Safety and Accident Prevention

Journal of Occupational and Organizational Psychology (2015), 88, 295–321

© 2015 The British Psychological Society

www.wileyonlinelibrary.com

Special section paper

Decision inertia: Deciding between least worst outcomes in emergency responses to disasters

Laurence Alison*, Nicola Power, Claudia van den Heuvel, Michael Humann, Marek Palasinksi and Jonathan Crego Centre for Critical and Major Incident Psychology, University of Liverpool, UK

This study demonstrates how naturalistic decision-making (NDM) can be usefully applied

to study ‘decision inertia’ – Namely the cognitive process associated with failures to execute action when a decision-maker struggles to choose between equally perceived

aversive outcomes. Data assessed the response and recovery from a sudden impact

disaster during a 2-day immersive simulated emergency response. Fourteen agencies

(including police, fire, ambulance, and military) and 194 participants were involved in the

exercise. By assessing the frequency, type, audience, and content of communications, and

by reference to five subject matter experts’ slow time analyses of critical turning points

during the incident, three barriers were identified as reducing multiagency information

sharing and the macrocognitive understanding of the incident. When the decision

problem was non-time-bounded, involved multiple agencies, and identification of

superordinate goals was lacking, the communication between agencies decreased and

agencies focused on within-agency information sharing. These barriers distracted teams

from timely and efficient discussions on decisions and action execution with seeking

redundant information, which resulted in decision inertia. Our study illustrates how

naturalistic environments are conducive to examining relatively understudied concepts of

decision inertia, failures to act, and shared situational macrocognition in situations

involving large distributed teams.

Practitioner points

� Researchers can use NDM to explore the cognitive processing associated with failures to act/decision inertia.

� Complexities in the decision-making environment of a multiteam system (e.g., non-time-bounded choice, large team size, and lack of strategic goals) are associated with decision-making failures.

� Barriers cause decision inertia as teams focus on redundant intra-agency information seeking rather than cooperative interagency communications.

� Strategic direction is especially important for shifting multiteam system communication towards interagency discussions on action execution.

If the behavioural implementation of making a decision is the execution of action

(Lipshitz, Klein, Orasanu, & Salas, 2001; Yates, 2003), then failing to make a decision (e.g.,

making an executive choice) is when action execution fails. Traditional decision-making

*Correspondence should be addressed to Laurence Alison, Department of Psychological Sciences, Eleanor Rathbone Building, Bedford Street South, University of Liverpool, Liverpool L69 7ZA, UK (email: [email protected]).

DOI:10.1111/joop.12108

295

research has helped psychologists to understand the conditions under which decisions

are likely to be actioned in a static, closed environment, where decision-makers have the

opportunity to analyse the choice context and select an optimal course of action (e.g.,

expected utility theory; Von Neumann & Morgenstern, 1947). Yet in the ‘real world’, decision-makers rarely have the luxury of having unambiguous information or sufficient

time to analyse choices, and thus, their rationality is bound by cognitive constraints

(Simon, 1956). Whilst early research treated bounded rationality as maladaptive in causing

cognitive biases and faulty heuristic processing (Kahneman & Tversky, 1979), ‘naturalistic

decision-making’ (NDM) takes a more pragmatic perspective.

The NDM paradigm NDM rejects the notion of ‘right’ or ‘wrong’ decisions and instead seeks to understand the

cognitive processes associated with choice implementation by studying decision-making

‘in the wild’ (Gore, Banks, Millward, & Kyraikidou, 2006; McAndrew & Gore, 2013). NDM

emphasizes the importance of real-world contexts, domain-specific expertise, and

macrocognition in distributed and sociotechnical teams (Stanton,Wong, Gore, Sevdalis, &

Strub, 2011). The aim of NDM is ‘to understand how people make decisions in real-world

contexts that are meaningful and familiar to them’ (Lipshitz et al., 2001, p. 332). It

juxtaposes earlier decision-making research that aimed to explore decision-making errors (Kahneman & Tversky, 1979) and instead takes a positive approach by focusing on the

processes that enable individuals to make decisions (Gore et al., 2006). This is why NDM

research is often grounded in the notion of ‘expertise’ as researchers seek to uncover the

declarative knowledge that is implicitly stored in the cognition of experts (Klein, 1997).

