Expert Witness Credibility as a Function of Eye Contact Behavior and Gender

Abstract

The effect of eye contact on credibility was examined via a 3 (low, medium, high eye contact) x 2 (male, female) between-groups design with 232 undergraduate participants. A trial transcript excerpt about a defendant’s recidivism likelihood was utilized as the experts’ script. A main effect was found: experts with high eye contact had higher credibility ratings than in the medium and low conditions. Although a confound precluded comparisons between the genders, results indicated that males with high eye contact were more credible than males with medium or low eye contact. The female experts’ credibility wasn’t significantly different regardless of eye contact. Eye contact may be especially important for males: male experts should maintain eye contact for maximum credibility. Copyright Sage Publications, 2008. DOI: 10.1177/0093854808325405

Full text

Neal, T.M.S. & Brodsky, S.L. (2008). Expert witness credibility as a function of eye contact behavior and
gender. Criminal Justice and Behavior, 35, 1515-1526. DOI: 10.1177/0093854808325405
Copyright Sage Publications, 2008.
Link to article: DOI: 10.1177/0093854808325405
Final version may have some differences from this pre-published version. This is not the copy of
record.

Expert Witness Credibility 2
Running head: Expert Witness Credibility and Eye Contact
Expert Witness Credibility as a Function of Eye Contact Behavior and Gender
Tess M.S. Neal and Stanley L. Brodsky
The University of Alabama

Expert Witness Credibility 3
Abstract
The effect of eye contact on credibility was examined via a 3 (low, medium, high eye contact) x
2 (male, female) between-groups design with 232 undergraduate participants. A trial transcript
excerpt about a defendant’s recidivism likelihood was utilized as the experts’ script. A main
effect was found: experts with high eye contact had higher credibility ratings than in the medium
and low conditions. Although a confound precluded comparisons between the genders, results
indicated that males with high eye contact were more credible than males with medium or low
eye contact. The female experts’ credibility wasn’t significantly different regardless of eye
contact. Eye contact may be especially important for males: male experts should maintain eye
contact for maximum credibility.

Expert Witness Credibility 4
Expert Witness Credibility as a Function of Eye Contact Behavior and Gender
During a trial, jurors listen to the testimony of various witnesses, interpret and make
sense of what they have heard, and subsequently make decisions that can affect people’s lives.
To do so, they must evaluate the witnesses from whom they hear testimony, and they may be
specifically instructed to “judge” credibility (Boccaccini, 2003).
Experts are often trained in nonverbal as well as verbal skills to effectively deliver
testimony (Boccaccini & Brodsky, 2002; Boccaccini, 2003). Research has shown that women
and men are perceived differently when their nonverbal behaviors are held constant (Aguinis &
Henle, 2001; Eagly, Makhijani, & Klonsky, 1992). Societal expectations for women and men
often shape the way people interpret their behaviors (Carli, LaFleur, & Loeber, 1995; Richmond
& McCroskey, 2004; Spangler, 1995). Men are expected to behave in more assertive and direct
ways than women (Aguinis & Henle, 2001). When women behave in stereotypically masculine
ways, (e.g., masculine styles of leadership), they are perceived negatively compared to other
women (Aguinis & Henle, 2001; Eagly & Karau, 2002; Shuller et al., 2001; Spangler, 1995).
This literature suggests that witness preparation should consider the differential perceptions of
male and female nonverbal behaviors and should tailor the preparation accordingly, so that
lawyers, trainers, and witnesses themselves may understand how their behaviors may be
interpreted.
Memon and Shuman (1998) stated that attorneys may use more male than female experts
because they believe that jurors regard the men as more credible. Others have concluded that
such gendered practices are consistent with the view that people often associate men with
positive traits not attributed to women (Eagly et al., 1992; Schuller, Terry, & McKimmie, 2005).

