The Tyranny of On-line Choice
Back in the bad old days, Canadian Social Studies teachers who wanted to use Web resources in the classroom had a problem: there was very little on-line content related to the topics they were teaching. Over the past fourteen years, this situation has changed dramatically. Today, teachers are faced with an overwhelming task in sorting through a vast array of on-line resources in an attempt to identify items that will meet their classroom needs. Pressured for time, with an ever-increasing set of demands from all sides, teachers frequently become frustrated in their quest to find suitable on-line resources, and either give up or make use of resources that are less than ideal.
The Problem With Web Search Engines
A Google™ search on the term “fur trade” returns about 1,300,000 results. How can the teacher of a grade 3 class in downtown Toronto or the teacher of the grade 11 class in rural Saskatchewan determine which of these resources will be most useful in achieving her learning objectives in the classroom? Without plowing through all 1.3 million resources, how does she know which ones address the aspects of the fur trade that are relevant for her class, which are written at an appropriate reading level, which resources come from credible sources, which resources are historical documents or lesson plans, which are video dramatizations? What if the most relevant document for her lesson does not use the term “fur trade”, but instead refers to “the trade in furs” or “beaver pelts”?
Addressing the Challenge
Starting in October 2005, The Historica Foundation and Simon Fraser University’s 7th Floor Media decided to tackle this challenge. With funding from the Department of Canadian Heritage’s Canadian Culture Online program, the team set itself the task of understanding the thought processes used by Canadian Social Studies teachers when they are looking for resources to use in the classroom, and of embedding those thought processes in an on-line gateway that would provide teachers with meaningful and contextualized access to quality on-line resources from credible Canadian content organizations. The result was access.ca (http://www.access.ca) an on-line environment, or “teaching space,” that allows teachers across Canada to access authoritative Canadian resources, evaluated, tagged and organized for easy retrieval according to their curricular needs.
Access.ca provides context and meaning for diverse content from many sources by creating interconnections and interrelations. At the heart of access.ca is a specialized search engine. By tying the partner assets together with extensive, intelligent and consistent metatagging, access.ca allows teachers to retrieve materials not only through keyword search or the kinds of subject matter organization normally provided by the content creators but also through key curricular themes, such as “exploration & discovery” or time period, and through reading level, format and content types, provincial curriculum correlation, language and accessibility criteria. This multidimensional tagging and cross referencing offers unprecedented coherence for teachers and will allow the content to grow over time through a common and open framework. Access.ca includes tools that encourage teachers to create their own collections of resources, to evaluate and to rate assets and to share their comments, ratings and collections with other teachers and students. An automated system analyses the patterns of resource usage by teachers, in order to improve search results by providing recommendations for related resources.
In this paper, we will discuss the design and technical decisions we made in developing access.ca, and the process through which we arrived at those decisions.
Understanding the Needs of Social Studies Teachers
Software applications that work well reflect the natural ways in which users perform the tasks they want to accomplish. Our first requirement, therefore, in designing access.ca was to understand how Canadian Social Studies teachers think when they are attempting to find resources for their classrooms. What characteristics of a resource were important to a teacher in determining its suitability for her class? In order to answer this question, we brought together twelve teachers to engage in a role-playing game.
Requirements Gathering Through Role Play
In the role play, teachers worked in groups of three. In each group, one teacher played the “Information Seeker”, one played the “Information Provider”, and one was the Note Taker. The Information Seeker was asked to think of a type of resource that he or she would like to find to use in the classroom. The Information Seeker was then asked to describe the resource to the Information Provider, and the Note Taker was asked to write down everything that the Information Seeker said. The Information Provider was instructed that he or she could make suggestions of specific resources that might meet Information Seeker’s description, but could not ask questions. The Information Seeker was free to add more details to the description of the resource at any time. Each member of the group had a turn to play each of the three roles.
After each group had completed its role play, it was asked to look at the Note Takers’ records and to list first those terms (or search criteria) used by all three Information Seekers, then those search criteria used by two out of three Information Seekers, and finally those search criteria used by only one Information Seeker.