Expert ‘macrocognition’ goes beyond behavioural decision-making and encompasses a

number of key cognitive elements that distinguish the expert from the novice, which

include mental models, perceptual skills, sense of typicality, routines, and declarative

knowledge (Klein & Militello, 2004). In addition to furthering interest in positivist research, NDM has also developed

naturalistic methods for data collection. For example, ‘cognitive task analysis’ is uniquely

designed to unpack macrocognition and includes various, and largely qualitative,

techniques such as the ‘critical decision method’ interview protocol (for more detail, see

Crandall, Klein, & Hoffman, 2006). An overreliance on ecologically invalid laboratory-

based settings is inappropriate (Schneider & Shanteau, 2003) as the real world is

characterized by ill-structured problems, uncertainty, poorly defined goals, multiple

feedback loops, time constraints, high stakes, multiple players, and conflict between personal ideals and contextual requirements (Orasanu & Connolly, 1993). Thus, this is the

setting in which research should be conducted. Lipshitz (1993) recognized that although

no unified NDM theory exists, they all have a common set of assumptions: Cognitive

processing in the real world varies, situation assessment is critical, mental imagery is

important, the decision-making context must be specified, decision-making is dynamic,

and research should focus on how decision-makers actually function rather than how

they ought to function.

An unparalleled advantage of NDM is the active encouragement for researchers to remain flexible and open minded when exploring large and naturalistic data sets. This

facilitates the discovery of novel or previously missed psychological phenomena through

coincidental observations. For example, the notion of expertise and ‘recognition-primed

decision-making’ was a product of NDM research (Klein, 1997). A series of cognitive

interviews conducted with expert firefighters revealed how skilled decision-makers were

296 Laurence Alison et al.

able to rapidly select a course of action by ‘pattern matching’ between the environment

and the available responses they had cognitively stored. Crucially, Klein (2008) found that

experts tend not to compare options to find the best response but instead evaluate each

option in singular succession until they find one that is good enough. These conclusions on expert cognition have since been replicated in a wide range of domains and have had a

significant impact on the understanding of expertise. For example, Jenkins, Stanton,

Salmon, Walker, and Rafferty (2010) found that experts will ‘leap’ or ‘shunt’ between

decision stages to facilitate rapid and intuitive action, whereas novice decision-makers are

more analytical and linear in their decision-making. Experts are also able to ‘reflect in

action’ during the decision-making process to reduce situation uncertainty and ensure

that their actions remain consistent with the changing environment (Cristancho,

Vanstone, Lingard, LeBel, & Ott, 2013; Sch€on, 1987; van den Heuvel, Alison, & Power, 2014).

Taking the above into account, the topic of interest for the current study was also a

product of naturalistic observations. The authors, after having used NDM to research

critical incident decision-making in the domain of emergency service workers for over

15 years, noticed that when faced with a challenging decision, even experienced

decision-makers often fail to take any action at all. In other words, they appeared to rely

on a cognitive heuristic that suggests when the solution to a problem is unclear, the

safest option is to do nothing. Rather than commit to a choice, they actively (and with considerable cognitive effort) try and delay the implementation of a choice (Eyre,

Alison, Crego, & McLean, 2008; van den Heuvel, Alison, & Crego, 2012). This is distinct

from decision avoidance (Anderson, 2003) where decision-makers refuse to evaluate

choice through passive inaction (e.g., ‘I choose not to decide for the time being’).

Instead, we have observed how decision-makers fail to act through ‘decision inertia’, the

cognitively active and redundant deliberation of choice despite there being a low (or

no) chance of discovering any new information to inform the decision (e.g., ‘I am still

thinking about whether I will commit to, refuse or avoid this choice’). This is important in dynamic crisis environments where often the ‘most-worst’ outcome is the failure to

act as ‘more is missed by not doing than not knowing’ (Byrnes, 2011, p. 28). If a decision

is optimal, then the team can perform well; if the decision is non-optimal, then the team

can re-evaluate, revise, and respond, yet if the decision is not made, then the team drifts

along a trajectory of uncertainty, inaction, and inertia. Not only will this study provide

an in-depth analysis of this novel approach to failures to act, but we will also highlight

how engagement with the NDM paradigm provided a contextualized, meaningful, and

applicable methodology to research and discover the cognitive processes associated with inaction.

