Thursday, April 30, 2015

Good thinkers know when to work, when to relax.

S. Krashen, April 2015

Csikszentmihalyi, Rathunde, and Whalen (1993)  monitored concentration among "talented" teenagers (grades 9 and 10 in accelerated classes and nominated by teachers) and comparison teenagers while doing various activities by interrupting them with electronic pagers and asking them various questions (p. 52-53), including "How well were you concentrating?"

They found that talented teens did not study more, did not read more, and in fact watched significantly more TV than others (page 87), but their levels of concentration were different: "Talented students reported relatively higher levels of concentration than their average peers when invovlved in classwork, studying, reading, and sports and games. But in less demanding activities, such as household chores, socializing and watching TV, their concentration levels dropped well below those of average students" (p. 97).

Apparently, they knew when to concentrate and when to relax.  Relaxation may have real value, promoting the emergence of new ideas and solutions to problems (Krashen, 2001).

Csikszentmihalyi, Rathunde, and Whalen (1993)  Talented Teenagers: The Roots of Success and Failure. Cambridge: Cambridge University Press.
Krashen, S. 2001. Incubation: A neglected aspect of the writing process. ESL Journal 4(2): 10-11.  (Available at:

Monday, April 27, 2015

Standardized tests versus grades

Sent to the Seattle Times, April 27.

Randy Dorn claims that "Testing can tell if students need help before college and career," (April 27).
I wonder if Mr. Dorn is aware of research evidence showing that high school grades in college prep courses are an excellent predictor of college success, and that standardized test scores (the SAT) do not provide much more information than grades alone.
Is there any evidence that the Smarter Balanced Tests are a better predictor than grades?

Stephen Krashen

Original article:

Bowen, W., Chingos, M., and McPherson, M. 2009. Crossing the Finish Line: Completing College at America's Universities. Princeton: Princeton University Press.
Geiser, S. and Santelices, M.V., 2007. Validity of high-school grades in predicting student success beyond the freshman year: High-school record vs. standardized tests as indicators of four-year college outcomes. Research and Occasional Papers Series: CSHE 6.07, University of California, Berkeley.

No need to test every student

Sent to the Wall Street Journal, April 27, 2015

The House version of the new education law will continue annual testing of reading and math for every student in grades 3-8 and once in high school  ("Education Law Vote Is Pushed Back," April 27).

There is no need to test every student every year. We can get the same information from low-pressure testing of small samples of students, each student taking only a part of the test, and extrapolating the results to larger groups, as is now done with the NAEP test.  This will save money, reduce anxiety, and give teachers more time to teach.

When you go to the doctor, they don't take all your blood. They only take a sample.

Stephen Krashen

Original article:

Friday, April 24, 2015

Reading for Pleasure Can Close 'Vocabulary Gap' at Any Age

Published in Education Week, May 13, 2015

Children of poverty clearly have slower vocabulary development, and this appears to be related not only to the quantity but also the quality of their interaction with parents ("Research on Quality of Conversation Holds Deeper Clues Into Word Gap," April 22, 2015).
   Rather than intervene and give parents "conversation training," as is described in your article, we might consider simpler solutions.
   First, despite the fact that Susan Neuman, a New York University professor of education and department chair, has misgivings about read-alouds, there is substantial evidence that even a modest effort to provide books and basic guidance in read-alouds has a substantial effect on vocabulary growth. Especially interesting are a series of studies using the methods undertaken by the literacy nonprofit group Reach Out and Read, in which books and guidance are provided during well-child clinic visits.
   Second, we can encourage self-selected free voluntary reading as soon as children can read independently. A recent study by Alice Sullivan and Matt Brown of the Center for Longitudinal Studies at the University of London confirmed that we can increase our vocabulary by reading at any age: The impact of reading on vocabulary development in older readers is independent of the level of poverty of their parents.
   So-called "late intervention" is powerful. A child of poverty who becomes a dedicated pleasure reader will rapidly close not only the vocabulary gap, but the literacy gap in general.
Stephen Krashen

original article:

This letter posted at:

some sources (not included in published letter):
Krashen, S. & McQuillan, J. 2007. Late intervention. Educational Leadership 65 (2): 68-73.
Krashen, S. 2011. Reach out and read (aloud). Language Magazine 10  (12): 17-19.
Sullivan, A. & Brown, M. 2014. Vocabulary from adolescence to middle-age. Centre for Longitudinal Studies
Institute of Education, University of London.
Trelease, J. 2013. The Read-Aloud Handbook. New York: Penguin. Seventh edition.

