Shadow

Of course my lab is not the only ReadingLab around.  Xerox owns the domain theReadingLab.com, which actually has some interesting (though outdated) information on some of their projects, for example,

The Reading Lab:Listen Reader

 

The Listen Reader is a personal interactive reading experience that combines the look and feel of a real book - a beautiful cover, paper pages and printed images and text - with the rich, evocative quality of a movie soundtrack. It is designed to preserve and even heighten the experience of immersive reading. It explores the idea of multi-modal reading: the use of background sound to provide a sense of place, and to add affect to the experience of reading a book without interrupting the flow of the story.

The Listen Reader also preserves the beauty of the book as token object. The design uses no LCD or pixeled screen, but traditional, richly colored printed pages and a classic immersive reading environment: a comfortable chair, a polished hardwood swing-arm reading stand, a pool of light in an otherwise shadowed corner. The soundtrack is controlled by the motion of the reader’s hands above the page; any motion within eight inches of the sensors is read as a volume, pitch, or pan control. Thus the natural gestures of the reader turning the pages provide a simple soundtrack; more complex, deliberate gesturing allows the reader to create richer soundscapes.

The technology is designed to be unobtrusively embedded in the pages and binding of the book. RFID (radio frequency identification) tags embedded in each page sense what page the book is open to; capacitive field sensors measure human proximity to the pages. Proximity measurements (hand gestures) control volume and other expressive parameters of the sounds associated with each page. In this instance, the page-ID data controls which set of sounds is presently being heard. The sound is multi-track and includes music and sound effects (but *not* a spoken version of the words on the page; that would interfere with the reader’s own ability to absorb the text.)

A technical paper on the Listen Reader.

 

If you followed my last a couple of posts (here, and here) back in May, you are due for an update. Now the interim report is superceded by this final report, which found something positive to report. 

By that I don’t mean the finding that RF schools spend more time on reading or are more likely to refer students to reading coaches. That’s part of the ReadingFirst package, and that’s supposed to happen regardless its effect on students. More time on reading instruction, more reading coaches, few switching of textbooks may sound good, but there is no guarantee that they will lead to better reading.

So … what have we in the report with regard to student performance? If you need more details, here’s the 20+M pdf. 

Report Highlights: Reading First Implementation Evaluation Final Report

Based on analyses of states’ reading assessment scores, there is limited but statistically significant evidence that successive cohorts of third- and fourth-grade students in RF schools improved their reading performance over time more quickly than did their counterparts in non-RF Title I schools.
Third-Grade Reading Performance. Average effect sizes across 24 states indicate that RF schools gained between 2 and 3 percentage points more, on average, from pre-to post-RF implementation than non-RF Title I schools on the proportion of students meeting standards on states’ third-grade reading assessments, a statistically significant yet small difference (p < .001). In 12 of 24 states, the improvement in third-grade reading scores among RF schools was statistically significantly larger than in non-RF Title I schools for at least one of the four methods used to define pre- and post-RF implementation years. In the other 12 states, there were no statistically significant differences between the two groups of schools.
Fourth-Grade Reading Performance. Average effect sizes across 17 states indicate that RF schools gained between 2 and 3 percentage points more, on average, from pre- to post-RF implementation than non-RF Title I schools on the proportion of students meeting standards on states’ fourth-grade reading assessments, a statistically significant yet small difference (p < .001). In six of 17 states, the improvement in fourth-grade reading scores among RF schools was statistically significantly larger than in non-RF Title I schools for at least one of methods (described earlier) used in the analysis. In 11 states there were no significant differences between the two types of schools.
There is a positive and statistically significant relationship between only one of four measures of RF and non-RF Title I schools’ implementation of RF-aligned activities, as measured through surveys, and their levels of third-grade reading achievement.
The study team analyzed the relationship between schools’ third-grade reading scores on state assessments and four composite measures constructed from survey data that characterize teachers’ RF-aligned activities: classroom reading instruction; strategies to help struggling readers; participation in professional development; and uses of assessment to inform instruction. ^[  1  ] For every increase of one standard deviation unit in the struggling readers implementation composite score, the probability of being in the top quartile increased by 15.6 percentage points, for the average school (p <.001).

