Response Measurement, Interobserver Agreement,
and Experimental Design
Happiness, as described by Green and Reid
(1996), was defined as any vocalization or facial
expression that could be considered typical of
happiness, which includes smiling and laughing.
Residents were scored as unhappy if yelling,
frowning, or crying was observed, and no affect
if neither condition was observed (Moore et al.,
2007). The effects of gambling on these
happiness measures were evaluated under four
conditions each day: baseline, simulated gambling,
10 min after gambling, and 30 min after
gambling. Levels of happiness during simulated
gambling and 10 min and 30 min after
gambling were compared to baseline in a fourcomponent
analysis (see Rapp, 2006). Each
participant was engaged in the simulated
gambling activity for 5-min, 10-min, and 20-
min intervals. Observers used 10-s partialinterval
recording to collect data.
Two independent observers recorded data for
33% of the sessions. Interobserver agreement
for happiness measures was determined by
dividing the number of agreements of emotional
affect by the number of agreements plus
disagreements and converting the ratio to a
percentage. Mean interobserver agreement was
99% (range, 96% to 100%).