Testing Reading
Originally uploaded by panta rhei.
On the one hand testing for reading seems easy. Take a passage, ask a few questions about it, and you have a test. On the other hand the test may not test what you want tested. A problem for a reading tester is to produce behavior that can somehow be measured. When people write and speak we see and hear but nothing overt happens when people are reading silently. So, for reading, what behaviors can be measured?
To attempt an answer to that question I will try and begin with myself. We are all different so I can only claim that this is one way to think about testing reading. When I was a philosopher, I read an “argument” slowly and carefully many times. That is the way it is in philosophy. The famous French philosopher Rene Descartes is known for the expression, “I think; therefore I am.” Like others I stopped to ponder his meaning. “I am a thinking thing.” But “What is the thing that thinks?” Is it mental or is it physical? If it is mental, how does it impact the physical? Ancient Greek philosophers believed people thought with their hearts. They correctly noticed that when a person’s heart stopped, the person tended to stop thinking. Philosophy can make everything on Heaven and Earth strange and illusory. I know! It is not a good area to explore for the testing of reading as the process is slow and laborious by its nature.
When I am not going crazy with philosophy or language acquisition issues, I read mysteries to relax. I like Michael Connelly and Ian Rankin and, unlike philosophy, I can read an entire mystery in an evening or two. I read newspapers, magazines, other types of fiction and non fiction as well. I may even read a new essay on how to teach reading. So I read different types of material. That is normal. I am going to assume reading mysteries and newspapers are typical reading exercises while philosophy reading is not. That is, my observations will have to do with the former material.
So what can I learn about myself from reading a mystery or a news article from a magazine or paper? I can quickly establish main points, determine the structure of the text, and decide what is relevant. So I have what theorists call the skill of being able to skim. I can find information quickly. I could also “scan,” another semi-technical word which means I can find dates, percentages, expressions, and names fairly quickly on a page. These are all items where speed is of some importance. I do such matters quickly and I find they are worth teaching as good readers have such skills.
I can also do careful reading operations. I can interpret complex sentences, interpret topic sentences, outline the organization of the text, distinguish general statements from examples, identify explicitly or implicitly stated main ideas, recognize a writer’s intention fairly accurately, recognize the writer’s attitudes, distinguish fact from opinion, and make a variety of inferences. I can do these behaviors with various degrees of success as they require more care than the first group listed in my previous paragraph. You can do all these things as well, but the challenges differ depending on the reading material. If – for example - Nell is Emily’s sister, I can infer Nell is a female but if I know nothing about American names, I can make no inference about Emily. This doesn’t mean I can make all inferences of the same difficulty. One has to be careful in testing. For example, if there are more men than the amount of hair on any one man’s head, I can infer at least two men have the same amount of hair. But that sort of inference probably comes from reading philosophers for fifty years and no sane person would ask questions like that! So testing is a matter of sensitivity, obviously. It is a matter of respecting the student and the material to be tested.
I try and test both categories; areas that require speed and areas that require care. Curiously, I have said nothing about vocabulary, a staple of the classic reading test. That is correct. I am seldom interested in testing for vocabulary when I am testing for reading and so I am in an extreme minority. I am interested in reading as a process. I am sure that I sometimes miss vocabulary words in my mystery books. I survive and I still read pretty well. What does vocabulary have to do with behaviors I just cited? So, I must admit to a minority position. The following six items constitute material I pick carefully to bring about testing behaviors that are of interest to me as an instructor.
1. I try to select representative samples of what the student has been reading.
2. I choose texts of appropriate length for the level of student language.
3. In order to test qualities such as skimming and scanning, I may look for passages with discrete information.
4. I make sure that the text has a clearly recognizable structure if I plan to ask organization questions.
5. I try to choose something of interest to the students.
6. I don’t choose culturally-laden texts to ensure I am testing reading and I also do not choose material the students are overly familiar with for the same reasons.
I tend not to use multiple choice questions. I have commented about this elsewhere, the biggest issue being the same size fits all testing mentality. After all such tests test only recognition knowledge and therefore will quite possibly provide misleading information about a person’s reading ability. True and false questions allow for 50% guess work. If you are using short answer question, one has to ensure only one answer is correct. I remind myself I am not testing writing. I do tend to use these short answer questions to determine if the student understands fact from fiction. In other words, such answers will do to test certain types of organization questions. If I want to know whether the student grasps the main idea, or more difficult items, I tend to use gap filling questions (as the main idea may be complicated). I must admit I practiced this method with my own mysteries and it takes some time in preparing the correct questions to elicit the information a tester wants to check.
As I can go on and on with this exciting area of research, let me bring matters to a conclusion to satisfy blog requirements. I never used the magic phrase “reading comprehension.” Perhaps I don’t really know what that is, I really don’t understand what “reading comprehension” means in a testing context. Some say “My student has low reading comprehension” this tells me very little and I have to ask many questions. So it appears I do the exact opposite of almost everyone else. I don’t want the material my students read to be beyond them at all. I want the responses to make minimum demands on their writing ability as well. Whatever the students read for a test must be well within their capabilities. How else will I get an accurate picture of their measurable reading ability? How else do I, as their instructor, know what I need to do to help them? I understand it is somewhat strange for someone testing reading who argues against testing reading comprehension. Before you say that I have been reading too much philosophy and should seek help, please stop to think what it is you are really testing when you are testing reading.
– Dr. Paul Schneider, Director of Teacher Education Programs, WAL