 I am quoting the closing paragraph, which I wish I could use to conclude a paper I am writing.

Although storybook reading remains the single most extensively researched aspect of the field of emergent literacy, still much is to be discovered about how exactly children acquire knowledge during this activity. Taylor’s (1986) prose home movie, basically an interpretive transcript of a mother reading with her three children, shows wonderfully the "blooming, buzzing confusion" (James, 1953, p. 76) surrounding storybook reading with parents who read to their children know well. Given this environment, it is all the more remarkable as some (Ninio & Bruner, 1978; Snow and Ninio, 1986) have shown that from this seeming chaos children extract knowledge about the written language environment and eventually msot construct an accurate picture of how the written symbol system works. However, as Teale (1984) pointed out early on, it is the particulars of storybook reading, what goes on during the activity, that is most neglected by researchers. Hopefully, the research in this paper has provided a closer look at the actual dynamics of parent-child interaction with a view toward discovering the ebb and flow of literacy information even as it is taking shape. 

References:

• Ninio, A., & Bruner, J. (1978). The Achievement and Antecedents of Labelling. Journal Of Child Language, 5, 5-15.
• Snow, C. E., & Ninio, A. (1986). The contracts of literacy: What children learn from learning to read books. Emergent literacy: Writing and reading, 116-138.
• James, W. (1953). The philosophy of William James. New York: Modern Library
• Taylor (1986) is in the same book as Snow & Ninio.

I am all for individualized instruction. After all, it’s Confucius’ teaching.

The follow paragraphs set up a strong expectation for more individual time between teachers and students. However, right after this, the authors turned to praise Lexia for adaptive testing. I am ok with adaptive testing. But I just hope students indeed received personalized instruction, i.e., someone is making an educated and sensible decision — along with the child — about what s/he needs and what’s the best way to achieve it. 

Maybe it’s just me, but I felt it’s a little ironic that the authors likened post-industrial schools to an assembly line, and meanwhile see reading as a set of skills to drill and test. Individulization seems to be limited to the schedule of learning, but students and teachers still do not have the choice of what to learn and how. A computerized assembly line, after all? 

Forbes.com - Magazine Article

Creative Disruption
Teaching To The New Test
Clayton Christensen, Michael Horn and Curtis Johnson 09.15.08, 12:00 AM ET
The art and science of testing children to see what they have learned can and should change. Here’s how. 
When K–12 education in the U.S. was done in one-room schoolhouses, most instruction occurred at individualized rates. Then an explosion in the student population in the early 20th century forced schools to adopt one-size-fits-all instruction. They borrowed from factories the concept of batch processing, with a fixed time spent in each stage of the process of assembling an educated person. Repair, rework and reject became a costly element of the system, just as it did in assembly plants.

We estimate that at least 80% of a typical teacher’s time is now spent in monolithic activity: preparing to teach, teaching and then testing an entire class. Less than 20% of that time is available to help students one-on-one. A profession whose work primarily was in tutoring students individually became one in which some of the most important skills are keeping order and commanding attention.

For those who know me, I look like a person who’d get my hands dirty the first day Google Chrome came out. Which I did.

So far Chrome is my favorite browser — fast and small — other than times when I needed Firefox plug-ins for CiteULike, Zotero, DukeLibraryX, and Amazon. 

Reliability? Let me show you

Got the above when I was trying to post the last blog entry, which had a few mixed up <blogquote> tags. And today I have got enough shares of this for posting this blog.

I know Google is busy updating my Chrome installation in the background. Hope they’ll secretively fix this by tomorrow. By the way, this was posted using the good’o Firefox 3.01

See this?

Where is it?

• Port Moresby, Papua New Guinea 
• Kiev, Ukraine 
• Managua, Nicaragua 
• Havana, Cuba
• Hawaii, USA  

That’s essentially the game of Flitta, an online experiment deviced by Chris and popularized around the world by Sujai. Chris — we should do this with eye-tracking.

