For instance, participant 19 had an average BSITER score of 89.22%,
with a confidence interval of +/- 5.84%. Therefore, we have 95%
certainty that their score falls between 83.38% – 95.06%. Because
this range overlaps with the 85% threshold for Episodic memory, we
cannot conclude with 95% confidence that Participant 19’s BSITER score
occurs above threshold, and in result we cannot be confident that they
are knowledgeable of crime-relevant information. Therefore, we suggest
that the most appropriate label for this participant is
“indeterminate”.
Note that our assessment differs from the traditional perspective where
a participant who scored 89.22% would typically be considered
“guilty/knowledgeable” when tested on Episodic, crime-relevant
information. However, we note that “indeterminate” labels may be more
acceptable for research purposes compared to practice. Indeed, there are
applications in diagnostic psychophysiology where an “indeterminate”
label is unacceptable (e.g., insurance claims, or investigative work)
and a decision must be rendered immediately with the (only) available
data. If a classification must be made, then it is important to
understand the probability of one decision state or another given the
subject’s data. The rBS method is also well-suited for this purpose.