Content uploaded by Adnan Fazlić
Author content
All content in this area was uploaded by Adnan Fazlić on Feb 05, 2025
Content may be subject to copyright.
NBP 2025, Vol. 30, Issue 2, pp.
https://doi.org/10.5937/nabepo30-54463
Original scientic paper
eISSN 2620-0406
Citation: Fazlić, A., & Deljkić, I. (2025). Face recall and face recognition:
Description accuracy does not indicate eyewitness identication accuracy. NBP.
Nauka, bezbednost, policija, 30(2), pp . https://doi.org/10.5937/nabepo30-54463
The Face Recall and Face Recognition: Description Accuracy
Does Not Indicate Eyewitness Identication Accuracy
Adnan Fazlić1, Irma Deljkić2
University of Sarajevo, Faculty of Criminal Justice and Security Studies
Bosnia and Herzegovina
Submitted: 2024-10-30 • Accepted: 2025-01-19 • Published: 2025-02-03
Abstract: Eyewitness testimonies are frequently considered crucial in criminal investigations.
e impact of the initial description provided by an eyewitness on subsequent identication ac-
curacy remains an understudied area. We employed a slightly modied version of Schooler and
Engstler-Schooler’s experimental approach to assess whether description quantity and quality are
related to lineup identication accuracy. e sample comprised 99 undergraduate students from
the University of Sarajevo. More than 50% of our participants made an accurate lineup identica-
tion, whereas they recalled and described only the general physical characteristics with minimal
detail. No signicant relationship between the quantity/quality of descriptions and identication
accuracy was observed in terms of the total number of words, total number of details in the de-
scription, and individual characteristics that participants mentioned in the description. In other
words, neither an extensive nor a detailed description eectively predicts lineup identication
accuracy. Finally, our ndings suggest that the recognition task is signicantly less challenging
for participants than the recall and description tasks because descriptions provide a general im-
pression of an individual and lack detailed information. e ndings of this study indicate that
a more detailed and extensive description does not serve as a reliable indicator of lineup identi-
cation accuracy.
Keywords: perpetrator description, lineup identication accuracy, eyewitness identication, eye-
witness testimony.
INTRODUCTION
Eyewitness testimony is frequently considered crucial in criminal investigations and may
occasionally be the only evidence for identifying perpetrators (Wells & Olson, 2003).
Eyewitness statements are commonly used as compelling evidence and historically re-
garded as the gold standard of proof (Ling et al., 2021; Albright & Garrett, 2022). Howev-
er, practitioners generally consider this evidence inaccurate and unreliable, particularly
1 Corresponding author: afazlic@n.unsa.ba • https://orcid.org/0000-0002-4419-1828 • Phone: +387 61
38 22 66
2 ideljkic@n.unsa.ba • https://orcid.org/0000-0001-5699-6095
ONLINE FIRST
NBP. Nauka, bezbednost, policija
2
NBP 2025, Vol. 30, Issue 2, pp.
for eyewitness identication (Wells, 2020). e importance of eyewitness testimony has
led to numerous studies showing that various factors can aect perpetrator description
accuracy (e.g. Berkowitz et al., 2020; Lockamyeir et al., 2020; Anakwah et al., 2020; Marr
et al., 2021; Baić et al., 2022), as well as eyewitness identication accuracy (Fazlić et al.,
2020; Bull Kovera & Evelo, 2020; Seale-Carlisle et al., 2022; Wixted & Wells, 2017). ese
ndings have resulted in best practices for conducting eyewitness interviews and line-
ups to maximise the reliability of eyewitness evidence (Wells et al., 2020). Despite these
advancements, the impact of the initial description provided by an eyewitness on subse-
quent identication accuracy remains an understudied area, warranting further research
(Fahsig et al., 2004).
Police ocers routinely collect detailed suspect descriptions from eyewitnesses, covering
attributes such as age, sex, race, height, physique, and distinctive features (Launay et al.,
2021). ese descriptions help to narrow the range of potential suspects and guide line-
ups (Mickes, 2016). However, the verbal overshadowing eect indicates that describing
a perpetrator can impair memory and reduce identication accuracy (Schooler & Engs-
tler-Schooler, 1990; Meissner & Brigham, 2001; Alogna et al., 2014). Some studies have
replicated this eect (e.g. Marmurek et al., 2021; Bacharach & Baker, 2024), while others
have not or found a reversal (Sauerland et al., 2008; Vredeveldt et al., 2015). Despite these
mixed ndings, researchers continue to explore ways to mitigate the potential negative
eects of verbal descriptions on eyewitness memory, while preserving their investigative
value.
Building on these conicting results, researchers have begun to apply alternative ap-
proaches to enquire about the inuence of verbal descriptions on eyewitness memory.