(and there is an obvious hack to cheat …) 

I have a long interest in setting up something like CiteSeer for Psychology. In fact I contacted Steve Lawrence way back when, and still have their alpha version server scripts they graciously shared with me (but I was never able to make it work).  

Fast forward, and CiteSeer has grown to be CiteSeerX:

CiteSeer was developed in 1997 at the NEC Research Institute, Princeton, New Jersey, by Steve Lawrence, Lee Giles and Kurt Bollacker. The service transitioned to the Pennsylvania State University’s College of Information Sciences and Technology in 2003. Since then, the project has been led by Lee Giles with technical and administrative direction by Isaac Councill.

After serving as a public search engine for nearly ten years, CiteSeer, originally intended as a prototype only, began to scale beyond the capabilities of its original architecture. Since its inception, the original CiteSeer grew to index over 750,000 documents and served over 1.5 million requests daily, pushing the limits of the system’s capabilities. Based on an analysis of problems encountered by the original system and the needs of the research community, a new architecture and data model was developed for the "Next Generation CiteSeer," or CiteSeer^x, in order to continue the CiteSeer legacy into the foreseeable future.

To use its MyCiteSeer functions — e.g., to correct citations — you need to log in, which I haven’t done yet. Not sure how this part of the site compares to something like CiteULike.

I will be meeting with an independent study student tomorrow morning to go over some paper related to our spelling project. One of the papers is

The sbelling of sdops: Preliterate children’s spelling of stops after /s/
Authors: Hannam, Rachel; Fraser, Helen; Byrne, Brian¹

Source: Reading and Writing, Volume 20, Number 4, June 2007 , pp. 399-412(14)

Abstract:
Newly literate children have a tendency to spell s-stop sequences in words like spin,stop, sky with B, D, G (SBIN, SDOP, SGY), rather than with standard P, T, K. This observation potentially has implications for theories of English phonology as well as of language and literacy acquisition. Understanding these implications, however, requires data about the spelling preferences of preliterate children. In this study, a training-and-transfer design was used to test these spelling preferences in preliterate children. Results confirm that these children relate words with stops after /s/ to words with initial /b, d, g/ rather than to words with initial /p, t, k/. The paper outlines several possible interpretations: that preliterate children have a different phonemic analysis from adults, that they believe spelling represents archiphonemes that they believe spelling represents allophones, and that their early spelling attempts track the phonetic surface. The data suggest rejection of the second interpretation and in our view favour the last over the remaining interpretations. Several theoretical issues are raised that need to be resolved before a full account of the data can be offered.

I wanted to update myself on this, and the following seems to be a very concise summary. 

LINGUIST List 15.1928: Aspiration in English sCC Clusters

On the other hand, I thought I read some optimality theory accounts some time ago, but I can’t seen to find anything directly relevant. There are lots of work, particularly by Robert Carlisle, on the L2 acquisition of /sC/ and /sCC/ clusters.  Most OT work on /sC/ has to do either with syllabification or with language acquisition. 

How do we know the underlying phoneme is a voiceless stop in the case of STAR, other than the fact it is spelled that way? Does SDAR simply look bad, or is it ‘really really’ bad linguistically?

The authors anticipated this question on page 408-409:

One solution would be to suggest that the standard account is based on a spelling effect, and may be in need of revision. We saw above, however, that although stops after /s/ have been extensively discussed over many decades, few phonologists have seriously proposed that stops after /s/ should be assigned to /b, d, g/. This is because phoneme assignment depends not just on the phonetic description of sounds, but on the way the sounds pattern in the language. As Treiman (1985) notes, throughout the English phonological system, clusters of stops and fricatives tend to share voicing. For example, in syllable-final position, only voiceless stops occur after /s/ (e.g., host); plural and past tense morphemes are adjusted to conform to the voicing of the preceding consonant. Indeed, this is common in all languages (Hockett, 1955). Even in a language like Italian, which represents clusters orthographically with SB, etc., the /s/ is pronounced as [z] so as to share voicing with the stop (Giannelli & Cravens, 1997). This in itself strongly supports assignment of stops after /s/ to the voiceless phonemes. Further support is given by the fact that a process of deaspiration of stops after /s/ is phonetically plausible and widely attested in languages of the world whereas the opposite is not (Ladefoged & Maddieson, 1996).