Using NDM to research macrocognition in a multiagency emergency team responding

to a simulated sudden impact major disaster

Emergency service workers must operate in highly dynamic, uncertain, and complex

decision-making environments where interoperable team coordination is essential to

facilitate action (House, Power, & Alison, 2014). These decision-making environments are highly unstable, rapid-changing, and unpredictable (van den Heuvel et al., 2014), and

thus, effective team processing requires that ‘members’ interdependent acts convert

inputs to outcomes through cognitive, verbal, and behavioural activities directed towards

organizing task work to achieve collective goals’ (Marks, Mathieu, & Zaccaro, 2001, p.

357). To clarify, macrocognition is the study of ‘how teams move between internalisation

Decision inertia 297

and externalisation of cognition to build knowledge in service of problem solving’ (Fiore

et al., 2010, p. 203). It considers both social and technical influences on team cognition

during the decision process (Schubert, Denmark, Crandall, Grome, & Pappas, 2012) to

increase understanding of collective group cognition during sense making (Stanton et al., 2011). Teams are not only made up of a collection of individuals, but also consist of

‘subteams’ (or ‘agencies’). In multiteam systems, individuals will share and utilize

information to assist both individual (intra-agency) decision-making goals and more

strategic joint (interagency) decisions and actions (Ancona & Chong, 1996). They need to

shift attention between within-group, intrateam processing, and between-group,

interteam coordination (Marks et al., 2001). For example, police officers operate with

their ‘intra-agency’ police colleagues along with their ‘interagency’ emergency response

colleagues (e.g., fire and ambulance services), who have both cohesive and conflicting goals depending on the incident and required outcome.

Teams struggle to make decisions due to poor communication, inconsistent situation

awareness, and conflict of interests (Chen, Sharman, Rao, & Upadhyaya, 2008). Rusman,

van Bruggen, Sloep, and Koper (2010) examined virtually distributed teams and found

that ‘trust’ between team members was based upon judgements of communality, ability,

benevolence, internalized norms, and accountability. Not only does this show the

importance of placing objective trust in another’s ability during teamwork (Mayer, Davis,

& Schoorman, 1995), but it also highlights the importance of subjective ratings of benevolence, norms, and communality that will be derived from past experience with one

another. Keyton and Beck (2010) highlighted how the type and meaning of communi-

cations in teams is based upon an ever-evolving context derived from past interactions

within the team. Past experiences of operating within a specific team influences an

individual’s understanding of their role and team culture (Bearman, Paletz, Orasanu, &

Thomas, 2010) which can influence their decision-making behaviour at both an individual

and collective levels (Marks et al., 2001). Thus, research needs to be sensitive to social

dynamics and embrace subjective and personal experiences. NDM methods are ideal for enabling a deeper exploration of both intra- and intergroup processing at both cognitive

and socially constructed levels.

Exploring how time pressure, team size, and strategic direction interact with

decision-making in emergency service decision-making teams

Emergency incident contexts provide a unique domain in which to apply NDM to study

multiteam systems, cognitive processing, and failures to act. Over the last decade, a considerable number of natural and man-made major emergencies have occurred, which

include flooding, tsunamis, nuclear accidents, and toxic chemical spills. The effects of

these emergency incidents are widespread, affecting a large population, disrupting

infrastructure, and incurring huge costs (emotional, physical, and commercial) in both the

short and long term. Although society recovers, post-incident criticisms of such major

disasters often suggest that there was inadequate pre-planning and delayed responding.

For example, Japan’s Nuclear Accident Independent Investigation Commission empha-

sized how delays in taking action contributed to the inappropriate response that occurred during the Fukushima nuclear disaster, which lead to confusing and conflicting messages

at both a response and civilian level (Kurokawa, 2012). As a result, the magnitude of the

initial damage was greater, and the amount of time taken to recover was longer, due to the

inherent ambiguity and complexity associated with the challenging environment (van

den Heuvel et al., 2014).

