SCIENTIFIC METHODS IN

PSYCHOLOGY
After studying this module, you should be able to:
1. Discuss the importance of replicable results.
2. Define burden of proof.
3. Explain why scientists seek the most parsimonious explanation of any result.
4. Explain why most psychologists are skeptical of claims of extrasensory perception.
 PSYCHOLOGICAL SCIENCE

vThe word science derives from a Latin word meaning “knowledge.” psychologists insist
that their field is a science. This is true for two reasons: its history and ethics that govern
psychology.
vAs opposed to the history of several academic disciplines, psychology, began as a
deliberate attempt to start a new science, applying the methods of the natural sciences to
some of the questions of philosophy. The early psychology professors frankly didn’t have
much to teach, other than what biologists had already discovered about the sense organs.
Secondly, the issue of ethics. chemists can do almost anything they want to a jar of
chemicals, as long as they don’t blow up the building. psychologists dealing with people
have stringent limits.
 GATHERING EVIDENCE

• Science is a search for knowledge based on carefully observed, replicable data. let’s first
consider data collection, and then examine that word replicable.
• Research starts with careful observation. a great deal of scientific research consists of
observing and measuring.
• Good observations and measurements often suggest a pattern that leads to a hypothesis,
which is a clear predictive statement, often an attempt to explain the observations.
• A test of a hypothesis goes through the series of steps described in the following four as
shown (Feist & Feist, 2009; Kalat, 2017).
Hypothesis
• A hypothesis can start with observations, such as noticing that some children who watch much televised
violence are themselves aggressive.
• The hypothesis would be: that watching violence leads to violence.
• A hypothesis can also be based on a more general statement, such as ‘children tend to imitate the behavior they
see.’
• A good hypothesis leads to predictions. For instance, ‘if we let children watch violent television, they will
behave more aggressively,’ or ‘if we decrease the amount of violence on television, the crime rate will
decrease.’
Method
• Any hypothesis could be tested in many ways. One way to test the effects of televised violence would be to
examine whether children who watch more violent programs are more violent themselves. When we find that
on average they are, that result is only the start.
• It does not tell us about cause and effect: does watching violence lead to violence? or Is it simply that people
who are already violent like to watch violence?
• A better method is to take a set of children, such as those attending a summer camp, randomly assign them to
two groups and let one group watch violent programs while the other group watches nonviolent programs, and
see whether the two groups differ in their violent behaviors (Parke, Berkowitz, Leyens, West, & Sebastian,
1977). The limitation is that researchers control what people watch for only a few days.
• Because any method has strengths and weaknesses, researchers vary their methods.
• According to Greenwald (2012), if studies using different methods all point to the same conclusion, we
gradually increase our confidence in the conclusion. a single study is almost never decisive.

Results
• Fundamental to any research is measuring the outcome.
• A phenomenon such as ‘violent behavior’ is tricky to measure. (Do threats count? Does verbal abuse? When
does a push or shove cross the line between playfulness and violence?)
• It is important for an investigator to set clear rules about measurements. After making the measurements, the
investigator determines whether the results are impressive enough to call for an explanation or whether the
apparent trends might have been due to chance.
Interpretation
• The researchers’ final task is to consider what the results mean. If the results contradict the hypothesis,
researchers should abandon or modify the original hypothesis.
• If the results match the prediction, investigators gain confidence in the hypothesis, but they also should
consider other hypotheses that fit the results.
•
 REPLICABILITY
• Most scientific researchers are scrupulously honest in stating their methods and results. A major reason is that anyone who
reports a scientific study must include the methods in enough detail for other people to repeat the procedure and, it is hoped
to get similar results. Someone who reports results falsely runs a risk of being caught, and therefore distrusted from then on.
• Replicable results are those that anyone can obtain, at least approximately, by following the same procedures. Scientists do
make certain allowances for small effects. for example, one method of teaching might work better than another, but only
slightly, so the advantage might not appear in all studies, especially those with a small number of participants.
• When researchers try to verify a small effect, they use a meta-analysis, which combines the results of many studies as if they
were all one huge study. A meta-analysis also determines which variations in procedure increase or decrease the effects.
However, if no one can find conditions under which the phenomenon occurs fairly consistently, it is not taken seriously. This
rule may seem harsh, but it is the best defense against error.
• An example of a nonreplicable result is this. In the 1960s and early 1970s, several researchers trained rats to do something,
chopped up the rats’ brains, extracted certain chemicals, and injected those chemicals into untrained animals. The recipients
then apparently remembered what the first group of rats had learned to do. From what we know of brain functioning,
Theoretically this procedure shouldn’t work, but if it did, imagine the possibilities. Some people proposed, semiseriously,
that someday you could get an injection of European history or introduction to calculus instead of going to class. alas, the
results were not replicable. When other researchers repeated the procedures, most of them found no effect from the brain
extracts (Smith, 1975).
• Psychological researchers have become increasingly concerned about the replicability of their findings.
QUESTION
• How does a meta-analysis relate to replicability?

