Some Recent Questions, Explicit and Implied

5/22/2013

I'm not sure what that quote means. Take a look at the 2011 PIRLS:

http://timss.bc.edu/pirls2011/international-results-pirls.html

U.S. kids did better reading literary text than informational text (as was true for some other countries: Ireland, for example). Then there are the countries that do well in both (Finland) and those that do better with informational text (Hong Kong). Generally, kids did less well with information than literary text (and because these are not on the same difficulty scale); the U.S. had one of the bigger differences.

5/23/2013

I get it that US students don't do quite as well with informational text. I simply ask for evidence that increasing the proportion of informational text will help.

As for what I quoted, I think it's pretty clear what the quote means: that on the 2006 PIRLS students who reported reading informational text more frequently did worse on the reading test. In other words, this data says pretty much the opposite of what you're telling me the data says.

I'm still looking for data that shows that it's important to read more informational text. Strangely enough, I'm more skeptical now than when I started!

5/23/2013

The research for more than 40 years now has shown that kids don't get the opportunity to read much informational text (especially compared to literary text) and a much smaller body of research has shown that they don't read such text as well as they read literary text. You are correct that there is no study showing that increasing the amount of the reading of informational text has had a clear positive unambiguous impact on reading achievement or student knowledge. (In fairness, however, there is no such research on literary text either--if you think the lack of such research is a good argument for the status quo, then show the level and quality of research that you demand in support of the status quo).

6/9/2013

I am not arguing that the status quo is better (as I said, my working hypothesis is that the key is overall reading volume, but my sense is that the evidence is not very strong at this point one way or another). I am only questioning implications and assertions that there is clear evidence that trying to make students read more informational text will lead to greater proficiency with informational text. In my first comment I asked you if there was such evidence. You seemed to answer yes (your words were: "Actually there is quite a bit of research showing that if you want students to be able to read expository text, you have to have them read (or write) expository text"). Now you seem to be admitting that your answer was wrong. I am genuinely confused as to why you answered wrongly.

As far as I can tell, you are right that there's a strange lack of clear evidence that reading literary text is important. This is very interesting, and it raises huge flags for me about the state of educational research and the value of advertising anything as "evidence-based"--but that's another topic.

6/9/2013

No, I think the basic problem here is with your understanding of research and how causal claims are put forward. First, not everything is going to be studied. Is third grade really necessary? No one is going to pursue a question that specific. We know that schooling is beneficial and that getting more schooling is better than less schooling, but when asked something as specific as whether third grade matters we don't have a "research based" answer that you would accept. That is not proof that having a research basis for policy questions is meaningless.

In terms of informational text (or expository and argumentative texts, since they fit into that category), I know of no study that has tested the benefits of giving students particular dosages of such tex. But in fairness, I know of no such studies of narrative text either. We know that kids benefit from reading instruction and that reading instruction must include text.

Given that how can we recommend an increase in informational text exposure? (1) Studies show that American kids are doing better in comprehending literary texts than informational ones. (2) Studies show that American schools are giving kids substantially more experience in reading literary texts than informational ones (the link between these two statements is an inference--an interpretive leap: since kids are getting less exposure to such text maybe that is why they aren't as good at reading it; (3) studies of text do show many differences, both in the texts and how they are comprehended (for example, causal structure interpretations differ); given these differences it might be necessary to provide students with direct experience with both; (4) studies show that comprehension can be taught successfully with expository text. Most of us put those four facts, substantially proven by research, into a logic model that suggests that it would be wise to provide students with opportunities to learn from such text.

Additionally, there are many analyses of student knowledge suggesting that American kids lack much information about science, history, current events, etc. There is certainly a greater opportunity to build up students' cultural literacy using materials that focus on such information. There is no question that students who possess such knowledge read better.

Someone could do studies in which they divide students into various groups that would receive different dosages of literary and informational text. However, I would not wait for such results before increasing the amount of informational text available to students in American classrooms (given the substantial evidence note above).

6/14/2013

Thanks for your thoughtful reply. I appreciate hearing a bit more about your thought process, but I think you are misunderstanding my position. I agree, of course, that we should take empirical evidence into consideration, and that our working hypotheses should fit with the empirical evidence that we have; the difference between us is that we have different working hypotheses, and you seem unable to imagine that someone could take the empirical evidence and come up with a different interpretation. As I've said, my working hypothesis is that overall reading volume is much, much more important than the type of text that's read. I don't imagine reading volume is all that's important ( I imagine cultural capital and genetic ability also play very large roles), but I think reading volume is way more important than whether kids read literary text or informational text, and I think you are way too casual about the relationship between the evidence and your conclusions. Your statement was, "Actually there is quite a bit of research showing that if you want students to be able to read expository text, you have to have them read (or write) expository text." That statement seems like a stretch to me, but maybe we have different understandings of the word "showing."