In analysing post-identication statements, two factors – description quality and amount
of detail – can predict identication accuracy with 70% identication accuracy (Short
& Dalby, 2007). Despite divergent ndings (e.g. Meissner et al., 2007; Demarchi & Py,
2009; Pozzulo et al., 2009; Sheahan et al., 2017; Areh & Walsh, 2020; Handler & Frühholz,
2021), research shows that quantity and quality of description do not consistently aect
identication accuracy. Also, Pozzulo et al. (2013) found that in this respect there are
no dierences between adults and adolescents (Generation Z). Inconsistent relationships
between descriptive characteristics and identication accuracy impede the assessment of
eyewitness reliability. e inuence of verbal descriptions on accuracy remains complex
and may depend on various factors.
is study explored the link between the quantity and quality of verbal descriptions and
eyewitness identication accuracy, with the understanding that recalling and recognising
human faces may involve dierent cognitive processes (Wells, 1984). As the rst of its kind
in Bosnia and Herzegovina, it aims to: (1) assess the quantity and quality of the description
and its relationship to the accuracy of the lineup identication; (2) explore whether the
eyewitness description features (quantity and quality of description) inuence their ability
to identify a suspect in a lineup accurately; (3) compare this study’s ndings globally, ex-
ploring cultural socialization’s impact on memory processes and generational dierences,
particularly the cognitive challenges digitalization poses to Generation Z.
ONLINE FIRST
NBP. Nauka, bezbednost, policija
3
NBP 2025, Vol. 30, Issue 2, pp.
METHODS
E
e present study employed a slightly modied experimental design from Schooler and
Engstler-Schooler (1990), which comprised a series of six experiments. eir ndings
demonstrated a detrimental eect of verbal description on subsequent identication of
a target face. is phenomenon is referred to as verbal overshadowing. More specically,
the present study replicates the experimental (face verbalisation) conditions of School-
er and Engstler-Schooler’s Experiment 2. Using a between-subjects experimental design
with counterbalancing, this study explored the impact of eyewitness description quality
and quantity (independent variables) on eyewitness identication accuracy (dependent
variable). is experimental design is used to assess whether the quantity and quality of
descriptions inuence the accuracy of eyewitness identication, whilst controlling for or-
der eects by counterbalancing the order of description and the ller task. It comprised
two experiments, each consisting of four steps. In Experiment 1, participants rst watched
a 44-second mock bank robbery video and then spent ve minutes describing the robber.
Next, they completed a 20-minute ller task (an easy crossword puzzle), and nally, they
identied the robber from a photo of eight people. Experiment 2 swapped the second
(ve-minute description) and third (20-minute ller task) steps. e study used the origi-
nal Schooler and Engstler-Schooler’s (1990) critical video and test photos, but the instruc-
tions and ller task were in the Bosnian language.
P
e sample comprised undergraduate students from the University of Sarajevo – UNSA
(N = 99; 57 – male, 42 – female), with a mean age of 19.14 years (AM = 19.14 years; SD =
.808) who participated individually but in groups of up to ten (for course credit). All the
participants were rst-year undergraduate students. Participants were randomly assigned
to two experiments: Experiment 1 (N = 49) and Experiment 2 (N = 50). e characteristics
of the two groups are presented in Table 1. e experimental groups were homogeneous in
terms of sex, year of study, and age. Informed consent was obtained from all participants.
Table 1. Characteristics of the Participants
Variable Category Tota l Experiment 1 Experiment 2
N % N % N %
Gender Male 57 57.6 26 53.1 32 62.0
Age
Female 43 42.4 23 46.9 19 38.0
M 19.14 18.94 19.34
SD .808 .659 .895
Min 18 18 18
Max 22 21 22
Note. M – mean; SD – standard deviation; Min – minimum value; Max – maximum value; N – number of respondents.
ONLINE FIRST
NBP. Nauka, bezbednost, policija
4
NBP 2025, Vol. 30, Issue 2, pp.
M P
e procedure for conducting Experiment 1 necessitated that the participants, before
commencement, be informed that the experiment comprised several tasks. Initially, they
were presented with a 44-second video depicting a simulated robbery. Subsequently, the
participants were requested to describe the appearance of the perpetrator and provide
their responses in written form on a specically designed document. A ve-minute du-
ration was allocated for the completion of these tasks. Aer three minutes, participants
in this experiment were instructed in all instances to continue describing to provide a
more comprehensive description. e second task commenced upon the expiration of the
time allocated to the aforementioned task. is task involved solving crossword puzzles
for which each participant received a copy of the crossword form. e duration allocated
for this task was 20 minutes. Upon completion of the previously mentioned task, the par-
ticipants performed the nal task, which involved identifying the person from the video
in a set of photographs. Accordingly, eight photographs (1–8) of potential perpetrators
from the video in the rst task were presented to the participants via a monitor or projec-
tor. Participants were instructed to circle one number from 1 to 8, corresponding to the
photograph that they believed depicted the perpetrator. Additionally, if any participant
believed that the perpetrator was not represented in any of the photographs, they had the
option of circling the number 9.
e procedure for Experiment 2 involved the same tasks and periods for their comple-
tion; however, this procedure included a permutation of the second (describing the per-
petrator) and third (solving the crosswords) tasks from Experiment 1. e nal phase of
Experiment 2’s procedure was identical to that outlined in the procedure for Experiment
1 and involved identifying the perpetrator from one of the eight presented photographs.