The lists produced by each group were then posted on the wall for discussion by all. Those search criteria that had been used by all four groups was identified first, then those that had been used by three out of four groups, and so on. Where similar but different terminology had been used, the participants discussed whether these terms reflected the same or different criteria. For those search criteria that had been used by only a few, the group discussed whether these were important criteria. At the end of the day, we were able to identify teacher preferences with respect to organization and presentation of resources, as well as for a range of functionality on the site.
Analyzing the Canadian Social Studies Curriculum
At the same time as we were working with teachers to identify ways in which to provide access to resources, we reviewed the different K-12 provincial Social Studies curriculum descriptions. We analyzed these descriptions and established a list of major topics based on those curricula. We also created a breakdown of time periods that corresponded with the way in which subjects were grouped chronologically within curricula across the country.
Having created a list of major topics, we then created a chart showing the subjects from each curriculum that fell within each topic heading, along with the province and grade levels at which they are taught. Finally, we determined the Canadian Subject Headings (CSH) or, in some cases, the Library of Congress (LOC) Subject Headings for each of the subjects, rationalizing duplicate subjects under these headings.
Designing the Metadata Structure and Metadata Exchange System
In order to enable teachers to search for resources based on the criteria that they had identified as meaningful, it was necessary for us to obtain metadata about partners’ content in an appropriate format. To accomplish this, we identified standards for the exchange of metadata that reflected the scope of the content that our partners were providing and that would be easily assimilated into a system for searching and filtering for teachers. A metadata structure for Canadian social studies content (in particular history and First Nations content) provided the basis for meaningful categorization and searching of content and correlation of that content to Canadian social studies curricula.
In dealing with the content providers in the proof-of-concept phase, during which buy-in was tentative, we focused on a subset of unqualified Dublin Core as the primary standard for metadata exchange. However, the ultimate goal is to establish a more thorough and precise standard based on qualified Dublin Core or Cancore/LOM, allowing for more automation and less human intervention in the process of exchanging metadata.
We established a data exchange methodology based on the Open Archives Initiative that will allow content providers to expose their data to the access.ca search engine. We offered the content partners the option of implementing an OAI server or of providing an OAI Static Repository that would be processed by an OAI gateway on our side. We also created a Web-based administrative tool to allow for the augmentation and refinement of metadata in the context of Canadian curricula, as represented by CSH and LOC Subject Headings.
Our early success in obtaining buy-in from content partners in the implementation of our metadata standards and data exchange methods was mixed. To some extent, this was a bit of a “chicken and egg” dilemma. While we obtained philosophical buy-in to our proposed metadata standards and data exchange protocol, some content partners were reluctant to undertake full implementation of the metadata standards and data exchange methods until they had seen the site and were convinced that it would truly be a valuable tool for teachers. In this initial implementation, the majority of the content partners opted for the simpler option of providing an OAI Static Repository. Cost/benefit was obviously a factor in these decisions. As with any search engine, the value to the content provider of having its content included is a function of the number of users. As the number of users increases, the value of inclusion begins to outweigh the cost of implementing technical changes. Conversely, in the early stages the costs of implementing technical changes generally outweigh the perceived value of inclusion. The seeming catch twenty-two is that a deep and broad collection of content is required in order to attract large numbers of users, but a large number of users is required to encourage content providers to commit resources to participation.
The Proof-of-Concept
In April 2007, we launched access.ca with initial content from five organizations: The Historica Foundation, Library and Archives Canada, the McCord Museum, White Pines, and 7th Floor Media. With this initial content, we are able to demonstrate the power of access.ca as a tool for Canadian Social Studies teachers. Below are descriptions of some of the key features of access.ca.
Search
Basic Search (Fig 1) functionality allows teachers to quickly and easily find a wealth of authoritative Canadian on-line resources that are relevant to their needs. Teachers can:
- search by keyword, topic, and time period;
- narrow their search to materials suitable for certain reading levels or to materials relevant to specific provincial curricula;
- restrict their search to particular formats (text, images, audio, video); and
- restrict their search to particular types of resources (articles, lesson plans, historical documents, artifacts, photographs, art, maps, dramatizations, documentaries, on-line activities)
Additional options allow teachers to:
- specify the language for the resource;
- include only material accessible to visually impaired or hearing impaired people; and
- exclude material requiring a high speed connection.
