Synopsis:

Every day you make decisions. You decide to take IH-35, rather than MoPac, to drive to school, because you think it will provide you a quicker, safer, and/or happier trip. You base this decision on some data you have collected from your previous experience, or from information people have told you, or from information gleaned from a map, or from radio and TV reports. Or maybe you just have a feeling.

During that drive to school, and likely before, and certainly after, you will hear or read many, many claims.

• "Crest makes your teeth brighter."
• "Our candidate will improve Austin traffic. "
• "Taking this course will help you be a better information scientist."
• "I like you."
• "This is a better way to design your web site."

In one of the required textbooks for this course, the author Vincent Dethier asserts that, “An experiment is [mankind’s] way of asking Nature a question.” As an information scientist, you may wish to ask Nature a question (e.g., “Will this intranet user interface enable all our users to carry out their tasks”?). You will CERTAINLY read of many, many answers other information scientists have inferred from questions they have asked of Nature.

The focus of this course will be on behavioral science – the design of experiments that enable us to acquire new information about human behavior. We will address qualitative research methods, historical research, and other forms of research, to place quantitative research in context. But the focus will be unmistakably on quantitative research. In the arena of experimental design we’ll cover sampling, the control of variables, the choice of within- and between-subject designs, experimental vs. field study, and this will lead us into statistics as we consider hypothesis testing. In the arena of statistics we’ll cover probability, descriptive statistics (measures of central tendency, measures of dispersion, correlation) and inferential statistics including tests of statistical significance.

Unprepared information scientists – indeed, unprepared citizens – are forced to consider the torrent of claims they hear every day, and either accept or reject them based on faith. Prepared scientists/citizens can, instead, consider the methods used to gain the information on which the claims are made, and evaluate for themselves the likely goodness of the claims.

Expect a course with a bias (heh heh) towards quantitative research, but flavored by an awareness that there are various ways to conduct research. Expect two tests where you have a chance to demonstrate that you understand the basics of experimental design and statistics, and know WHY one experiment is better than another, to answer a particular question. Expect some lecture, some discussion, and some hands-on designing of some research. Expect to be surprised how interesting (and painless) this stuff can be, regardless of how math phobic you may be. Expect to know how many socks you need to pull out of your sock drawer, in the dark, to be assured of having a pair of the same color. Expect to come out of the course being able to evaluate whether a piece of research you read about was well conducted and appropriate.