Attributes from the verbal descriptions were extracted and coded by a single researcher.
e coder was familiar with the coding procedure, study objectives, and specic criteria
for evaluating the quality and quantity of descriptions. Each description was coded ac-
cording to the established criteria (1 = correct, 2 = partly correct, 3 = incorrect) to ensure
consistency in the evaluation of the quantity and quality of the descriptions. To ensure
coding reliability and consistency, despite having one coder, several measures were imple-
mented. ese included recoding a subset of descriptions (approximately 10–20%) by the
same individual aer a time period to evaluate temporal consistency. e coder regularly
examined the rubric to maintain adherence to coding criteria and reduce potential bias.
S
Initial data entry into the database preserved the original form of the participants’ respons-
es. ey were then classied into various distinct groups based on the perpetrator’s char-
acteristics, including hair, forehead, ears, eyebrows, eyes, nose, facial hair, mouth shape,
jaw shape, clothes, gender, build, race, face shape, and teeth. Following this, each category
of individual description for which information was gathered was evaluated against the
actual description of the perpetrator.
Responses were categorised according to their level of accuracy in describing the oend-
er’s actual appearance. A correct description was assigned a code of 1, a partly correct de-
ONLINE FIRST
NBP. Nauka, bezbednost, policija
5
NBP 2025, Vol. 30, Issue 2, pp.
scription was given a code of 2, and an incorrect description was coded as 3. Participants
who failed to provide any description were deemed unable to characterise certain aspects
of the oender and were omitted from the analysis. In the photo identication lineup, the
correct position is 6.
Statistical analyses were performed using IBM SPSS Statistics (Version 21.0) (IBM Corp,
2012). Descriptive statistics for each variable/category was calculated, including the mean,
standard deviation (SD), minimum (Min), and maximum (Max). e analysis focused on
the descriptions provided by participants regarding the individual depicted in the video.
Specically, the term “words pertaining to facial features” refers to all words used to de-
scribe the facial appearance of the person shown, regardless of whether the description
was accurate. To assess the number of correct details, a exible approach was adopted. For
example, since the individual in the video had dark brown hair and eyes, any description
approximating the hair colour (e.g., “dark hair”, “dark brown hair”, “black hair”, “brown
hair”) or eye colour (e.g., “dark eyes”, “brown eyes”) was considered correct. In contrast,
incorrect descriptions were those that did not reasonably describe the person’s appearance
(e.g., “light hair”, “blonde hair”, “grey hair”, “blue eyes”, “green eyes”). is approach was
consistently applied when analysing the descriptions in both experiments.
An independent samples t-test was used to compare the mean values of the quantity of
descriptions between the groups of participants who made accurate and inaccurate lineup
identications. In this context, the mean values of the total number of words in the de-
scription, the total number of words in the description pertaining to the facial features,
the total number of correct details in the description, and the total number of incorrect
details in the description were considered. Additionally, a chi-square test of independence
was used to examine the relationship between the accuracy of the description and lineup
identication.
RESULTS
Descriptive statistical analysis was specically directed towards the quantication of de-
scriptions in terms of four key parameters: a) the number of words in the description, b)
the number of words pertaining to facial features, c) the number of accurate details, and d)
the number of inaccurate details. e results revealed that, on average, participants used
35.1 words to describe the individual, with 18.1 words, on average, dedicated to the facial
characteristics of the person. In terms of accuracy, the mean number of correct details
was 7.6, while the mean number of incorrect details was 10.5. ese ndings indicate that
approximately half of the words in the descriptions were related to facial characteristics.
Moreover, the data suggest that participants tended to include slightly more incorrect de-
tails than correct ones in their descriptions (see Table 2).
ONLINE FIRST
NBP. Nauka, bezbednost, policija
6
NBP 2025, Vol. 30, Issue 2, pp.
Table 2. Quantitative Analysis of the Description
Description quantity MSD Min Max
e number of words in the description 35.1 18.9 6 79
e number of words in the description
pertaining to the facial features 18.1 12.1 0 60
e number of accurate details 7.6 5.3 0 37
e number of inaccurate details 10.5 9.3 0 50
Note. M – mean; SD – standard deviation; Min – minimum value; Max – maximum value.
Furthermore, as part of the analysis of the descriptions obtained in the experiment, atten-
tion was also given to the facial features most frequently included in the descriptions by
participants from the experimental groups. e analysis revealed that participants most
oen described the following ve characteristics: gender, hair, facial hair, clothing, and
eyes (see Table 3).
Table 3. Description Accuracy
Variable
Frequency Accuracy (%)
N % Correct Partly
correct Incorrect
Hair 89 89.9 39.3 58.4 2.2
Forehead 6 6.1 66.7 33.3 0.0
Ears 3 3.0 0.0 33.3 66.7
Eyebrows 15 15.2 40.0 40.0 20.0
Eyes 41 41.4 58.5 31.7 9.8
Nose 14 14.1 14.3 50.0 35.7
Facial hair 83 83.8 50.6 42.2 7.2
Mouth shape 5 5.1 60.0 20.0 20.0
Jaw shape 2 2.0 50.0 0.0 50.0
Clothes 81 81.8 11.1 74.1 14.8
Gender 97 98.0 100.0 0.0
Build 12 12.1 58.3 8.3 33.3
Race 6 6.1 83.3 0.0 16.7
Face shape 9 9.1 11.1 44.4 44.4
Teeth 0 0
Note. N = number of descriptions including corresponding attribute.
ONLINE FIRST
NBP. Nauka, bezbednost, policija
7
NBP 2025, Vol. 30, Issue 2, pp.
An independent samples t-test was used to compare the total word count in descriptions
with the lineup identication accuracy. e dierence in word count between participants
who made accurate identications (M = 31.9; SD = 18.1) and those who made inaccurate
identications (M = 38.4; SD = 19.3) was not statistically signicant (t [97] = –1.72, p >
0.05). Similarly, no signicant dierence was found in the word count for facial charac-
teristics between the accurate (M = 17.7; SD =12.9) and inaccurate groups (M = 18.6;
SD = 11.3) (t [97] = –.37, p > 0.05). For the qualitative aspect of the verbal description, no
signicant dierence was found in the correct details (t [97] = –1.06, p > 0.05) between
those who accurately identied the target (M = 7.1; SD = 5.8) and those who did not
(M = 8.2; SD = 4.9). Additionally, there was no signicant dierence in incorrect details
(t [97] = .09, p > 0.05; correct identication: M = 10.6; SD = 9.9; incorrect identication:
M = 10.4; SD = 8.7). ese results indicate that the quantity and quality of verbal descrip-
tions did not dier signicantly between accurate and inaccurate identications.
e statistical relationship between the description and lineup identication accuracy was
subsequently examined. Specically, a chi-square test of independence was conducted to
ascertain whether correct, partly correct, or incorrect descriptions of certain characteris-
tics could serve as indicators of correct or incorrect identication in the lineup. e results
did not reveal a statistically signicant relationship (Table 4).
Table 4. Relationship Between the Description and Lineup Identication Accuracy
Described
feature
Description
accuracy
Identication accuracy
Test of dierences
Total
Accurate
N (%)
Inaccurate
N (%)
Hair
Correct 19 (21.3) 16 (18.0)
χ2(2) = 2.23, p = .33Partly correct 27 (30.3) 25 (28.1)
Incorrect 2 (2.2)
Forehead
Correct 2 (33.3) 2 (33.3)
χ2(1) = 0.94, p = .76Partly correct 2 (33.3)
Incorrect
Ears
Correct
a
Partly correct 1 (33.3)
Incorrect 2 (66.7)
Eyebrows
Correct 1 (6.7) 5 (33.3)
χ2(2) = 5.97, p = .05Partly correct 2 (13.3) 4 (26.7)
Incorrect 3 (20.0)
Eyes
Correct 11 (26.8) 13 (31.7)
χ2(2) = 1.65, p = .44Partly correct 5 (12.2) 8 (19.5)
Incorrect 3 (7.3) 1 (2.4)
ONLINE FIRST
NBP. Nauka, bezbednost, policija
8
NBP 2025, Vol. 30, Issue 2, pp.
Nose
Correct 1 (7.1) 1 (7.1)
χ2(2) = 0.58, p = .97Partly correct 3 (21.4) 4 (28.6)
Incorrect 2 (14.3) 3 (21.4)
Facial hair
Correct 18 (21.7) 24 (28.9)
χ2(2) = 1.48, p = .48Partly correct 19 (22.9) 16 (19.3)
Incorrect 2 (2.4) 4 (4.8)
Mouth shape
Correct 1 (20.0) 2 (40.0)
χ2(2) = 2.22, p = .33Partly correct 1 (20.0) 0 (0.0)
Incorrect 1 (20.0)
Jaw shape
Correct 1 (50.0)
χ2(1) = 0.00, p = 1.00Partly correct
Incorrect 1 (50.0)
Clothes
Correct 4 (4.9) 5 (6.2)
χ2(2) = 2.22, p = .33
Partly correct 29 (35.8) 31 (38.3)
Incorrect 8 (9.9) 4 (4.9)
Gender Correct 48 (49.5) 49 (50.5) a
Incorrect
Build
Correct 5 (41.7) 2 (16.7)
χ2(2) = 2.01, p = .37Partly correct 1 (8.3)
Incorrect 2 (16.7) 2 (16.7)
Race
Correct 1 (16.7) 4 (66.7)
χ2(1) = 0.00, p = 1.00Partly correct
Incorrect 1 (16.7)
Face shape
Correct 1 (11.1)
χ2(2) = 1.12, p = .57Partly correct 2 (22.2) 2 (22.2)
Incorrect 3 (33.3) 1 (11.1)
Note. N – number of respondents; a – the lack of variability in respondents’ answers precluded the possibility of
examining discrepancies in the frequency of answers in the identication accuracy.
No statistically signicant relationship between the description and lineup identication
accuracy was observed, even when the results were analysed through the lenses of Experi-
ments 1 and 2. e temporal interval between description and photo lineup identication,
as well as the delay between viewing the video and providing the description, did not
demonstrate a statistically signicant inuence on the outcomes.
ONLINE FIRST
NBP. Nauka, bezbednost, policija
9
NBP 2025, Vol. 30, Issue 2, pp.
DISCUSSION
e primary nding of the present study indicates that participants found the human
face recognition task to be easier than the recall and description tasks. is observation
supports the idea that face recall and face recognition are distinct cognitive processes
with diering demands. While recall requires the active generation and retrieval of facial
identity, recognition relies on a familiarity judgment of the currently viewed face (Grin
& Motta-Mena, 2021). ese results rearm that the human face is a highly informative,
complex, and non-verbal stimulus. Describing a face involves recalling detailed features
and translating them into words, a process that is cognitively demanding and constrained
by memory capacity. In contrast, human faces are processed holistically, meaning they are
perceived as unied wholes rather than as a collection of individual features. is holistic
processing enables the brain to quickly and subconsciously assess familiarity, making rec-
ognition easier. Unlike recall, recognition relies on matching a face to an existing mem-
ory, even when that memory lacks precise details. In the present study, just over 50% of
participants accurately identied the perpetrator in the lineup, while their descriptions
contained only general physical characteristics with minimal detail. Given this relatively
low descriptive detail and the fact that participants were describing an unfamiliar face, it
is reasonable to infer that they relied on relative judgment processing rather than absolute
processing during the lineup identication task (Wells, 1984). is suggests that recogni-
tion tasks engage dierent and less eortful cognitive processes compared to recall and
description tasks.
e perpetrator descriptions in our study primarily reect a general impression, with only
a small portion of the descriptions focusing on details specically related to the perpetra-
tor’s identity. is nding complements previous research (e.g., Fahsig et al., 2004; Pozzu-
lo et al., 2018). Despite a slightly increased word count in the descriptions, the additional
quantity did not result in higher-quality or more detailed descriptions. ese results are
consistent with prior studies showing no correlation between the quantity and quality of
descriptions (e.g., Demarchi & Py, 2009). Importantly, our ndings suggest that neither
the extensiveness nor the detail of a description signicantly inuences lineup identica-
tion accuracy. is lack of relationship between the quantity or quality of descriptions and
identication accuracy persists both in terms of the total number of words and the specic
characteristics mentioned in the descriptions. ese ndings may be explained by the
verbal overshadowing eect (Schooler & Engstler-Schooler, 1990), which occurs when de-
scribing a human face impairs subsequent face recognition, irrespective of the quality or
quantity of the description provided. Additionally, the attractiveness and distinctiveness
of a face may play a signicant role in face recognition. For example, Yamaguchi and Sug-
imori (2024) found that attractive and distinctive faces are remembered and recognized
more eectively than faces lacking these traits. In our study, participants did not comment
on the perceived attractiveness or distinctiveness of the perpetrator’s face. Future research
should incorporate these factors to better understand their impact on both facial descrip-
tions and identication accuracy.
e cultural background of witnesses plays a signicant role in shaping memory reports
(Anakwah et al., 2020). erefore, the ndings of our study were interpreted within the
cultural context of Bosnia and Herzegovina. While there is no consensus in the literature,
ONLINE FIRST
NBP. Nauka, bezbednost, policija
10
NBP 2025, Vol. 30, Issue 2, pp.
some studies suggest that Bosnia and Herzegovina retains the elements of collectivism, a
legacy of its socialist past in the former Yugoslavia, distinguishing it from predominantly
individualistic cultures in Europe (Klarin et al., 2012). Based on this, we hypothesized
that the results of our study might exhibit unique cultural characteristics. However, our
ndings did not dier signicantly from the studies conducted in individualistic cultures.
