Chapter 1. Concepts
1.1 Introduction
Concepts
How good are you at recognizing patterns, particularly patterns that vary to some degree? Think of the different fonts for the letter a: A, A, A . There are two phases in this experiment, a learning phase and a testing phase. First, of course, is the learning phase. In this experiment you will be presented with a pattern of dots. You have to guess which group the dots belong to. At first you can only guess, but you will be given feedback as to whether you are correct and, if you are wrong, what group the dots belong to. Use the feedback to determine your future responses. Pay close attention to the feedback. The testing phase is similar, only you will not get any feedback.
1.2 Experiment Setup
1.3 Instructions
Instructions
There are two phases in this experiment: the learning phase and the testing phase. The two phases will run exactly the same way, except that in the testing phase there will be some new random patterns as well as the three prototypes used to make all of the variations.
You will be presented with a pattern of dots drawn from one of three possible patterns. Your job is to indicate which of the three original patterns the dots came from. You may use the buttons below the pattern or the keyboard responses. If you do not know the correct answer, you still need to make a guess. During the learning phase, after you respond, you will get feedback as to whether your response was correct.
| Key | What Response Means |
|---|---|
| A | Pattern A |
| B | Pattern B |
| C | Pattern C |
| Spacebar | Random Pattern (in Testing Phase) |
1.4 Experiment
Begin Experiment
1.5 Results
Results
1.6 Debriefing
Debriefing
One of the hallmarks of human cognition is our ability to work with abstract concepts. Concepts, even concepts like justice, derive from real world examples, like learning to be fair as children. To create a concept, we have to form some idea that can relate to multiple real situations. As a result, it is easier to study how we form concepts using a visual stimulus. Take the letter A, for example. Given both upper and lower cases and all the available fonts, it is remarkable that we can so quickly and easily recognize an A as an A. One way to think about this ability is that we store a prototype for the letter A and use this prototype to understand all the individual examples. A prototype in this context is the standard representation from which the examples are derived.
Let us review the experiment. There were two phases; in the first phase, the learning phase, you had to categorize different variants of three different prototypes. You never say the prototype. That feature of the experiment is very important. In the testing phase, you saw the prototypes, the old variants, new variants, and random patterns not derived from any prototypes. The interesting trials are those that are derived from the prototypes, so we will not consider the random trials right now. Posner and Keele (1968) found that people were slightly faster and more accurate when responding to the prototypes than to even the old variants, though the response to the old variants was both faster and more accurate than responding to the new variants. The difference between the old and new variants is understandable. You have seen the old variants before and can make your choices based upon memory. The interesting result is the speed and accuracy of the responses to the prototypes of the patterns. You have not seen the prototype patterns before and yet you respond better than to the old patterns and much better than the responses for the new patterns. One way to understand these results is to consider what is being stored in memory. Focusing on the performance to the prototypes, these results suggest as we perceive patterns, we generate an abstract representation of all the patterns that we group together. So, as we learn to categorize the A dot patterns, we also create a sort of standard that lacks the variations of the individual A dot patterns. We create a representation that is like the prototype. When the prototype is presented in the testing phase, there is already a memory representation for this prototype, so it aids our performance.
This abstract representation can be called a concept or schema. In other words, when presented with several versions of the same object, like faces at different angles, or different examples of the same concept, like justice, we generate an abstract version of the object or concept that allows us to generalize to new situations or versions of the stimulus.
References:
Posner, M. I., & Keele, S. W. (1968). On the genesis of abstract ideas. Journal of Experimental Psychology, 77(3, Pt.1), 353-363.
Tunney, R. J., & Fernie, G. (2012). Episodic and prototype models of category learning. Cognitive Processing, 13(1), 41-54.
1.7 Quiz
Quiz