I don’t know what to make of the language typology argument — the only language I speak other than English forbids sC clusters. I am also not sure I’d buy the Italian SB example. If underlyingly it’s /sb/ and realized as /zb/ because of voicing harmony, then we got ourselves an example where the underlying representation is not harmonic. Then why can’t the underlying of STAR be /sd/ and save us some transformations?

Consider the following language game: STAR -> Suh/Tuh/A/Ruh. Familiar? That’s how teachers sound out words. But if you were to do this without thinking about the spelling, for example, in singing, would you say /t/ or /d/? 

I find myself doing what 5-years did in this paper — I said /d/.  

How we know our own minds: The relationship between mindreading and metacognition"

 by  Peter Carruthers

ABSTRACT: Four different accounts of the relationship between third-person mindreading and first-person metacognition are compared and evaluated. While three of them endorse the existence of introspection for propositional attitudes, the fourth (defended here) claims that our knowledge of our own attitudes results from turning our mindreading capacities upon ourselves. Section 1 introduces the four accounts. Section 2 develops the “mindreading is prior” model in more detail, showing how it predicts introspection for perceptual and quasi-perceptual (e.g. imagistic) mental events while claiming that metacognitive access to our own attitudes always results from swift unconscious self-interpretation. It also considers the model’s relationship to the expression of attitudes in speech. Section 3 argues that the commonsense belief in the existence of introspection should be given no weight. Section 4 argues briefly that data from childhood development are of no help in resolving this debate. Section 5 considers the evolutionary claims to which the different accounts are committed, and argues that the three introspective views make predictions that aren’t borne out by the data. Section 6 examines the extensive evidence that people often confabulate when self-attributing attitudes. Section 7 considers “two systems” accounts of human thinking and reasoning, arguing that although there are inrospectable events within System 2, there are no introspectable attitudes. Section 8 examines alleged evidence of “unsymbolized thinking”. Section 9 considers the claim that schizophrenia exhibits a dissociation between mindreading and metacognition. Finally, Section 10 evaluates the claim that autism presents a dissociation in the opposite direction, of metacognition without mindreading.

 Please click here for the PDF file

Andrew Ho argues that the use of percentage of proficient students (PPS) is problematic both statistically and as a policy tool. There are some nice demonstrations of how one can play up or down various gaps by choose the cutpoint strategically without changing student performance. 

The paper is freely available at  http://www.aera.net/uploadedFiles/Publications/Journals/Educational_Researcher/3706/09EDR08-351.pdf

TITLE: The propositional nature of human associative learning

AUTHORS: Chris J. Mitchell, Jan De Houwer, and Peter F. Lovibond

ABSTRACT: The past 50 years have seen an accumulation of evidence suggesting that associative learning depends on high-level cognitive processes that give rise to propositional knowledge. Yet many learning theorists maintain a belief in a learning mechanism in which links between mental representations are formed automatically. We characterize and highlight the differences between the propositional and link approaches, and review the relevant empirical evidence. We conclude that learning is the consequence of propositional reasoning processes that cooperate with the unconscious processes involved in memory retrieval and perception. We argue that this new conceptual framework allows many of the important recent advances in associative learning research to be retained, but recast in a model that provides a firmer foundation for both immediate application and future research.

KEYWORDS: Conditioning, associative link, association, human associative learning,
dual-system, awareness, automatic, propositional, controlled

FULL TEXT: http://www.bbsonline.org/Preprints/Mitchell-08092007/Referees/

Cantonese and other Southan Chinese dialects use characters that are not part of the official Chinese character set. In many cases they are invented quite recently, and the ways in which they are created are interesting.

