Cory's Senior Thesis

The XO laptop (flagship product of the OLPC project) has a lot going for it, but it still suffers from a lack of software. It's Linux core means that it can't run Mac or Windows software. It's lack of common run-time environments means it can't run things like Java or Flash. Taken together with it's limited processing power, means that most software has to be written specifically for it. For this reason, the amount of software available for the platform is limited. This is especially true for Activities targeted towards teaching Creative Writing.

The goal of this thesis is to compare the existing software on the XOs that can be used to write creative short stories with a new piece of software that I create. The regular pen-and-paper method used in class can be the baseline for comparisons. Using pre- and post- surveys, students will evaluate the different methods. Finally, the results of the surveys will be compared with each other to determine what features are effective and what features are distracting.

Writings toward research question:

writing on the xo is currently through the writing activity. otherwise, non-technology solutions with pen-and-paper are used.

How could the power of the xo be used more effectively to create more effective ways to teach writing at CCCS?

where more effective means: more creativity in stories, increased enthusiasm about writing, more technical proficiency in the writing

Could give a writing prompt before and after intervention

Get some baseline measures of writing before use your tool

focus groups - interviews - enthusiasm, motivation to write, how helped them

State of Delaware/PA rubrics for assessing writing - objective measure of student writing

My current idea for the software is similar to an old program from my youth named Hollywood. Users could write a script that cartoon characters would then act out (using animations and text-to-speech). I'm thinking of something similar to this, crossed with Scratch, but emphasizing the story development process. Users should first develop a plot, characters, setting, and then a plot outline before writing their story (although these steps can be skipped around if the student already has an idea of what they want to do).

This is a rough mock-up of the program. I am not an artist.

In the most expected case, program flow would go as follows:

1. Student writes a one line description of their story's plot. A “brainstorm” button nearby would provide story prompts, plot ideas, etc. if the student is having a hard time getting started.
2. Student picks a setting for their story. A wide range of pre-loaded backgrounds should be provided, so that students aren't limited. Example venues include: haunted house, deserted island, a back alley, a pirate ship, etc.
3. Student adds characters to their story. In the original Hollywood program, characters were anthropomorphic animals with distinctive looks (a hippy-ish weasel, a fat cat, business-y anteater); I'm attached to this idea since it avoids any potential problems of race. However, besides choosing how their characters will look, students need to give them a name, a job, motivations, backstory, etc.
4. Student outlines their story. Here they will be asked to elaborate on their one-sentence plot summary. What is the conflict? How will the conflict be overcome?
5. Student writes the story. Tiles can be dragged and dropped onto the story panel to add dialogue, have characters change emotions, have the narrator describe something, have characters move around the screen, etc.
6. Student plays the finished story. Either by displaying subtitles on the screen or by using the built-in text-to-speech, the story is read aloud while the characters act out.
7. Student exports the finished story to Write for further editing, or saves it directly as a text file.

Speaking generally, a good body of literature demonstrates that, when done right, games are great at helping kids learn. Students are more engaged, retain information better, and are able to tackle much more difficult material.

In terms of my specific program, I'm synthesizing elements from a number of existing pieces of software. A big thing that my program will emphasize is iterative development of your story: you start with a one-sentence description, flesh out story elements, then write an outline, and finally write the actual story. This iterative development process is used in programs like Story Wizard and Kidspiration . In fact, this method is already used as is in the classroom via “idea webs” (also able to be created through mind-mapping software such as Mindomo) and outlines; I'll be using outlines instead of idea webs.

Another large component of my software is the visual, interactive media it offers. Besides it's inspiration Hollywood, programs such as Scratch and Alice can be used in a similar capacity (albeit with considerable less emphasis on the story telling). Scratch in particular has a tile-based layout that will be emulated to some degree with my program.

