PSYCHOLOGICAL ASPECTS OF MAN’S ENVIRONMENT 1031
been extensively validated against certain criteria, such as
psychiatric diagnosis. However, the criteria themselves are
open to argument in many cases. Also the criteria used for
validation are seldom those of interest if a test is to mea-
sure temporary changes in emotional state due to changed
environment. The tests often appear likely to be insensitive
to the types of mood changes which may be of interest to
environmentalists. Projective personality tests appear, on the
surface, to be more sensitive for this purpose, for they permit
much freer response than objective tests, and they have been
used, for example, in studies of aggression. However, they
are exceedingly difficult to validate, even against the type
of criteria used for objective personality tests, and their pre-
dictive value for other types of behavior is almost entirely
unknown.
Comments on Psychological Methods In evaluating these
procedures for use in the environmental field, there are some
general considerations to keep in mind. First, the psycho-
physical procedures are generally reliable, if carefully done,
and generalization to other individuals from a small sample
is good. This work is often done with two or three observ-
ers, and the results are repeatable. The extreme sensitivity of
the sensory systems means that very precise control over the
stimulus is needed, which is often expensive. Certain stim-
uli, especially odors, are difficult to handle; special equip-
ment, such as odorometers, is necessary. Timing down into
the millisecond range is typically needed in visual and audi-
tory experiments.
Other types of behaviors tend to be more variable. Only
in simple conditioning situations are such small numbers of
experimental subjects likely to be found, and even here, for
most purposes, more than three or four subjects are neces-
sary. The requirements for stimulus control are somewhat
less rigid, in most cases. In verbal learning studies, timing
to .01 sec is usually adequate; less precision is often toler-
able. However, the substantial variability of organisms
presented with even relatively simple learning tasks, such
as mazes or simple discrimination learning, makes a larger
sample desirable. Individual experimental conditions may
be tested on 10–12 subjects, or more; it is not uncommon to
find verbal learning experiments which use several hundred
subjects. Generalization to other populations is sometimes
more of a problem and the equivalence of stimuli in different
experiments is more difficult to establish. However, within
these limitations, reliable results can be obtained in simple
conditioning situations and most other simple learning tasks,
including simple verbal learning tasks.
Both classical and operant conditioning procedures are
very sensitive to the action of certain agents. Effects of levels
of ionizing radiation have been reported as low as 5.0 r in con-
ditioning experiments. Behavioral effects of several drugs of
environmental importance, such as alcohol and amphetamines,
can be found for relatively low dosages. While the informa-
tion on the sensitivity of conditioning procedures to agents of
interest such as SO 2 and NO 2 is sparse, there is no reason for
believing that behavioral measures will not be equally appli-
cable to many of these substances. Behavioral studies done in
Russia emphasize the use of classical conditioning techniquesin studying various environmental factors. In such research
behavioral effects for low levels are frequent, though it is dif-
ficult to evaluate due to the obscurity of much of this literature
by American standards of reporting.
Problems arise when it is desired to interpret behavioral
measures in a more refined fashion than simply as indicators
of some physiological activity of the substances. Temporary
threshold shifts from noise exposure have been extensively
investigated, but their implications for permanent damage is
still being debated (see Noise). Similar problems exist for
other measures. The problems which have faced psycho-
pharmacologists in interpreting effects of drugs on animal
behavior in terms of drug effects on humans have their coun-
terpart in the environmental field; granted that a certain level
of ionizing radiation can serve as an aversive stimulus in
operant conditioning experiments on saccharin preference,
for example, how are the results to be interpreted in terms of
public health standards? How long lasting must such effect
be in order to be indicators of possible physical damage? If
no physical damage is to be expected, there is still the ques-
tion of the costs of possible functional impairment. Does
impairment of operant conditioning imply that the agent will
also impair more complex functions in humans, with a pos-
sible increase in errors or accidents as the result? At present,
answers are not available to these questions.
More complex tasks raise more serious problems of reli-
ability. It is not uncommon in educational research for several
experiments to yield contradictory results, for reasons which
are often obscure. The complexity of the materials and other
variations in subjects and procedures which are difficult to
control or characterize are probably responsible. Thus, a
single study done on the environmental effect of a substance
on complex learning should be viewed with caution. Only
when a series of experiments done under varying conditions
converge on a result should the result be accepted.
Certain special problems which often arise are the
effects of novelty (the so-called “Hawthorne” effect) and
habituation effects. Individuals may perform well under a
certain test procedure, which, when put into practice proves
no better than the alternative. The problem is that subjects
on whom new procedures are tried may make a special effort
which is not sustained under routine conditions. A similar
effect arises in regard to habituation. Individuals may
respond badly to a new environment, but adjust to it over
time without final decrement in performance. Since adapta-
tion for some tasks can take months, adaptation effects are
often inadequately tested.
Most research involving the possibility of exposure
to physically hazardous conditions is done on animals.
However, a good deal of research which involves physical
discomfort or emotional stress is done on humans. Also, in
some research, personal information of a private nature may
be obtained. Concern over possible abuses of subjects has
arisen in recent years, and several professional societies have
or are developing codes of ethics to deal with these prob-
lems. There are also legal aspects to the use of human and
animal subjects of which any potential research worker in
these fields should make himself aware.C016_012_r03.indd 1031C016_012_r03.indd 1031 11/18/2005 11:03:54 AM11/18/2005 11:03:54 AM