next up previous contents
Next: A design challenge for Up: Modgits: When triggers aren't Previous: Authoring question viewers and

Notebook modgits

Besides questions and answers, there is another category of model objects that needs visual representation. Answers and tests results are summarized for the student in a format called ``points'' (simply strings of text or pictures usually preceded by a 49#49). These points represent the relevant parts of the information contained in the result or the expert help. Authors define points in the model as just text and pointers to pictures when entering ASK answers or experimental results.

Students need to have a searchable place on screen where their experimental and ASK evidence points are recorded as they collect them. We reified this in INDIE as an on-screen notebook; each action students take that gives them more information about the case or about the GBS's domain is summarized in the experimental notebook as a list of bulleted points.9.3


  
Figure 4.8: A notebook, as seen in the interface.
50#50

The experimental notebook is simply a list of all the points students have found so far, and the points are organized chronologically.

When students are ready to make their claim and support it, they need to be able to find relevant points from the experimental notebook and move them to other categories of points that support or disconfirm a claim. By moving a point into one of these new categories, a student is semantically saying that they believe this point is relevant to this category, so a student might move a point that says ``No steam puffs were visible in the crater today'' into the category that supports the claim ``I think this volcano is unlikely to erupt.'' These categories are called lists, even though, as we will see in the next chapter, the order of the points in them doesn't matter.


  
Figure 4.9: A claim with evidence supporting it.
51#51

INDIE interfaces let students do these actions by letting students drag points (lines of text or pictures) from their experimental notebook to other notebooks that are visually located near the claims they support or disconfirm.

Since points don't need to be in any particular order, students only need to be able to drag points from the experimental notebook to the support/disconfirm notebooks for each claim. They need to be able to scroll through their evidence, since students can collect between 30 and 200 points per scenario in various INDIE applications. They also need a way to remove points from a claim's category.

In the INDIE interface, we represent all of the notebooks, regardless of purpose, as different ``notebook modgits.'' Authors can drag out a notebook modgit and assign it a role as either an experimental notebook or a supporting evidence notebook. When a point is made visible, a draggable text item is added to the experimental notebook. Support/disconfirm notebooks accept drags from the experimental notebook, and students can drag points from the support/disconfirm notebooks to a special ``Trash modgit'', which is simply a location authors define that removes a point from a notebook.

In INDIE-speak, these notebook modgits are called ``evidence viewer modgits,'' since a each point is called an evidence point. Figure 4.10 shows three different notebooks.


  
Figure 4.10: Rembrandt's critiquer. Students drag information from the notebook at left into the argument-constructor at right. Points that have already been dragged are highlighted.
52#52

Notebooks, aside from having editable properties like font and size and being repositionable, automatically support scrolling and any restrictions on dragging and dropping between notebooks that authors might want to define. For example, authors might want a list of points students can refer to, but that aren't needed to prove a case, so that notebook should not allow any dragging to or from it.


next up previous contents
Next: A design challenge for Up: Modgits: When triggers aren't Previous: Authoring question viewers and
Wolff Dobson
1998-07-28