questions and keyword concept mapping were particularly useful in dealing with ill-structured problems. However, students felt the need to develop proficiency in these tasks and hence more training.

The results of the quantitative components of the study also demonstrate that different problem types demand different search capabilities and strategies. There was a significant improvement in efficiency, relevance and depth of the information obtained using the strategies outlined in the framework.

Recommendations
The research findings indicate that internet information search skills need to be systematically and formally taught to enable students to become better problem solvers. I agree with Muthukumar that the suggested framework can be incorporated into ICT training planned for students. While the model has been tested in a polytechnic, problem based learning environment, its application in other environments such as schools and colleges needs further research.
Information literacy also plays a crucial role here as once the search results are obtained, students must be able to evaluate them for relevance, authority, accuracy, currency and objectivity.

References
21st Century Literacies. (2002). Retrieved 25 Oct, 2008, from http://www.kn.pacbell.com/wired/21stcent/information.html
Muthukumar, L. (2008). Information seeking strategies in engaging in problem solving. Unpublished doctoral thesis, Macquarie University, Australia.

Rivlin, E., Botafogo, R., & Shneiderman, B. (1994). Navigating in hyperspace: Designing a structure-based toolbox. Communications of the ACM, 37(2), 87-96.

No comments: