Information Age Education
   Issue Number 153
January, 2015   

This free Information Age Education Newsletter is written by Dave Moursund and Bob Sylwester, and produced by Ken Loge. The newsletter is one component of the Information Age Education (IAE) publications.

All back issues of the newsletter and subscription information are available online. In addition, four free books based on the newsletters are available: Understanding and Mastering Complexity; Consciousness and Morality: Recent Research Developments; Creating an Appropriate 21st Century Education; and Common Core State Standards for Education in America.

This is the 6th IAE Newsletter in a new series on Credibility and Validity of Information.

Credibility and Validity of Information
Part 6: Computer and Information Science

David Moursund
Emeritus Professor of Education
University of Oregon

In August 1945, while Grace Hopper and some associates were working at Harvard on an experimental [computing] machine called the Mark I, a circuit malfunctioned. A researcher using tweezers located and removed the problem: a 2-in. long moth. Hopper taped the offending insect into her logbook. Says she: “From then on, when anything went wrong with a computer, we said it had bugs in it.” (Google Doodle, n.d.)

“To err is human, to really foul things up requires a computer.” (Bill Vaughan, April 2, 1969, newspaper column.)

From Previous Newsletters in the Credibility/Validity Series

Credibility focuses on a belief that the person who made an allegation about a phenomenon is believable and can indeed be trusted. It is common to talk about such a person and what the person writes and/or says as being credible and believable.

This newsletter series has not yet explored the situation in which the speaker or writer is an artificially intelligent computer and there is no specific person who can be investigated to determine his or her credibility.

Validity is an important component of research. The word tends to be used in two somewhat different ways:
  1. Validity is the quality of being logically and factually sound. Validity is the extent to which a concept, conclusion, or measurement is well founded. Others repeating the research can test results produced by valid research. Additional evidence of validity is produced when research is designed to find any contradictions to the results, and no such contradictions are found.

  2. In education, a research instrument or test is valid if it accurately measures what it is purported to measure.
In summary, one can think of credibility having a subjective base and validity having an objective base. The performance of a gymnast or dancer is evaluated by subjective methodologies, while mathematics and the sciences use objective methodologies to support their claims.

Computer and Information Science (CIS)

“Computer Science is no more about computers than astronomy is about telescopes.” (Edsger Dijkstra; Dutch computer scientist; 1930-2002.)

The discipline of Computer and Information Science (CIS) is usually categorized as a science. However the quotation from Edsger Dijkstra given above captures the essence of the situation. Computers are general-purpose tools. I like the analogy of CIS being somewhat akin to reading and writing. Literacy is far more than writing symbols and the human mind learning to decode these symbols. CIS is now a ubiquitous aid to the human mind in every discipline of study.

Computers

A computer is a machine designed for the input, storage, automatic processing, and output of information. The “automatic processing” is done by the computer hardware following a detailed step-by-step set of instructions called a computer program (the computer software). Software reflects the thinking and problem-solving abilities of the programmers. However, the area of CIS called Artificial Intelligence or Machine Intelligence has progressed to the level that quite a bit of the “thinking” that goes into developing both hardware and software is aided by computers (Markoff, 12/15/2014). In a sense, computers are helping to develop their own hardware and software, and they are getting better at it!

Many of today’s college and university CIS departments had their beginning in Mathematics and/or Engineering departments. The software and information science aspects of CIS are closely related to mathematics, and computer hardware is well rooted in Engineering and Physics. Perhaps results produced by computers should have a level of credibility and validity approximately the same as those we expect from mathematics, engineering, and the natural sciences?

However, that is not the case. I strongly believe that part of a modern education is learning to assess the credibility and validity of results produced by computers.

Computers have become more reliable and more powerful over the years—but most people have come to understand that computers are not infallible. Indeed, in our everyday use of computers we have come to expect that computer software and hardware may contain “bugs” that can lead to the computer producing incorrect results.

Computer hardware can be designed to have a very high level of validity (very nearly but not completely error free). This topic is discussed later in the newsletter.

While some computer programs are bug free, we routinely use computer programs that are not bug free. So, use of these programs on even the very best computer hardware does not guarantee valid results.

Incorrect (invalid) computer results can be caused by bugs in the hardware, bugs in the software, cosmic rays that change a bit of data stored in a computer memory (Dunietz, 5/9/2014), incorrect data/information, keyboarding errors, and/or the misreading, misunderstanding, or misuse of the results.

With all of these potentials for producing invalid results, how much faith should you place in the work computers perform for you and the information they provide to you? Just because a computer is used to help solve a problem does not guarantee the results are valid or credible.

Validity in Computer and Information Science

Many of today’s college and university CIS departments had their beginning in Mathematics and/or Engineering departments. The software and information science aspects of CIS are closely related to mathematics, and computer hardware is well rooted in Engineering and Physics. Perhaps results produced by computers should have a level of credibility and validity approximately the same as those we expect from mathematics, engineering, and the natural sciences?

The answer is no. In the IAE Newsletter about credibility and validity in mathematics, we differentiated between results in “pure” math and the results when math is used to represent and help solve problems outside the discipline of math (Moursund, 12/15/2014). We noted that strictly within the discipline of math, the math results have a very high level of validity. However, results produced by using math in other disciplines, such as applying statistics in educational research, do not gain automatic validity just because math is being used.

The next four sections explore various types of threats to the validity of results produced when computers are used.

Information Science

As noted earlier in this newsletter, a computer is a machine designed for processing information. It is common to use the term information to refer to a combination of data, information, and knowledge (Moursund, 2013). There is nothing in these terms that suggests the data, information, or knowledge being processed is correct (valid).

Quoting from the Wikipedia:

An early definition of Information Science (going back to 1968, the year when the American Documentation Institute renamed itself as the American Society for Information Science and Technology) states:

"Information science is that discipline that investigates the properties and behavior of information, the forces governing the flow of information, and the means of processing information for optimum accessibility and usability. It is concerned with that body of knowledge relating to the origination, collection, organization, storage, retrieval, interpretation, transmission, transformation, and utilization of information. This includes the investigation of information representations in both natural and artificial systems, the use of codes for efficient message transmission, and the study of information processing devices and techniques such as computers and their programming systems. It is an interdisciplinary science derived from and related to such fields as mathematics, logic, linguistics, psychology, computer technology, operations research, the graphic arts, communications, library science, management, and other similar fields. It has both a pure science component, which inquires into the subject without regard to its application, and an applied science component, which develops services and products." [Bold added for emphasis.]

As a pure science, validity of results in this field of study are similar to the validity of results in mathematics and the natural sciences. However, when information science is applied to real world problems, we find the same types of difficulties as when math is used to help solve real world problems.

I routinely make use of a search engine to look up information on the Web, despite knowing the Web contains a great deal of information that is not correct. The statement Garbage In, Garbage Out (GIGO) applies. For example, I make frequent use of the Wikipedia. But, I do so with full knowledge that it contains errors and biases. I evaluate what I read in terms of my knowledge and understanding. If I encounter something that just doesn’t seem plausible or right, I check other sources. I am reminded of Thomas Jefferson’s statement, “The price of freedom is never ending vigilance.”

Errors in Computer Hardware

I use a medium priced desktop computer that is a couple of years old. The electronics of this hardware can carry out more than a billion operations (such as addition or multiplication) per second. Suppose such a computer makes a hardware error an average of once per 10 trillion operations it carries out. Hmm. An hour is 3,600 seconds. In 10 hours this machine could carry out well over 30 trillion operations—and would be quite likely to make an error. Are you comfortable about the thought of flying in an airplane that is on automatic pilot, or riding in a driverless car, guided by a computer with this error rate?

Of course, there are ways to reduce hardware errors. The combination of hardware and software can be designed to detect and/or circumvent many of the possible errors.

Moreover, the hardware can contain considerable redundancy. For example, suppose that the automatic pilot in an airplane or rocket ship actually contains three completely independent computers, each doing all of the computation. If all three agree on an action to be taken, one can have a great deal that no hardware has occurred. If two of the three computers agree, the “trio” computer system can be programmed to go ahead and use the result, but also to inform the human pilot that something may be going wrong, and that human intervention is advised.

In summary, hardware errors can and do occur in computers. However, in situations in which an error might produce a catastrophically disastrous result, hardware and software can be designed so that such errors are very infrequent. Absolute perfection cannot be guaranteed.

You may have noticed that I focused on errors in the computational hardware. Suppose that you go to an ATM machine to make a cash withdrawal. The ATM machine contains a large supply of $20 bills, and it automatically counts off the correct number of bills from its supply. Such “paper handling” machines are quite accurate, but they can make mistakes. However, suppose an error does occur and you get too few or too many $20 bills. This is not what one would call a catastrophically disastrous result. Typically, the error rate in such machines is far lower than the error rate of an average human clerk in a store counting out change. In addition, a bank has good procedures in place to handle such “minor” problems.

Bugs in Data and Data Input

My hand-eye coordination leaves much to be desired. When I am using a simple handheld calculator, I often make reading and/or keyboarding errors. Even when the calculator hardware and software work perfectly, the results from processing incorrect input are usually incorrect output.

GIGO describes this situation. Calculator and computer users need to be constantly alert to the possibility that they are making input errors and/or that the data itself is incorrect. Mental estimation of plausible/reasonable results is a valuable aid to detecting such errors.

Now, let me provide a variation on this problem. With a handwritten check signed with my name and “for deposit only,” I go to an ATM machine. The machine takes my check, finds and reads the numbers giving the amount of the check, and deposits it in my account. Hmm. How good is the handwriting of the dollar amount on the check, and how good is the machine’s number recognition system? Poor handwriting and/or poor handwritten number recognition skills can lead to an error.

Fortunately, the ATM machine shows you the results of its check reading and asks you to verify if its result is correct. This provides an excellent example of a human/machine interface in which both the human and the machine bear joint responsibility for the results that are produced.

However, now consider a voice input system that takes natural language input and changes it into printed text. A very good system might well produce 95 percent accuracy. That means that if you speak a short paragraph, say 50–60 words, an error is quite likely to occur. The voice input system may well include a spelling checker so there are no incorrectly spelled words in what it produces for you. But, the correctly spelled word may not be the word you said. You know about homonyms, words that sound nearly the same, so this is an easy error to have occur. Also, the fact that people have widely varying accents can cause added difficulty in converting speech to text.

This type of problem is considerably exacerbated if you are using a voice input language translation system. Language translation by computers is a very difficult task. Voice input and language translation systems are good enough to be quite useful. However, they are far from perfect. My message is: “User beware. Computer language text input and voice input translation systems now widely used make frequent errors.”

Errors in Computer Software

I am pleasantly surprised when I spend a day using my computer and do not have a “crash” in an application I am using and/or in the operating system. I am using very long, complex programs, and such programs are very likely to contain bugs.

When the word processor I am using crashes, I am very, very annoyed. Of course, the computer system is often able to recover most of my document, and I can instruct my word processor to automatically make a backup copy every few minutes. So, a combination of proper programming of the computer and proper actions on my part can usually alleviate most of the pain from such crashes.

In some sense, a computer program is like a math proof. But, operating systems and a number of computer application programs are millions of instructions in length. Such a program may involve a hundred or more programmers and program testers working on its development over a period of years. I have considerable confidence in my statement that “they contain errors.”

Final Remarks

Computer and Information Science is a powerful mind tool. Today’s children are learning quite a bit about CIS because of its prevalence in their everyday lives. But, you can make the same statement about reading, writing, arithmetic, science, and the other areas of human knowledge and skills that are part of our K-12 school curriculum. One does not acquire the needed contemporary level of knowledge and understanding of these disciplines through osmosis. Organized teaching and learning help!

Our schools need to integrate appropriate use of CIS into each discipline students are studying. Schools need to place considerable increased emphasis on the credibility and validity of the course content and the tools—such as CIS—of the disciplines being studied.

The various types of media and communication made possible by the use of CIS are exacerbating the challenge of dealing with problems in the credibility and validity of the information that we use while interacting with each other, making personal decisions, voting, purchasing goods and services, and so on.

References

Dunietz, J. (5/9/2014). The cosmos’ attack on computers. Science Non Fiction. Retrieved 1/2/2015 from http://sciencenonfiction.org/2014/05/09/the-cosmos-attack-on-computers/.

Google Doodle (n.d.). Retrieved 1/2/2015 from http://newsfeed.time.com/2013/12/09/google-doodle-honors-grace-hopper-early-computer-scientist/.

Markoff, J. (12/15/2014). Innovators of intelligence look to the past. The New York Times. Retrieved 1/2/2015 from http://www.nytimes.com/2014/12/16/science/paul-allen-adds-oomph-to-ai-pursuit.html?ref=science&_r=1.

Moursund, D. (12/15/2014). Credibility and validity of information. Part 4: The discipline of mathematics. IAE Newsletter. Retrieved 1/2/2015 from http://i-a-e.org/newsletters/IAE-Newsletter-2014-151.html.

Moursund, D. (2013). Computational thinking. IAE-pedia. Retrieved 1/2/2015 from http://iae-pedia.org/Computational_Thinking.

Author

David Moursund is an Emeritus Professor of Education at the University of Oregon, and coeditor of the IAE Newsletter. His professional career includes founding the International Society for Technology in Education (ISTE) in 1979, serving as ISTE’s executive officer for 19 years, and establishing ISTE’s flagship publication, Learning and Leading with Technology. He was the major professor or co-major professor for 82 doctoral students. He has presented hundreds of professional talks and workshops. He has authored or coauthored more than 60 academic books and hundreds of articles. Many of these books are available free online. See http://iaepedia.org/David_Moursund_Books. In 2007, Moursund founded Information Age Education (IAE). IAE provides free online educational materials via its IAE-pedia, IAE Newsletter, IAE Blog, and books. See http://iaepedia.org/Main_Page#IAE_in_a_Nutshell.

Email: moursund@uoregon.edu.

Reader Comments

We are using the Disqus commenting system to facilitate comments and discussions pertaining to this newsletter. To use Disqus, please click the Login link below and sign in. If you have questions about how to use Disqus, please refer to this help page.


Readers may also send comments via email directly to moursund@uoregon.edu and bobsyl@uoregon.edu.

About Information Age Education, Inc.

Information Age Education is a non-profit organization dedicated to improving education for learners of all ages throughout the world. Current IAE activities and free materials include the IAE-pedia at http://iae-pedia.org, a Website containing free books and articles at http://i-a-e.org/, a Blog at http://i-a-e.org/iae-blog.html, and the free newsletter you are now reading. See all back issues of the Blog at http://iae-pedia.org/IAE_Blog and all back issues of the Newsletter at http://i-a-e.org/iae-newsletter.html.