The Sociotechnical Nature of Mobile Computing Work

Copyright © 2005, Idea Group Inc. Copying or distributing in print or electronic forms without writtenpermission of Idea Group Inc. is prohibited.

International Journal of Technology and Human Interaction, 1(3), 1-14, July-September 2005 1

The Sociotechnical Nature ofMobile Computing Work:

Evidence from a Study ofPolicing in the United States

Steve Sawyer, The Pennsylvania State University, USA

Andrea Tapia, The Pennsylvania State University, USA

ABSTRACT

In this article we discuss the sociotechnical nature of mobile computing as used by threepolicing agencies within the United States. Mobile devices, access, and service was providedvia a third-generation wireless network to a focal application, Pennsylvania’s Justice NETwork(JNET), a secure Web-based portal connecting authorized users to a set of 23 federated criminaljustice and law enforcement databases via a query-based interface. In this study we conceptualizemobility and policing as a sociotechnical ensemble that builds on the social-shaping oftechnology perspective and the tradition of sociotechnical theorizing, focusing on the co-design of work practices and technologies to support work. Drawing from the social informaticstradition, we turn a critical, empirical, and contextual lens on the practices of mobility andwork. Our analysis of the data leads us to observing that the social and the technical are stillconsidered separately in the context of mobile work. This simple view of social and technical asrelated, but distinct, often leads to problems with collecting and interpreting evidence of ICT-based systems’ design and use. We further note that this over-simplification of sociotechnicalaction is likely to continue unless more viable analytic approaches are developed and theassumptions of the current techno-determinist approaches are challenged more explicitly.

Keywords: field trial; mobility; policing; sociotechnical analysis

INTRODUCTION

One of the many alluring possibilitiesof mobile computing is that people will beable to access computing resources whileon the move. In organizational contexts,mobile computing (or mobility as we referto it here) engenders scenarios of increased

productivity through instant access to com-puting resources at any time from any-where. Here we explore thesociotechnical nature of this envisionedfuture for mobility. In the social informaticstradition, we turn a critical, empirical, andcontextual lens on the practices of mobility(Kling, 1999, 2000; Sawyer & Eschenfelder,2002).

2 International Journal of Technology and Human Interaction, 1(3), 1-14, July-September 2005


We first explain why policing is anappropriate domain in which to exploremobility and work. We then conceptualizemobility as a sociotechnical ensemble. Insubsequent sections we lay out the re-search, outline our data collection andanalysis, and then present and discussseven findings. We conclude by focusingon implications regarding sociotechnicalanalysis.

Why Focus on Policing?There are at least three reasons why

policing is an appropriate domain for study-ing mobility. First, police officers’ work hasalways been highly mobile. It is also knowl-edge-intensive and pervasive (for more onthis, see Manning, 2003). Second, therecontinues to be great interest in using ICTto better support police officers’ informa-tion needs. For example Manning (1996),in his study of cellular phone take-up amongpolice, reported on the long-standing dis-parity between police officers’ informationneeds and the abilities of the ICT used toprovide them that information.1 Third, po-licing and criminal justice have long been afocus of academic study; that provides uswith extensive literature on police work, thesocial norms, informal and formal organi-zational governance mechanisms, and anunderstanding of their technological basis(see Manning, 1977; Klockars &Mastrofski, 1991; Manning, 2003)2.

Current research findings providecontrary views as to whether the take-upof ICTs is driving the organization andstructure of police departments, or if it isthe reverse (Manning, 2003; Lin, Hu, &Chen, 2004; NASCIO, 2003; Taylor, Epper,& Tolman, 1998). Evidence is clear thatthe uptake of new computer-based systemsand uses of mobile technologies (beyondthe nearly omnipresent radio communica-tions suite in most cars and with most po-

lice officers in the U.S.) is accelerating inthe U.S. (Nunn, 2001). Partly, this atten-tion comes in response to the country’s in-creased attention to Homeland Security(Rudman, Clarke, & Metzel, 2003), thoughefforts to improve policing through ad-vanced computing pre-date current atten-tion (Northrup, Kraemer, & King, 1995).The limited functionality and advanced ageof many criminal justice and police systemsfurther magnify this attention (Brown,2001).

Contemporary research also suggeststhat police are open-minded about newtechnologies (wireless and otherwise) andgenerally view favorably the potential ofnew technologies to change policing (Nunn& Quinet, 2002; Lin et al., 2004). In fact,the evidence shows that most police de-partments across the United States haveone- to three-year plans either to imple-ment wireless technology or have alreadyimplemented some form of wireless tech-nology (Dunworth, 2000). To support theseefforts, both the United States departmentsof Homeland Security (DHS) and Justice(DoJ) provide a range of grants to supportinformation technology innovations in po-lice departments throughout the nation. Inaddition, there is funding by local jurisdic-tions and a variety of other sources, includ-ing internally generated revenue, such asfines, to support technological innovation.

