CHAPTER 02:H OW D O W E FIND OUT ?T HE L OGIC,A RT ,AND E THICS OF SCIENTIFIC
DISCOVERY (P G.29-57)
The difference between descriptive and prescriptive laws (i.e., the difference between what people
ought to do and what people actually do)
Scientists and laypeople alike often have preconceived notions about the way to discover what is
o These preconceived notions sometimes get in the way of accurate scientific discoveries.
T HE L OGIC OF SCIENTIFIC D ISCOVERY
LAWS ,THEORIES,AND H YPOTHESES
“He who pays any attention to the syntax of things will never wholly kiss you” – Poet E.E. Cummings
o Seemed to mean that it is easy to ruin a wonderful experience by overanalyzing it.
o Scientific research has confirmed it
Example: Taste testers who think too much about the basis of their preferences while
sampling jellies have a lot of trouble telling the difference between really succulent
and really sucky jellies.
o Does not mean we shouldn’t seek scientific explanations for important experiences
o Means that we shouldn’t allow what we learn about the nature and causes of these important
experiences to get in the way of the enjoyment these experiences can bring us
Some critics argue that although it would be highly desirable to understand things like infatuation or
gustation, it simply is not possible.
o No one can predict exactly what another person will do.
o Psychologists can predict the behaviour of large groups of people just as meteorologists can
predict the weather (loosely speaking, the behaviour of large group of clouds).
o Example: Researchers who study interpersonal attraction couldn’t tell you for sure whether
Rupert is going to fall in love with Rita tomorrow, but they could identify a long list of
conditions that would make this much more likely than usual.
On the basis of scientific studies of interpersonal attraction, we know that people
gravitate toward other people who:
Are physically attractive
Are familiar rather than unfamiliar
Highly competent (but not too competent)
Research also shows that we are strongly attracted to others whose earlobe shapes
closely resemble our own.
The goal of every branch of science is to discover laws
o Sir Isaac Newton
Scientist when he wasn’t busy inventing calculus or the reflecting telescope, he was
busy developing his three famous “laws of motion”
Second law of motion The acceleration of an object is directly proportional to the
net force acting upon it and inversely proportional to its mass
Knowledge of this law has allowed scientists to put telecommunication
satellites into orbit around the earth
Newton proposed this law without any qualification because he was confident
that it would predict the acceleration of any object in any circumstance
Law: a precise, coherent, well-developed account of the relation between two or more variables. A
universal statement that allows reliable predictions of future events.
If we could identify laws of human behaviour, we would not necessarily be able to predict every
detail of a person’s behaviour. A well-trained psychologists should be able to make some predictions that untrained people would
be unable to make.
The psychologist’s ability to make correct predictions about someone should increase (1) to the
degree that the behaviour in question was well studied (e.g., psychologist’s know more about human
memory than they do about human taste in music) and (2) to the degree that the psychologist was
given a great deal of relevant data about someone.
A theory is a general statement about the relation between two or more variables.
o Good theories should:
Be logically and orderly
Emphasizes the systematic causal relations between variables
Be empirically grounded they should generate predictions about readily observable
Be parsimonious a simple and concise statement that allows you to predict a wide
variety of conceptually similar behaviors
The main difference between laws and theories is their breadth or universality.
o Laws are comprehensive, fundamental statements about reality
o Theories have boundary conditions there are plenty of times when they do not apply
It is possible for more than one theory to be true.
The best of theories offers accurate prediction only in certain circumstances.
o Example: A psychologist who develops a theory of aggression usually knows that this theory
will not predict all forms of aggression in all situations.
If theory is a good one, it will offer reliable prediction of aggression under a
reasonable, and specifiable set of circumstances.
A theory may apply well to physical aggression among boys, but it may need
some careful revision to explain the aggressive behaviour of girls.
It is possible for more than one theory to be true.
o When one theory predicts behaviour in one set of circumstances (or for one group of people)
and another theory predicts the same kind of behaviour in a different set of circumstances (or
for a different group of people)
A good theory not only states specific relations between specific variables, but it also states the
conditions under which these relations do and do not apply.
o Two or more correct theories will sometimes cancel one another out.
Example: “Exposure to violent media images produces aggression”
If a person grew up watching a lot of violent TV but never drinks alcohol and is
not currently frustrated, this person’s nonviolence does not invalidate the
theory that exposure to violent TV causes aggression.
Sometimes, the friend or stranger will criticize the theory by noting that they know someone who
does not usually behave in the manner suggested by the theory
o Exceptions do little to invalidate a theory.
o Theory falls short of being a law
Human behaviour usually has multiple determinants this is known as the principle of equifinality
– the notion that the same behaviour is often produced by many different causes.
o Because of equifinality, psychologists sometimes disagree about the conditions favoring one
theory over another, and so good theory testing will often take the form of determining each
theory’s proper domain of application.
o This sort of theory testing is especially important when different theories lead to
It is hard to predict behaviors in a specific situation just means that numerous factors (including
some we surely did not think of (may contribute to this specific decision) This does not mean that
behaviour doesn’t follow meaningful rules. Laws and theories do not account for every kind of statement scientists can make about the relation
between two or more variables
Hypotheses: Predictions about specific events that are derived from one or more theories
o To test a theory, a person should use that theory to derive specific predictions that can be
For hypotheses to serve as tests of specific theories, they must follow in a clear and logical fashion
from the theories in question, preferably under conditions clearly specified by the theory.