Expert Witness Credibility 5
Eye contact. The eyes are the most important of facial features in the human
communication process (Boccaccini, 2003; Richmond & McCrosky, 2004). Eye contact has been
found to be salient for perceptions of communicator credibility (Brooks, Church, & Fraser, 2001;
Kleinke, 1986), as well as a critical element of observing people on videotapes (Brooks et al.,
2001; Wheeler, Baron, Michell, & Ginsburg, 1979). Brodsky (1999) noted that law students are
taught to block eye contact between the witness and the jury during cross examination, but to
encourage their witness to make eye contact with the jury during direct examination. He
concluded that this is based on a sound principle: “we believe people who look us in the eye”
(143).
Research has shown that the frequency and duration of eye contact affects perceived
potency. Frequency of eye shifts and duration of eye contact were the only two nonverbal
behaviors found to be mediators of the perceptions of an individual’s intelligence during an
interview (Wheeler et al., 1979). A study by Brooks et al. in 2001 found that as eye contact
increases, people are perceived as more assertive, dominant, decisive, and aggressive. These
researchers randomly assigned participants to different conditions of duration of eye contact (5
vs. 30 vs. 50 seconds) and gender of model. Participants were asked to rate the model after
watching a 60 second videotape during which the model was viewed listening to directions given
by a speaker. Increasing eye contact from 5 to 30 seconds in a 60 second period yielded
significant changes in perceptions, but the changes tapered off after 30 seconds and there was not
a significant difference between the 30 and 50 second conditions. Only the male model was rated
as higher in leadership qualities as eye contact increased.
Additional research has shown that avoidance of eye contact negatively affects the
credibility ratings of witnesses (Hemsley & Doob, 1978). The simulated videotaped testimony of

Expert Witness Credibility 6
an alibi witness who either looked directly toward or slightly away from the lawyer while
testifying was viewed by participants. Results indicated that observers perceive a witness as
being less credible when he or she fails to look toward the target of his or her communication.
Eye Contact and Gender Implications. In 2001 Aguinis and Henle conducted a study in
which participants viewed a female employer interacting with a male employee. Eye contact was
manipulated to be either direct or indirect (looking directly at the employee vs. looking around
the room with an occasional glance at him). Direct eye contact by the female employer, as
compared with indirect eye contact, led to significantly higher ratings for a negative source of
power; specifically, coercive power. However, findings from the 1998 study by Aguinis,
Simonsen, and Pierce showed that when two male employees were interacting, direct eye contact
led to significantly higher credibility ratings than indirect eye contact.
The expectancy violation model as described in Bond, Omar, Pitre, Lashley, Skaggs, and
Kirk’s (1992) study explains how people judge when others are lying. The model suggests that
when there is a match between people’s nonverbal behavior and the observer’s expectations,
there is little need for the observers to find an explanation for the behavior. The authors describe
how violations of normative nonverbal expectations might raise suspicion and demand
explanation, which, in turn, might cause perceivers to scrutinize motivational factors. Aguinis
and Henle’s 2001 finding that women who maintained direct eye contact were perceived as
having a coercive source of power might be interpreted through the expectancy violation model.
The Present Study
Women and men are perceived differently when their nonverbal behaviors are equivalent
(Aguinis & Henle, 2001). Much of the research on the impact of differential perception of
nonverbal behaviors between the genders has been in human resources and employee-employer

Expert Witness Credibility 7
relations (Aguinis & Henle, 2001; Spangler, 1995). The present study examined the differential
perception of nonverbal behaviors on the witness stand when enacted by each gender and its
impact on juror decision making.
Eye contact has been found to be a salient nonverbal behavior that has a substantial
impact on receivers (Brooks et al., 2001; Hemsley & Doob, 1978; Wheeler et al., 1979). Eye
contact has been shown to be differentially perceived when men and women enact the same
looking-behaviors: only men have been perceived more positively as the amount of eye contact
increased (Aguinis & Henle, 2001; Brooks et al., 2001). The current study sought to investigate
the differential effect of eye contact behaviors for male and female expert witnesses on
credibility and impact on jurors’ decisions relative to the likelihood of future violence in a capital
case.
Hypotheses
1. It was hypothesized that women in the medium eye contact condition would rated as
more credible (and elicit higher percentages of substantive agreement) than women in
both the high and low eye contact conditions.
2. It was expected that credibility ratings (and agreement percentages) between the
medium and high eye contact conditions would not differ significantly for men, but that
both the medium and high eye contact conditions would have significantly higher
credibility ratings (and agreement percentages) than the low eye contact condition.
3. It was hypothesized that women would be rated as less credible than men in the high eye
contact condition. Likewise, percent of substantive agreement was hypothesized to be
lower for women than men in the high eye contact condition.

Expert Witness Credibility 8
Exploratory Analyses
There are no data to suggest a directional difference in credibility ratings between low
and high eye contact conditions for female expert witnesses. An exploratory analysis
(with no predicted direction) was conducted for these conditions.
Method
Pilot Study
Prior to conducting the primary study, pilot research was conducted to ensure successful
manipulations, to control for extraneous variables, and to evaluate participants’ understanding of
the study instructions and measures (N = 135). Results showed support for a) differential ratings
of eye contact between conditions, b) participants’ perception of the expert witnesses as
equitably credible and attractive, and c) participants’ understanding of the study materials. A
total of four witnesses (two men and two women) were filmed in each of the three manipulated
conditions of eye contact. Based on the results from the pilot study that indicated equitable
credibility and attractiveness ratings between the two female witnesses and between the two
male witnesses, one male and one female were dropped from further data collection. The
attractiveness and eye contact ratings were equitable between the male and female expert
witnesses utilized in the main study. Credibility ratings were not compared at this point, since
they were the primary dependent variable for the study.
Design
This study utilized a 3 x 2 factorial design, with 3 levels of eye contact (low vs. medium
vs. high) and 2 levels of expert witness gender (male vs. female). Eye contact and gender were
between-subjects manipulations and ratings of credibility and percent of substantive agreement
with testimony were the dependent variables.

Expert Witness Credibility 9
Participants
Participants consisted of 232 undergraduates at a large, southeastern university in the
United States who were at least 18 years of age. The mean age was 19.35 years old (SD = 2.57).
The sample was 63% female and 37% male, 77% percent of whom were Caucasian, 16% of
whom were African American, and 7% of whom were from other racial backgrounds. They were
recruited from Introductory Psychology courses as partial completion of research requirements.
At least 33 participants were assigned to each condition. This number was calculated via a power
analysis using an effect size of r = .23 (an average of the effect sizes of studies included in the
literature review for this study), a significance of p = .05, and a power of .80 (l-beta).
Stimuli
Based on Brooks, Church, and Fraser’s 2001 study that used conditions of 5, 30, and 50
seconds of eye contact in a 60 second period, we approximately replicated these ratios during a
five-minute video. The total measure of eye contact consisted of duration and frequency; that is,
the length of each contact (duration) between eye shifts (frequency) was totaled for the “amount
of eye contact.” Low eye contact for this study was operationally defined as eye contact lasting
for a total of thirty seconds out of the five minute video. Medium eye contact was defined as
totaling two and one-half minutes, and high eye contact was defined as four minutes. These
ratios were consistent with the ratios used in the Brooks et al. (2001) study (e.g., ~10%, ~50%,
and ~80%).
In each video, participants saw either a female or male expert witness on the witness
stand. Based on Boccaccini and Brodsky’s (2002) findings that most (62%) survey respondents
are likely to believe a defendant who makes eye contact with both the attorney and the jurors
while testifying, the expert witness’s total eye contact was measured when looking either at the

Expert Witness Credibility 10
attorney or the jurors for the total measure. Hemsley and Doob (1978) presented their witness as
looking “steadily in the direction of the lawyer (i.e., directly into the camera)” (p. 140). Brooks
et al., (2001) had the camera pointing over the shoulder of the speaker while pointing at the
model. They note that “one shoulder and the back of the head of the interviewer were clearly
visible, and it was obvious when the model was gazing at the face of the interviewer” (p. 72).
This study combined the two methods into one camera angle; the camera pointed at the witness
over the shoulder of the attorney so that the attorney’s right shoulder was visible on the very
bottom of the screen and part of the back of the attorney’s head was visible on the left side of the
screen. The witness’s face and chest were clearly visible in the middle of the screen, and the
witness box behind which the witness was seated was visible in the lower third of the screen. The
witness made eye contact with the attorney (i.e., eyes averted slightly to the left of the camera),
and also with the mock jurors via a subjective camera method (i.e., looking directly into the
camera). The lengths of time the witness made eye contact with the attorney, as well as eye
contact with the camera (i.e., mock jurors) were totaled for the three manipulated “eye contact”
conditions.
The expert witnesses were matched before filming on demographic variables, including
age, education, and ethnicity. The witnesses underwent training by the researchers to match tone
of voice and pace of speech for each condition, as well as coaching regarding the amount of eye
contact for each condition. Dacy and Brodsky (1992) conducted a study in which men and
women were presented in photographs with manipulated formality of clothing. Findings
indicated that more formal attire was positively associated with credibility. Thus, the men and
women filmed for this study were clothed similarly and professionally during filming to control
for this factor.

Expert Witness Credibility 11
Each of the experts was filmed for each of the three controlled conditions of eye contact,
using the same script in each condition. This script portraying a forensic expert witness testifying
about her or his evaluation of a convicted murder’s likelihood of committing future violent acts
was developed by Krauss and Sales (2001) and was adapted for this study. The expert testifies
that he/she believes the defendant will commit violent acts in the future and describes the
defendant as a psychopath. The expert is then effectively cross-examined in this script. Other
studies have used this script (e.g., Krauss & Lee, 2003).
Measures
Witness Credibility Scale. The Witness Credibility Scale (Griffin, Brodsky, Blackwood,
Abboud, & Flanagan, 2004) was used to assess the credibility of the expert witness. The scale
contains 20 bi-polar adjectives on a 10-point Likert scale, for a total 200 possible points. Higher
scores indicate greater credibility ratings. A few examples of these bi-polar adjectives include:
unkind (anchored at 1) to kind (anchored at 10); dishonest (1) to honest (10); and shaken (1) to
poised (10). Four separate, robust domains of trustworthiness, confidence, likeability, and
knowledge were found upon factor analysis (Griffin et al., 2004). Alpha coefficients have been
reported for each domain as follows: confidence (.88), likeability (.86), trustworthiness (.93),
knowledge (.86), and overall credibility (.95).
Sentencing ratings. Participants, serving as mock jurors, were asked to estimate a
percentage of probability that the defendant would commit criminal acts of violence in the
future. This estimate was used to assess the degree of substantive agreement with the testimony
of the expert witness by participants, as the expert testified about the predicted likelihood the
defendant would commit criminal acts of violence in the future.

Expert Witness Credibility 12
Shyness Scale. Cheek and Buss (1981) conducted a study in which they found that shy
participants averted their gaze more than non-shy participants. The Shyness Scale (Cheek &
Buss, 1981) was included in this study to assess the degree of the participants’ discomfort and
inhibition in the presence of others and with direct eye contact. This scale was used to assess the
impact of shyness and decreased eye contact preference in professional interactions. People who
score high on this scale (i.e., more shy) engage in less eye contact in dyadic interactions (Cheek
& Buss, 1981, as cited in Robinson et al., 1991).
Demographics. Participants completed a demographic questionnaire, inquiring about
such items as age, gender, ethnicity, socioeconomic status, and religious orientation.
Manipulation check. The following open-ended questions adapted from Hemsley and
Doob’s 1978 study were included as a manipulation check for eye contact: “(a) Was there any
particular piece of information that had the most influence on your decision regarding the
defendant? (b) Did you notice anything of significance in the manner in which the witness
behaved while giving his/her testimony? (c) Did you notice where the witness was looking while
he/she gave his/her testimony?” (p. 140). Then, a more specific question was included, which
asked the participant to estimate the percentage of time the expert maintained eye contact during
the testimony (0% -100%). In addition, a ten-point Likert-type scale was included to assess
perceptions of the expert witness’s attractiveness, with one being “not at all attractive” and ten
being “extremely attractive.” This measure was included to control for the biasing effect of
attractiveness. Hosoda, Stone-Romero, and Coats published a meta-analysis in 2003 in which
they calculated a weighted mean effect size of d = .37 for all studies included. The overall result
of the meta-analysis was that attractive people fared significantly better in terms of job-related

Expert Witness Credibility 13
outcomes than did unattractive people. Other research has shown that attractive people are
judged more favorably than unattractive people (Ritts, Patterson, & Tubbs, 1992).
Procedure
Upon arrival, participants were greeted and given a copy of the participant information
sheet to keep while the researcher read the sheet aloud and solicited questions. The participant
information sheet included information about the participant’s rights as research participants as
well as a brief introduction to the study. The brief introduction was as follows: “This study seeks
to investigate mock jurors’ perceptions of an expert witness who testifies about the likelihood of
the defendant re-offending in a short videotaped testimony. If you decide to be in this study, you
will be asked to watch a short mock court video of an expert witness testifying about a defendant
and then fill out some questionnaires about the video and about yourself.” Participants then
viewed a randomly assigned video condition. At the completion of the five-minute video,
participants were asked to complete a packet of questionnaires. The packet contained the Witness
Credibility Scale, a demographics questionnaire, and a manipulation check for the pilot study.
The primary study used these same three questionnaires and also included a sentencing rating
and the Shyness Scale. A debriefing document was given to each participant as they left, which
included contact information for the researchers. It also included the variables being investigated
and their suspected relation, as well as how the participants could obtain results once the analysis
and write-up were completed.
Results
Statistical Analyses
Before conducting any analyses, the parametric assumptions, including equal variance,
normal distribution, and independence were checked to ensure that the assumptions were not

Expert Witness Credibility 14
violated. They were not. Analysis of participant’s responses indicated that the manipulation was
successful; that is, participants perceived a significant difference in eye contact between the
conditions, F(1, 227) = 44.45, p < .001, ηp2 = .286. Examination of the LSD post-hoc test
revealed that the high condition had significantly higher percentage ratings of estimated eye
contact (M = 43.26%, SD = 26.91) than both the medium (M = 34.52%, SD = 27.07) condition, p
= 0.001, d = .32, and the low condition (M = 11.47%, SD = 13.58), p < 0.001, d = 1.49. In
addition, the medium condition had significantly higher percentage estimates than did the low
condition, p < 0.001, d = 1.08. These estimated percentages are interesting in that they reflect the
fact that the conditions are different, but also that the estimates for the medium and high
conditions are much lower than the actual percentages of eye contact for these conditions.
A multivariate analyses of variance (MANOVA) was conducted to assess the effect of
level of eye contact on ratings of credibility and substantive agreement with the expert’s
testimony. LSD post-hoc comparisons were conducted for any significant main effects. Although
the MANOVA originally included participant shyness, gender, and ethnicity as control variables,
none were found to contribute significantly to the overall model. The “best fit” model was a
MANOVA with the one independent and two dependent variables included.
It should be noted that the study was designed to include expert witness gender in the
model as one of the independent variables; however, only the total amount of time the experts
made eye contact with the attorney or with the camera was held constant across gender.
Unfortunately, the frequency of eye shifts was not held constant between the genders. This is a
potential confound in interpreting the results of a MANOVA with both independent variables
included. Thus, the analyses were conducted with eye contact condition as the only independent
variable.

Expert Witness Credibility 15
Primary Analyses
The test of the overall MANOVA with witness credibility as the dependent variable was
found to be significant, F(2, 194) = 6.12, p = .003, ηp2 = .06. Specifically, experts in the high eye
contact condition had significantly higher credibility ratings (M = 147.32, SD = 32.23) than
experts in the medium (M = 135.20, SD = 31.86) and low (M = 127.32, SD = 34.72) conditions, p
= .037, d = .38 and p = .001, d = .60, respectively. The credibility ratings between the medium
and low conditions did not significantly differ (p = .17) (see Table 1). The test of the overall
MANOVA with percent of substantive agreement with the expert’s testimony was not
significant, F(2, 194) = 0.35, p = .71 (see Table 2).
The first a priori hypothesis predicted that the medium eye contact condition for the
female would be rated as more credible than the high and low eye contact conditions. The data
from the female expert conditions were analyzed through an ANOVA with eye contact condition
as the independent variable and witness credibility as the dependent variable. The overall model
was not significant, indicating that the credibility ratings between the three eye contact
conditions for the female expert were not significantly different from one another, F(2, 97) =
6.34, p = .07 (see Table 1). Thus, the hypothesis was not supported.
It was expected that credibility ratings between the medium and high eye contact
conditions would not differ significantly for the male expert, but that both the high and medium
eye contact conditions would have significantly higher credibility ratings than the low eye
contact condition. Results indicated that at least two of the three conditions of eye contact were
significantly different for the male expert, F(2, 97) = 6.34, p = .003, ηp2 = .12. Both the high eye
contact condition (M=136.73, SD=33.10) and the medium eye contact condition (M=126.27,
SD=31.44) had significantly higher credibility rating than the low condition (M=108.85,

Expert Witness Credibility 16
SD=31.77), for the male expert, p = .001, d = .86. and p = .029, d = .55, respectively. There was
no significant difference in credibility ratings for the male expert in the medium and high eye
contact conditions (p = .20). These results support the hypothesis.
The third hypothesis predicted that in the high eye contact condition, men would be rated
as more credible than women. This hypothesis was not tested, since the potential confound of
frequency of eye shifts between the genders was identified. The first exploratory analysis to
examine the differences in credibility ratings between male and female expert witnesses in low
and medium eye contact conditions was also not tested due to the potential confound.
Discussion
Judgments of Credibility
The expectancy violation model (Bond, Oman, Pitre, Lashley, Skaggs, & Kirk, 1992)
may help explain many of the findings from this study. A main effect for eye contact condition
showed that experts in the high eye contact condition had significantly higher credibility ratings
than experts in the medium and low eye contact conditions. Additional specific hypotheses
regarding the simple effects of eye contact also were consistent with the expectancy violation
model. For instance, the male expert witness in the high and medium eye contact conditions had
significantly higher credibility ratings than in the low condition. These results suggest that eye
contact is especially important for male experts: male experts should maintain high levels of eye
contact for maximum credibility.
Other simple effects were less consistent with the expectancy violation model. It was
hypothesized that the woman in the high eye contact condition would be rated as less credible
than in the medium condition based on prior research. Previous studies have shown that when
women behave in stereotypically masculine ways, they are perceived negatively (Aguinis &

