stop, look, listen, and think?: what young children really do when crossing the road

Download Stop, look, listen, and think?: What young children really do when crossing the road

Post on 02-Jul-2016




1 download

Embed Size (px)


  • Accident Analysis and Prevention 34 (2002) 4350

    Stop, look, listen, and think?What young children really do when crossing the road

    M. Suzanne Zeedyk a,*, Linda Wallace b, Linsay Spry a

    a Department of Psychology, Uniersity of Dundee, Dundee DD1 4HN, UKb Tayside Police, Police Headquarters, West Bell Street, Dundee DD1 9JU, UK

    Received 10 February 2000; received in revised form 15 August 2000; accepted 13 September 2000


    This study sought to provide basic information about childrens behaviour in realistic traffic situations. Most literature in thisarea has focused on childrens knowledge about road safety or has assessed their behaviour in simulated traffic environments.However, until more is known about what children actually do in traffic environments, our ability to identify the importantelements for inclusion in educational programmes remains limited. Fifty-six children, aged 56 years, took part in a treasure trailactivity in which they were confronted with two road crossings, one at a T-junction with a moving car and the other betweenparked cars. Childrens performance was videotaped and coded for relevant behaviours such as stopping at the kerb, looking fortraffic, direction of gaze, and style of crossing (i.e. walking vs. running). Results revealed that performance was extremely poor.Sixty percent of the children failed to stop before proceeding from the kerb onto the road. Looking for oncoming traffic wasexhibited by no more than 41% of the sample, dropping to as low as 7% in some instances. When looking did occur, it wasinitially as likely to be in the inappropriate direction (i.e. to the left) as in the appropriate direction (i.e. to the right).Consideration of individual performance revealed the existence of individual differences within the sample; such differences wererelatively stable across the two road crossings. These findings, based on controlled naturalistic tasks and detailed observationalmethods, build on earlier studies that are generally able to provide only estimated rates of childrens behaviour. 2001 ElsevierScience Ltd. All rights reserved.

    Keywords: Childrens behaviour; Realistic traffic situations; Controlled naturalistic tasks; Detailed observational methods

    1. Introduction

    Concern for childrens safety as pedestrians has beenapparent within the UK for at least 25 years. Pedestrianaccidents are still widely regarded as one of the mostserious of all health risks facing children in developedcountries (Thomson et al., 1996), with Britain leadingEurope in the rate of child pedestrian fatalities (IR-TAD, 1995). Numerous educational initiatives at thenational level, such as the Green Cross Code, the TuftyClub, and the Childrens Traffic Club, as well as at thelocal level (e.g. Footsteps programme, 1996; Walkwiseprogramme, Davies et al., 1993), have attempted toaddress the issue by teaching children skills for dealing

    with the traffic environment. The Department of theEnvironment, Transport and the Regions recentlyfunded research programmes designed to obtain a bet-ter understanding of childrens pedestrian skills and todevelop new educational initiatives on that basis (De-partment of Transport et al., 1996). All of these pro-grammes, regardless of whether or not they eventuallyproved to be effective, are commendable in that theyhave taken the issue of road safety seriously.

    However, such programmes continue to ignore abasic observation made 20 years ago by Grayson: oneshould know how children do behave before tellingthem how they should behave (Grayson, 1981, p. 172,italics added). We may know a bit more today than weonce did about how children deploy their attention,about the ways in which their physiological develop-ment limits their ability to cope with traffic, and aboutthe separate component skills that comprise the

    * Corresponding author. Tel.: +44-1382-344611; fax: +44-1382-229993.

    E-mail address: (M.S. Zeedyk).

    0001-4575/01/$ - see front matter 2001 Elsevier Science Ltd. All rights reserved.PII: S00 0 1 -4575 (00 )00101 -9

  • M.S. Zeedyk et al. / Accident Analysis and Preention 34 (2002) 435044

    pedestrian task (Thomson et al., 1996: 4). Ironically,though we still know very little about the basic questionof what children actually do on roads. What aspects ofthe traffic environment do they pay attention to? Howgood are they at integrating various pieces of informa-tion? Under what circumstances are they distracted?Thomson and his colleagues recently reinforcedGraysons earlier comments, when they observed thatonly if we have a clear idea of what children actuallydo look out for in the traffic environment will we be ina position to improve their attention to those featureswhich they ought to focus on to render them safe roadusers (Foot et al., 1999, p.400, italics in original). Thus,the purpose of the present study was to address this gapin the literature by gathering basic information aboutwhat children actually do on the road, through detailedobservation of their behaviour in two realistic road-crossing situations.

    Behavioural assessment, particularly in realistic situa-tions, is notoriously difficult. The risk of placing chil-dren in real traffic situations is high. Little supportwould be gained for implementing the questionablekind of programme described by Downing (1981: 99),in which Scandinavian parents were encouraged totrain children by hiding a toy across the road and thento stop their child once they had start(ed) to run intoand across the road. Instead, simulated environmentsand techniques have gained popularity, such as thepretend road (Lee et al., 1984; Young and Lee, 1987),kerbside judgements (Demetre et al., 1993), and trafficgardens (Sandels, 1975). Unfortunately, employing suchtechniques makes the issue of road safety salient forchildren, perhaps encouraging them to demonstratetheir best practice rather than more typical behaviour.The dimensions, timescales, and other characteristics ofthe tasks also often do not reflect those of real situa-tions accurately; indeed, children involved in such tasksmay even draw inaccurate or detrimental conclusionsabout the complexity of crossing the road.

    There have been some studies that observed childrenin real-life environments and coded their actions on-line, including the research conducted by the ScottishDevelopment Department (1989) which observed 10 000children within a region of Scotland. Other studies havevideotaped childrens actions inconspicuously (e.g.Schioldborg, 1976; Valavuo, 1976; Molen, 1983; allcited in Thomson et al., 1996). Although these methodsare superior to simulations in helping us to know whatchildren do in naturalistic settings, one drawback isthat on-line coding can be done only once and cannotbe subsequently verified. The vantagepoint of the ob-server in naturalistic settings may also obscure many ofthe childs behaviours, particularly in regard to lookingfor traffic. Relevant characteristics, such as age, culturalbackground and experience with traffic can at best onlybe estimated, the absence of which is important because

    a range of individual differences has been shown toexist among children (e.g. Whitebread and Neilson,1999). In effect, researchers take what they can get inthe way of data, but the sampling technique can beincomplete.

    The present study was designed to try to overcomesome of these problems and to provide detailed infor-mation about childrens abilities by inconspicuouslyfilming the childrens behaviour when they were con-fronted with realistic traffic situations. Children wereengaged in an outdoor treasure trail, the route ofwhich presented them with two different types of roadcrossings. The traffic environment they faced was thesame as they would encounter in their daily lives;junctions, parked cars, and moving traffic. The environ-ment was, in reality, a safe one, for traffic flow wasbeing regulated by police officers out of sight of thechildren. Because the children were not aware of theseprecautions, their behaviour allowed us to monitortheir actions in what was, for them, a very real trafficenvironment.

    The context within which the assessment was con-ducted contained several features that were intended toreplicate circumstances common to child pedestrianaccidents. For example, the most frequent activity pre-ceding traffic accidents involving young children is play(Grayson, 1975; Sandels, 1975; Christie, 1995), withchildren running into the road without looking fortraffic, generally because their attention is focused onsomething else (Scottish Development Department,1989). By embedding the road crossings within anotheractivity (i.e. the treasure trail), we created the equiva-lent of such circumstances. The fact that the activitiesoccurred in the presence of adults (i.e. school andresearch staff) also recreated the circumstances com-mon to between one-third and one-half of child pedes-trian accidents, which occur when children are in thepresence of adults (Grayson, 1975; Sandels, 1975;Christie, 1995). Finally, although children were facedwith a road crossing, they could choose to cross ontheir own or to seek adult assistance. Thus, the activityis equally as applicable to children who have alreadygained autonomous experience of roads, because theirparents allow them such independence, as to childrenwho have gained only supervised experience of roads.All children today will be familiar with traffic situationssuch as arriving at school with their parents, unloadinggroceries from the car with their parents, or catchingsight of a friend standing across the street. They mustmake decisions about how to cope with these types ofsituationsassume adults are monitoring the situationand proceed without confirming that this is the case?Wait until an adult arrives to help with crossing? Checkoncomin