Proceedings of the International MultiConference of Engineers and Computer Scientists 2013 Vol II,

IMECS 2013, March 13 - 15, 2013, Hong Kong

Traffic Sign Comprehension: a Review of

Influential Factors and Future Directions for
Research
Tingru Zhang, Alan H.S. Chan


Abstract— Traffic signs are effective only when users clearly age, gender, driving experience, education background etc.
understand their meaning. This paper reviews recent studies might significantly influence comprehension level. More
concerning traffic sign comprehension and finds that recently, the effects of sign cognitive features including
unsatisfactory comprehension is a common problem for drivers
in many countries. The influence of user characteristics and sign
familiarity, concreteness, simplicity, meaningfulness and
cognitive features on understanding levels are summarized and semantic distance have been explored. This paper reviews
discussed. While the positive relation between comprehension research on the ability of drivers to understand traffic signs
level and educational background is confirmed by various and summarizes techniques used in evaluation. Work from
researchers, conclusions about effects of other factors are not previous studies on the effects of both user and sign
unanimous. Finally, possible future traffic sign comprehension characteristics are also analyzed and summarized. Some
research topics and directions for future research are proposed
to fill the present knowledge gap. possible factors and the future directions of research on this
topic are also proposed.
Index Terms—traffic sign comprehension, user characteristic,
sign cognitive feature II. COMPREHENSION LEVEL AND MEASUREMENT
Accurate understanding is a dominant factor for the
effectiveness of a traffic sign system. With poor
I. INTRODUCTION comprehension, memory for sign meanings and the likelihood

T RAFFIC signs use symbols with different shapes, colors

and sometimes with the assistance of words and labels to
regulate road traffic. They are the most commonly used
of prompting effective driver reactions will decrease [6].
Some unfamiliar and ambiguous signs may even give drivers
an impression that they are allowed to take certain actions
devices for controlling traffic [1]. Well-designed traffic signs although in fact it may be dangerous to do so [7]. Lack of
with clear symbols can communicate instructions quickly and comprehension or misunderstanding may reduce the intended
are useful to road users with reading difficulties [2]. benefits of traffic signs or, even worse, may actually cause a
According to Dewar, the most important factor for a traffic sign to become the main reason for an accident.
well-designed sign is the understandability of the sign [3]. According to ISO 3846 (2002), signs are considered
Pline pointed out that to be effective, signs should satisfy the acceptable when a level of at least 67% accuracy is obtained
following requirements: fulfill a need, command attention, in a comprehension test [8]. The American National Standard
convey a clear and simple meaning, command respect of the Institute (ANSI) is stricter and has a criterion of 85%
road users and give adequate time for proper response [4]. accuracy. Based on these criteria, studies have shown that
Among these requirements, the third one (convey a clear and lack of traffic sign comprehension is a serious problem in
simple meaning) emphasizes the need for signs to be many countries. For example, Ogden et al. interviewed 205
intelligible. So, for successful application of traffic signs they motorists in the US and found that the respondents had some
must be accurately comprehended by users. difficulty in interpreting both word and symbol messages on
Research concerning traffic sign comprehension dates back signs [9]. Another study showed that out of all the 85 standard
to 1966. Early studies focused on evaluating user traffic symbols used in the U.S., only 16 were understood by
understanding levels of local traffic signs and most of the more than 95% of motorists and the rest could not be
results indicated that the general comprehension performance comprehended well by most drivers [10]. Even worse, 10
was far from satisfactory [5]. Some research further signs were understood by less than 40% of motorists. Similar
investigated individual differences in performance on comprehension problems were found in Turkey, which is an
comprehension test and proposed that user characteristics like important transport hub between Europe and Asia [11]. A
survey found that of the 30 signs investigated, only 12 were
considered "well known" eliciting over 70% correct
Manuscript received December 10, 2012.
T. Zhang is with Department of Systems Engineering and Engineering
responses; 13 of the signs were either "partially known" or
Management, City University of Hong Kong, Hong Kong (phone: +852 produced "no comments". Surprisingly, there were five signs,
59832370; e-mail: [email protected]). for example, "Minimum speed limit", where more than 10%
Alan H.S. Chan is with Department of Systems Engineering and
of drivers interpreted the meaning as opposite to the intended
Engineering Management, City University of Hong Kong, Hong Kong
(e-mail: [email protected]). meaning. Such opposite interpretation may create situations

ISBN: 978-988-19252-6-8 IMECS 2013

ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
Proceedings of the International MultiConference of Engineers and Computer Scientists 2013 Vol II,
IMECS 2013, March 13 - 15, 2013, Hong Kong

more dangerous than not knowing the meaning of a sign, information, used in various previous research.
because such drivers may perform undesirable and prohibited
actions. Since a lot of research has shown that drivers exhibit III. USER CHARACTERISTICS
unsatisfactory comprehension of many signs, it seems The fact that comprehension level varies among
imperative that traffic administration institutes and/or individuals encouraged researchers to explore the relations
governments must take action such as reviewing sign design, between understanding and characteristics that are associated
training courses, and publicity and promotional events, in with the users themselves. Factors like age, gender, driving
order to strengthen driver comprehension of signs. experience, culture and education background have most
Both open-ended questions and multiple-choice questions often been analyzed, while marital status, monthly income,
(MCQs) can be used to measure driver comprehension of and last time driving have also been tested [11, 15, 16].
traffic signs. Open-ended would allow drivers to verbally However, results concerning the influence of user
describe the meaning of the sign and then several judges with characteristics have not been unanimous due to the use of
adequate sign knowledge should independently judge the different study sample populations, diverse experimental and
correctness of the answers. In some studies, responses were analytical methods or other various uncontrolled conditions.
classified into one of the four categories: correct and Nevertheless, the following sections will review studies on
complete, partially correct, incorrect or opposite meaning age, gender, driving experience, education and cultural
[11-13]. The conventional format of MCQs includes three background.
components: the stem, the correct answer, and several
incorrect but plausible answers or distractors [14]. The format A. Age
characteristics of a multiple-choice question such as quality of It has been generally found that age does not affect
the wrong answers, number of options or order of the correct comprehension of traffic signs and even where there is some
answer usually influence the difficulty and discrimination of impact, there is no consensus concerning the trend of the
MCQs. Compared with open-ended questions; the impact. Hawkins Jr et al. found that misunderstanding of over
multiple-choice question method can reduce time, cost and two thirds of the signs that they tested were not related to age
analysis effort, and help the respondents who have difficulty [17]. No significant differences among three age groups
to better communicate their ideas. Table I is a summary of (18-27 years; 28-37 years; 38-57 years) in comprehension
different testing methods, together with other experimental performance were observed in a survey done by Ng and Chan
TABLE I
SUMMARY OF TESTING METHODS USED IN TRAFFIC SIGN COMPREHENSION RESEARCH