ANSWER
• If some studies replicate an effect and others do not, a researcher may
conduct a meta-analysis that combines all studies as if they were one large
study.
 E VA LUATING S CIE NTIF IC THE ORIE S

• If replicable data support some hypothesis, eventually researchers propose a theory. a

scientific theory is more than a guess. It is an explanation or model that fits many
observations and makes accurate predictions.
• According to Feist and Feist (2009), theories are tools used to generate research and organize
observations, but neither “truth” nor “fact” has a place in a scientific terminology. It is a set of
related assumptions that allows scientists to use logical deductive reasoning to formulate
testable hypotheses
• A good theory starts with as few assumptions as possible and leads to many correct
predictions. in that way, it reduces the amount of information we must remember. The
periodic table in chemistry is an excellent example: from the information about the elements,
we can predict the properties of an enormous number of compounds. Five points are worth
noting:
 E VA LUATING S CIE NTIF IC THE ORIE S
1. First, a theory is a set of assumptions. A single assumption can never fill all the requirements of an adequate theory. A
single assumption, for example, could not serve to integrate several observations, something a useful theory should do.
2. Second, a theory is a set of related assumptions. Isolated assumptions can neither generate meaningful hypotheses nor
possess internal consistency—two criteria of a useful theory.
3. A third key word in the definition is assumptions. The components of a theory are not proven facts in the sense that their
validity has been absolutely established. They are, however, accepted as if they were true. This is a practical step, taken so
that scientists can conduct useful research, the results of which continue to build and re- shape the original theory.
4. Fourth, logical deductive reasoning is used by the researcher to formulate hypotheses. The tenets of a theory must be
stated with sufficient precision and logical consistency to permit scientists to deduce clearly stated hypotheses. The
hypotheses are not components of the theory, but flow from it. It is the job of an imaginative scientist to begin with the
general theory and, through deductive reasoning, arrive at a particular hypothesis that can be tested. If the general
theoretical propositions are illogical, they remain sterile and incapable of generating hypotheses. Moreover, if a researcher
uses faulty logic in deducing hypotheses, the resulting research will be meaningless and will make no contribution to the
ongoing process of theory construction.
5. Fifth, the definition includes the qualifier testable. Unless a hypothesis can be tested in some way, it is worthless. The
hypothesis need not be tested im- mediately, but it must suggest the possibility that scientists in the future might develop the
necessary means to test it.
•
 WHAT MAKES A THEORY USEFUL?