Expert Witness Credibility 17
Henle, 2001; Eagly & Karau, 2002; Schuller, Terry, & McKimmie, 2001; Spangler, 1995). These
women may face a trade-off between their organizational role and their gender role expectations,
where they may violate expectations one way or the other (Aguinis & Henle, 2001; Eagly et al.,
1992). Therefore, it was expected in this study that women would be perceived more positively
and would receive higher credibility ratings if they fulfilled normative nonverbal expectations of
eye contact (e.g., maintaining direct but not-to-powerful eye contact). The finding in this study
that the differences in credibility ratings between the three conditions of eye contact did not
reach statistical significance was unexpected. The finding that women who behaved in a more
assertive manner that may have violated societal expectations (e.g., maintained a high level of
eye contact) and the woman who fulfilled societal expectations (e.g., made non-assertive eye
contact in the low and medium conditions) were rated as having similar levels of credibility may
reflect a cultural trend toward accepting both assertive behaviors and passive behaviors from
women. Another key variable that may account for this trend is that expert witnesses may have
high perceived competence to begin with, and expectations regarding normative eye contact may
generalize to “experts” regardless of their gender or eye contact behavior.
Judgments of Sentencing Ratings
The sentencing rating was included to assess the degree to which participants agreed with
the expert witness’s testimony. It was expected that participants would be more in agreement
with experts who did not violate expectations of eye contact than experts who did violate the
normative expectation.
Results indicated that this prediction was not supported; the substantive agreement was
not lower for experts who violated expectations of eye contact. There was no main effect found
for eye contact condition, meaning that there were no significant differences between sentencing

Expert Witness Credibility 18
ratings regardless of the amount of eye contact the expert witness maintained. Examination of
the means and standard deviations in Table 2 reveals that participants consistently rated the
defendant as highly likely to commit future violence. The mean percentages ranged from a low
of 67.18% to a high of 78.67%. One explanation that may account for this finding is that
participants may have distorted perceptions regarding recidivism.
An alternative explanation may be that the sample was drawn from a conservative state in
the American South with more policies reflecting crime control than rehabilitation of offenders.
These participants may have a predisposition toward control-oriented policies and thus may have
a high threshold of estimated risk. A third possible explanation is that even though the expert
was less credible in the low conditions, he or she was still credible: the lowest mean credibility
rating in any condition was 108.85 out of 200. Even while avoiding eye contact the expert was
possibly perceived as credible enough for participants to take what they said into account in
making their judgments. One other possible explanation may be that the use of the word
“psychopath” in the stimulus materials may have had a strong influence on participant’s
judgments. Examination of the qualitative data for open-ended manipulation check questions
revealed that many participants said that the expert’s use of the word “psychopath” to describe
the defendant had an impact on their decisions.
Judgments of Eye Contact Behavior
Information regarding participants’ perception of the eye contact manipulation was
gathered as an informal manipulation check. Although participants perceived significant
differences in estimation of eye contact, the estimates were much lower than the actual amount
of eye contact for the medium and especially for the high condition. The actual percentage of eye
contact in the low condition was about 10%, and participant’s estimated the eye contact for this

Expert Witness Credibility 19
condition at 11.5%, a close estimate. However, the actual percentage of eye contact for the
medium condition was about 50%, but participants estimated 34.2%. The actual percentage of
eye contact in the high condition was about 80%, but participants estimated it at only 43.3%.
This finding may be explained in that low, medium, and high eye contact were defined based on
ratios reported in prior studies, ratios that may have elicited stronger obvious recognition of
nonnormative eye contact if they had been more salient.
General Discussion
There are several strengths of the current study. First, the eye contact manipulations were
successful and well developed. Second, there was good statistical power. Third, the study had
good internal and external validity. To increase internal validity, standardized scripts were
followed to ensure all participants were exposed to comparable conditions and procedures.
Several techniques were employed to enhance external validity. The expert witness was
portrayed by a forensic clinical psychologist who had experience testifying in court. The video-
taped conditions were set in a realistic witness stand and in the Witness Research Lab to provide
an authentic courtroom context.
There are also limitations to the current study. Although external validity was enhanced
by the previously mentioned methods, several important components of a capital trial were not
included. First, the dynamic of jury deliberation was not accounted for in the study. It is possible
that had the mock jurors been allowed to discuss their decision-making processes, the effect of
the eye contact manipulation may have been mitigated or strengthened. Also, had the study
defined low, medium, and high eye contact more saliently than ratios reported in prior research
(e.g., high condition at 100%), it may have elicited stronger results.