Tuesday, April 21, 2015

Reading aloud to dogs: What it does and what it doesn't do

Published in American Libraries, March/April 2015, vol 46 #3/4: 5-6

(submitted November, 2014)

Appeared with this title: Indirect (dog) therapy

I would like to suggest that reading aloud to dogs ("Dog Therapy, November/December, 2014) does not help children improve their reading ability directly, but it may have positive indirect benefits.

Research on reading consistently supports one conclusion: Children improve their reading ability by reading books that are comprehensible and interesting, when they understand and are interested in what is on the page.

There is no scientific evidence that children improve by reading aloud to dogs (or to humans). Reading aloud is rarely reading for meaning. Only reading for meaning, understanding the message on the page, promotes literacy development.

I suggest that reading to dogs helps young readers indirectly: As the article states in the first sentence, reading to animals may help children "get comfortable" with reading. The few studies done so far support this: they show that children who read to dogs regularly improve in "fluency," that is, reading speed. This is not the same as improving in the ability to understand texts. Increased comfort with reading, and associating reading with pleasure, however, could lead to more interest in books and more reading for meaning, which in turn means more literacy development.

Stephen Krashen, Ph.D.
Professor Emeritus, University of Southern California

Sources (not included in published letter).
Lane, H. B., & Zavada, S. D. W. (2013). When reading gets ruff: Canine-assisted reading programs. Reading Teacher 67, 87-95.
Paddock, C. 2010 Dogs helped kids improve reading fluency.
Smith, Corrione Serra 2008. An Analysis and Evaluation of Sit Stay Read: Is the Program Effective in Improving Student Engagement and Reading Outcomes? Doctoral dissertation, National Louis University.
Smith, M. and Meehan, C. Canine buddies help youth develop reading skills. No date.

Tuesday, April 14, 2015

Accelerated Reader research misses the point, again

Accelerated Reader research misses the point, again
Stephen Krashen April, 2015

Still another study has been published claiming that Accelerated Reader (AR) works (Shannon, Styers, Wilkerson, and Peery, 2015).  And once again, the study fails to provide a real comparison group. *

All we are told is that the comparison group had "traditional" instruction that included independent reading, but we have no idea of how much reading time was set aside.  Those in the AR group (grades 1 through 4) averaged more than 30 minutes per day self-selected reading over 24 weeks "as a supplement to their core reading program" (p. 24).  ("Classrooms in both the treatment and comparison condition had the same amount of class time devoted to reading and language arts instruction" p. 24).

As pointed out elsewhere (Krashen, 2003, 2004, 2005, 2007), it is no surprise that AR produces gains in reading achievement. The important question is whether the tests and awards contribute to these gains. To determine this, AR needs to be compared not with traditional programs, but with programs in which equal time is spent in self-selected reading, without the AR tests and prizes.  This has not yet been done.

*The study was paid for by Renaissance Learning, the parent company of Accelerated Reader.


Krashen, S. (2003). The (lack of) experimental evidence supporting the use of accelerated reader. Journal of Children’s Literature, 29 (2), 9,16-30.
Krashen, S. (2004). A comment on Accelerated Reader: The pot calls the kettle black. Journal of Adolescent and Adult Literacy 47(6): 444-445.
Krashen, S. (2005). Accelerated reader: Evidence still lacking. Knowledge Quest, 33(3), 48-49.
Krashen, S. (2007). Accelerated reader: Once again, evidence still lacking. Knowledge Quest 36 September/October. Available at:
Shannon, L. C., Styers, M. K., Wilkerson, S. B., & Peery, E. 2015. Computer-assisted learning in elementary reading: A randomized control trial. Computers in the Schools 32(1): 20-34

Monday, April 13, 2015

Opt out - NOT anti-testing

Sent to the New York Post, April 13, 2015

David Bradford ("Opt out of tests – kids will suffer," April 13) thinks that parents who opt their children out of the current tests are opposed to assessment. Wrong. They are opposed to excessive and inappropriate assessment.