Year Author Sample Size Country Tested Signs Equipment Testing Method

2012 Kirmizioglu 1478 drivers Turkey 39 signs with new and old Colored pictures Open-ended questions that can
version of the same signs on on questionnaires be answered either verbally or in
different questionnaires A and B writing. Answers were evaluated
by one person.
2012 Ou & Liu 30 Taiwanese Taiwan 203 traffic signs: not mentioned Verbally describe sign meanings
and 30 50 warning and answers were evaluated by
Vietnamese 75 prohibition three scorers
drivers 63 indicatory
15 auxiliary
2010 Ng & Chan 109 Hong Kong Hong 21 signs using only symbols Color and square Four-option multiple-choice
drivers Kong paper display questions
2008 Lesch 43 drivers U.S. 92 safety symbols not limited Software-controll True-false questions
to transportation ed display on
computer
2003 Shinar et al. 250 for each Canada, 31 highway signs with some Card display Verbally described sign
country Finland, unique to one country while meanings and answers were
Israel others used in more than one evaluated by experimenters
and country
Poland
2002 Al-Madani et al. 4774 drivers 5 Arab 28 posted signs: Colored pictures Multiple-choice questions on
counties 18 regulatory on questionnaires testing comprehension and
10 warning short-answer on personal
characteristics
1994 Dewar 480 drivers U.S. 85 traffic signs in US Projected color Open-ended questions
slides
1993 Hawkins Jr et al. 1745 drivers U.S. 38 signs: Videotape Multiple-choice questions
13 regulatory
18 warning
7 pavement markings
1990 Ogden et al. 205 drivers from U.S. Work zone signs Videotape Choice-based questions &
farm-to-work interviews
area
1988 Richards & 176 drivers U.S. Railroad grade crossing traffic Pictorial 16 multiple-choice questions
Heathington 35 police officers control devices (signs and
markings)
1979 Hulbert &Fowler 3164 motorists U.S. 16 traffic control signs, Motion picture Multiple choice questionnaires
signals and markings test film

ISBN: 978-988-19252-6-8 IMECS 2013

ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
Proceedings of the International MultiConference of Engineers and Computer Scientists 2013 Vol II,
IMECS 2013, March 13 - 15, 2013, Hong Kong

[12]. However, the "old" group defined in this study was negatively correlated with years licensed.
much younger than the "old" participants in other similar
C. Gender
research, so in order to have a deeper and better
understanding of the age effect, the performance of more Many previous studies on gender have shown that males
drivers of advanced age should be analyzed. were generally better than females with regard to traffic sign
Some studies have shown that young drivers have an comprehension. Al-Madani et al. found that male driver
advantage in understanding signs. For example, Dewar found comprehension of posted signs was better than that for
that approximately 39% of US traffic signs were better females in Arab as well as European and American countries
understood by young drivers both before and after [15]. In research reported by Hawkins Jr et al., men were
modifications to some of the symbols [10]. This was also true better in correctly identifying one fifth of warning signs [17].
in Saudi Arabia, where Al-Sharea found that younger drivers The causes of such phenomenon may be complex. There were
possessed much more traffic safety knowledge than older no explanations given by the authors. It may be due to the
drivers [18]. Possible reasons for the better performance of congenital difference in space perception or intellectual
younger drivers include: they may have passed the driving test functioning or different cultures and exposure rates to signs
relatively recently so they should have a fresh memory of the between males and females. For example, in some Middle
meanings of the signs; also, young people have generally Eastern countries females have very much more restricted
better information processing capabilities as well as better lives than males and in some places are not allowed to drive at
vision than old drivers [19]. all.
Opposing results concerning an age effect have also been When education background is controlled for, the
reported. Richards and Heathington conducted a survey of differences between males and females may become less
motorist comprehension of railroad grade crossing traffic remarkable. A study has demonstrated that when European
control devices (signs and signals) and found that both very and American male and female drivers with at least
young drivers (under 19 years) and elderly drivers (over 54 undergraduate degrees were compared, gender had no
years) had difficulty in understanding and recognizing such statistically significant impact [1]. It was found in a recent
devices [16]. Hulbert et al. studied the understanding that study by Ng and Chan, that for males and females of similar
over 3000 drivers had of eight traffic signs in the U.S. and education levels, there was no significant difference between
concluded that there were significant differences in them for traffic sign guessability [22].
comprehension among different age groups: comprehension D. Education background
level was 70% on average for young drivers (under 24 years), Several attempts have been made to explore possible
79% for middle-aged drivers (between 24 and 50 years old) associations between comprehension and educational level. A
and 72% for the old group (over 50 years) [20]. recent study by Al-Madani et al. found that drivers with
In summary, clear differences exist between research bachelor’s or higher degree had better understanding of post
results regarding age-influence on driver understanding signs than those with lower educational background [15]. A
capabilities of traffic signs probably because of the similar conclusion was reached by Ng and Chan [12]. In an
differences in types of signs used, study methods, samples, experiment investigating Hong Kong driver comprehension
definitions and limitations. of local traffic signs, Ng and Chan found that comprehension
B. Driving experience scores were higher for those with university education or
Before discussing the influence of driving experience, it is above than for those without a university education or above.
necessary to consider the ways in which it is measured This might be reasonably explained since in general people
because there are no commonly agreed criteria. Some studies with high education level are more likely to have had
have used a definition from Simpson, that driving experience experiences to encourage them to be quick learners, use their
can be taken as number of years licensed [21]. Al-Madani memories and have better information processing abilities.
reported that there was a small increasing trend of sign E. Culture background
comprehension with years of driving and those with at least Among some of the different nationality groups tested it
two decades of experience performed significantly better than has been found that drivers from Europe and America scored
those with at most five years of experience [7]. However, as highest while those from Arabian States performed worst [15].
there might be a high correlation between age and experience, This may be due to the higher educational qualifications of the
the effect of experience with age constraint involved was European drivers compared to others even when age and
further explored in a follow-up study by Al-Madani et al. [1]. education are controlled. Using 31 traffic signs from four
They only explored one age group and the results showed that countries (Canada, Finland, Israel and Poland), Shinar
experience had no significant influence when drivers were 45 conducted a cross-culture traffic sign comprehension study
or older. Ng and Chan proposed that actual years of active [23]. The sample consisted of 250 unpaid volunteers in each
driving and hours of driving in the 12 months prior to the country and they were tested on comprehension performance
study should be additional measurement for driving on both local and non-local traffic signs. As expected, drivers’
experience, as there were situations where licensed drivers performance on local signs was remarkably better than
rarely drove after obtaining a driving license [12]. Using the non-local signs: 77.9% and 32.4% were perfectly identified
three different indicators, they concluded that comprehension separately. ANOVA analysis also indicated that the main
level was not related to the factors of years of active driving effect of country was significant on comprehension
and hours of driving in past 12 months, but was found to be

ISBN: 978-988-19252-6-8 IMECS 2013

ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
Proceedings of the International MultiConference of Engineers and Computer Scientists 2013 Vol II,
IMECS 2013, March 13 - 15, 2013, Hong Kong

performance. Ou and Liu studied the comprehension of characteristics separately. However, possible interactions
Taiwanese traffic signs by two user groups: Taiwanese and may exist among those factors. This offers the research
the second biggest foreign group in Taiwan, the Vietnamese community a challenge to determine the interactive effects of
[13]. The results showed that Taiwanese had better factors such as age, driving experience or educational level
understanding compared to Vietnamese participants and a and examine their combined influence on the comprehension
possible reason, according to the authors, was the different abilities of drivers.
traffic sign design principles in different cultures. However, Currently, cross-country, or cross-border driving is
both groups showed similar performance immediately after a becoming more and more common as the world is becoming a
training programme, which partially demonstrated the global village due to increasing multinational co-operation
effectiveness of training. and communication (e.g. Hong Kong and Mainland China). In
most cases, drivers will not be trained to deal with foreign
IV. SIGN CHARACTERISTICS traffic system before driving in a new environment; therefore
Five characteristics were proposed by Mcdougall as it is likely that their driving activity may not fully comply with
contributing to icon usability: familiarity, concreteness, the instructions of traffic signs. Few attempts have been made
complexity, meaningfulness and semantic distance [24]. to investigate visiting driver comprehension of foreign signs.
Familiarity measures the frequency of user encounters with A survey comparing driver comprehension levels for local
the symbols. Concrete symbols depict objects, people and and non-local signs would no doubt help detect possible
materials that are familiar to people in the real world while problems in understanding foreign signs.
abstract symbols use simple lines, shapes, etc. to There is no doubt understandability varies among different
communicate information. Many studies have found that signs. Some signs can be easily understood without much
concrete symbols are better in representing information, as training while the others are hard to comprehend. Current
they appear to be more consistent with the obvious visual research has not clarified the key design factors that make a
images compared to abstract symbols [25, 26]. Complex signs sign easy to comprehend. So, the effects of sign design
contain a lot of details while simple signs do not. As features on comprehension needs to be investigated further.
extraneous decorative parts are likely to confound Also, very little research attention has been given to the
understanding of signs, it is suggested that drivers will possible influence of words on the understandability of signs.
perform better with simple signs than complex signs. Therefore, further studies focusing on the effect of words on
Meaningful signs can convey certain messages that make sign comprehension, whether or not words should be added to
users aware of what to do or what not to do. The more traffic signs, the best number of words used or the influence of
meaningful a sign is, the more likely that it will be understood bilingual words is necessary and would be beneficial.
correctly by users. Semantic distance is used to measure the
relatedness or closeness of a symbol and what it is intended to VI. CONCLUSION
represent. This paper attempted to give a systematic summary of
Prospective drivers in Hong Kong with no traveling influential user factors and sign characteristics on traffic sign
experience in Mainland China were asked to give subjective comprehension. Generally, drivers with higher educational
ratings on these five characteristics for 120 Mainland China background performed better in comprehension tests, but
traffic signs. The results showed that the Hong Kong there is no consensus on the influence of age, gender or
participants perceived the test signs as moderately unfamiliar, cultural background. Clearly, drivers understand familiar
concrete, simple, meaningful and semantically close [27]. signs better than those less frequently encountered. Further
Pearson correlation analysis was used to explore the investigation is required to clarify the effects of other sign
interrelationships between the sign features and it was found features. With increasing communications among countries,
that while concreteness had high and positive correlation with studies concerned with the interactions among user
meaningfulness, semantic closeness and familiarity; characteristics and the comprehension issues of cross-border
simplicity only correlated with familiarity. A similar drivers are necessary.
experiment was conducted by the same researchers to further
examine relationships between sign characteristics and REFERENCES
understandability using Hong Kong traffic signs [12]. [1] Al-Madani, H., Al-Janahi, and A. Rahman, Assessment of drivers'
Surprisingly, only familiarity showed a significantly positive comprehension of traffic signs based on their traffic, personal and
social characteristics. Transportation Research Part F: Traffic
relation with the testing score. This conclusion conflicts with Psychology and Behaviour, 2002. 5: p. 63-76.
the result from a more recent study. When studying the effects [2] Michael, S.W., J. Russell, and J.W. Brelsford, Comprehension and
of sign design features on Taiwanese and Vietnamese user retention of safety pictorials. Ergonomics, 1997. 40(5): p. 531-542.
comprehension of traffic signs, Ou and Liu concluded that all [3] Dewar, R., Criteria for the design and evaluation of traffic sign
symbols1988.
five characteristics were correlated with comprehension for [4] Pline, J.L., Traffic engineering handbook1992: Prentice-Hall.
both groups with only one exception: simplicity had no [5] Johansson, G. and K.R. RUMAR, Drivers and road signs: a
significant influence on the performance of the Taiwanese preliminary investigation of the capacity of car drivers to get
information from road signs. Ergonomics, 1966. 9(1): p. 57-62.
[13]. [6] Charlton, S.G., Conspicuity, memorability, comprehension, and
priming in road hazard warning signs. Accident; analysis and
V. FUTURE DIRECTIONS FOR RESEARCH prevention, 2006. 38: p. 496-506.

