Training Perceptual Experts: Feedback, Labels, and Contrasts
Rachel A. Searston and Jason M. Tangen
The University of Queensland
Are strategies for learning in education effective for learning in applied visual domains, such as
fingerprint identification? We compare the effect of practice with immediate corrective feedback
(feedback training), generating labels for features of matching and mismatching fingerprints (labels
training), and contrasting matching and mismatching fingerprints (contrast training). We benchmark
these strategies against a baseline of regular practice discriminating fingerprints. We found that all 3
training protocols—feedback, labels, and contrasts—resulted in a significantly greater ability to discrim-
inate new pairs of prints (independent of response bias) than the baseline training protocol. We also found
that feedback and labels training produced significantly lower rates of bias (i.e., learners in these groups
were less likely to overcall matches) compared with baseline training. Our results demonstrate 3 different
ways to boost expertise with matching prints, and have direct application to training perceptual expertise.
Keywords: perceptual expertise, categorisation, feedback, transfer, forensic science
There is a growing literature aimed at understanding the study
behaviours and metacognitive abilities (and illusions) of learners
(Bjork, Dunlosky, & Kornell, 2013; Dunlosky & Lipko, 2007)—as
well as a surge of research aimed at pinpointing the best and most
generalisable ways to practice (Bjork et al., 2013; Dunlosky,
Rawson, Marsh, Nathan, & Willingham, 2013; Hattie, 2009).
Several robust learning strategies—typically referred to as desir-
able difficulties— have been identified, which result in superior
performance during transfer and after a delay (e.g., interleaving,
retrieval practice, elaborative interrogation; Bjork & Bjork, 2011;
Schmidt & Bjork, 1992). Other work has focused on the role of
feedback (Butler, Karpicke, & Roediger, 2008; Hattie & Timper-
ley, 2007; Pashler, Cepeda, Wixted, & Rohrer, 2005). A large
portion of these studies, however, have relied on stimuli geared
toward education in schools or material intended to be memorised
(e.g., English-Swahili word pairs). Here, we extend these learning
strategies to the applied visual domain of fingerprint identification.
Fingerprint examiners typically refer to their expertise as being
based on “training and experience” with matching and mismatching
fingerprints (Busey & Parada, 2010); they spend their days determin-
ing whether a fingerprint collected at a crime scene belongs to the
same finger or different fingers as a candidate print. Their years of
training and experience has been the benchmark for courts to accept
fingerprint evidence. Remarkably, this benchmark has existed with
very little pressure to empirically demonstrate the quality and effec-
tiveness of current training programs to produce genuine expertise
(Tangen, Thompson, & McCarthy, 2011). Several prominent scien-
tific bodies have now encouraged the development of research pro-
grams on human observer performance in forensic examinations
(Campbell, 2011; National Research Council, Committee on Identi-
fying the Needs of the Forensic Science Community, 2009; National
Institute of Standards and Technology (NIST), 2012). In particular,
these reports highlight a need to establish improved, empirically
validated, and standardised training programs for forensic examiners.
We test the effect of three learning strategies for developing finger-
print expertise, comparing them to “individuation training” or practice
with matching prints without feedback, labels, or any additional
categorical information.
Feedback
Immediate feedback is often prescribed, alongside deliberate prac-
tice, as an effective training tool for developing expertise (Ericsson,
Krampe, & Tesch-Römer, 1993; Ericsson & Lehmann, 1996). In the
education and learning literatures, however, the specific role of im-
mediate feedback is less clear (Hattie & Timperley, 2007). Some
researchers have found learning interventions to be more helpful
when feedback is gradually reduced during practice (Wulf & Schmidt,
1989), when learners have a choice to skip feedback in favour of more
retrieval practice (Hays, Kornell, & Bjork, 2010), and when immedi-
ate feedback is provided only after responding incorrectly (Karpicke
& Roediger, 2007, 2008; Pashler et al., 2005), responding correctly
with little confidence (Butler et al., 2008), or responding correctly
with a delay (Kulhavy & Anderson, 1972; Smith & Kimball, 2010;
but see Metcalfe, Kornell, & Finn, 2009). Immediate feedback has
also been shown to have an undesirable effect on learners’ metacog-
nitive judgments (Kornell & Rhodes, 2013).
These studies show that the effect of feedback varies substantially
across conditions, but few studies have examined its use for devel-
oping visual expertise (see White, Kemp, Kemp, Jenkins, & Burton,
2014 for one exception), and we were unable to find any previous
studies that have directly compared the effect of feedback training
with other learning strategies in an applied visual domain. We com-
pared the effect of feedback training to two other training protocols (in
addition to our baseline, no feedback protocol) as a tool for learning
to distinguish fingerprints. White et al. (2014) found immediate cor-
rective feedback to be more effective for learning to distinguish
Rachel A. Searston and Jason M. Tangen, School of Psychology, The
University of Queensland.
Correspondence concerning this article should be addressed to Rachel A.
Searston, School of Psychology, The University of Queensland, St Lucia
QLD, Australia 4072. E-mail: rachel.searston@gmail.com
Canadian Journal of Experimental Psychology / Revue canadienne de psychologie expérimentale © 2017 Canadian Psychological Association
2017, Vol. 71, No. 1, 32–39 1196-1961/17/$12.00 http://dx.doi.org/10.1037/cep0000124
32