1. Write (on XO): A simple rich text editor for the XO. Features the ability to add graphics (not provided) and tables, and also has some simple word processing. While a useful tool, it does not offer students any direction when it comes to writing (beyond spell check).
2. Story Builder (on XO): This is a Learning Activity designed to help users write creative, short stories. Besides a place to type in text and a title, users can also drag-and-drop (poorly) animated gifs onto a colorful background. “Completion has been halted” on this program "because of a bug in the olpcgames wrapper". After some investigation, I would rephrase their reasoning as “we're giving up because we're too much of a sissy to dive into source code when it's necessary”. Not that I'm being judgmental or anything. The sissies.
3. Labyrinth (on XO): “Labyrinth is a lightweight mind-mapping activity that supports text, images, and simple drawings.” This program might be useful in addition to Write or pencil-and-paper for planning out stories.
4. Story Wizard: This program is for Windows only. It is designed to help students write stories. Based on the website, the program seems to follow a method of developing a simple plot description into a full story. There doesn't seem to be any emphasis on graphics for the user, relying more on the brainstorming elements.
5. Webspiration: Browser-based (Flash) collaborative writing tool. The Flash-based part means it's not appropriate for the XO's, but it's pretty impressive tool. It seems to let teachers comment directly on students work from their own computer, and watch them as they develop ideas before they have to commit to any writing. Still, this seems to be focused more on essay writing rather than story writing.
6. Mindomo: A browser-based (Flash) mind-mapping tool. Mind-mapping tools are great for getting ideas onto paper (or screen in this case) and outlining a story, so it has some strong points. However, it doesn't have any tools for the actual writing process.

The program will be designed specifically for the XO laptops as a Learning Activity, which will include integration with XO features. The primary programming language will be Python, and graphics will be handled by the Pygame library. Finally, GUI elements will be done using a customized PGU installation.

As demonstrated above, I am not an artist. This program needs graphics, as it is currently described. Here are the options as I see them:

1. Hand drawn scribbles by myself (Worst case).
2. Use photos and motion capture of either humans or claymation (not unreasonable, but not simplest)
3. Find open source graphics (could be difficult)
4. Partner with a VC student (can be very hard to find reliable ones)
5. Hire professional artist via scholarship money (could be expensive)

Depending on human availability and the caprices of software development, all dates subject to change.

1. September 16: Finish first version of program
2. September 30: Test program on teachers and small group of students
3. October 16: Finish second version of program
4. October 31: Test program on teachers and small group of students, finish third version.
5. November 16: Test program alongside other, already available software, gather results
6. December 6: Progress report due via web form to Thesis committee
7. December 16: Finish first draft of thesis paper (actual deadline is April 4)
8. April 18: deadline to set date for thesis defense
9. May 7: Senior thesis symposium
10. May 20: Deadline to be finished thesis defense
11. May 27: Deadline to submit printed copy of thesis

Based on the following:

1. Reactions: Did the student enjoy using the program? “At this level, we measure the customer satisfaction of the learner. Many TBLs can integrate this level of evaluation directly into their learning management systems in the form of simple online surveys. Key questions include: Did the learner like the training? What aspects of the training could be improved? Would they recommend the training to others?”
2. Learning: Did the student learn how to write a story better? “At this level, we measure the mastery of learning content, skill advancement using pre- and post-tests and self-assessments. TBL courses include quizzes and more formal online tests that measure learning progress. LMSs track progress over time and provide feedback to learners and to instructors. Typical research questions at this level include: What knowledge was retained? What skills were developed?”
3. Behavior (Transfer): Is the student more excited about writing stories in the future? “At this level we measure the impact of training on behavior. The goal is to learn how well learners are transferring their knowledge back into their workplace. This measurement rarely occurs directly as part of TBL initiatives and usually requires additional evaluation efforts. Key questions at this level include: What behaviors changed? How did this training affect on-the job performance?”
4. Results: Was the student and teacher satisfied with the story? “At this level, we measure business results that can be tied back to the training event. It is often described as the bottom-line measurement and may include a cost-benefit calculation. For accurate results, evaluators may use a control group who do not receive the same training or who do not receive any training at all.”