Grants:IdeaLab/Improving paragraphs - teaching literacy, improving Wikipedia
What is the problem you're trying to solve?
People are not always very good at writing. One way of teaching writing skills (including Wikipedia writing) is to ask students to differentiate and rank sets of texts on the same topic. While finding samples of good, or indeed poor, writing is not so difficult, it can be challenging to find a range of response qualities to illustrate the features which move a text from being 'poor' to 'good'. If teachers wish to run tasks in which students are asked to order paragraphs of this nature by their quality, they often have to write them themselves (or use peer creation models). The tasks make this worthwhile, but it is time consuming, and not always appropriate.
What is your solution?
We have a huge corpus of gradually improving texts, which sit in categories. By extracting paragraphs at different rev-points, we can present these to students. The paragraph-rev-WikiTrust-scores (for example) could be used to test the assumption that the paragraph text is 'improving' in quality over time. Students would be asked to engage in rank ordering of the texts, and possibly feature extraction to identify features salient to text quality. This potentially contributes to our wider understanding of text quality, but moreover it is intended as an instructional exercise, and to support diagnostic assessment of student's understanding of written work (e.g. a failure to identify citations as key indicates that this area should be given further study).
A final task, either for every text selection, or just for targeted ones could be "Could you improve this paragraph?". In that way the project could contribute back to the encyclopaedia, and act as an onboarding mechanism.
- Deeper understanding of features salient to 'quality' in Wikipedia
- Reduce barriers to access open knowledge (through improved digital literacy)
- Improve quality of open knowledge (the tasks could be used as an onboarding strategy to encourage students to edit)
Welcome, brainstormers! Your feedback on this idea is welcome. Please click the "discussion" link at the top of the page to start the conversation and share your thoughts.