There are a variety of ways we know something.
Personal Experience We know something from Personal Experience, from our life events. For example I know Travis Tritt's song "It's a great day to be alive" is good because it makes me feel good. The problem with using personal experience to generalize is that individuals have different personal experiences which affect their reaction to life so you cannot come up with a general understanding of human behavior based on personal experiences.
Authority We know something from Authority if a respected source states it; for example "Jesus loves me this I know for the Bible tells me so." or "JLo is great because all my friends say she is." The problem with Authority as a basis for knowledge is that each individual has different Authority figures so any universal truth cannot be arrived at.
Deduction We know something from Deductive Logic if a conclusion logically follows a true premise; for example, "(1)All men die, (2)Socrates is a man, (3)Socrates is (or will be) dead." For psychology the task is to test if the premise is true by seeing if the predicted conclusion does occur. For example, take the premise of observational learning theory (the basis for the commonsense saying, "monkey see, monkey do.") One premise of observational learning is, "(1)People tend to copy admired others." This, (1), is a theoretical hypothesis which the psychologist could design a study to test by collecting data. Following the deductive model we might predict, "(2)People will copy Jennifer Lopez" (the most admired person in the 18-30 age group in a Dec, 2002 CNN poll). Then we could choose a criteria for copying. For example, "(3) The sales of Jennifer Lopez's clothing line will correlate with her popularity rating." This is the most controversial step in the application since in this type of field study it is uncertain what is the best dependent variable to choose to test the hypothesis. The idea here is that Ms. Lopez's appearance is an important part of her admirable qualities so consumers would want to purchase items to look like her. If the data supported the conclusion (3) we would have more evidence on the scope and power of observational learning theory.
The problem psychology has with the purely deductive approach is that it
does not lead to new knowledge. Notice it just draws out the conclusions that already
implicitly exist in the statements since pure deduction assumes the major premise (1) is
correct. However, as shown above, deductive logic is used as one tool
in psychology to make
sure the findings of psychological studies are properly tested, interpreted, and related
to relevant theories.
Induction To say we know
something from Empiricism means that (using Inductive Logic) we can go from the
total set of individual facts (like data from a number of individuals ) to a
generalization about behavior. For example, if restaurant servers who introduce
themselves by name get larger tips (23% average tips) compared to those who
don't introduce themselves by name (15% average tips) we have evidence that name
introductions cause higher tips. First we have to see if this finding could just
be chance or is it real? If our test of
statistical significance shows it is likely real, then we are ready to
consider how this result adds to psychological science. This inductive
approach is our preferred way of finding knowledge
in science because it avoids the problems the other methods have. It has a built-in self corrective method since any trained scientist can replicate a study,
present it in a public forum and have the results confirmed or disconfirmed by
further studies. In fact, psychological science consists of the consensus from
the accumulation of all the studies done. So induction builds psychological
science on two levels. At the individual study level, we summarize individual's
responses (usually using statistics like an average) to reach a conclusion. If
we put the results of a group of individual studies together we can summarize the results of a number of studies
(using a statistical approach called meta-analysis) to reach an overall conclusion about a
topic.
HOW DOES A PSYCHOLOGIST START A RESEARCH STUDY?
A psychologist's research can start in two different ways. One approach is to
start with
a theory to develop hypotheses to predict
empirical observations like in the JLo example where we took a premise of
observational learning theory as a start. The other approach is to start
with observations of behavior. For example, if you were a server in a restaurant and
noticed that some servers got bigger tips than others and tried to figure out
why, you would be using induction by looking at the specifics about servers that
varied with tip size. You might reach a tentative generalization that servers
who introduced themselves by name got bigger tips. You would then formulate a
hypothesis (1) "Name introductions lead to higher tips." and design an experiment
that will either confirm or disconfirm this hypothesis. So here the
psychologist uses induction to analyze all the specific facts and come up with a
tentative hypothesis. Then, using deduction, the psychologist figures out the
specific facts that would occur if the hypothesis, the (1) element in a
deductive syllogism, is true and designs a test for
the hypothesis (as shown in the JLo example.) The interplay between these two modes of reasoning is called the
double movement of reflective thought. If your name introduction
hypothesis
was confirmed you might see how far you can extend it by testing the broader
hypothesis that "Personalization leads to higher tips." Then you would
design a study to confirm or disconfirm that theory's predictions.
So, to recap, starting with a theory the major logical tool initially used is deduction to develop the specific study to test the theory's prediction Or if starting with some specific observations, the logical tool of induction is used to arrive at some testable generalization and that is the double movement of reflective thought.