It is important to note that our participants were the members of Generation Z, and fu-
ture research should explore potential generational dierences, particularly between older
generations and Generation Z in Bosnia and Herzegovina.
e generational aspect was also considered in interpreting our ndings. Previous re-
search indicates that Generation Z, as digital natives, may possess specic cognitive traits
(Szymkowiak et al., 2021). We hypothesised that participants from this generation might
display unique patterns in their responses. However, our results align with those of Poz-
zulo et al. (2013), particularly regarding the average number of details in descriptions and
the characteristics most frequently mentioned (e.g., clothing, hair). Similarly, earlier stud-
ies involving older generations (e.g., Kuehn, 1974) reported comparable results, suggest-
ing that generational aliation may not signicantly inuence eyewitness descriptions.
is study has notable limitations that must be addressed. One limitation is the experi-
mental structure, particularly the shorter time delays compared to real-life scenarios. In
police investigations, eyewitness identication oen occurs aer substantial temporal de-
lays – measured in days, weeks, or even months – between the crime, the description, and
the lineup identication. In our study, all participants provided descriptions and complet-
ed identications within 25 minutes of viewing the video. While our analysis found no sig-
nicant eect of the time interval, these short delays do not reect real-world conditions,
leaving open the possibility that results might dier with longer, more realistic intervals.
Additionally, the method of data collection poses a limitation. Unlike police interviews,
our study did not involve questioning participants to elicit descriptions, which may have
inuenced the level of detail provided. Future research should address these limitations
by incorporating longer time delays and methods more closely aligned with actual police
practices. Such adjustments could oer deeper insights into the reliability and accuracy of
eyewitness descriptions and identications in real-world contexts.
CONCLUSIONS
is study provides valuable insights into the cognitive processes underlying face recall
and recognition and highlights their distinct demands. e ndings revealed that face
recognition tasks are signicantly easier than recall and description tasks, supporting the
notion that these are separate cognitive processes. Human faces, as complex non-verbal
stimuli, are processed holistically, enabling rapid familiarity assessments during recogni-
tion. Participants in the experiment appeared to rely on relative judgment processing dur-
ing lineup identication, engaging in less eortful cognitive strategies than those required
for recall and description tasks.
e study also found no signicant inuence of the quantity or quality of descriptions on
lineup identication accuracy. Although conducted in Bosnia and Herzegovina with Gen-
eration Z participants, the ndings did not dier signicantly from those of the studies
ONLINE FIRST
NBP. Nauka, bezbednost, policija
11
NBP 2025, Vol. 30, Issue 2, pp.
conducted in other cultures or older generations. is suggests that cultural and gener-
ational factors may have less of an impact on eyewitness descriptions and identications
than previously assumed.
ese ndings have signicant implications for law enforcement and the criminal justice
system, emphasising the complex nature of eyewitness testimony. ese results suggest
that caution should be exercised when using detailed witness descriptions as predictors of
lineup identication accuracy. Instead, such descriptions may be utilised more eectively
to narrow the pool of suspects and construct lineups. Understanding the limitations of
eyewitness accounts and their underlying cognitive processes is critical for developing
more eective eyewitness identication procedures. is approach has the potential to
enhance the accuracy of suspect identication and reduce the risk of wrongful convictions
resulting from unreliable eyewitness testimonies.
Despite the limitations mentioned earlier, this study advances our understanding of the
cognitive processes involved in face recognition, recall, and description, as well as their
implications for eyewitness testimony. Future research should explore factors such as face
attractiveness, distinctiveness, and potential cultural and generational dierences in larger
and more diverse samples. Such eorts will rene understanding of these complex pro-
cesses and enhance their practical applications in investigative contexts.
ACKNOWLEDGMENT
Authors declare that they have no conicts of interest.
REFERENCES
Albright, T. D., & Garrett, B. L. (2022). e law and science of eyewitness evidence. Boston
University Law Review, 102(2), 511–630. http://dx.doi.org/10.2139/ssrn.3675055
Alogna, V. K., Attaya, M. K., Aucoin, P., Bahník, Š., Birch, S., Birt, A. R., Bornstein, B. H.,
Bouwmeester, S., Brandimonte, M. A., Brown, C., Buswell, K., Carlson, C., Carlson, M., Chu,
S., Cislak, A., Colarusso, M., Collo, M. F., Dellapaolera, K. S., Delvenne, J.-F., … Zwaan, R.
A. (2014). Registered replication report: Schooler and Engstler-Schooler (1990). Perspec-
tives on Psychological Science, 9(5), 556–578. https://doi.org/10.1177/1745691614545653
Anakwah, N., Horselenberg, R., Hope, L., Amankwah-Poku, M., & van Koppen, P. J.
(2020). Cross-cultural dierences in eyewitness memory reports. Applied Cognitive Psy-
chology, 34(2), 504–515. https://doi.org/10.1002/acp.3637
Areh, I., & Walsh, D. (2020). Own-gender bias may aect eyewitness accuracy of perpe-
trators’ personal descriptions. Revija za kriminalistiko in kriminologijo, 71(4), 247–255.
http://dx.doi.org/10.6084/m9.gshare.13703806
Bacharach, V. R., & Baker, M. A. (2024). Verbal overshadowing and decision criterion
eects on recognition memory for faces. Journal of Cognitive Psychology, 36(8), 881–897.
https://doi.org/10.1080/20445911.2024.2419888
ONLINE FIRST
NBP. Nauka, bezbednost, policija
12
NBP 2025, Vol. 30, Issue 2, pp.
