Reading Notes - Active comprehension (1982 article)
This post is my quick note-taking about the following 1982 article:
Singer, Harry and Donlan, Dan. Active comprehension: Problem-solving schema with question generation for comprehension of complex short stories.Reading Research Quarterly, pp. 166–186, 1982.
online link (Not open access)
I came across a reference to this article while reading Wang et.al. (2017) paper on joint learning of question answering and generation earlier this week.
The primary question explored in this article is: if students are taught through direct instructions on typical elements in a story (e.g., characters, goals, obstacles, outcome etc) and examples of generic questions that would elicit these elements, can they learn to pose specific questions about the stories and learn to comprehend better? (Long convoluted way of writing - but that is how I understood the main point).
Since this kind of experimental study is so far away from what I usually read about or work on, I will only summarize my key takeaways or thoughts, not their actual experiment.
- I have never thought much about what is the point of instruction. Here, in this article, there is this sentence that made me think on that:
“However, the goal of instruction is to have readers acquire not only knowledge but also a process for learning how to learn (Buswell, 1956)”
I have been interested in automatic question generation as a means of assessing reader comprehension. However, while reading this article, it occurred to me that questions can also be posed before reading, to direct the reader towards better comprehension of the key ideas in the text. I have read something about questions posed for supporting writing (Liu et.al., 2010) and ofcourse, we all know about these “guiding questions” given in school for reading. However, I never thought about question generation this way.
- I liked this idea of grouping questions into “schema general” and “story specific”. “Deep Questions without Deep Understanding” paper from Microsoft Research that we read in a reading group early last year (or was it in late 2015?) seems to be one of the examples of “schema general” automatic question generation.
(Jekyll markdown is driving me crazy. A colon in the post title gave me so much of trouble for this post!)