Expert Witness Credibility 20
Though it is possible that using a college student sample may have limited the
generalizability of these results, some research has suggested that concerns about the external
validity of jury research may be overstated. In a review of jury-simulation research, Bornstein
(1999) found that little research has obtained differences between different trial media or
different mock juror samples. He concluded that the use of students as mock jurors is not
necessarily a cause for concern. In a meta-analysis assessing the external validity of jury
simulation research, Dunn (2003) found that different methodologies accounted for only 1.7% of
the variance between ecological validity and external validity.
The primary limitation of the study was that, although the total amount of eye contact
was controlled, the quality of eye contact was not (e.g., the number of eye shifts). Specifically,
the female expert maintained steady gazes of eye-contact, alternating with steady gazes in the
distance for “no eye contact.” The frequency of her eye shifts was lower than that of the male
expert. The male expert’s eye shifts were more numerous, where his contact was brief but
frequent. When he looked away, he tended to look up, down, and around the room more
frequently and rapidly than the female expert did. It should be noted that there are conflicting
findings in the literature regarding perception of people with frequent eye contact shifts. Some
prior research has shown a negative relation; that as the number of eye contact shifts increased,
positive attributions decreased (Wheeler et al., 1979). However, other research has found the
opposite; that as the number of eye shifts increased, positive attributions also increased (Brooks
et al., 2001).
Future research requires inquiry into the quality of eye contact (e.g., frequency of eye
shifts, duration of eye contact) and how it relates to status, to gender, and to their interactions.
Research should also focus on the status of people being evaluated: are there different

Expert Witness Credibility 21
expectations of nonverbal behavior based on contexts? Are expert witnesses expected to
maintain a greater percentage of eye contact, with fewer eye shifts, than the general public?
Future studies should evaluate gender as well: are there different expectations of nonverbal
behavior based on gender? Prior research suggests this is the case; however, the current research
seems to refute some of the prior evidence. Has there been a cultural shift in expectations of
nonverbal behavior based on gender?
The interaction of status and gender should be further explored with regards to
expectations of eye contact behavior. Are women in professional roles expected to have different
eye contact behaviors than women in traditional roles? Are expectations for men’s eye contact
behavior the same regardless of their role? Finally, future research should explore the
effectiveness of steady gazing versus quicker glances in the context of expert testimony.

Expert Witness Credibility 22
References
Aguinis, H., & Henle, C. A. (2001). Effects of nonverbal behavior on perceptions of a
female employee’s power bases. Journal of Social Psychology, 141, 537-549.
Aguinis, H., Simonsen, M.M., & Pierce, C.A. (1998). Effects of nonverbal behavior on
perceptions of power bases. Journal of Social Psychology, 138, 455-469.
Brodsky, S. L. (1999). Power and control: Gaze and eye contact. In Testifying
in Court: Guidelines and Maxims for the Expert Witness (pp.143-146). Washington
D.C.: APA.
Boccaccini, M. T. (2003). Impact of witness preparation training on the use of targeted
testimony delivery skills, perceived credibility, and evaluations of guilt (Doctoral
dissertation, The University of Alabama, 2003). (UMI No. 3101735).
Boccaccini, M. T., & Brodsky, S. L. (2002). Believability of expert and lay witnesses:
Implications for trial consultation. Professional Psychology: Research and Practice, 33,
384-388.
Bond, C. F., Omar, D., Pitre, U., Lashley, B. R., Skaggs, L. M., & Kirk, C. T. (1992).
Fishy-looking liars: Deception judgment from expectancy violation. Journal of
Personality and Social Psychology, 63, 969-977.
Bornstein, B. H. (1999). The ecological validity of jury simulations: Is the jury still out?
Law and Human Behavior, 23, 75-91.
Brooks, C. I., Church, M. A., & Fraser, L. (2001). Effects of duration of eye contact on
judgments of personality characteristics. Journal of Social Psychology, 126, 71-78.
Carli, L. L., LaFleur, S. J., & Loeber, C. C. (1995). Nonverbal behavior, gender, and
influence. Journal of Personality and Social Psychology, 68, 1030-1041.