Baić, V., Oljača, M., & Tasić, M. (2022). e Inuence of the period of retention on the re-
liability of episodic memory in the context of testimony. NBP. Nauka, bezbednost, policija,
26(3), 7–24. https://doi.org/10.5937/nabepo26-35280
Berkowitz, S. R., Garrett, B. L., Fenn, K. M., & Lous, E. F. (2020). Convicting with con-
dence? Why we should not over-rely on eyewitness condence. Memory, 30(1), 10–15.
https://doi.org/10.1080/09658211.2020.1849308
Bull Kovera, M., & Evelo, A. J. (2021). Eyewitness identication in its social context. Journal
of Applied Research in Memory and Cognition, 10(3), 313–327. https://doi.org/10.1016/j.
jarmac.2021.04.003
Demarchi, S., & Py, J. (2009). A method to enhance person description: A eld study. In
R. Bull, T. Valentine, & T. Williamson (Eds.), Handbook of psychology of investigative in-
terviewing: Current developments and future directions (pp. 241–256). John Wiley & Sons.
Fahsig, I. A., Ask, K., & Granhag, P. A. (2004). e man behind the mask: Accuracy and
predictors of eyewitness oender descriptions. Journal of Applied Psychology, 89(4), 722–
729. https://doi.org/10.1037/0021-9010.89.4.722
Fazlić, A., Deljkić, I., & Bull, R. (2020). Gender eects regarding eyewitness identi-
cation performance. Criminal Justice Issues, 20(5), 31–41. https://doi.org/10.51235/
kt.2020.20.5.31
Grin, J. W., & Motta-Mena, N. V. (2021). Face recall. In T. K. Shackelford, & V. A.
Weekes-Shackelford (Eds.), Encyclopedia of evolutionary psychological science (pp. 2884–
2887). Springer. https://doi.org/10.1007/978-3-319-19650-3_3454
Handler, A., & Frühholz, S. (2021). Eyewitness memory for person identication: Pre-
dicting mugbook recognition accuracy according to person description abilities and sub-
jective condence of witnesses. Frontiers in Psychology, 12, Article 675956. https://doi.
org/10.3389/fpsyg.2021.675956
IBM Corp. (2012). IBM SPSS Statistics for Windows (Version 21.0) [Computer soware].
IBM Corp. https://www.ibm.com/products/spss-statistics
Klarin, M., Pororoković, A., Šašić, S. Š., & Arnaudova, V. (2012). Some characteristics of
social interactions among adolescents in Croatia, Bosnia and Herzegovina, and Macedo-
nia. Psychology Research and Behavior Management, 5, 163–172. https://doi.org/10.2147/
PRBM.S36389
Kuehn, L. L. (1974). Looking down a gun barrel: Person perception and violent crime. Per-
ceptual and Motor Skills, 39(3), 1159–1164. https://doi.org/10.2466/pms.1974.39.3.1159
Launay, C., Py, J., Brunel, M., & Demarchi, S. (2021). Beyond investigation-relevant in-
formation: A content analysis of police questioning. Police Practice and Research, 22(4),
1341–1355. https://doi.org/10.1080/15614263.2020.1869002
Ling, S., Kaplan, J., & Berryessa, C. M. (2021). e importance of forensic evidence for
decisions on criminal guilt. Science & Justice, 61(2), 142–149. https://doi.org/10.1016/j.
scijus.2020.11.004
Lockamyeir, R. F., Carlson, C. A., Jones, A. R., Carlson, M. A., & Weatherford, D. R. (2020).
e eect of viewing distance on empirical discriminability and the condence–accuracy
ONLINE FIRST
NBP. Nauka, bezbednost, policija
13
NBP 2025, Vol. 30, Issue 2, pp.
relationship for eyewitness identication. Applied Cognitive Psychology, 34(5), 1047–1060.
https://doi.org/10.1002/acp.3683
Marmurek, H. H. C., Rusyn, R., Zgardau, A., & Zgardau, A. M. (2021). Verbal overshad-
owing at an immediate Task-Test delay is independent of Video-Task delay. Journal of
Cognitive Psychology, 34(2), 243–248. https://doi.org/10.1080/20445911.2021.1981916
Marr, C., Otgaar, H., Sauerland, M., Quaedieg, C. W. E. M., & Hope, L. (2021). e eects
of stress on eyewitness memory: A survey of memory experts and laypeople. Memory &
Cognition, 49(3), 401–421. https://doi.org/10.3758/s13421-020-01115-4
Meissner, C. A., & Brigham, J. C. (2001). A meta-analysis of the verbal overshadowing
eect in face identication. Applied Cognitive Psychology, 15(6), 603–616. https://doi.
org/10.1002/acp.728
Meissner, C. A., Sporer, S. L., & Schooler, J. W. (2007). Person descriptions as eyewitness
evidence. In R. L. C. Lindsay, D. F. Ross, J. D. Read, & M. P. Toglia (Eds.), e handbook of
eyewitness psychology: Vol. 2. Memory for people (pp. 1–34). Routledge.
