Greeno (1978) - three basic classes.
Problems of inducing structure. Series completion and analogy problems. Problems of arrangement. String problem and anagrams. Often solved through insight. Problems of transformation. Hobbits and orcs problem. Water jar problem.

You are going to create an intelligence test. Administer it, and then crunch the numbers, which means being familiar with statistics, graphing, and go back and look at chapter 2 to refresh yourself. Also look at Appendix B, which has a nice breakdown of stats. You will not have to come up with a hypothesis -- you're just running an intelligence test.

Turing test is valid.
-- Have a bot that the person can talk with on the computer. And then another one where a person can be on the other end.

Problem solving vs. decision making
Problem solving: you do not know your options.
Decision making: you have an array of options.
You have a challenge and you have to go through a process -- that's problem solving (simulations). Decision making is comparisons, analyzing, it's much more nebulous, like heuristics.

1) Problems of inducing structure
-- Like filling in a gap. Or analogy problems. Hand is to glove as hat is to head. 
2) Problems of arrangement
-- Shifting things around to come up with the right answer. Anagram.
--- You get a word where the letters are all mixed up. You are shuffling things.
--- Some people can just look at anagrams and know the answer. This may be due to context-sniffing.
3) Problems of transformation
- Mathematics includes problems of transformation.

Barriers to effective problem solving:
-- Irrelevant information
-- Functional fixedness
--- Using tools in interesting new ways.
-- Mental set
--- Your attitude or belief system that prevents you from taking certain steps, like if you have always solved this sort of problem a certain way. Call this being stubborn.
-- Unnecessary constraints
--- Time constraints are usually meaningless.


Well-defined v. ill-defined problems
-- insecurity with ill-defined problems? Not everybody experiences this.
--- The more you do, the better you are, at least with those ill-defined Creativity Problems.

Algorithms for problem solving - step by step. Recipes. Programs. Sequences. And when you take a shortcut in recipes, you're taking a heuristic. The problem with heuristics is that you can be wrong. 

Heuristics
-- Shortcuts
-- No guaranteed solution.
--- Forming subgoals
--- Working backwards (reverse engineering)
--- Searching for analogies
--- Changing the representation of the problem (like translating a maze into a graph ...)
==== Is there a way to translate a "find the path through the maze" problem into a graph?? Structures of mazes, possible ways.

Decision making -- you are not doing the steps of the problem. 

The Theory of Bounded Rationality (Simon, 1957) - you can limit your ability to think through problems without even knowing it. You bring in your own outcomes, your own biases, subjectivities, you can limit yourself. 

Making choices:
* Additive strategies
* Elimination by aspects
* Risky decision making
** Expected value
*** What you think you are going to get out of it
*** Emotional value
** Subjective utility
*** How effective are you? Are you a good observer of your ... ?
** Subjective probability
*** Just because it's me, this won't happen. (WRONG)
*** Odds of being in an accident if you're a drunk driver - 1 in 3. 
**** That's historical analysis / statistics, not based off of the probability due to your brain structures.

Heuristics in judging probabilities
- Availability heuristic. When making your shortcut, you use information at hand. You might be misinformed, or underinformed.
- Representativeness. You're using a prototype. The exception messes up your process.
- Tendency to ignore base rates.
- Conjunction fallacy. We tend to think that because things happen together, that they will always happen together. So we start building belief systems on top of this. It's conditioning, and sometimes not conditioning.  
- Alternative outcomes effect. We start thinking about all of the possibilities.

Gambler's fallacy - it'll pay off at some point, RIGHT ?!
Overestimating the improbable - "That'll never happen." 
Confirmation bias / belief perseverence - You already have a bias on your decision, and you go research, but you only research things that support your decision. Who wants to research the other stuff? Belief perseverence. 
Overconfidence effect - Napolean and other generals tend to make this. They figure, since we have won everything, we will continue to win. What a terrible way to run an army.