Expert Witness Credibility 23
Cheek, J. M., & Buss, A. H. (1981). Shyness and sociability. Journal of Personality and
Social Psychology, 41, 330-339.
Dacy, J. M., & Brodsky, S. L. (1992). Effects of therapist attire and gender.
Psychotherapy, 29, 486-490.
Dunn, K.F. (2003). Assessing the external validity of jury simulation research: A meta-
analysis. (Doctoral dissertation, University of Nebraska-Lincoln, 2003). Dissertation
Abstracts International, 63, 6141.
Eagly, A. H., & Karau, S. J. (2002). Role congruity theory of prejudice toward female
leaders. Psychological Review, 109, 573-598.
Eagly, A. H., Makhijani, M. G., & Klonsky, B. G. (1992). Gender and the evaluation of
leaders: A meta-analysis. Psychological Bulletin, 111, 3-22.
Griffin, M. P., Brodsky, S. L., Blackwood, H., Abboud, B., & Flanagan, C. B. (2004).
The Witness Credibility Scale. Unpublished Document. The University of Alabama.
Hemsley, G. D., & Doob, A. N. (1978). The effect of looking behavior on perceptions of
a communicator’s credibility. Journal of Applied Social Psychology, 8, 136-144.
Hosoda, M., Stone-Romero, E. F., & Coats, G. (2003). The effects of physical
attractiveness on job-related outcomes: A meta-analysis of experimental studies.
Personnel Psychology, 56, 431-462.
Kleinke, C. L. (1986). Gaze and eye contact: A research review. Psychological Bulletin,
100, 78-100.
Krauss, D. A., & Lee, D. H. (2003). Deliberating on dangerousness and death: Jurors' ability to
differentiate between expert actuarial and clinical predictions of dangerousness.
International Journal of Law & Psychiatry, 26, 113-137.

Expert Witness Credibility 24
Krauss, D. A., & Sales, B. D. (2001). The effects of clinical and scientific expert
testimony on juror decision making in capital sentencing. Psychology, Public Policy, &
Law, 7, 267-310.
Memon, A., & Shuman, D. W. (1998). Juror perception of experts in civil disputes: The
role of race and gender. Law & Psychology Review, 22, 179-197.
Richmond, V. P., & McCrosky, J. C. (2004). Nonverbal Behavior in Interpersonal
Relations, Fifth Edition. Boston: Pearson Education, Inc.
Ritts, V., Patterson, M. L., & Tubbs, M. E. (1992). Expectations, impressions, and
judgments of physically attractive students: A review. Review of Educational Research,
62, 413-426.
Robinson, J. P., Shaver, P. R., & Wrightsman, L. S. (Eds.) (1991). Measures of
Personality and Social Psychological Attitudes: Volume 1 in Measures of Social
Psychological Attitudes Series (pp. 182-184). San Diego, CA: Academic Press, Inc.
Schuller, R. A., Terry, D., & McKimmie, B. (2001). The impact of an expert’s gender
on jurors’ decisions. Law and Psychology Review, 25, 59-79.
Schuller, R. A., Terry, D., & McKimmie, B. (2005). The impact of expert testimony on
jurors’ decisions: Gender of the expert and testimony complexity. Journal of Applied
Social Psychology, 35, 1266-1280.
Spangler, L. (1995). Gender-specific nonverbal communication: Impact for speaker
effectiveness. Human Resource Development Quarterly, 6, 409-419.
Wheeler, W. R., Baron, J. C., Michell, S., & Ginsburg, H. J. (1979). Eye contact and the
perception of intelligence. Bulletin of the Psychonomic Society, 13, 101-102.

Expert Witness Credibility 25
Table 1
Means (and Standard Deviations) for Credibility Ratings by Experimental Condition
Expert Witness Gender
Expert
Witness
Eye Contact
Male
Female
Total
Low
108.85 (31.77)
145.79 (27.12)
127.32 (34.72)
Medium
126.27 (31.44)
144.12 (30.15)
135.20 (31.86)
High
136.73 (33.10)
158.25 (27.76)
147.32 (32.23)
Total
123.95 (33.82)
149.30 (28.78)
136.56 (33.81)

Expert Witness Credibility 26
Table 2
Means (and Standard Deviations) of Sentencing Ratings by Experimental Condition
(Percent agreement with the expert’s testimony that the defendant is likely to reoffend)
Expert Witness Gender
Expert
Witness
Eye Contact
Male
Female
Total
Low
67.18% (19.01)
78.67% (13.26)
72.92% (17.26)
Medium
68.21% (18.78)
74.70% (20.13)
71.45% (19.59)
High
72.39% (15.86)
75.69% (15.37)
74.02% (15.59)
Total
69.26% (17.90)
76.36% (16.43)
72.79% (17.51)