MOBILE COMPUTING ASA SOCIOTECHNICALENSEMBLE

Sociotechnical perspectives focusboth conceptual and analytical attentionon three concepts: that which is social,that which is technical, and their inter-relations. In our study of mobile accessto computing resources for police work,the sociotechnical perspective helps us



to highlight that mobility is a complex andinterdependent set of relations amongpeople (workers and managers), their or-ganizational rules and roles, and variouscomputing resources (such as the techni-cal aspects of the mobile infrastructure,devices used, information sources, and ap-plications accessed). Following Orlikowskiand Iacono (2001), we conceptualize mo-bile access to computing resources as anensemble comprising the wireless network,access devices, applications being used,information and data (both structures andcontent), procedures followed, norms ofbehavior (relative to events, systems, andothers), governance structures, and bothinstitutional and environmental constraints.

Conceptualizing mobility as asociotechnical ensemble helps highlightthe nuanced and multi-faceted interde-pendencies uniting people, what they dowith computing resources, and how theyare designed and used. We further arguethat what is social and what is technicalare engaged in certain times and placesand in certain ways. Thus, we build onthe work of policing by focusing on spe-cific events and situate these events inspecific times and places. This contex-tual frame provides us the means toground the analysis of the sociotechnicalinteractions.

The particular interactions amongthese constructs will likely vary by situa-tion. For example, in a routine3 event suchas a traffic stop, these constructs are tiedtogether in a prescribed way. There arepolicies regarding the use of the car andpersonal (attached to the officer’s uniform)radio, a standard set of practices that guidethe set of interactions the officer has withboth the police dispatcher and with thedriver of the car being stopped, particularrules about the information needed frompolice resources (such as registration, li-

cense plate numbers, car details, and evendata on the driver based on the driver’slicense proffered to the officer), and whatdata the officer can and should collect.Escalation procedures are proscribed, andthese vary based on time of day, assess-ments of the local situation, and other op-erational considerations.

For instance, imagine that a sergeant4

sees a pick-up truck speeding down abreakdown lane to avoid stopped traffic inthe travel lanes and gives a chase. The driv-ers of the truck see the police car chasingthem and, as is customary in the U.S., pullsover to the side of the road. The sergeantsees that the driver is agitated to the pointwhere he is cursing out the vehicle’s win-dow; the truck is shaking from “omnidirec-tional fury,” and the sergeant and calls forbackup from his car radio. While waitingfor backup, the officer puts on black leathergloves (in case they scuffle), unsnaps hisweapon’s securing strap (in case it goesbeyond scuffling), calls in to police dispatchwith vehicle information, and then switchesto his body radio, talk activated. With theradio live (and all other officers on that fre-quency quiet, and the police dispatcher dis-passionately updating time until backup ar-rives)5, the sergeant approaches the upsetdriver and starts the (relatively prescribed)process of gathering particular informationon the driver’s identity as the first step inwriting up a traffic citation. The backupofficer arrives while the sergeant is con-fronting the driver, pulls up diagonally infront of the stopped pickup (to reduce thepossibility of a “drive-off”), and stands inplain view and direct line of sight to thedriver, weapon at the ready.

A more common traffic stop will haveless drama for the driver (but perhaps someirritation), may not require backup or bringout the visible presence of force, and likelydoes not escalate until the driver receives



multiple citations. But both traffic stopsengage the same set of devices, applica-tions, network, common information, anddata flows; draw on the same governancestructures; follow the same set of proce-dures (albeit down differing paths, butpaths stemming from the same proceduralguides); and reflect common and well-developed norms of policing behavior(norms both explicitly taught through ex-tensive training, and also learned and re-inforced by doing policing).

Conceptualizing mobility and polic-ing as a sociotechnical ensemble builds onthe social-shaping of technology perspec-tives developed by Bijker (1995), Law andBijker (1992), and Bijker et al. (1987). Inmaking this point, we acknowledge thatthere are several active streams ofsociotechnical research/theorizing (seeHorton, Davenport, & Wood-Harper,forthcoming). For example, the Europeantradition of sociotechnical theorizing,which we build on here, takes a socialshaping of technology (SST) perspective.The SST perspective highlights that thematerial characteristics and actions of anytechnology are shaped by the social actionsof the designers, the specific uses of thattechnology, and the evolving patterns of useover time. A second, work-studies tradi-tion of sociotechnical theorizing focuses onthe co-design of work practices and tech-nologies to support work. This co-designperspective has been taken up in NorthAmerica and evolved in two ways. Thefirst is a benign neglect of the interactionbetween what is social and technical, lead-ing to an evocation of the concepts withouta concomitant analytical activity (seeScacchi, 2004, for a critical discussion). Thesecond, an SST approach, is more recentand reflects social informatics in that theefforts are focused on developing specificanalytic approaches that make explicit as-

pects of the social, the technical, and theirinteraction (see Kling, McKim, & King,2001).

Rather than focusing on a specifictheoretical approach to examining thesociotechnical action of policing and mo-bility, we use Bijker’s (1995) principles ofsociotechnical change theory to illustratethe generic goals of this approach, and todiscuss the theoretical tensions that existin sociotechnical IT research. These ten-sions provide a range of possibilities forspecific sociotechnical research efforts.Here we use them as orienting principlesfor our conceptualization of mobility andthe consequent design of our research, datacollection, and analysis.

Bijker’s (1995) four principles ofsociotechnical change theory are derivedfrom work in the sociology of technology.These four principles provide a set of goalsfor any theory that strives to take asociotechnical perspective: the seamlessWeb principle, the principle of change andcontinuity, the symmetry principle, and theprinciple of action and structure. Theseamless Web principle states that anysociotechnical analysis should not a prioriprivilege technological or material expla-nations ahead of social explanations, andvice versa. The principle of change of con-tinuity argues that sociotechnical analysesmust account for both change and continu-ity, not just one or the other. The symmetryprinciple states that the successful work-ing of a technology must be explained as aprocess, rather than assumed to be the out-come of “superior technology.” The actorand structure principle states thatsociotechnical analyses should address boththe actor-oriented side of social behavior,with its actor strategies and micro interac-tions, and structure-oriented side of socialbehavior, with its larger collective and in-stitutionalized social processes.



While Bijker’s principles provide a setof ideals for sociotechnical research tostrive for, in practice they illustrate tensionsto be managed in the research process.Given the space limitations, in the analysisto follow we focus on highlighting findingsrelative to our concepts and not specificallyexamining how the four principles guide thiswork.

EVIDENCE FROM A FIELDTRIAL OF POLICING,COMPUTING, ANDMOBILITY

To explore the sociotechnical per-spective on productivity and the effects onwork due in part to pervasive access tocomputing resources, we report on a fieldstudy6 of police officers’ uses of an inte-grated criminal justice system accessedvia the public wireless data network fromlaptops and personal digital assistants(PDAs) provided to the participants.Each element of our field trial is dis-cussed below.

Mobile access for this trial was donevia a third-generation (3G) data network.In the U.S., 3G networks are rolling out(typically based on population density) andmirror the cellular phone network in termsof coverage. However, 3G networks useInternet protocols, packet switching (and,thus, digital packets), spread-spectrumtransmission (which is inherently more se-cure than cellular and 2G standards), andcan sustain throughput speeds of up to 150kilobits per second. The 3G data networksin the U.S. are private, and multiple pro-viders compete directly in each market.While wireless coverage is extensive, noone carrier provides complete coverage ofthe geography of the U.S., and there maybe gaps in service within covered areas.Moreover, collectively, all providers’ cov-

erage does not cover the geography of theU.S., and a service gap in one provider’scoverage is not alleviated by the coverageof a second. The major carriers in the U.S.have deployed their 3G networks in differ-ent ways and at different rates7. Gener-ally, though, they have focused on deploy-ing in areas where that are most populated(cities and suburbs) and most traveled(along major highways). Costs, reliability,and coverage vary greatly in all other ar-eas (Federal Communications Commission,2002).

The focal application wasPennsylvania’s Justice NETwork (JNET)8,a secure Web-based portal connecting au-thorized users to a set of 23 federated crimi-nal justice and law enforcement databasesvia a query-based interface. The JNETarchitecture is characterized by four ele-ments. First, and as noted, for the user itacts as a portal to the criminal-justice-re-lated databases of the Commonwealth ofPA (and the U.S. Federal government).The data are owned by the relevant stateor Federal agency (e.g., Pennsylvania’sDepartment of Transportation, orPennDOT, maintains driver’s licenserecords and a picture database), and JNETprovides query-based access to the driver’slicense photos. Second, JNET is a securesystem. Users are carefully vetted beforethey get access, their use is tied to specificroles, and these roles grant them varyinglevels of access to the range of data avail-able. Further, use is tied to secure connec-tivity (enabled through encryption and vir-tual private networks); this requires sev-eral forms of identification to be used9.Users must also re-authenticate periodicallyduring their sessions in order to assure se-curity during use. Furthermore, re-authen-tication is required when accessing certainspecific databases through JNET. Until thefield trial we report on here, there was no



mobile access: thus, security was done viafixed lines and desktop computers. Third,JNET also provides electronic messaging,e-mail, and reporting functions for users.These functions serve as both a commonmessage board across all criminal justicepersonnel in PA. The e-mail alerts providea means for people to keep track of activi-ties where they have some interests. Forexample, it is possible for a parole officerto set up a query on a particular name, so-cial security number, or case number(s). Ifthat name or those numbers come acrossthe message board, she will be alerted andcan more easily follow up on the parolee.Fourth, JNET has been operational sinceearly 2000, and it supports thousands ofqueries each month (and use has grown bynearly 10% per month since inception)(JNET, 2004)10.

The third part of the mobile access toJNET is the device being used to providemobile access to JNET (and to the Internetmore broadly). This device must have aspecial 3G modem card and needs to bemobile. Most police cruisers have an inte-grated laptop, making this seemingly a trivialeffort (put in the wireless modem card, loadon the security software, and use abrowser). However, there were a numberof operational and legal issues that madethis a nontrivial effort. For example, manyof the laptops are not equipped with spaceto load the modem card. Battery draw onpolice cruisers is substantial, and this fur-ther limits laptop use (and the 3G modemcards draw substantial power to run theantenna and maintain connectivity). More-over, some cruisers’ laptops have othersoftware whose security and operational/licensing requirements precluded additionalapplications from being added.

For officers not in a cruiser, the mo-bile device must be carried on their per-son. Again, this is not a trivial effort, con-

sidering that almost every square inch ofthe average police person’s body is cov-ered by some piece of gear. Moreover, thecombination of current equipment (includ-ing communications, weapons, body armor,etc.) is nearly 25 pounds. This means thatthe mobile device must often displacesomething the officer already carries. Wereturn to this discussion later in the article.

FIELD TRIAL DESIGN,DATA COLLECTION, ANDANALYSIS

The field trial’s design focused atten-tion to collecting data on the wirelessnetwork’s use, device uses, JNET andother applications’ uses, information anddata sharing, changes or alterations topolice officers’ work practices (particu-larly changes to infield operations), socialnorms on computing/uses (particularly re-garding the value and importance of bothmobile access and JNET), and the officer’soperational governance (particularly therole of dispatch). As we noted at thearticle’s outset, in focusing on criminal jus-tice, we leverage the extensive knowledgeof policing and also partially control for in-dustrial (extra-organizational) factors bystaying within one work domain.

The field trial also served as an inter-vention: mobile workers11 were providedwith either a laptop or a personal digitalassistant (PDA)12 and secure access to thepublic 3G network. This was done in twophases for pragmatic reasons. The firstphase lasted three months, included fiveparticipants, and focused on laptop usage.The small number allowed us to refine datacollection protocols, and ensure that wecould meet the technological demands ofsupporting the access, security, and appli-cation use demands of a demanding op-erational environment. The second phase



began directly after the first phase’scompletion, involved 13 participants, lastedthree months, and focused on PDA usage.All five of the participants in the first trialwere part of the second trial. This providedus with a subset of users who were en-gaged in mobile access to JNET for sixmonths. The two-phase trial’s six-monthduration was guided by practical constraintsof users’ ability to participate in a trial whiledoing their normal policing and official du-ties. The number included in the trial wasconstrained by the costs of providing de-vices, connectivity, and support to the of-ficers.

Participants in both trials were policeand other criminal justice officers from threeorganizations (one county level and twolocal level) located within one Pennsylva-nia county. Two incentives were used tomotivate participants. First, we promisedthat all participants could keep the mobiledevice(s) they were given to use (late-model laptops and high-end PDAs, bothequipped with 3G modem cards; and in thecase of the PDA, an external sleeve andbattery pack to support the modem card).Second, we made it clear that the partici-pants’ input would be used to drive the de-sign of JNET for criminal justice uses, par-ticularly for mobile access. Participantsmentioned that both were important to theirdeciding to engage. In addition, we workedwith the department heads and unit policechiefs to ensure that officers were givenofficial recognition for engaging in the fieldtrial. Participating department heads andunit police chiefs were both enthusiastic andsupportive.

We used seven forms of data collec-tion. First, we did pre- and post-interviews(at the beginning and end of each trial peri-ods) of all users. In phase one these wereface-to-face, open-ended, and semi-struc-tured interviews that lasted from 60 to 90

minutes. In phase two, we used a morestructured, self-administered survey inplace of some of the open-ended user in-terviews and followed up with a phone dis-cussion. Second, we led focus groups ofusers following the trials. These were vol-untary, and only two participants did notparticipate (for schedule reasons). Third,all users completed a one-week time diaryof work behavior during the field trial.Fourth, members of the research team didride-alongs with users. We chose to ride-along with both police and court officers,and with both supervisors and patrol offic-ers. Fifth, we gathered documents duringall interviews, observations, and visits (anddid extensive Web and library research tosupport the field work). Sixth, we engagedin informal weekly interactions (via phone,e-mail, and in person) with users. Finally,we gathered data about laptop uses, wire-less data transmission, and JNET usage viaunobtrusive means (such browser logs,server logs, and telecom activity logs). Datafrom the first six sources were either tran-scribed into digital format or collected atsource in digital format. Data from the us-age logs came in digital format.

Our analysis focused on identifyingissues with the 3G network’s connectivity/reliability, speed and access, uses of JNET(and other sources/applications), informa-tion and data access, and the roles of themobile devices. This was done throughanalysis of data drawn from the trouble-ticketing log, analysis of time use (drawnfrom the logs) regarding connection via 3Gnetworks, volume of data transfer and time/usage of JNET, and through a series of topi-cal analyses of the texts created from thesix forms of intensive data collection.

Analysis of data regarding informa-tion and data sharing, work practices, so-cial norms, and operational governance fol-lowed traditional qualitative data analysis



approaches (see Miles & Huberman, 1994).In particular, we used three techniques: in-terim analysis of the data to guide both fu-ture data collection and its interpretation,explanatory even matrices, and contentanalysis of the transcripts, logs, and fieldnotes.

FINDINGSWe present and discuss seven find-

ings. We find that police officer’s uses of3G wireless networks is dependent moreon coverage and reliability of access thanon speed (bandwidth). Certainly, higherthroughput speeds are better than lowerspeeds (particularly when transferringdriver’s license photos, as we discuss be-low). However, if coverage is not certain,then officers either forget to access thenetwork, or become frustrated and activelychoose to NOT access the network. More-over, if an officer takes the time, cognitiveenergy, and effort to connect, and the ac-cess attempt fails (for any number of rea-sons), it appears they quickly cease trying.

We find that the police officers in ourstudy do not value laptops as access de-vices. They do, however, appreciate thesedevices for other activities (such as writ-ing up their incident reports and other tasksthat did not require them to have wirelessaccess). Police officers valued PDAs toan even greater degree. Again, these de-vices are valued for personal informationmanagement and not as connective devicesto the 3G network. We did not attempt totrial pen-based or tablet computers: wesuspect that these may combine the port-ability of a PDA with the power and screensize (an important issue for officers) of alaptop.

The mobile access to and uses ofJNET and other applications was diffi-cult to assess for two reasons. First, the

low reliability of the network coveragemade it difficult for officers to access theseapplications. The officer had to becomevery familiar with coverage patterns (thatis, where they could and could not gainaccess) and then be able to adjust theirwork patterns to accommodate this cover-age. Second, authentication and securityoverhead in access complicated the logonprocedures and caused connection drops.The two factor logon procedures made itdifficult for officers in the field to manageboth connection and conduct their work.The design of JNET (which asks for up-dates on passwords and re-authenticationas different databases are searched) meantthat it was easy for JNET to shut down thesession unless the officer devoted consid-erable attention to managing the interac-tion. This considerable attention to JNEThad to come at the expense of attention toother aspects of the officer’s work. In anyoperational event (such as a traffic stop)the officer would not make this commit-ment.

Despite this difficulty, officers valueJNET for its ability to provide them infor-mation about drivers, particularly thedriver’s license photos and drivers’ records.On this (and limited evidence of this) alone,officers prized mobile access to JNET andfound value in mobility. We did not see anychanges in information and data shar-ing for at least two reasons. First, the de-sign of JNET for mobile access is to pro-vide it to officers, and not through policedispatch. Most all other information anddata sharing, however, goes through policedispatch (both in a controlled voice-basedinteraction and via current text-based sys-tems that come to the police vehicle’sonboard laptop).

We saw little changes to police offic-ers’ work practices. Perhaps this is notsurprising — the operational environment



of policing is harsh, and sometimes fatal.Police train extensively, continually, andwith great care to develop procedures totake an ambiguous situation and make itless so. Changes in operational proceduresare, thus, slow to come, painstakinglythought out, and must be demonstrable im-provements. If not, police are unlikely torisk their lives.

The great enthusiasm and interest onthe uses of computing to improve policingseems to be one of the strong social normsthat police carry forward (Manning, 2003).However, when confronted with changesto operational procedures and concernswith the computing system’s reliability, thesocial norms of policing operations such assafety, professionalism, and force projec-tion overwhelm the potential value of mo-bile access to computing resources thatcannot be consistently demonstrated.

The trial of mobile access to JNETand other computing resources amplifiedthe institutional embeddedness of the com-mand and control structures in policing. In

particular, the critical social, organizational,and technical roles that the police dispatcherplays came clear during this trial. The de-sign of JNET for individual access doesnot work well within police officers’ op-erational governance. Were JNET to bea dispatch-based access model, however,governance and information sharing wouldlikely change more quickly.

A SOCIOTECHNICALACTION PERSPECTIVE

In this section we draw on Bijker’s(1995) four principles of sociotechnicalchange theory to help us reflect on and in-terpret these seven findings (see Table 1).Building on this reflection and interpretation,we raise three points: that the sociotechnicalprinciples were supported by these findings;that simplistic approaches to engaging thesociotechnical nature of mobility may makeit hard to interpret the results; and that thereis a dire need for more substantivesociotechnical analysis techniques.

Table 1. Sociotechnical analysis

Findings Principles Comments Coverage and reliability of access more important than speed/bandwidth

Seamless Web Technological features (bandwidth) were seen as more central than operational needs of officers (operational reliability).

PDA valued for personal use, not for mobile access

Symmetry Take up of the device is a social decision, shaped by technical characteristics, and often made for personal needs.

JNET and other applications are used when mobile

Change and continuity The expectation that JNET would be valuable for mobile officers (as it has been for officers via fixed access) was borne out in the study.

Officers value information drawn from JNET

Change and continuity The expectation that information received while mobile would be valued was borne out in the study.

No changes in information and data sharing

Actor and structure Social and operational structures seemed to be resilient to new technologies of access and use.

No changes to police officer’s work practices and social norms

Actor and structure Work practices seemed to be resilient to new technologies of access and use.

No changes to work governance Actor and structure Governance structures seemed to be resilient to new technologies of access and use.



The premise of this field trial was thattechnological factors, such as mobile con-nectivity and higher bandwidth, would becentral to taking up mobility. This violatesthe seamless Web principle. The findingssuggest that the institutional structureswhich help to govern the work of policingserve as powerful moderators to both tak-ing up and taking advantage of mobility. Itappears, however, that the technologicalbelief of connection and bandwidth wereprivileged, relative to these institutionalstructures. It appears from these findingsthat moving out access to high-value re-sources (from fixed to mobile connection)is valued, supporting the principle of changeand continuity. However, the structural andinstitutional forces severely constrain ac-tion, and the promised performance of thedevices, mobile access, and informationsharing require more agentic effort thanpolice officers have.

Second, we observe that the currentprofessional practice of evaluating new ICTdoes not seem to engage sociotechnicalprinciples. For example, the failure to fullyengage sociotechnical principles when de-signing and trialing mobile access to JNETreflects a naïve view of sociotechnical ac-tion: that social and technical are distinctof one another (and that change in oneleads to change in the other). The find-ings we note above are not surprising;current institutional structures in policingwere not considered (or, worse, ignored— as was the case with dispatch) whendesigning new work technologies. And, thetechnological elements must be consideredon par with social elements — had thisbeen more carefully considered, bandwidthwould not have been the focus; it wouldhave been reliability.

The field trial design reflects the col-laboration between wireless service pro-viders, device manufacturers, local and

state police and information technologyleaders, and faculty. That the resulting trialunderplayed the sociotechnical issues leadsus to theorize that organizational decisionmakers, users, and technology evaluators’orientation towards problem solving willmake it attractive to focus on matchingtechnical features with work and organi-zational needs. In doing this they are notlikely to address the systemic interactionsor to consider extended interdependencies.In essence, this simplification in analysiscomes at the cost of accuracy in imple-mentation.

Sociotechnical approaches, such asBijker’s (1995) four principles, appear morelikely to be applied in post-hoc analysis.They become a comfortable frame forscholars to use. However, they are at besta weak analytic structure to base proac-tive action. That is, the principles are use-ful to frame and interpret evidence, but aredifficult to use in guiding specific designs.What is missing are the intermediate-levelguidance linked to specific technologies orspecific social actions. In the absence ofthis intermediate-level guidance, the prin-ciples are difficult to apply proactively.

Building on this, it seems important, ifnot imperative, that sociotechnical modelsprovide more intermediate guidance. Bythis we mean support for constraints andenablers tied to particular social actions orthat highlight elements of particular tech-nologies. This intermediate level ofsociotechnical knowledge is likely to be rep-resented as contingent or localized mod-els. In doing this, such localized models willhelp academics and practicing profession-als more directly to dominant patterns ofinteractions and consequences, and makethese findings available in ways that moredirectly influence ICT/systems design andorganizational decision making.

Our final observation from this analy-



sis is that the over-simplification ofsociotechnical action is likely to continueunless more viable analytic approaches aredeveloped and the assumptions of the cur-rent techno-determinist approaches arechallenged more explicitly. Given this view,it seems likely that organizational decisionmakers, users, and ICT designers will havetrouble making sense of evidence drawnfrom failed attempts to implement and useICT based on their simple views of ICTuse, cause and effect. We believe the in-ability to understand this data is driven bythe unsound approach of invoking directeffects of ICT use, not by the measure-ments taken or instruments used to gatherevidence (e.g., Sawyer, Allen, & Lee,2003).

While the research literature focusedon the effects of ICT highlights on the indi-rect and often nuanced relationships amonguse of ICT and performance, professionalpractice continues to press for the directeffects model of ICT value. This suggeststhat more robust system or contingencymodels of ICT effects are needed (e.g.,Avgerou, 2002). This is one of the mostactive areas of scholarship in IT, and thisactivity needs to enter the texts, teachingcases, and classrooms of the nextgeneration’s IT leaders, organizational man-agers, and technology developers. For ex-ample, those who have focused specificallyon the roles of mobile and fixed locationuses of ICT in policing all note that the op-erational value derived from using new ICT-centric information systems is minimal, ifdiscernible (Ackroyd, Harper, Hughes,Shapiro, & Soothill, 1996; Dunworth, 2000;Meehan, 2000).

What seems important to us is a morefocused effort to engage the principles ofsociotechnical action in direct comparisonto the bases of direct effects models (e.g.,Kling & Lamb, 2000). They develop a com-

parative analysis of tool and Web modelsof computing relative to organizational ac-tivity. In doing this, they highlight both theseamless Web principle (privileging neitherthe social nor the technical) and the prin-ciple of action and structure by highlightingthe concept of a social actor — one thathas agency, but constrained by institutionalstructures (Lamb & Kling, 2003). Buildingon these two principles, in the work re-ported here we provide a means of repre-senting the principle of change and conti-nuity by explicitly linking elements of thetechnical structure of JNET, the institutionalstructures of police work, and the actionsof police.

REFERENCESAckroyd, S., Harper, R., Hughes, J.,

Shapiro, D., & Soothill, K. (1996). Newtechnology and police work.Buckingham: Open University Press.

Avgerou, C. (2002). Information systemsand global diversity. Oxford: OxfordUniversity Press.

Bijker, W. (1995). Of bicycles, bakelites,and bulbs: Toward a theory ofsociotechnical change. Cambridge,MA: The MIT Press.

Bijker, W., Hughes, T., & Pinch, T. (1987).The social construction of technologi-cal systems. Cambridge, MA: The MITPress.

Brown, M.M. (2001). The benefits andcosts of information technology innova-tions: An empirical assessment of a lo-cal government agency. Public Perfor-mance & Management Review, 24(4),351-366.

Dunworth, T. (2000) Criminal justice andthe information technology revolution. InHorney (Ed.), Policies, processes anddecisions of the justice system (vol-ume 3, pp. 372-426). Washington, DC:National Institute of Justice/Office of



Justice Programs.Horton, K., Davenport, E., & Wood-Harper,

T. (2005). Exploring sociotechnical in-teraction with Rob Kling: Five ‘big’ ideas.Information, Technology and People,(in press).

JNET. (2004). Usage statistics. Retrievedfrom: http://www.pajnet.state.pa.us/pajnet/site/default.asp

Kling, R. (1999). What is social informatics,and why does it matter? D-Lib Maga-zine, 5(1). Retrieved from: http://www.dlib.org:80/dlib/january99/kling/01kling.html

Kling, R. (2000). Learning about informa-tion technologies and social change: Thecontribution of social informatics. TheInformation Society, 16(3), 217-232.

Kling, R. & Lamb, R. (2000). IT and orga-nizational change in digital economies:A sociotechnical approach. In B. Kahin& E. Brynjolfsson (Ed.), Understand-ing the digital economy: Data, toolsand research: Cambridge, MA: TheMIT Press.

Klockers, C. & Mastrofski, S. (Eds.).(1991). Thinking about police: Con-temporary readings. New York:McGraw-Hill.

Lamb, R. & Kling, R. (2003).Reconceptualizing users as social actorsin information systems research. MISQuarterly, 27(2), 197-235.

Law, J. & Bijker, W. (1992). Technology,stability and social theory. In W. Bijker(Ed.), Shaping technology/buildingsociety (pp. 32-50). Cambridge, MA:The MIT Press.

Lin, C., Hu, P., & Chen, H. (2004). Tech-nology implementation management inlaw enforcement. Social Science Com-puter Review, 22(1), 24.

Manning, P. (1977). Police work: The so-cial organization of policing. ProspectHeights, IL: Waveland Publishing.

Manning, P. (1996). Information technol-ogy in the police context: The ‘sailor’phone. Information Systems Research,7(1), 275-289.

Manning, P. (2003). Policing contingen-cies. Chicago: University of ChicagoPress.

Meehan, A. (2000). The transformation ofthe oral tradition of policing through theintroduction of information technology.Sociology of Crime, Law and Devi-ance, 2, 107-132.

NASCIO (National Association of StateChief Information Officers). (2003).Concept for operations for integratedjustice information sharing version 1.0.Retrieved from: https://www.nascio.org/publications/index.cfm

Northrup, A., Kraemer, K.L., & King, J.L.(1995). Police use of computers. Jour-nal of Criminal Justice, 23(3), 259-275.

