EXPERIMENTAL RESEARCHentrenched in some areas. The experiment’s appeal
was its objective, unbiased, scientific approach to
studying mental and social life in an era when the
scientific study of social life was just gaining broad
public acceptance. Four trends sped the expansion
of experimental social research: the rise of behav-
iorism, the spread of quantification, the changes
in research participants, and the method’s practical
applications. Let us briefly consider each trend.1.Behaviorismis an approach in psycho-
logy founded by the American James B. Watson
(1878–1958) and expanded by B. F. Skinner
(1904–1990). It emphasizes creating precise mea-
sures of observable behavior or outward manifesta-
tions of inner mental life and advocates the
experiment to conduct rigorous empirical tests of
hypotheses.
2.Quantification,or measuring social phe-
nomena with numbers, expanded between 1900 and
- Researchers conceptualized social constructs
as quantified measures and jettisoned other non-
quantifiable constructs (e.g., spirit, consciousness,
will) from empirical research. An example is mea-
suring mental ability by using the IQ test. Originally
developed by Alfred Binet (1857–1911), a French-
man, researchers translated the test into English and
revised it by 1916. It soon had widespread use and
appeal as a way to represent something as subjec-
tive as a person’s mental ability with a single score
and became an objective, scientific way to rank
people. Between the years of 1921 and 1936, more
than 5,000 articles were published on intelligence
tests.^3 Many scaling and index techniques were
developed in this period, and social researchers
began to use applied statistics.
3. Over time, the people used as participants
changed. Early social research reports contained the
names of the specific individuals who participated
in a study, and most were professional researchers.
Later reports treated participants anonymously and
reported only the results of their actions. Over time,
there was a shift to use college students or school-
children as research participants. The relationship
between a researcher and the people studied became
more distant. Such distancing reflected a trend for
the experimenters to be more detached, remote, and
objective from the people under study. Researchers
saw reducing emotional engagement with research
participants in their studies as becoming more neu-
tral or value-free and truly “scientific” in a positivist
sense.- As researchers became aware of an
experiment’s practical applications, businesses,
governments, health care facilities, and schools in-
creasingly used experimental methods for applied
purposes. For example, the U.S. Army adopted in-
telligence tests during World War I to sort thousands
of soldiers into different military positions. The leader
of the “scientific management” movement in facto-
ries, Frederick W. Taylor (1856–1915), advocated
using experiments in factories. He worked with man-
agement to modify factory conditions as a way to in-
crease worker productivity. In the 1920s, educational
researchers conducted many experiments on teaching
methods andthe effect of class size on learning.
By the 1950s and 1960s, researchers became
more concerned with possible sources of alterna-
tive explanations, or confounding variables, that
might slip into experimental design. Researchers
designed experiments to reduce such potential
errors and increasingly used statistical procedures in
data analysis. A turning point in the increasingly rig-
orous design of social science experiments was a
book by Campbell and Stanley (1963), who defined
basic designs and issues in experimental methods.
By the 1970s, researchers increasingly evalu-
ated the methodological rigor of studies. A related
trend was the increased use of deception and a cor-
responding rise in concern about ethical issues. For
example, the now common practice of debriefing
did not come into use until the 1960s.^4 Over the last
three decades, the trend has been to use more
sophisticated experimental designs and statistical
techniques for data analysis.
Experiments and Theory
We conduct two types of social science experi-
ments: empirically based and theory-directed
(see Willer and Walker, 2007a, 2007b). The practi-
cal process of doing an experiment differs little, but
each type has different purposes. Most studies are
empirically based.