Literature Review

This project raises three challenges:

How best to teach students in STEM disciplines to write
How to train future faculty to teach technical writing
Automated text analysis and classification

Sub-pages (links below and in top menu) explore each of these challenges in detail.

Challenges of Teaching Technical Writing

Over 50% of colleges and universities have some form of Writing in the Disciplines, Write to Learn, or Writing Across the Curriculum (WID/WTL/WAC) as part of their general education programs. There are well-established WAC/WID teaching principles for non-technical writing based on >40 years of evidence and experience, but we know much less about teaching technical writing. What do we know, and what can we import from the WAC/WID community? What do we NOT know?
https://adanieljohnson.github.io/default_website/techwriting.html

Challenges of Training Future Faculty

One major barrier to incorporating more writing in STEM courses is limited instructor awareness of evidence-based training models. Many college STEM teachers have never been introduced to proven writing training strategies or trained to give effective, context-specific feedback. How does this impact instruction more generally?
https://adanieljohnson.github.io/default_website/futurefaculty.html

Challenges of Applying Automated Text Analysis to Student Learning

How well does automated text analysis support student learning? Can it extend what instructors do beyond current capabiities, or does it only replicate what instructors already do? Students claim to want more opportunities for feedback and revision, but do they actually act on automated feedback?
https://adanieljohnson.github.io/default_website/aacr.html

Text Classification 1 - Feature Selection and Engineering

What is the general strategy for solving a text classification problem? What features of texts are important?
https://adanieljohnson.github.io/default_website/Text_Classifier_Models.html

Text Classification 2 - Topic Categorization and Modeling

Establishing meaningful text classification categories depends on the goals of the project or the question being asked. Are relevant categories known already? If not, how are they established?
https://adanieljohnson.github.io/default_website/Topic_categorization.html

Text Classification 3 - Selecting an Appropriate Classifier Method

There are three general classes of text classification strategies: pattern matching, algorithmic, and neural networks. Each is described briefly.
https://adanieljohnson.github.io/default_website/Choosing_classifier.html

Text Classification 4 - Evaluating and Improving the Classifier

If there are many ways to classify texts, how do we know which one is most accurate? How can we improve the classification process? How do we make comparisons and evaluate the outcomes of a text classification problem?
https://adanieljohnson.github.io/default_website/Improving_classifier.html