Mickes, L. (2016). e eects of verbal descriptions on eyewitness memory: Implications
for the real-world. Journal of Applied Research in Memory and Cognition, 5(3), 270–276.
https://doi.org/10.1016/j.jarmac.2016.07.003
Pozzulo, J., Bennell, C., & Forth, A. (2018). Forensic psychology (5th ed.). Pearson Canada.
Pozzulo, J. D., Dempsey, J. L., Crescini, C., & Lemieux, J. M. T. (2009). Examining the rela-
tion between eyewitness recall and recognition for children and adults. Psychology, Crime
& Law, 15(5), 409–424. https://doi.org/10.1080/10683160802279625
Pozzulo, J. D., Dempsey, J., & Pettalia, J. (2013). e Z generation: Examining perpetrator
descriptions and lineup identication procedures. Journal of Police and Criminal Psychol-
ogy, 28(1), 63–74. https://doi.org/10.1007/s11896-012-9107-5
Sauerland, M., Holub, F. E., & Sporer, S. L. (2008). Person descriptions and person identi-
cations: Verbal overshadowing or recognition criterion shi? European Journal of Cogni-
tive Psychology, 20(3), 497–528. https://doi.org/10.1080/09541440701728417
Seale-Carlisle, T. M., Grabman, J. H., & Dodson, C. S. (2022). e language of accurate
and inaccurate eyewitnesses. Journal of Experimental Psychology. General, 151(6), 1283–
1305. https://doi.org/10.1037/xge0001152
Schooler, J. W., & Engstler-Schooler, T. Y. (1990). Verbal overshadowing of visual mem-
ories: Some things are better le unsaid. Cognitive Psychology, 22(1), 36–71. https://doi.
org/10.1016/0010-0285(90)90003-M
Sheahan, C. L., Pica, E., Pozzulo, J. D., & Nastasa, C. (2017). Eyewitness recall and identi-
cation abilities of adolescent and young-adults. Journal of Applied Developmental Psychol-
ogy, 53, 86–95. https://doi.org/10.1016/j.appdev.2017.09.008
Short, J. L., & Dalby, J. T. (2007). Analysis of identication accuracy: Determining the ac-
curacy of eyewitness identications using statement analysis. Europe’s Journal of Psycholo-
gy, 3(3). https://doi.org/10.5964/ejop.v3i3.406
ONLINE FIRST
NBP. Nauka, bezbednost, policija
14
NBP 2025, Vol. 30, Issue 2, pp.
Szymkowiak, A., Melović, B., Dabić, M., Jeganathan, K., & Kundi, S. G. (2021). Infor-
mation technology and Gen Z: e role of teachers, the internet, and technology in the
education of young people. Technology in Society, 65(6), 101565. https://doi.org/10.1016/j.
techsoc.2021.101565
Vredeveldt, A., Tredoux, C. G., Kempen, K., & Nortje, A. (2015). Eye remember what hap-
pened: Eye-closure improves recall of events but not face recognition. Applied Cognitive
Psychology, 29(2), 169–180. https://doi.org/10.1002/acp.3092
Wells, G. L. (1984). e psychology of lineup identications. Journal of Applied Social Psy-
chology, 14(2), 89–103. https://doi.org/10.1111/j.1559-1816.1984.tb02223.x
Wells, G. L. (2020). Psychological science on eyewitness identication and its impact
on police practices and policies. American Psychologist, 75(9), 1316–1329. https://doi.
org/10.1037/amp0000749
Wells, G. L., Kovera, M. B., Douglass, A. B., Brewer, N., Meissner, C. A., & Wixted, J.
T. (2020). Policy and procedure recommendations for the collection and preservation of
eyewitness identication evidence. Law and Human Behavior, 44(1), 3–36. https://doi.
org/10.1037/lhb0000359
Wells, G. L., & Olson, E. A. (2003). Eyewitness testimony. Annual Review of Psychology,
54(1), 277–295. https://doi.org/10.1146/annurev.psych.54.101601.145028
Wixted, J. T., & Wells, G. L. (2017). e relationship between eyewitness condence and
identication accuracy: A new synthesis. Psychological Science in the Public Interest, 18(1),
10–65. https://doi.org/10.1177/1529100616686966
Yamaguchi, M., & Sugimori, E. (2024). Perceived facial attractiveness and distinctive-
ness aect face recognition. Psychologia, 66(2), 162–178. https://doi.org/10.2117/psys-
oc.2023-A224
ONLINE FIRST