A theory can clearly and obviously lead to a specific hypothesis in a given circumstance.
A lack of support for the hypothesis translates into a lack of support for theory that produced it.
Researchers may be uncertain about the conditions under which the theory applies.
o Test hypotheses in specific circumstances to see whether these circumstances fall within the
boundary conditions of the theory.
Two different theories might make opposing predictions in the same conditions, and researchers
might then outline two rival hypotheses.
o If results support one theory to the detriment of the other, the theory that lost out may prove
to be false.
o It is possible that this particular set of circumstances simply favored the theory that seemed
to win out.
T HE SCIENCE OF O BSERVATION
The notion that we should base what we know on what we observe now seems pretty obvious, but
development of this concept represents a major advance in the history of science.
Early part of the 17 century- Francis Bacon proclaimed that scientists can uncover hidden truths
from everyday observation.
o To benefit from careful scientific observations, Bacon argued, you must apply the method of
Making use of induction means making many observations under controlled
conditions and arriving at a general statement about how things are.
Induction reasoning from specific instances to general principles.
We engage in induction all the time
In scientific circles general conclusions drawn in this way (e.g., “dropped objects always fall toward
the earth”) usually come to be known as theories, and these theories can often be tested against new
observations that are made in a variety of new situations.
o If new observations are consistent with the statement, then the statement survives.
o If not, the statement is either discarded or revised.
If the statement is revised, it is then tested against new observations, and the whole
process starts over again.
With the benefit of continuous revisions, the general statement should become more
and more accurate.
The inductive technique Bacon supported requires careful, systematic observation, and in scientific
circles it usually leads to a more or less continuous cycle of revisions based on precisely these
o Felt that Bacon was dead wrong
o Argued that observation is not the appropriate dividing line between science and nonscience
o Problem of induction:
Careful induction requires scientists to make a great number of observations
o According to Hume, you never make enough observations to be sure that the law is true.
Even if you have made hundreds of observations, all of which are consistent with your
general conclusion (i.e., your law or theory), it is always possible that the very next
observation you make will prove you wrong by violating your conclusion. Example illustrating the problem of induction:
o Social psychologists who study person perception once believed that there is a fundamental
human tendency to attribute a person’s actions to that person’s enduring personality (rather
than attributing the actions to situational factors)
Lee Ross (1977) was so impressed with this tendency for people to favor dispositional
over situational explanations that he labeled this tendency the fundamental attribution
Ross’s use of the word “fundamental” suggested that this attributional tendency
was a basic feature of human social judgment.
Miller (1984) collected cross-cultural data suggesting that people from India do not
make the fundamental attribution error.
Social psychologists though that they had uncovered something basic to human
judgment because they had run numerous carefully designed studies showing this
effect. Miller’s findings changed the whole picture.
Deduction: Reasoning from the general to the specific (e.g., drawing a conclusion about human
memory by logically deriving it from a higher-order principles about human cognition.
o In science, deduction occurs when a general statement (a theory) is used to develop
predictions (hypotheses) that are then tested against observations.
o Karl Popper (the same guy who promoted logical positivism) was particularly fond of the
Popper argued that good theories have to be falsifiable, which means being open to
Big fan of subjecting theories and laws to empirical tests even though he strongly
believed that no theory or law could ever be proven.
It is possible to prove that a theory is false, but it is not possible to prove that it is true.
o We can never prove anything with absolute certainty, but we can sometimes show that a
general idea is supported under a wide variety of circumstances.
If we’re willing to live with a little bit of uncertainty and trust science to correct itself, we don’t need
to worry about the fact that very few, if any, scientific statements are ever completely true
A great deal may be learned from a failure (from falsifying a theory)
o Sometimes it is more useful to know what is not true than to know what is true.
o From this viewpoint, scientists do not travel forward toward truth so much as they walk
backwards away from falsehood.
o The debate over what is scientifically knowable is partly a matter of how you choose to frame
Most scientists realize that their erroneous theories, if they receive attention from other scientists,
will serve as the inspiration fro the superior theories that someday replace them.
T HREE A PPROACHES TO H YPOTHESIS T ESTING
Positive Test Bias: the tendency for hypothesis testers to attempt to confirm rather than disconfirm a
hypothesis. The bias in hypothesis testing is similar to (1) behavioral confirmation in social interaction
and (2) validation as an approach to scientific hypothesis testing.
o Example: Research in social psychology also suggests that when people are testing social
hypotheses (e.g., “Is Zoe an extrovert?”), they are inclined to seek out evidence that is consistent
rather than inconsistent with their preexisting expectations.
