Cognition: Mental Representations, Problem Solving, and Decision Making - Foundations for Designing User-Centered Systems - page 180

They can be quite simple and isolated, or they can occur within a string of
decisions, perhaps while problem solving.
Decision making is studied by a wide range of disciplines, partly because there
are many kinds of decisions, and partly because they affect all areas of human
behavior. Interesting work on decisions are found in psychology (e.g., Gigerenzer
et al. 1999 ; Kahneman et al. 1982 ; Evans 1990 ), business (e.g., Simon 1997 ),
advertising (e.g., Ries and Trout 1986 / 2000 ), human factors (e.g., Cacciabue et al.
1992 ; Hoffman et al. 1998 ; Lehto 1997 ), HCI/marketing (e.g., Fogg 2003 ), and
economics (e.g., Levitt and Dubner 2005 ). Here we focus on some of the gener-
alities that are applicable to designing systems.
6.4.1 Decision Making is Often Not Rational
One of the fundamental truths in this area, and the first to note, is that often the
most rational choices to an outside observer or experimenter are not chosen.
People making decisions do not always take account of a lot of potentially relevant
information when making decisions, and their decision making processes have
general, knowable biases. That is, there are systematic ways that decisions are
badly made.
Simon ( 1997 ) argues that, rather than trying to find the optimal answer, people
typically satisfice. So, for example, if users were trying to find the cheapest price for
something online they would choose a good enough price given the constraints of
doing better which may involve looking at the price across 10 sites, rather than 500.
In some sense, they are factoring in the price of further searching and the extra time
needed to find a better price.
There are other problems with the way that people make decisions as compared
to the way computers typically make decisions. These problems can often be
related back to the cognitive processing capabilities introduced in earlier chapters.
The rest of this chapter describes some of these systematic limitations, starting
with simple decisions (Kahneman 2013 ).
6.4.2 Simple Decisions: Hicks Law and Speed-Accuracy
Trade-Offs
As noted earlier, signal detection theory provides a simple way to categorize
decisions in terms of their likelihood of being correct. It is also important, how-
ever, to know how long it takes to make a decision.
Pressing the corresponding button when a particular light illuminates involves a
simple decision. It takes around 150-400 ms, depending on factors such as the
strength of the light and the distance to the button. These numbers are useful in
system design, because factors such as the time to press the brake pedal in a car