An example is 冇, which means the opposite of 有 (to have, to possess). I am sure you can tell why.

Chinese Character 冇 (mou5 | mao3; mou3) : do not have - CantoDict

冇

mou5        mao3; mou3

粵][客] have not; did not; don’t have | 文: 無 | 漢: 沒; 沒有 Spoken Cantonese or informal Cantonese writing only 粵 Stroke count: 4 Level: 3 Radical: 冂 (#13)

冇

Apparently the character for "having not" has to come from the character for "having". You just take out what you already have — in this case the two trokoes that are in your safe.

In the study of character etymology (see English version) this would be a case of 指事. 说文叙：“指事者，视而可识，察而见意，上下是也。” The idea is that you look at 冇, and its meaning impresses upon you.

That maybe true, but how do you know the pronunciation? Well, that’s not too hard. 冇 means 没有 ("mei2 you3", not having). Take the onset of the first syllable and the rime of the 2nd, you get /mou3/ (pinyin).

Phonologically speaking, there are a handful of characters like this: 甭 (/beng2/, not necessary, no need) is visually a combination of 不用 (/bu2 yong3/, not + use/need), and phonologically /b+ong/, except that’s not allowed in Mandarin, and you change it to /b+eng/. 

But the similarity ends there. While etymologically 甭 is of the 会意 (ideogrammic compounds) type, 冇 is not a compound (unless you allow negative compounds). 冇 would have to be a simple ideogram, though it was clearly not created from nothing. It was created from "something (有) ". 

(And plus, the aforementioned dictionary says its Radical is 冂. I doubt the latter has anything to do with the character. It’s simply the left over when you take everything out of ’something’.)

A way to put citations online in a multi-facet interface. Now sure what problem this solves, but nonetheless cool. 

Citeline

 Welcome to Citeline, a service to facilitate the web publishing of bibliographies and citation collections as interactiveexhibits and facilitate the sharing of this type of data.
Here are some of our featured interactive publication exhibits:

David Karger’s Bibliography 907 publications Who is he?

Erik Demaine’s Selected Publications 317 publications Who is he?

A patent filing described by AppleInsider | Apple looks to take multi-touch beyond the touch-screen shows that Apple plans to use the iSight webcam to track the "gaze vector". What’s interesting is that they will fuse it with other input such as hand gestures. 

Gaze Vector Fusion

Similarly, iSight cameras could also serve to record gaze vector data, where operations on a computer screen are partially determined by the direction in which a user directs his eyes or head position.

For example, if the user wishes to bring forward a window in the lower left corner of a screen, which is currently underneath two other windows, the user would direct his gaze to the window of interest and then tap a specific chord on the multi-touch surface. 

Daniel Akst is a successful writer with at least 3 books in print and numerous articles on NYT, WSJ, and LAT. So when he speaks about parenting and the cure of education, it has a distinct tone of a successful person.

I will refrain from making comments here — trying not to be too cynical.  I am tempted to assign this paper in my seminar on reading development. Not that this is anything extraordinary, but precisely because it represents (arguably) the loudest voices in the media. It should be an interesting exercise to bring to light some of the unstated assumptions in it.

Raising the Bar: How Parents Can Fix Education - WSJ.com

At the start of yet another school year, it’s time for some radical change in your local schools — a specific change that only parents can bring about. … It’s simple. All you have to do is to start insisting that your children fully apply themselves to their studies — and commit yourself to doing your part. That means making sure they do all the work expected of them as well as their abilities allow. It also means making sure everything at home stands behind these principles and supports the idea of learning.

According to Akst, SES doesn’t matter (and in the ideal world, it shouldn’t):

One great thing about this statement is that income should not matter, since almost any family can insist that conscientious schoolwork be Job One. The stereotype, of course, is of frantic upper-middle-class parents bombarding their precious little ones in utero with Mozart and then hectoring teachers and hiring tutors right up until the Harvard application essay.

