Abnormal Psychology

188 CHAPTER 5

SUMMING UP

Summary of Using the

Scientifi c Method to

Understand Abnormality

Researchers use the scientifi c method to un-

derstand and study psychopathology. In doing

so, they observe relevant phenomena, identify

a question to be answered, develop a hypoth-

esis that might answer the question, collect

new observations to test the hypothesis, draw

on the evidence to formulate a theory, and test

the theory.

When conducting experiments, research-

ers systematically manipulate one or more

independent variables and observe possible

changes in one or more dependent variables.

Researchers evaluate the possible contribu-

tion of confounds by using control groups or

control conditions. To minimize unintentional

bias, they randomly assign participants to

groups. Experiments should have both inter-

nal and external validity. Researchers may use

a quasi-experimental design when random as-

signment is not ethical, desirable, or possible.

Correlational research is used when in-

dependent variables cannot or should not be

manipulated. Such studies allow researchers

to investigate the relationship between two

variables, specifically, whether a change in

one variable is associated with a change

in the other. However, a correlation does not

imply causation; it only indicates that the two

variables are related. Statistical significance

indicates that the obtained measurement is

greater than would be expected to occur by

chance alone. Longitudinal studies of psycho-

pathology often use correlational data.

Case studies allow a clinician or re-

searcher to examine one individual in detail.

Single-participant experiments (often using

an ABAB design) provide information about

how one variable affects another. However,

the results of case studies may not general-

ize to other people. Meta-analysis allows

researchers to aggregate the results of a num-

ber of studies that address the same question

in order to determine the relations among cer-

tain variables.

Thinking like a researcher

Dr. Xavier studies compulsive gambling; she

believes that people who begin gambling

compulsively in their teens and early 20s

have a different type of problem than those

who begin gambling compulsively in their 40s

or later. Based on what you have read about

research methods, how would you state her

question? What type of research design would

you suggest she use (experimental, quasi-

experimental, or correlational) and why? What

might be some confounding variables for

which she should try to control? How would

Dr. Xavier go about ascertaining whether the

study’s results were statistically signifi cant?

What types of information would you want to

If you wanted to test the effi cacy of your grief box therapy, you would again use

the scientifi c method, perhaps creating a treatment manual and organizing a RCT (in

accordance with the guidelines for experimental treatments and with IRB approval).

You might have four treatment groups: grief box only, encouraging social interactions

only, grief box plus social interactions, and CBT (which serves as a nonexperimental

comparison treatment). In this way, you could determine the relative effects of each

component of your therapy, alone and in combination, and see how they fare against

an established treatment. Depending on your results, you might refi ne your therapy.

Key Concepts and Facts About Researching Treatment

• When studying biomedical treatments, researchers should try
  to determine the extent to which a placebo effect—rather than
  a true drug effect—infl uences the results. Researchers should
  also take attrition rates into account when examining the effects
  of a treatment.


• Many of the challenges that arise in studying biomedical treat-
  ments, such as placebo effects and attrition, also arise when
  studying psychological treatments. A treatment may be effec-
  tive because of common factors, as well as because of specifi c
  factors unique to that treatment. Control groups and analogue
  studies can allow researchers to examine specifi c factors and to
  rule out possible effects of some confounding factors. In gen-
  eral, research has demonstrated that therapy is more helpful
  than no treatment. Although the Dodo bird verdict found no one
  form of therapy superior overall, some specific therapies are
  better for some particular disorders.


• Randomized clinical trials (RCTs) are designed to investigate
  the effi cacy of specifi c factors or treatments. RCTs typically in-

volve manual-based treatment; some RCTs may have limited

generalizability because of their exclusion criteria, homogenous

samples, and other factors. Results of well-designed and well-

conducted studies may indicate whether a particular treatment

is empirically supported for a specifi c disorder. The results of

research generally suggest a dose-response relationship. The

code of ethics for psychologists includes specifi c guidelines for

research on experimental treatments.

Researching Treatments That Target Social Factors

matching patients and therapists by ethnicity, gender, or age

does not systematically alter the effectiveness of therapy. For

patients with a strong preference, however, matching may lead

to a better outcome. Cultural forces infl uence whether a treat-

ment (including placebo) is effective.

• Any successful treatment, whatever type of neuropsychosocial
  factor it targets, also affects the other factors through feedback
  loops, inducing positive change. This is also true of changes
  that arise because of the placebo effect.

ll dd tt ll llididitit RR hh ii tt thth ll MM tt ll ii ll

f i h