Nunn, S. (2001). Police information tech-nology: Assessing the effects of com-puterization on urban police functions.Public Administration Review, 61(2),221-234.

Nunn, S. & Quinet, K. (2002). Evaluatingthe effects of information technology onproblem-oriented-policing: If it doesn’tfit, must we quit? Evaluation Review,26(1), 81-108.

Orlikowski, W. &. Iacono, S. (2001). Des-perately seeking the “IT” in IT research— A call to theorizing the IT artifact.Information Systems Research, 12(2),121-124.

Rosenbach, W. & Zawacki, R. (1989).Participative work redesign: A field studyin the public sector. Public Administra-tion Quarterly, 43, 111-127.

Rudman, W., Clarke, R., & Metzel, J. (2003,July 29). Emergency responders:Drastically underfunded, danger-ously unprepared. Report of an inde-pendent task force sponsored by the



Council on Foreign Relations. Re-trieved from: http://www.cfr.org/pdf/Responders_TF.pdf

Sawyer, S. & Eschenfelder, K. (2002).Social informatics: Perspectives, ex-amples, and trends. In B. Cronin (Ed.),Annual review of information scienceand technology (volume 36, pp. 427-265). Medford, NJ: Information TodayInc./ASIST.

Sawyer, S., Allen, J., & Lee, H. (2003).Broadband and mobile opportunities: Asociotechnical perspective. Journal ofInformation Technology, 18(2), 11-35.

Sawyer, S., Tapia, A., Pesheck, L., & Dav-enport, J. (2004). Observations on mo-bility and the first responder. Communi-cations of the ACM, 47(2), 62-65.

Taylor, M., Epper, R., & Tolman, T.(1998). Wireless communications andinteroperability among state and locallaw enforcement agencies. Report168945 of the National Criminal JusticeClearinghouse, Washington, D.C.

ENDNOTES1 Manning (1996) focused on the take up

and uses of cellular phones by police.Personal cellular phone ownership anduse is now common among criminal jus-tice officers. While the take up and useof a cellular phone is beyond the scopeof this article, two attributes are worthnoting. First, the officers use their own(personal) cellular phones and do notconsider them as part of their profes-sional equipment. Second, personal usehas made officers aware of issues withwireless coverage, reliability, and use.

2 Given the extensive literature on polic-ing, in this article we draw from but donot develop or discuss principle findings.Instead, we refer the interested readerto anthologies of such work (listed in ourreferences and cited here). The inter-

ested reader can also find courses incrime, law, and justice offered in mostsociology departments, and the exten-sive material on the Web in locationssuch as the U.S. Department of Justice,the UK Home Office, and the Interna-tional Association of Chiefs of Police.

3 Perhaps one of the more difficult partsof a police officer’s job is to rememberthat even a seemingly common thingsuch as stopping a speeding car may leadto armed confrontation. Thus, training isfocused on preventing common from be-coming routine.

4 Policing in the United States is organizedalong paramilitary lines. Thus, sergeantsare senior/experienced officers, typicallywith both patrol and supervisory respon-sibilities.

5 Most police in the United States workalone, which means: 1) they rely on theradio as a link to others and 2) the policedispatcher is a critical node in this link-age. The radio stays on, and no one elsespeaks so that all can listen for a gun-shot or the words “officer down.”

6 Our research design here builds on pre-vious public-sector field studies of work(Rosenbach & Zawacki, 1989).

7 Details of the debate and key issues inwireless network deployment, coverage,access,and use are beyond the scope ofthis article.

8 For more information about JNET, seewww.pajnet.state.pa.us.

9 Security in the trial was done via “two-factor” identification. This means hav-ing a physical key, called a dangle bythe officers, that stores a digital recordidentifying the owner that is tied to a logi-cal password that must be entered whenthe physical key is connected (via USBport) to the computer.

1 0 JNET is one of the earliest and mostvisible examples of a small and growing



number of these integrated criminal jus-tice information systems that are a fo-cus of homeland security efforts in theUnited States. Others include the Capi-tal Area Wireless Integrated Network(CAPWIN, see www.capwin.org), theautomated regional justice administrationsystem (ARJIS, see www.arjis.org),anda fast-growing number of municipal ef-forts such as systems in Chicago,IL;Montgomery County, MD; and LosAngeles County, CA.

1 1 Participants included one sergeant (orshift supervisor), nine patrolmen, andthree deputies (of the court). Participantswere male from ages 28 to 45. The av-erage work experience was 11 years;the most experienced officer had 18years of work and the least experiencedofficer had seven years of work.

1 2 Laptops and PDAs were provided to of-ficers as an incentive to participate andwere paid for by the research team.

Please provide current bios

The Sociotechnical Nature of Mobile Computing Work

Documents

Transcript of The Sociotechnical Nature of Mobile Computing Work