Here’s the prescription you will need:

• set education as your family’s highest priority

The first thing we did was to tell our kids that we had no doubt they could do well, and that in fact we expect it of them. We declared that their education is our family’s highest priority, and that during the school year everything in our home will revolve around their success in school. We reiterate these messages regularly, and we communicate them to teachers and administrators, making clear that we want to be kept well-informed.

• Become an academic consultant 

With some effort, we resist the impulse to "help" our boys much with their homework. Would doing push-ups for them strengthen their arms? The same applies to schoolwork, whatever it is — including science projects. But we make sure homework is done early, without loud music or other distractions. We’re available for consulting, and while they’re still young we review their work nightly.

• No game, no TV, and watch some 1942 movies

We keep a tight lid on media. Computer time is limited, there’s no gaming system, and during the school week virtually no television. Extracurricular reading is constantly encouraged, and we choose movies with care. For years now we’ve made a family project out of classic cinema, most of which is highly suitable for kids (and pleases grown-ups as well). "To Be or Not to Be" (1942), in which Carole Lombard and Jack Benny hilariously foil the Nazis, was recently a huge hit with our boys. They can have their jarring music, as long as there’s no foul language or misogyny, but during family meals — which we never miss — they can get used to Mahler or Miles Davis.

• Cut back on sports and other "perks" if they don’t do well in school

We’re also conscious that incentives matter. Like most kids, ours have spending money, cellphones and most other perks of prosperity. But none of these things are mandatory, and all parties understand that blowing off school will have a high cost. Extracurricular activities hinge on school performance too. Recently I heard from a friend that his teenage son, a superlative athlete, was getting poor grades, so I asked if they’d considered cutting back on sports. "I could never do that to him," my friend said, and I couldn’t help thinking: "How could you not do it for him?"

There is a fine line between commonsense and nonsense — it depends on which side of the line you are standing on. 

That is the most entertaining word I have heard from Microsoft so far.

BlueScreen

BlueScreen Screen Saver v3.2

By Mark Russinovich

Published: November 1, 2006

Introduction

One of the most feared colors in the NT world is blue. The infamous Blue Screen of Death (BSOD) will pop up on an NT system whenever something has gone terribly wrong. Bluescreen is a screen saver that not only authentically mimics a BSOD, but will simulate startup screens seen during a system boot.

• On NT 4.0 installations it simulates chkdsk of disk drives with errors!
• On Win2K and Windows 9x it presents the Win2K startup splash screen, complete with rotating progress band and progress control updates!
• On Windows XP and Windows Server 2003 it present the XP/Server 2003 startup splash screen with progress bar!

Bluescreen cycles between different Blue Screens and simulated boots every 15 seconds or so. Virtually all the information shown on Bluescreen’s BSOD and system start screen is obtained from your system configuration - its accuracy will fool even advanced NT developers. For example, the NT build number, processor revision, loaded drivers and addresses, disk drive characteristics, and memory size are all taken from the system Bluescreen is running on.

Use Bluescreen to amaze your friends and scare your enemies!

Download Download Bluescreen (64 KB)

 

I left a bait in an earlier post 伊辛巴耶娃 to 伊娃 :

… Doesn’t matter if you don’t read Chinese — each character is always a syllable (except for 1 character, which I will not tell here; let me know if you know the answer).

Kevin at timelysnow.com wants to know the answer. Here goes:

瓩 - Wiktionary

瓩 (radical 98 瓦+3, 8 strokes, cangjie input 一弓竹十 (MNHJ))

1. kilowatt
2. kilogram

[edit] References

• KangXi: not present, would follow page 748, character 8
• Dai Kanwa Jiten: character 21450
• Dae Jaweon: page 1156, character 14
• Hanyu Da Zidian: volume 2, page 1422, character 5
• Unihan data for U+74E9

 瓩 (pinyin qiǎn1wa (qian1wa3), qiǎn2wa (qian2wa3), Wade-Giles ³, ³)

It’s used quite frequently in Taiwanese publications, where in Mainland it’s written separately as two characters, 千瓦, also pronounced as qian1-wa3.

Interestingly, the same character exists in Japanese and Korean. In Japanese it actually means kilogram, not kilowatt. I suspect this one was made in Japan. It was not on Kangxi Dictionary — Watt was probably not born then. I suspect it was created in Japanese and then borrowed back into Chinese in the 1920’s to 40’s. But I have not way of proving now.

How did I find this out? It goes back to when I was bored in elementary school, and flipped through a half torn 新华字典 (New China Character Dictionary) and was stunned to find the character. I am pretty sure there is not another example like this in the modest New China Dictionary because I must have tried a thousand times to replicate my discovery.

Well, you probably argue that there is another case. For $38 USD you can buy one of these:

which is a pretty traditional paper-cut that you’d put on your windows during the Lunar New Year. What seems to be a single character is actually composed of 4 characters: 招财进宝 (but in traditional characters).

Good find, but I challenge you to find its unicode. 

Following up the previous post on the disyllable preference of Chinese words, here’s another example:

The title of the following article reads : "22 champs talk about Phlying-fish and Bolt, who is greater; most choose the former"

Who is Phlying Fish?

We all have seen his face way too many times in the past a couple of weeks. He has been called many things — Alien, Big Fish, and yes, the Flying Fish, or literally 飞鱼 (in pinyin: fei yu). The latter is very appropriate because of his favorite stroke, the fly. 

His last name, translated in Chinese, is 菲尔普斯 (in pinyin: fei-er-pu-si; WRT to /er/, Chinese does distinguish /l/ and /r/, unlike Japanese. But /l/ cannot end a syllable whereas /r/ can but only in /er/. So pretty much all unstressed /l/ and /r/ sounds become 尔. End of digression), which is 4 long syllables. Bad, bad, bad.

Coincidentally, the first syllable, "Fei" (close enough to "Phe") is written with the character 菲, but sounds the same as 飞. Phelps now morphs from 飞鱼 (fei yu) to 菲鱼 (fei yu). So is my transliteration, from Flying Fish to Phlying Fish.

Why is this a better Chinese (nick) name?

First thing first, it’s two syllables. Moreover, it provides hints of both the phonetics — fei for Phelps — and semantics — the fish. This is a primary mechanism for creating Chinese characters — by putting together a phonetic radical and a semantic radical. And it’s a common strategy to make Chinese loan words. A familiar example is 卡车 (ka-che), literally "bolted carts" or something like that. It actually means "truck." 卡 /ka/ is a phonetic translation of "car". 车 /che/ refers to any land-based vehicle with wheels. I don’t know the precise etymology of this translation, but at some point in the last century it settled on its current meaning of trucks. Again, two syllables. Of course, 菲鱼 could have been fish imported from the Philippines (菲律宾, pinyin: fei-lu:-bin), just like 菲佣 (pinyin: fei-yong) are Filipino housekeepers. But in this context, few would compare the fastest human to Pacific fish.

22冠军纵论菲鱼和博尔特谁更伟大 半数多人选前者_综合体育_NIKE新浪竞技风暴_新浪网

在22位北京奥运会金牌获得者中，有12人认为“菲尔普斯比博尔特更伟大”。其中即包括科比、莱斯利这样的菲尔普斯同胞，也有意大利击剑冠军维扎莉、德国铁人三项赛冠军弗勒德诺等欧洲选手。

At this point you may wonder what exactly the 22 Gold medalists had said about Phelps and Bolt. 12/22 choose the Phlying Fish, 7 liked Bolt, and 3 refused to pick. Among the pro-Phish champs is Kobe Bryant, know simply as 科比 (pinyin: ke-bi). What a beautiful Chinese name. Now you know why he is sooooo popular in Beijing. 

Reading the PISA 2006 report, I was surprised to find no mentioning of USA reading scores in the Reading section. They did participate, but according to

InternationalEd.org | PISA

The PISA has been given every three years since 2000 to fifteen-year-olds in the thirty member countries of the Organisation for Economic Co-operation and Development (OECD) and in twenty-seven partner countries. Students are tested in reading and in mathematical and scientific literacy to assess whether they have “acquired the knowledge and skills essential in everyday life,” according to the OECD. The 2006 results, which concentrate on students’ scientific literacy, also included math and reading, however, a printing error resulted in a decision not to include U.S. students’ reading results for 2006 in the 2007 report.

This brief from the ED gives additional details, although not much:

In the US, because of an error in printing the test booklets, some of the reading items had incorrect instructions and the mean performance in reading cannot be accurately estimated. The impact of the error on estimates of student performance is likely to exceed one standard error of sampling. No results are therefore reported for the US.

The PISA FAQ has this to say:

United States

In PISA 2006, in the United States an error in printing the test booklets, in which the pagination was changed and instructions for some reading items directed students to the wrong page, may have affected student performance. The potential impact of the printing error on student performance was estimated by examining the relative performance of students in the United States on the item set that was common between PISA 2006 and PISA 2003, after controlling for performance on the items that were not likely to be affected by the printing error.

The predicted effect of the printing error and the wrong directions on student mean performance on the reading test was up to 6 score points, and thus exceeds one standard error of sampling. Reading performance data for the United States are therefore excluded from this publication and the PISA database.

The predicted effect of the printing error on student mean performance on the mathematics and science tests was one score point. Mathematics and science performance data for the United States, therefore, have been retained.

Another report has more to say:

U.S. READING RESULTS ON INTERNATIONAL COMPARISON THROWN OUT DUE TO PRINTING ERROR: Alliance President Calls for Retest

 On November 19, officials from the National Center for Educational Statistics announced that the United States’ reading scores on the 2006 Programme for International Student Assessment (PISA) had been invalidated because of an error in the printing of the test. The other results from PISA, which tests fifteen-year-olds in fifty-seven countries in mathematics, science, and reading literacy, are still scheduled to be released by the Organisation for Economic Co-operation and Development (OECD) on December 4.

The error occurred because of a printing problem by RTI International, the firm with which the U.S. Department of Education contracted to administer the test. Originally, students who opened the test booklets were supposed to find a reading passage on the left page and a series of questions related to it on the right page. However, in printing the test booklets, RTI International noticed that the color from the cover of the test booklet bled through the first sheet. As a result, it decided to begin the test on the first right-hand page. Students opening the booklets were instructed to answer questions related to a story or passage on the "opposite page." Because of the new layout, those directions were incorrect.

Mark S. Schneider, the commissioner of the National Center for Education Statistics, which oversees U.S. participation in the exam, called the printing error a "major loss for the study" and an "embarrassment" to everyone involved.

I am surprised that 4th graders would be thrown off by something like this. Apparently they had similar problems in math and science but the consequences were small enough that they decided to include the results. Were kids who took the reading test somehow more literal in reading the instructions?

I am using the "segmented" package by  Vito M. R. Muggeo (2008, http://cran.r-project.org/doc/Rnews/Rnews_2008-1.pdf) to do some piecewise-linear regressions. Wanted to make a function but this won’t run. Looks like it’s — or I am — confusing objects in different scopes.

 

## need to load the package first
library("segmented")

 

## define the segreg3 function
################
## This function doesn’t work, because segmented() can’t find "d"
################
segreg3<-function(t, haz, graph=FALSE, para=c(80,180,260)) {
    d<-data.frame(time=t, hazard=haz)
    h.lm<-lm(hazard~time,data=d)
    summary(h.lm)
    h.seg<-segmented(h.lm,seg.Z=~time,psi=list(time=para),control=seg.control(display=FALSE,h=0.5))
    summary(h.seg)
    slope(h.seg)
    if (graph) {
        points(d, col="light blue")
        plot(h.seg, add=TRUE, col="light blue")
        lines(h.seg, term="x", col="light blue", k=15)
    }
}
## end function