• First, a theory generates a number of hypotheses that can be investigated through research, thus
yielding research data. These data flow back into the theory and restructure it. From this newly
contoured theory, scientists can extract other hypotheses, leading to more research and
additional data, which in turn reshape and enlarge the theory even more. This cyclic
relationship continues for as long as the theory proves useful.
• Second, a useful theory organizes research data into a meaningful structure and provides an
explanation for the results of scientific research. This relationship between theory and research
data. When a theory is no longer able to generate additional research or to explain related
research data, it loses its usefulness and is set aside in favor of a more useful one.
• In addition to sparking research and explaining research data, a useful theory must lend itself to
confirmation or disconfirmation, provide the practitioner with a guide to action, be consistent
with itself, and be as simple as possible. Therefore, we have evaluated each of the theories
presented in this book on the basis of six criteria: A useful theory (1) generates research, (2) is
falsifiable, (3) organizes data, (4) guides action, (5) is internally consistent, and (6) is
parsimonious (Feist & Feist, 2009).
1. Generates Research
• The most important criterion of a useful theory is its ability to stimulate and guide further research. Without an adequate
theory to point the way, many of science’s present empirical findings would have remained undiscovered. In astronomy,
for example, the planet Neptune was discovered because the theory of motion generated the hypothesis that the irregularity
in the path of Uranus must be caused by the presence of another planet. Useful theory provided astronomers with a road
map that guided their search for and discovery of the new planet.
• A useful theory will generate two different kinds of research: descriptive re- search and hypothesis testing. Descriptive
research, which can expand an existing theory, is concerned with the measurement, labeling, and categorization of the
units employed in theory building. Descriptive research has a symbiotic relationship with
2. Is Falsifiable
• A theory must also be evaluated on its ability to be confirmed or disconfirmed; that is, it must be falsifiable. To be
falsifiable, a theory must be precise enough to suggest research that may either support or fail to support its major tenets. If
a theory is so vague and nebulous that both positive and negative research results can be interpreted as support, then that
theory is not falsifiable and ceases to be useful.
• Falsifiability, however, is not the same as false; it simply means that negative research results will refute the theory and
force the theorist to either discard it or modify it.
• A falsifiable theory is accountable to experimental results. Science is distinguished from nonscience by its ability to reject
ideas that are not supported empirically even though they seem logical and rational. For example, Aristotle used logic to
argue that lighter bodies fall at slower rates than heavier bodies. Although his argument may have agreed with “common
sense,” it had one problem: It was empirically wrong (Feist & Feist, 2009).
 PROVE IT
• For the claim “every object falls,” we cannot expect anyone to demonstrate it to be true for
every object, and so the burden of proof is on someone who doubts the claim. (We continue to
believe the statement unless someone shows an exception.) for a claim such as “UFOs from
outer space have visited earth” or “some people have psychic powers to perceive things
without any sensory information,” the burden of proof is on anyone who supports these
statements. If they are true, someone should be able to show clear evidence.
• Instead of insisting that all research is an effort to falsify a theory, another approach is to
discuss burden of proof, the obligation to present evidence to support one’s claim. in a
criminal trial, the burden of proof is on the prosecution. if the prosecution does not make a
convincing case, the defendant goes free. The reason is that the prosecution should be able to
find convincing evidence if someone is guilty, but in many cases innocent defendants could not
possibly demonstrate their innocence.
3. Organizes Data
• A useful theory should also be able to organize those research data that are not
in- compatible with each other. Without some organization or classification,
research findings would remain isolated and meaningless. Unless data are
organized into some intelligible framework, scientists are left with no clear
direction to follow in the pursuit of further knowledge. They cannot ask
intelligent questions without a theoretical framework that organizes their
information. Without intelligent questions, further research is severely curtailed.
• A useful theory of personality must be capable of integrating what is currently
known about human behavior and personality development. It must be able to
shape as many bits of information as possible into a meaningful arrangement. If
a personality theory does not offer a reasonable explanation of at least some
kinds of behavior, it ceases to be useful (Feist & Feist, 2009, p. 10).
4. Guides Action
• A fourth criterion of a useful theory is its ability to guide the practitioner over the rough course of day-to-day
problems. For example, parents, teachers, business man- agers, and psychotherapists are confronted
continually with an avalanche of questions for which they try to find workable answers. Good theory provides
a structure for finding many of those answers. Without a useful theory, practitioners would stumble in the
darkness of trial and error techniques; with a sound theoretical orientation, they can discern a suitable course
of action.
• For the Freudian psychoanalyst and the Rogerian counselor, answers to the same question would be very
different.
• To the question “How can I best treat this patient?” the psychoanalytic therapist might answer along these
lines: If psychoneuroses are caused by childhood sexual conflicts that have become unconscious, then I can
help this patient best by delving into these repressions and allowing the patient to re- live the experiences in
the absence of conflict. To the same question, the Rogerian therapist might answer: If, in order to grow
psychologically, people need empathy, unconditional positive regard, and a relationship with a congruent
therapist, then I can best help this client by providing an accepting, nonthreatening atmosphere. No- tice that
both therapists constructed their answers in an if-then framework, even though the two answers call for very
different courses of action (Feist & Feist, 2009, p. 10).
5. Is Internally Consistent
• A useful theory need not be consistent with other theories, but it must be consistent with itself. An internally consistent
theory is one whose components are logically compatible. Its limitations of scope are carefully defined and it does not
offer explanations that lie beyond that scope. Also, an internally consistent theory uses language in a consistent manner;
that is, it does not use the same term to mean two different things, nor does it use two separate terms to refer to the same
concept.
• A good theory will use concepts and terms that have been clearly and operationally defined. An operational definition is
one that defines units in terms of obervable events or behaviors that can be measured. For example, an extravert can be
operationally defined as any person who attains a predetermined score on a particular personality inventory (Feist &
Feist, 2009, p.11).
6. Is Parsimonious
• When two theories are equal in their ability to generate research, be falsified, give meaning to data, guide the practitioner,
and be self-consistent, the simpler one is preferred. This is the law of parsimony. In fact, of course, two theories are
never exactly equal in these other abilities, but in general, simple, straightforward theories are more useful than ones that
bog down under the weight of complicated concepts and esoteric language.
• In building a theory of personality, psychologists should begin on a limited scale and avoid sweeping generalizations that
attempt to explain all of human behavior. That course of action was followed by most of the theorists discussed in this
book. For example, Freud began with a theory based largely on hysterical neuroses and, over a period of years, gradually
expanded it to include more and more of the total personality (Feist & Feist, 2009, p.11).

