Lecture October 30

Analyzing Change- continued – OCT. 30, 2002

Overview of Lecture:


1. Review of Oct. 28’s #4: does the computer have a built-in bias that favours a certain kind of analysis (of “technological change” in our case?)

(maybe, since the computer is a number cruncher, it favours quantifiable analysis of data)
(yet, we’ve been looking at theories that are not that kind of research done by social scientists)

2. MANY social scientists measure change through quantifiable analyses of data:
experimental and correlational research

3. The relationship between science and technology:

Homer-Dixon’s views and echoes of Postman’s views re” “techno-hubris”
*************

2. Overview of “number-crunching” social science research:

- desire to measure quantifiable changes to users’ abilities, attitudes, etc., in light of using the computer

examples of some kinds of questions:

* Do people’s abilities (in doing X) improve using computers?
* Have people’s attitudes to computers changed over time? (are they happier? more frustrated, etc.)
* Are people’s other activities influenced by time spent using computers?

But how exactly prove that changes have taken place?

2.1. Research Method #1: Experimental Study
Experimental Hypothesis: “Students will write better essays if they use a word processor”

What would you need to do to prove this conclusively?

--> would require independent variable that is subject to manipulation;
--> control that other potential variables are kept constant;
--> require random selection of subjects into control and experimental groups;
--> determine what makes a better essay, then do pre and post testing to see if use of independent variable had predicted effect.

Impossible to prove at this point because:

* you can’t control all other variables which would influence how well someone writes, and
* how would you find subjects who could be in the control group?

2.2. Research Method #2: Correlational Study

Hypothesis: Males will do better in computer science courses than Females

Independent variable: male or female
Dependent variable: marks

(Can’t say that success is DIRECTLY CAUSED BY ONE’S SEX, only that there might be a CORRELATION between one’s sex and success in computer science courses.)

--> BTW: at least one study disproved this correlation. And even if proven, it doesn’t answer the more important and interesting question, “why?”

What hypothesis might we want to test?:

“With increased computer usage, comes decreased television viewing.”

What questions would one need on a survey to show this correlation?

Over the past year, has the time you spend on the computer (circle your response):
1
2
3
4
5
increased alot
increased somewhat
stayed the same
decreased somewhat
decreased alot
Over the past year, has the time you spend watching t.v.:
1
2
3
4
5
increased alot
increased somewhat
stayed the same
decreased somewhat
decreased alot

More interesting question: “If there is a change to people’s leisure habits, what do you attribute this change to?”

Observations: Only empirical researchers might indeed be called social SCIENTISTS

3. Homer-Dixon on science and technology :
“We have placed a lot of faith in that ingenuity-producing powerhouse modern science…..and “Basic science must be the foundation of any new practical technology”

H-D’s comparison of technological change across 4 domains of technology from early 19c. to 20c:

Military explosives – billionfold increase in explosive power

Long-distance communication – trillionfold improvement in performance

Personal transport – 50 – 300fold improvement in transport speed

Agriculture – 4fold improvement

What does this tell us?
1. “astonishing changes technology has brought us recently”
2. improvements vary dramatically across domains: “technology solves some problems faster than others.”
3. there doesn’t seem to be any relationship between the problems most needing solving and the problems that get solved. (MS vs. advances in information technologies)
4. another factor explaining the delays between scientific discoveries and technical inventions, have to do with 4 factors that affect the pace of scientific discoveries:

- human cognitive limits: beyond our mental capabilities to comprehend the complexities around us
- the “intrinsic complexities of the research field’s subject matter
- nature of its scientific institutions
- science’s “social context”
- another factor affecting movement from science to technological invention is “right economic markets” (e.g., fax machine)

Where does he predict significant technological advances?:
Genetics and medicine – DNA, etc.

Information processing and computation
Microprocessors, genetic programming, AI
(see prediction re: information-gathering)
Materials engineering
Machine miniaturization

But his parting words echo Postman’s view of technology: if “miraculous machines become our gods… we could face “techno-hubris.”



This page last revised 9/17/02