I think your proposed study of whether third grade is useful is silly; as you say, there is a lot of evidence that more school is useful. But the question of whether imposing more reading of informational text is useful is far from silly, and I think someone should do it. Nell Duke, for example, has written several books and many articles, and she has gotten millions of dollars of grants for research; it seems like she might have been able to spend a couple of years doing a rigorous study to test her hypothesis. Then you and she would have a better answer for teachers like me, teachers who really want to know whose theory is right. It looks to me like there is way too little basic literacy research being done. Whenever I look into the evidence for recommendations (like for explicit vocab instruction), I am surprised to find that the evidence looks really, really thin to me (everyone refers back to a single 1982 article that I found very, very unconvincing, dealing as it does with instruction in only 100 words over six months, and then comprehension testing on text specifically selected to include those very 100 words!)

Anyway, thanks for the thoughtful reply.

6/14/2013

I think we are talking past each other. I am explaining to you the evidence that people (like Nell Duke) are using to argue for allowing children to have experience with a wider range of text. You are proposing a different issue (you'd rather see kids reading more of what they are already reading).

They are both legitimate issues, but I don't understand the mechanism by which your idea works. If the goal is to produce kids who can better handle the kinds of reading expected in college or in the workplace, how would the reading of more fictional stories do that. There is a long history of research on the concept of transfer in psychology. If you want to improve someone's performance effectively, it is important to make the practice as much like the outcome as possible (distant transfer rarely happens).

I'm sure none of the folks who are arguing for having kids read informational and disciplinary texts support the idea of having kids read less (in fact, most of us would like to see kids read more , too).

6/14/2013

Yes, we're talking past each other. First I thought I wasn't understanding you; now I fear you haven't understood me. You ask how reading fiction would help children handle the reading expected in college or the workplace. Wow! I am truly amazed you would ask this.

First, a theoretical proposition: fiction is similar enough to informational text that if you ead a lot of one of them, you will be able to learn the other extremely quickly. To use your language, the "transfer" is just not very "distant."

As for the “research”: when I ask my initial question, you say you have evidence, but don't cite any. I look it up, find the evidence unconvincing, and am surprised to find that one of the studies Duke cites to support your position (2006 PIRLS) contains data that is in direct contradiction. You admit there's no data directly supporting your position, but counter that there is none in favor of my position either, and suggest, gallingly, that I don't understand the nature of "research and how causal claims are put forward."

No, I don't understand, because I find your practice, like that of many Ed. researchers, confusing: in articles aimed at the public, scholars like you and Duke tend to argue for your preferred policies by asserting that the research ("quite a bit of research") shows that your preferred policies are important. In scholarly papers, or when pressed, you admit that "there is no study showing that increasing the amount of the reading of informational text has had a clear positive unambiguous impact on reading achievement or student knowledge" (Duke has similar qualifications in her scholarly work, though not in her articles for the popular press).

My position is coherent and consistent with the empirical data, such as it is, and I suspect that policies, like the CCSS, that enforce higher percentages of "informational text," will in practice reduce the overall volume of reading done by children like my daughter, who seem mainly to enjoy fictional stories, so I fear that yours, Duke's and the CCSS's emphasis on "informational text" will conceivably actually decrease student achievement. I am aware that there’s not "quite a bit of research" supporting my opinion. On the contrary, I am somewhat dismayed by the remarkable paucity of empirical studies on an important topic that you and Duke seem willing to take strong and confident positions on.

You say you’re not for children reading less. Of course! You’re an intelligent, erudite, well-meaning guy who’s spent his whole life trying to help children read better! Nevertheless, if helping children read more is the key (and while I think it is, I am by no means absolutely certain), it is possible that despite your best intentions the direction you are taking is not the most productive one.

Anyway, sorry for the long comment. I would long ago have just agreed to disagree, except that I really didn't understand what your position was. Is it possible that a leading scholars can think that informational text and literary text are so "distant" that "transfer" will "rarely happen"?! I think that if you did some empirical research (you wouldn't even need a controlled study, you could just interview competent readers of informational text!) you might find that reading fiction transfers quite well. If you interviewed me, for example, or if you interviewed Malcolm Gladwell, who reads airport thrillers. And so on! The CCSS ratios are nutty. Again, if you had convincing empirical evidence I'd be open to persuasion--but as far as I can tell, you don't.

Again, thanks for continuing to read and reply. I admire you in many ways, and this dialogue has been helpful. I think I am seeing more clearly the kind of logic that's driving your arguments--even if I don't agree.