People would ask a preponderance of questions that are designed to elicit extroverted
Behavioral Confirmation: the tendency for social perceivers to elicit behaviors from a person that are
consistent with the perceivers’ initial expectations about the person (e.g., expecting a person to be
aggressive and being more aggressive than usual with the person, which leads the person to behave in a
very aggressive fashion towards you).
o Example: Study by Snyder, Tanke, and Berscheid (1977) Researchers gave some men the hunch that women they were getting to know over a
laboratory “telephone” were either sociable or shy (by giving the men fake photos
depicting either extremely attractive or unattractive women)
Men who thought they were talking to highly attractive women later reported that they
had expected the women to be highly sociable and entertaining.
In the course of their conversations with the presumably attractive woman, the men made
these expectations come true
Men were more animated and entertaining themselves when they thought they
were talking with an attractive woman
Women on the other end of the phone (who knew nothing about the misleading
photographs) confirmed the men’s originally false expectations by behaving in a highly
sociable fashion themselves (as confirmed by raters who did not see the photos and only
listened to what the women were saying)
Research on self-fulfilling prophecies, experimenter bias and stereotyping tell a similar story
o Once we get an idea in our heads, most of us tend to engage in hypothesis-confirming behaviors
that may falsely convince us that the idea is correct.
o Consequence of confirmatory judgment biases is that:
People often believe that they have confirmed hypotheses that are not true
Once we have been exposed to some tentative evidence in support of our theories or ideas,
we also become very reluctant to give them up- even in the face of strong disconfirming
evidence that comes along later.
Dan Gilbert and his colleagues have gathered evidence suggesting that it may be impossible for human
beings to comprehend a statement without initially encoding the statement as true.
Psychologists (and other scientists) sometimes show the same bias as laypeople that are evaluating
Three general approaches that scientists seem to adopt when testing hypothesis include validation,
falsification and qualification.
Validation: An approach to scientific hypothesis testing in which a researcher attempts to gather
evidence that supports or confirms the hypothesis.
o Rough scientific equivalent of the positive test bias
Researchers would make implicit choices about what kind of data to examine and they may even
engineer laboratory situations that are highly conductive to supporting their theory or hypothesis
This problem is serious only if scientists pretend that their tests of their hypotheses are free of any bias.
o Example: Researchers who are testing a novel theory sometimes want to know if the theory is
true under any circumstances
Any conscious or unconscious biases toward selecting just the right sample or just the
right situation might be forgivable
If a phenomenon proves to be interesting future researchers will almost always try to
identify the limitations (i.e., the boundary conditions) of the theory
Festinger’s (1957) theory of Cognitive Dissonance
o Cognitive dissonance theory states that when a person simultaneously holds two beliefs that are
dissonant (i.e., logically inconsistent), the person will experience an aversive state of arousal.
The person will be highly motivated to reduce this aversive arousal (i.e., this dissonance)
by making two beliefs more compatible.
o Study (Festinger and Carlsomith, 1959):
Participants spent an hour engaging in an extremely boring peg-turning task.
Experimenter then convinced participants that they would be doing him a big favor if they
would tell an unsuspecting person (who was waiting to being the same task) that the
wearisome task was extremely interesting Experimenter made people feel personally responsible for their behaviour by
letting them know that, as much as he needed their help, participants had the
choice not to help him.
Virtually every one of the thousands of participants agreed to help the experimenter – and
harm the person waiting to being the study – by agreeing to tell the lie.
Experimenter was in control of whether people had an excuse for the glaring
contradiction between their direct experiences of boredom and their direct reports of
fascination because the experimenter offered people different amounts of reward for
“helping him out”
In one condition, the experimenter offered people the paltry reward of $1 to lie to
the potential peg-turners
In a second condition, experimenter offered people a $20 reward to tell the same
Presumably, $20 was enough to prevent people from experiencing a great deal of
dissonance about telling the lie.
$1 condition participants reduced this dissonance by deciding that the task was quite a
bit more enjoyable than they had originally thought.
Results: $1 paid participants justified their deceptive behaviour by changing their
attitudes to be more consistent with their behaviour.
In comparison with the well-paid participants (or an additional group of
participants who were not induced to lie at all), the $1 participants reported that
they felt the experimental task was quite interesting.
o Festinger and Carlsmith were looking for evidence that could confirm dissonance theory.
This didn’t prove to be a serious problem because many generations of future researchers
worked hard to identify the precise boundary conditions of this interesting theory.
Falsification is an approach to hypothesis testing in which researchers attempt to gather evidence that
invalidates or disconfirms a theory or hypothesis.
Theories or hypotheses that survive careful attempts at falsification are all the better for it.
o Theories that do not, according to Popper, should be discarded (hopefully in favor of better
Some aspects of falsification are simply an inherent part of scientific investigation
o Scientists have to justify and carefully describe their methods to other scientists they can’t
usually get away with biased approaches in hypothesis testing.
o Any study that involves careful, objective data collection can yield results that are the opposite of
what a researcher might have hoped for or predicted.
Some aspects of the scientific method promotes falsification whether a researcher likes it
or not scientists must follow a minimal set of rules
Theories are often scrutinized by a great number of scientists, including those whose biases run
contrary to those of a theory’s proponents.
o Psychology has a healthy tradition of adversarial theory development
o Researchers forms different “camps” of