Your Learn Math Easy Yourlearnmath Your Learn Math Easy You Learn Math Power to Learn

We need to find ways to help ourselves, colleagues, and students to begin to follow at least some of these rules. Introducing the rules of the Charter, however, must be done gently, for we don't want to hurt the feelings of those who have written long emails, sent unwanted attachments, felt it only polite to reply that a note was appreciated, or thought their colorful fonts were attractive.

(The EMAIL CHARTER rules deal specifically with time considerations, but other rules for any type of digital communication should be addressed with your students. We'll tackle those in our next column on email.)

You Learn Math

Your Learn Math Easy Yourlearnmath Your Learn Math Easy You Learn Math Power to Learn - The Technology of Testing

Your Learn Math Easy Yourlearnmath Your Learn Math Easy You Learn Math

The Technology of Testing
by James G. Lengel, Hunter College CUNY, 10/19/11

The psychometrics behind our current spate of standardized tests stem from the work of French psychologist Alfred Binet, who lived from 1857 until 1911. Binet was trying to find an easy way to identify students who needed special help in school. His first test involved practical tasks that the child would perform with an examiner, such as pointing to various body parts, or defining simple words. Binet's work was picked up after his death by Lewis Terman at Stanford University, who adapted Binet's methods to large-scale testing for the U.S. Army as it prepared soldiers for the World War I (WWI). The Army needed a way to quickly classify thousands of young men, sending the best to officer training school, and rejecting the very worst. Terman developed a multiple-choice, paper-and-pencil version of the Binet test that could be easily administered and scored. The questions on the test were crafted to produce a normal distribution of scores among young American men in the second decade of the 20th Century. Terman believed that his test was measuring IQ, an intelligence quotient, a mental characteristic that was inherited, unchangeable throughout life, and a strong predictor of success.

The questions on Terman's tests were selected in such a way as to produce a normal distribution of scores. It didn't really matter what the questions were about, only that they reliably divided the test-takers into a range of scores, and that they produced the same distribution over time, like this:

Here are some questions from the Army test used in WW I:

1. A company advanced 6 miles and retreated 2 miles. How far was it then from its first position?

2. A dealer bought some mules for $1,200. He sold them for $1,500, making $50 on each mule. How many mules were there?

3. Thermometers are useful because

They regulate temperature
They tell us how warm it is
They contain mercury

4. A machine gun is more deadly than a rifle, because it

Was invented more recently
Fires more rapidly
Can be used with less training

5. For these next two items, examinees first had to unscramble the words to form a sentence, and then indicate if the sentence was true or false.

happy is man sick always a
day it snow does every not

6. The next two items required examinees to determine the next two numbers in each sequence.

3 4 5 6 7 8
18 14 17 13 16 12

7. A portion of the Army Alpha required examinees to solve analogies.

shoe - foot. hat - kitten, head, knife, penny
eye - head. window - key, floor, room, door

8. In these next two examples, examinees were required to complete the sentence by selecting one of the four possible answers.

The apple grows on a shrub, vine, bush, tree
Denim is a dance, food, fabric, drink

As these kinds of tests proliferated in the U.S. and Europe, test administrators developed a shorthand for describing those who performed at the lowest end of the distribution, terms that found their way into the popular argot:

IQ Range        Classification
70-80               Borderline deficiency
50-69               Moron
20-49               Imbecile
below 20         Idiot

As long as one hundred years ago, we were looking for quick and easy ways to rank human beings.

The U.S. especially went on a rampage of testing. We administered these tests to military recruits, schoolchildren, factory workers, immigrants, and job applicants. They were easy to administer -- any idiot could score the tests -- and they produced a single number with which to classify everyone on the same scale.

The key to a reliable test was in the selection of questions. Some questions had to be easy, so that just about everyone would be able to answer them, such as #8 above. Other questions had to be difficult, such that only a few would get it right, such as #2 above. And the rest in between. Test-makers developed banks of questions that produced consistent results. They'd combine a selection of hard, easy, and medium questions into a test, then test the test with a sample of people to make sure it produced a normal distribution.

This is the technology of testing that's still used today. In the 1970's, I worked on a panel for the National Council for the Social Studies to help a test-maker develop a test for American history. Here is one of the questions they produced:

Who led American forces in a decisive naval victory at the Battle of Lake Erie in the War of 1812?

a. Perry Hazard Oliver
b. Oliver Hazard Perry
c. Commodore William Farragut
d. Matthew Perry

The test-makers liked this question because it reliably caused 75% of test-takers to get it wrong. (20% would get it right simply by guessing; 5% actually know the answer; and the rest got it wrong by choosing one of the other plausible answers.) The NCSS committee did not like this question because it did not assess the important aspects of this era in history, such as why the war was fought, or how its outcome shaped the economy and maritime influence of each side. Nonetheless, the question went on the test.

In that same year, I was teaching 5th grade in Vermont. Spring had sprung, and it was time for my students to take the Iowa Test of Basic Skills. As I walked around monitoring the classroom, I saw Ricky Bragg with his question book closed, happily coloring in the dots on the answer sheet with his #2 pencil according to a neat pattern: ABCDCBABCDCBA... and so forth, right down the page. I told him he could not do that. He persisted. I gave in. When the scores came back in the fall, Ricky placed at a grade level of 5.1 -- a bit down in the distribution, but not as bad as some. (We would have expected him to be at 5.8, fifth grade, eighth month.)

These experiences led me to learn more about the technology of testing. And not to trust the testing industry.

The good thing is that this century-old psychometric technology ensures that 50% of students score above average. And 50% below. And that if we repeat the test next year, we'll get exactly the same distribution. Perhaps this explains why, no matter how hard we try, we cannot get 80% of students to score above average as required by the No Child Left Behind act.

We need a new technology of testing, one that measures what's truly important to learn, and with which we can measure each student's progress toward well-defined, content-specific goals. Future articles in this series will propose some new approaches and technologies for this kind of assessment. (Commodore Perry, by the way, is buried in Newport, Rhode Island, near where I live. That's Oliver Hazard, by the way; his younger brother Matthew, also at the Battle of Lake Erie, later managed to find his way to East Asia where he and his well-armed fleet forced the opening of Japan to trade with the West.)

Filebusters
by James G. Lengel, Hunter College School of Education, 01/28/2008

Vignette #1

The students had completed their slide show tracing the dissemination of Islamic art forms through areas of Spain and France in the 11th - 13th centuries. Replete with animated maps and photographic examples, the slide show supported their well-researched spoken narrative on this topic. Now it was time to post the PowerPoint slide show to the class web site.

With the help of their professor, they uploaded the slide show...but it did not make it. The system told them it would take six hours to upload the file! (And so, of course, it would take anyone wishing to view the file the same six hours to download it.) This was not what they were aiming at.

Vignette #2

The kindergartners' beautifully-published books on animal habitats were a big hit at the PTA Curriculum Fair. Printed in full color on glossy paper in a hardback binding, they told the story, in words and pictures, of adaptation, predation, and protection. The students used iPhoto to create the book, based on extensive online research, original photography, group discussion, and serious composition. Now it was time to provide a copy for each student.
But not every family had the iPhoto application on their computer at home, nor did the school have a .Mac account that would have allowed easy uploading and viewing of the book over the Web.

Vignette #3

The three faculty members had worked long and hard to prepare the grant proposal. They each sent their narratives, supporting research papers, and curriculum vitae to the grants manager, all in the form of Microsoft Word documents. As the grants manager compiled the final copy for submission, he noticed that some of the tables looked a little odd, and he remembered seeing a pop-up window warning of some missing fonts. But he'd learned to ignore all those pop-up windows, and so thought nothing of it.
Their proposal was rejected, on the grounds that two crucial data tables were indecipherable to the grant-review committee. The main ideas on the proposal were quite sound, remarked the committee, but the garbled tables did not allow them to see the results of the previous research.

Who are you going to call?

All three of the educators described in these vignettes have problems with their files: they are either too big, too strange, or too messed up to be useful. What they need is the digital equivalent of Ghostbusters, perhaps called Filebusters, to come in and save the day. Most computer-using teachers and students have at one time or another confronted issues such as these, where the files just don't work for the intended educational purpose. And a few have discovered a solution that applies in many similar situations, called Portable Document Format, or PDF.

The PDF format was pioneered by the Adobe company to make it possible to publish a document that would be eminently readable, and nicely printable, no matter what kind of computer you displayed it on, or printer you printed it on, or software you used to view it. And once published by the author, a PDF document could not be altered by the reader. This format was based in part on on Adobe's patented PostScript technology, which is used in many printers and some computer displays.

Here's how PDF could have helped our three disabled digerati:

Had the students of Islamic art saved their slide show in a properly compressed PDF format, it would have been small enough for posting to and downloading from the school web site. That's because the PDF format saves only the information it needs to display the slides on a computer with standard resolution. PowerPoint, on the other hand, saves the full resolution of each image in the slide show, which can amount to many megabytes of unnecessary pixels. And just about everybody has a PDF reader on their computer -- most are free or built in. But not everyone has the latest version of PowerPoint, which must be purchased. So PDF is concise.
Had the kindergartners exported their iPhoto books in PDF format, they could easily have been distributed over the web or on CD, and displayed on any type of computer, with or without iPhoto. From the PDF file, the books could be printed at home, or read directly from the computer screen. In full color. Or emailed to grandma in Texas. PDF is compatible.
Had the faculty members submitted their grant application in PDF format, it would have been much less likely to become contaminated by subsequent reviewers, and much more likely to display exactly as desired no matter what kind of computer or printer was used by the reader. That's because PDF files are not alterable by most grant mangers or reviewers, as Word files are. PDF is consistent.

How to save in PDF

You may need to save your own publications in the PDF format. Here's how:

On Apple Macintosh, it's easy and built in. No matter which program you are using, choose from the menubar File --> Print. Then, in the Print dialog box, click the PDF button in the lower left corner. You'll get a choice of dispositions: Save as PDF, Compress PDF, and so forth. For the situations described above, Save or Compress would have been the best choices. This process creates a new file on your computer, in PDF format.
On Windows and Linux, you'll need to install a PDF-saving utility on your computer, and then follow its directions to convert your documents to the Portable Document Format. A search on PDF utilities for Windows will point you to several free and paid programs for this purpose.

Once saved in PDF format, these files can be distributed by all of the means at your digital disposal:

You can attach the PDF file to on email, and end it to your correspondents with the confidence that it's concise enough to pass the email file size censor, compatible enough to be read by all, in a consistent format.
You can copy the PDF file to a compact disc, or flash memory stick, and let your public copy them from there to their own computers with the same confidence.

You can post the PDF file to a web site, knowing at all web servers know how to send out this format, and all web browsers know how to send it to the PDF reader to display it. Just as you published it.