Students in New York and across the country are now being tested more than any time in history, far more than is necessary and far more than is helpful. In addition, the tests have been made arbitrarily too difficult, resulting in high failure rates that do not reflect reality, and are inaccurate measures of students' abilities.

The opt out movement is a rational reaction to overtesting, and unresearched, unreasonable standards. 

Stephen Krashen

original article

Friday, April 10, 2015

No need to test every student

Sent to the Los Angeles Times, April 10, 2015

I do not share the Times' cheerful interpretation of the Alexander-Murray compromise for the new education law ("A new No Child Left Behind," April 10).  As long as schools continue annual high-stakes testing for every student in grades 3-8 (and once in high school), the "fretful anxiety and teach to the test mentality" will remain.  The huge cost of testing will also remain, and will increase now that testing is done online. 

There is no need to test every student. We can get the same information from low-pressure testing of small samples of students, each student taking only a part of the test, and extrapolating the results to larger groups.  This will save money, reduce anxiety, and give teachers more time to teach.

When you go to the doctor, they don't take all your blood. They only take a sample.

Stephen Krashen

Original article:
On line title of article: No Child Left Behind overhaul should keep states accountable

Wednesday, April 8, 2015

Seattle Sun Times one of the few newspapers to say why opting out is taking place.

Submitted to:

Nearly every newspaper story I have read about opting out of common core tests fails to include the reasons why many parents across the country are refusing these tests for their children. "More Seattle students are expected to deny Common Core testing," (April 8) is a welcome exception. The tests are indeed too hard, with passing levels set much too high, resulting in high failure rates that do not reflect reality. Also, as the Sun Times points out, the tests are an inaccurate measures of students' abilities: The tests are based on a rigid, incompletely researched and often developmentally inappropriate set of standards.
Parents across the country are not protesting all testing; they are protesting the common core testing program.

Monday, April 6, 2015

Opting out of common core tests: The rest of the story

Sent to the Seattle Times, April 6, 2015.

Not mentioned in "More Seattle students opt out of new Common Core tests," (April 6) is the fact that the opt out movement is nation-wide: The article mentions only Washington State and New Mexico, but parents are refusing the common core tests for their children in record numbers throughout the country.
The article barely mentions why. Opting out is not simply because of a desire for local control or fears of children getting a low score: For many parents, it is because the common core has increased the amount of testing far beyond anything experienced before, and because the tests are based on a rigid, untested and often developmentally inappropriate set of standards. 
The parents are doing what the educational establishment should have done when the common core was first presented to the American public.

Stephen Krashen

original article:

Do education officials understand why the opt out movement is so strong?

Posted at:…/Education-commissioner-Students-can…
In response to: "Education commissioner: Students can't opt out of testing"
Missing from this article is why parents are refusing to allow their children to be tested: the parents are not against assessment but are against excessive and inappropriate assessment. This is the reason the opt-out movement is so widespread. I wonder if those imposing penalties for opting out understand this.
S Krashen

Friday, April 3, 2015

Opt out is not "anti-testing"

Sent to Education Week, April 3, 2015

Contrary to "States Seek Guidance in Face of 'Opt Out' Push" (April 1), the Opt Out movement ( is not "an anti-testing" movement. No education professionals and no parents are opposed to proper assessment. Rather, the opt-out movement opposes excessive and inappropriate testing. Students are now being tested more than any time in history, and the tests are based on unresearched, unreasonable standards. 

There is good reason why the Opt Out movement is so widespread.

Stephen Krashen

Original article: