on altruism a good enough sample of all possible
situations to allow psychologists to make generaliza-
tions about it? Are observations of hospitalized
patients about to undergo surgery general enough
to suggest how these patients might react to other
stressful situations?
Suggestions for Improving Reliability
and Validity of ObservationsThe following suggestions are offered as ways to
improve the reliability and validity of observational
procedures. Like similar suggestions made for inter-
views (see Chapter 6), they often cannot be fully
implemented in clinical situations. Nevertheless, an
awareness of these points may help focus the clin-
ician’s attention in directions that will improve the
validity of observations.
- Decide on target behaviors that are both relevant
and comprehensive. Specify direct and observ-
able behaviors that can be defined objectively. - In specifying these behaviors, work as much as
possible from an explicit theoretical framework
that will help define the behaviors of interest. - Employ trained observers whose reliability has
been established and who are familiar with the
objective, standardized observational format to
be used. - Make sure that the observational format is
strictly specified, including the units of analysis,
the form in which observers’ratings will be
made, the exact observational procedures, the
scoring system, and the observational schedule
to be followed. - Be aware of such potential sources of error in
the observations as bias and fluctuations in
concentration. - Consider the possibility of reactivity on the part
of those being observed and the general influ-
ence of awareness that they are being observed. - Give careful consideration to issues of how
representative the observations really are and
how much one can generalize from them to
behavior in other settings.
Contemporary Trends in Data Acquisition
There are many ways in which technological
advances have begun to change the face of behav-
ioral assessment methods that involve observation.
First, the availability of laptop and handheld com-
puters/tablets/smartphones facilitates the coding of
observational data by assessors. Second, handheld
computers etc. can be assigned to clients so that
clients can provide real-time self-monitoring data.
One advantage of using handheld computers is that
they can be programmed to prompt clients to
respond to queries at specified times of the day or
night. Third, data from handheld computers can be
loaded onto other computers that have greater
processing and memory capacity so that observa-
tions can be aggregated, scored, and analyzed.
Finally, a variety of other technologies are being
used to gather data on patients while they are in
their natural environment. These include biosensors
(Haynes & Yoshioka, 2007), audiorecordings (Mehl,
Pennebaker, Crow, Dabbs, & Price, 2001), and both
GPS devices and activity monitors (Intille, 2007).
It is clear that behavioral assessors will continue to
capitalize on future technological advances.
The use of these devices for data collection in
individuals’natural environment is often calledeco-
logical momentary assessment(EMA; Stone & Shiff-
man, 1994; Shiffman, Stone, & Hufford, 2008),
an important new research tool that has come
into its own over the past 10 years. In some EMA
research, for example, subjects carryelectronic diaries
(EDs) with them throughout the day in their natu-
ral environments. We briefly discussed electronic
diaries above. The EDs are programmed to prompt
or“beep”participants at random or quasi-random
moments throughout each day. These prompts sig-
nal participants to complete a self-report assessment
battery programmed into the ED. Typically, assess-
ment items ask about momentary states and beha-
viors (e.g., “What is your mood right now?”).
Thus, EDs are used as experience sampling tools.
EMA has many conceptual advantages over tradi-
tional research designs when investigators are inter-
ested in characterizing dynamic, clinically important272 CHAPTER 9