Most of the research to date analyzes the effects of user

ISBN: 978-988-19252-6-8 IMECS 2013

ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)
Proceedings of the International MultiConference of Engineers and Computer Scientists 2013 Vol II,
IMECS 2013, March 13 - 15, 2013, Hong Kong

[7] Al-Madani, H., Influence of drivers' comprehension of posted signs on

their safety related characteristics. Accident Analysis & Prevention,
2000. 32: p. 575-81.
[8] ISO, Graphical Symbols - Safety Colours and Safety Signs, in Part 1:
Design Principles for safety Signs in Workplaces and Public
Areas2002: Switzerland.
[9] Ogden, M., K. Womack, and J. Mounce, Motorist Comprehension of
signing applied in urban arterial work zones. Transportation Research
Record, 1990(1281).
[10] Dewar, R.E., D.W. Kline, and H. Swanson, Age differences in
comprehension of traffic sign symbols. Transportation Research
Record, 1994(1456).
[11] Kirmizioglu, E. and H. Tuydes-Yaman, Comprehensibility of traffic
signs among urban drivers in Turkey. Accident; analysis and
prevention, 2012. 45: p. 131-41.
[12] Ng, A.W.Y. and A.H.S. Chan, The effects of driver factors and sign
design features on the comprehensibility of traffic signs. Journal of
Safety Research, 2008. 39(3): p. 321-328.
[13] Ou, Y.-K. and Y.-C. Liu, Effects of sign design features and training on
comprehension of traffic signs in Taiwanese and Vietnamese user
groups. International Journal of Industrial Ergonomics, 2012. 42: p.
1-7.
[14] Considine, J., M. Botti, and S. Thomas, Design, format, validity and
reliability of multiple choice questions for use in nursing research and
education. Collegian: Journal of the Royal College of Nursing
Australia, 2005. 12(1): p. 19-24.
[15] Al-Madani, H., Al-Janahi, and A. Rahman, Role of drivers' personal
characteristics in understanding traffic sign symbols. Accident;
analysis and prevention, 2002. 34: p. 185-96.
[16] Richards, S.H. and K.W. Heathington, Motorist understanding of
railroad-highway grade crossing traffic control devices and associated
traffic laws1988.
[17] Hawkins Jr, H., K. Womack, and J. Mounce, Driver comprehension of
regulatory signs, warning signs, and pavement markings.
Transportation research record, 1993(1403).
[18] Al-Sharea. Behavior of the drivers and safety measures (in Arabic). in
Proc., 3rd IRF Middle East Conference. 1998. Saudi Arabia.
[19] Vallesi, A., et al., Age-related differences in processing irrelevant
information: evidence from event-related potentials.
Neuropsychologia, 2009. 47(2): p. 577-586.
[20] Hulbert, S., J. Beers, and P. Fowler, Motorist's understanding of traffic
control devices. 1979.
[21] Simpson, H.M., The evolution and effectiveness of graduated
licensing. Journal of Safety Research, 2003. 34(1): p. 25-34.
[22] Ng, A.W.Y. and A.H.S. Chan, The guessability of traffic signs: Effects
of prospective-user factors and sign design features. Accident Analysis
& Prevention, 2007. 39(6): p. 1245-1257.
[23] Shinar, D., et al., Traffic sign symbol comprehension: a cross-cultural
study. Ergonomics, 2003. 46: p. 1549-65.
[24] Mcdougall, S.J.P., M.B. Curry, and O. de Bruijn, Measuring symbol
and icon characteristics: Norms for concreteness, complexity,
meaningfulness, familiarity, and semantic distance for 239 symbols.
Behavior Research Methods, 1999. 31(3): p. 487-519.
[25] Davies, S., et al., Safety pictograms: are they getting the message
across? Applied ergonomics, 1998. 29(1): p. 15-23.
[26] Nakata, A., J.L. Campbell, and J.B. Richman. Driver acceptance of
general vs. specific icons for in-vehicle information. in Proceedings of
the Human Factors and Ergonomics Society Annual Meeting. 2002.
SAGE Publications.
[27] Ng, A.W.Y. and A.H.S. Chan, Cognitive Design Features on Traffic
Signs. Recent Advances in Engineering and Computer Science
2007(Lecture Notes in Engineering and Computer Science Number
62), 2007. 62: p. 97-102.

ISBN: 978-988-19252-6-8 IMECS 2013

ISSN: 2078-0958 (Print); ISSN: 2078-0966 (Online)