•
Parsimony
What do we do if several theories fit the known facts? Suppose you notice that a picture on your
wall is hanging on an angle. you consider four explanations:
1. The ground shook when a big truck drove by.
2. A gust of wind moved the picture.
3. One of your friends bumped it without telling you.
4. A ghost moved it
 APPLYING PARSIMONY:
EXTRASENSORY PERCEPTION

• The possibility of extrasensory perception (ESP) has long been controversial in psychology.
Supporters of extrasensory perception claim that some people sometimes acquire information
without receiving any energy through any sense organ. Supporters claim that people with ESP can
identify someone else’s thoughts (telepathy) even from a great distance and despite barriers that
would block any known form of energy.
• Supporters also claim that certain people can perceive objects that are hidden from sight
(clairvoyance), predict the future (precognition), and influence such physical events as a roll of
dice by mental concentration (psychokinesis).
• Accepting any of these claims would require us not only to overhaul major concepts in psychology
but also to discard the most fundamental tenets of physics. What evidence is there for ESP?
 ANECDOTES

• Anecdotes are people’s reports of isolated events, such as a dream or hunch that comes true.
Such experiences often seem impressive, but they are not scientific evidence. Sooner or
later, occasional bizarre coincidences are almost sure to occur, and people tend to remember
them. at one point a company in North Carolina had two employees named Suresh c.
Srivastava. What are the odds against that? Well, this is the wrong question. The odds
against that particular coincidence may be high, but the chance of some strange coincidence
occurring is highly likely, given a long enough wait.
• Furthermore, we tend to remember, talk about, and sometimes exaggerate the hunches and
dreams that do come true and forget the ones that don’t. We could evaluate anecdotal
evidence only if people recorded their hunches and dreams before the predicted events.
• You may have heard of the “prophet Nostradamus,” a 16th-century French writer who
allegedly predicted many events of later centuries. After something happens, people
imaginatively reinterpret his writings to fit the event. (if we don’t know what a prediction
means until after it occurs, is it really a prediction?)
 SUMMARY OF SCIENTIFIC THINKING IN PSYCHOLOGY
What have we learned about science in general? Science does not deal with proof or certainty. all scientific
conclusions are tentative and are subject to revision. nevertheless, this tentative- ness does not imply a willingness
to abandon well-established theories without excellent reasons.
1. Steps in a scientific study. a scientific study goes through the following sequence of steps: hypothesis, method,
results, and interpretation. Because almost any study is subject to more than one possible interpretation, we
base our conclusions on a pattern of results from many studies.
2. Replicability. The results of a given study are taken seriously only if other investigators following the same
method obtain similar results.
3. Burden of proof. in any dispute, the side that should be capable of presenting clear evidence has the obligation
to do so.
4. Parsimony. all else being equal, scientists prefer the theory that relies on simpler assumptions, or assumptions consistent
with other theories that are already accepted.
5. Skepticism about extrasensory perception. psychologists care- fully scrutinize claims of extrasensory perception because the
evidence reported so far has been un-replicable and because the scientific approach includes a search for parsimonious
explanations.