Fig 1: Access.ca Search Page
By default, Search Results (Fig 2) are sorted by “closest match.” However, teachers can choose to sort results by “highest rated” (those resources that match the search criteria and have received the highest explicit ratings from other teachers appear first in the results) or “most frequently collected” (those resources that match the search criteria and that teacher have added most frequently to their access.ca Libraries appear first in the results).
Fig 2: Access.ca Search Results Page
For each resource identified in the Search Results, access.ca provides a Resource Summary Information Page (Fig 3), which provides teachers with quick access to the following information:
- Title and Source
- Current rating (based on ratings by access.ca users)
- Format and content type
- Grade level(s) (age appropriateness); Time period(s); Topic(s) (subjects or themes for which this resource is relevant)
- Description
- Link to reviews (evaluations of this resource by teachers)
- Links to recommendations (see discussion below)
From the Resource Summary Information page, the teacher can open the resource from its original site. The teacher can also instantly rate or review the resource and add the resource to a personal Library.
Fig 3: Access.ca Resource Summary Information Page
Library
A teacher can create Collections of resources in a personal Library (Fig. 4) on access.ca. For example, a teacher might create a Collection of resources for a unit on Canadian Confederation, and add an article on John A. Macdonald, Canada’s first prime minister. Teachers can create Collections purely for their own use. Or they can provide their students with the URL to a Collection for use in carrying out a school assignment. For each resource in a collection, the teacher may add notes for her own reference and/or notes for her students to read when they look at the student version of the Collection.
Fig
4: Access.ca Teacher’s Library Collection
Ratings, Reviews and Recommender Systems
While teachers informed us that they would be interested in reading reviews of resources written by other teachers, teachers also told us that they would be unlikely to write reviews of a resource unless is was “really good” or “really bad.” Thus, while it was relatively easy to allow Search Results to be sorted by ratings, it appeared that this information would be of limited value to users. Accordingly, we explored other ways we could identify resources that teachers had found useful so that we could expose this information to other users.
Our first step in going beyond an explicit rating and review system was to hypothesize that a teacher would, in all likelihood, only add a resource to an access.ca Library Collection if she considered that resource to be worthwhile. By tracking which resources teachers were adding to Library Collections, we were able to add a “sort” to the Search Results that prioritized those resources that had been most frequently used by teachers.
Our next step was to create a system that tracked each time one resource was collected with another resource. If a teacher saved one resource in the same collection as another resource, we reasoned, there was a probability that the two resources were related in some way. If a number of teachers saved those two resources together, the probability that the resources were related grew stronger. We therefore developed an automated system that would identify two resources as “related” once a threshold number of teachers had saved those two resources together in one of their collections. Once that threshold number has been reached, a teacher looking at the Summary Information for one of the resources will see that “other teachers who found this resource useful” also found the other resource useful.
When teachers save resources into their collections, they identify the Province and grade for which they will be using the resource. As the number of teachers creating Library Collections on access.ca increases, we may provide more detailed recommendations, along the lines of: “Teachers who found this resource useful for Grade 6 in Alberta also found the following resources useful.”
Fig 5: Access.ca determines that Resources D & E
are related because many teachers have placed them both in one of their
collections.
Conclusion
Access.ca demonstrates how a deep understanding of users’ thought processes, combined with appropriate metadata standards and an effective metadata exchange system, can expose content providers’ resources in a manner that makes them truly accessible and useful.
By simulating the tasks to be performed in a person-to-person setting, without imposing any preconceived technical parameters, we were able to clearly identify user needs. By allowing these user needs to drive technical development, we were able to develop technical requirements and standards that would provide the target users with quick and easy access to appropriate resources and on-line tools. And by probing the realities of teachers’ behaviour, we were able to devise systems to indirectly analyze teachers’ on-line actions in order to create an automated recommender system.
A coherent standard for metadata for Canadian K-12 Social studies and a method for exchanging content provider metadata now exist. The challenge is to foster wide adoption for these standards.
Cite as:
Zilber, J. and J. Marsh, Access.ca: Social Studies Resources for Canadian Teachers , in International Cultural Heritage Informatics Meeting (ICHIM07): Proceedings, J. Trant and D. Bearman (eds). Toronto: Archives & Museum Informatics. 2007. Published October 24, 2007 at http://www.archimuse.com/ichim07/papers/zilber/zilber.html
Search A&MI