Kassin, S.M., Goldstein, C.C., & Savitsky, K. (2003)
Behavioral confirmation in the interrogation room: On the dangers of presuming
guilt. Law and Human Behavior,9, 42-52.
Problem.
Using Inbau, Reid, Buckley, and Jayne’s (2001) training
manual as their basis, the present study works on the potential flaws in the
interrogation process. Kassin et al.
find two major potential flaws they would like to follow. First, there has been
no empirical evidence inferring that interrogators have any better ability to
distinguish between innocent suspects and guilty suspects denying their
crime. Second, because of the
interview/ interrogation paradigm, the investigators are predisposed to the
suspect and may make instinctual decisions (Imbau et al., 2001; Mortimer &
Shepherd, 1999).
The Study.
H1 – The prediction is that those
investigators given an expectation of guilt would allow that expectation to
influence their interrogation through increased-level aggressive interviews and
using more guilt-presumptive questions.
It is also predicted that this increase in aggression in the
investigators would cause an increase in the defensive behavior of the
suspects.
Phase I Procedure –
± ±
Participants were assigned status as either an investigator or suspect
into one of the four factorial cells (see table)
|
|
Interrogator Expectation |
||
|
Suspect Status |
|
Guilty |
Innocent |
|
Guilty |
GG |
GI |
|
|
Innocent |
IG |
II |
|
± ±
Interrogators received their expectation through base rate given by the
experimenter describing the suspect in the study: either 80% of the suspects
were said to be guilty (guilty expectation) or 20% of the subjects were said to
be guilty (innocent expectation).
± ±
Interrogators were given two goals to accomplish
o
o
Get a confession from the suspect
o
o
Determine the guilt or innocence of the suspect
± ±
Interrogators were given a packet containing the following
o
o
Excerpt from Inbau et al. (2001) that gave basics on interrogation
o
o
A list of 13 potential interrogation questions (some guilt-presumtive)
of which they were to chose 6 to use
o
o
A list of 13 interrogation techniques of which they were to chose 6 to
use
± ±
The interrogators were then given time to prepare a strategy.
± ±
Suspect in the guilt condition were told step by step how to carry out a
mock crime, which was then carried out.
Suspects in the innocent condition were told to knock on the door in the
area of the crime so as to familiarize the suspect with the area.
± ±
Suspects were told that their goal was to deny the crime and convince
interrogators that they were innocent.
± ±
Interrogations could last up to 10 minutes.
± ±
After interrogation, interrogators completed questionnaires with items
on the characteristics of the suspects and their own behavior. Suspect completed a parallel questionnaire.
Results –
± ±
The guilt–presumptive questions were chosen significantly more often by
the interrogators with guilty expectations (GE interrogators) than by
interrogators with innocent expectations.
± ±
On the techniques front, GE interrogators used significantly more
interrogation techniques at the outset (outset = first minute), but overall
there was no difference between the interrogator groups. There was, however, a difference in the
number of techniques used based on the guilt of the suspect, with a higher
technique usage rate by interrogators dealing with innocent suspects.
± ±
Interrogators rated themselves putting more pressure and trying harder
to get a confession from those suspects who were innocent.
± ±
Suspects who were innocent perceived their interrogators using more
pressure and trying harder to obtain a confession from the suspect.
H2 – GE Interrogators would cause behavior
in the suspects that independent observers would perceive as guilty behaviors.
Phase II Procedure –
± ±
Observers listened to tape of either interrogators only, suspects only, or
both. Each observer listened to one
tape from each of the four cells.
± ±
Observers were asked to rate the following after listening to each of
the tapes.
o
o
Guilt or innocence of victim and confidence in rating
o
o
Prediction of what guilt rating the interrogator had assigned to the
suspect and the observers confidence in that rating
o
o
Degree of presumed guilt observed in the interrogator
o
o
How hard the interrogator tried to get a confession and the amount of
pressure used
o
o
Anxiousness and defensiveness of the suspect, firmness of denial, and
believability of alibi
Results –
± ±
Guilty judgments of the observers occurred in 42% of the case with guilty
suspects, but they only occurred in 28% of the cases with innocent suspects.
± ±
GE interrogators were rated as guilt-biased 76% of the time by
observers, and IE interrogators were rated as guilt-biased 61% of the time, a
significant difference.
± ±
Observers also rated the GE Interrogators as more presumptive of guilt.
± ±
Observers rated interrogators as more presumptive of guilt when
questioning innocent suspects.
± ±
GE interrogators were seen as using more pressure on suspects.
± ±
Interrogators were observed as using more pressure on innocent suspects.
± ±
The previous two results were paralleled when assessing how hard the
interrogators tried to get a confession.
± ±
Suspects questioned by GE interrogators were observed to be more
defensive than those suspects interviewed by IE interrogators.
± ±
Observers found the stories of innocent suspects as more plausible than
the yarns of the guilty.
Discussion –
Kassin and
Co. found that the GE interrogators chose more guilt-presumptive questions from
the original list and they used more interrogation techniques. The GE interrogators were also more likely
to rate a suspect as guilty. It was
also found that suspects contributed to the guilt rating they received, as
suspects paired with GE interrogators were more defensive.
A very
interesting reaction was found when the suspect was innocent and paired with a
GE interrogator. Even though observers
rated the stories of the innocent suspects as more plausible, the interviews of
innocent suspects conducted by GE interrogators produced the highest-pressure
ratings by all groups. It appears to
Kassin and Co. that rather than believing the stories of the innocent suspects,
the GE interrogators increased their level of pressure based on their original
assumptions.
Observers
were also affected by the tomfoolery of the interrogators. While the observers rated GE interrogators
as more guilt-presuming than IE interrogators, the observers still tended to
judge more of the suspects guilty. This
is somewhat disturbing, as the observers should be able to make that
connection, a connection similar to those jurors have to make.
Kassin and
Co. conclude with a call for more research on this potent topic, and the
testing and education of the results to the proper law enforcement bodies.