Have a look through your textbook. There's a good chance you'll find a very high percentage of questions and tasks that deal with literal comprehension -- the very lowest level. How can we expect students to function at a higher level when we don't challenge them to?

I'll try to share some ideas for doing that.

Literal comprehension
Even though this is the lowest level, I am not suggesting it isn't important. It is. Understanding literal meaning is the first step toward a deeper meaning. Also, for many of our students, tests are or will be very important. Most comprehension tests focus on literal comprehension because it is the easiest to test. That doesn't, however, mean we are locked into the old "read and regurgitate" pattern of teacher questioning.

For example, if the learners have a reading passage and have to find an answer, make copies of the questions and the answer key -- about one copy for each group of five students. One learner in each group becomes "quizmaster." Everyone in the group turns their book face down on their desk. The quizmaster reads the question twice to make sure everyone understands, then says, "Go!" The other students look at their book and scan to find the answer. When they do, they show their partners where the answer is. This is useful since it encourages speed and scanning, two skills that help learners deal with real text and tests.

Reorganization
Jigsaws -- tasks where different students have different information or different parts of the same information -- can be a way of having learners reorganize information. Stories that are cut up or sequences of pictures can be a route in. Those pictures can then be rearranged to have the learners come up with their own stories. At the simplest level, you can write a sequence of events on the board. Write them in the wrong order. Have the students listen to or read a story and determine the sequence.

Inference
Inference is reading (or listening, or thinking) "between the lines." Learners are looking for information that is in the text even though it isn't stated directly. Inference is a higher level processing skill. For that reason, it is left out of many elementary materials. That's a big mistake. Although it is higher level thinking, it is at the elementary level when students still lack vocabulary, grammar and other linguistic knowledge that they most need to "mentally fill in the blanks."

In some ways, inference is easy to add to other tasks. Simply predicting the end to a story is inference. So it's making guesses about how a character would act or things they would and wouldn't say. Unless it is already stated explicitly, identifying the emotions of a character is usually inference. Regardless of what we do, the key is to get them to notice -- and identify -- the reasons they made the guess. This is usually the words or phrases that gave them the hints. By having students share those hints with each other, they both become more aware of what they are doing as learners and help other students who didn't pick up on the hints.

Evaluation
The term sounds sophisticated (and it can be). But often, this can be as simple as sorting fact from opinion, same/different and good or bad. Students can, for example, read a story and decide which character is the most like their own personality. Again, it is useful to go beyond the evaluation and ask the awareness questions: Why do you think so? How did you know? I know many teachers in Japan will roll their eyes at the idea of students actually stating their opinions. They can, but it takes issues they have opinions about. It can also be useful to provide task support. One way to do this is to make cards with 8-12 opinion phrases like

I think __________ because __________ .
I can see your point, but __________ .
I disagree. I think __________ .

Copy the cards and give a copy to each student. They put the cards on the table. They discuss the topic (one given or, ideally, one they have come up with or selected.). Other months, I've mentioned Language Planning as a way to prepare for activities. It is a good idea to give the students a minute or two of silence to think about what they want to say and how to say it.

Appreciation
This is the highest level of processing. That means one of the sophisticated, yet elegantly simple, questions you can ask is, "Did you like the story? Why or why not?" To be able to answer -- perhaps in English, perhaps in Japanese -- indicates a very deep level of understanding. This works with stories, song, and poems -- almost any narrative input. The students who do this are processing at a deep level. Far deeper that knowing where Helen Keller was sitting. They know where they are sitting.

I learned about Barrett's Taxonomy from Jack Richards. It is mentioned briefly in Alderson, C. & Urquhart (1984) Reading in a Foreign Language, ( Harlow : Longman). You can read an article on my web page that deals with it and the whole issue of comprehension questions.