298 Laurence Alison et al.

We selected three barriers that, with reference to the literature and anecdotal

evidence, were anticipated to reduce action implementation and instead cause time-

delayed inertia decision-making processes. We predicted that interagency communica-

tions would decrease when choices were non-time-bounded, involved multiple agencies, and lacked clear strategic direction. These ‘barriers’ were relevant to the decision-making

context of emergency incidents as they typically involve multiteam systems who often

hold competing goals and must operate under variable time pressures. Although it is

recognized that conducting NDM research at simulated training events offers the potential

to explore a vast array of other variables (e.g., expertise, decision strategies, behavioural

patterns), we chose to focus on three specific barriers in order to reduce and control the

data set, and to aid our understanding of the cognitive and social processes associated with

failures to act.

Time-bounded choice

A choice can be either time pressured to meet a deadline or non-time-bounded and thus

amenable to deliberation. When an individual is exposed to time-pressured deadlines,

they will try and adapt to time constraints by reducing their generation of potential

options (Alison, Doran, Long, Power, & Humphrey, 2013). Time pressure makes

individuals more risk-seeking and increases the likelihood of choice implementation (Young, Goodie, Hall, & Wu, 2012), but this does not always lead to increased decision

accuracy. NDM research has described how skilled decision-makers can adapt well to

time pressure using efficient and intuitive pattern matching (Alison, van den Heuvel

et al., 2013; Klein, 1997), whereas less skilled decision-makers will utilize faulty and

biased cognitive and heuristic shortcuts to speed up and bypass the decision-making

process (Ask & Granhag, 2007). Yet there has been relatively little research on how the

absence of time pressure influences choice, specifically with regard to how it

influences the likelihood and timeliness of action execution. This may be due to the traditional use of experimental paradigms whereby choices are made salient to the

decision-maker in order to prompt a response, thereby artificially inducing a pressure to

respond.

When a decision is not constrained to deadlines, this means that decision-makers have

the opportunity to avoid or redundantly deliberate on their choice. When decisions are

perceived to be difficult, then individuals will try to avoid committing to a course of action

(van den Heuvel et al., 2012) in order to avoid anticipated negative consequences such as

negative emotions (Anderson, 2003), and when operating within organizational settings, the potential for criticism is due to accountability (Waring, Alison, Cunningham, &

Whitfield, 2013). When someone expects to receive feedback on their actions in team-

based settings, they will tend to avoid choice (Zeelenberg & van Dijk, 1997) due to the

potential for anticipated loss (Crotty & Thompson, 2009), the anticipation of blame (Eyre

et al., 2008), or the perceived inability to personally justify their choice (Brooks, 2011). A

lack of deadlines means that individuals are more likely to use deliberative and alternative-

based processing of choice to establish whether any of their available options are

available/acceptable (Parker & Schrift, 2011), in place of more time-efficient cognitive search strategies, such as satisficing (Simon, 1956), which establishes which of the

available options is the ‘least worst’ or ‘acceptable’. Thus, a complex decision problem

that is ‘non-time-bounded’ has high potential for inertia, and we predicted that this will

principally manifest through reduced action execution and increased redundant

deliberation over already established intra-agency information.

Decision inertia 299

Team size

Communication and coordination are essential for facilitating a team’s macrocognitive

functioning (Keyton & Beck, 2010). We hypothesized that as the number of responding

agencies involved in the decision task increased, the ability to coordinate action would diminish. This is because although ‘interdependent’ tasks require the coordination of

information and resources from multiple players (van de Ven, Delbecq, & Koenig, 1976),

the increase in the number of people involved makes the decision environment more

complex (LePine, Piccolo, Jackson, Mathieu, & Saul, 2008). Team decision-making is

facilitated by shared mental models, where all players hold a common understanding of

the situation and associated actions (Mathieu, Marks, & Zaccaro, 2001). The key

macrocognitive function of a multiteam system is to reduce uncertainty and risk by

establishing and maintaining common ground and shared mental models (Schubert et al., 2012). In multiteam systems, shared mental models not only inform how players should

coordinate within their own team but also help a collective awareness of interagency

goals and actions (Mathieu et al., 2001). Multiteam systems must monitor both inter- and

intrateam processes to provide relevant feedback and instruction to one another

(Dickinson & McIntyre, 1997). Team monitoring and coordination behaviours that help

facilitate action should occur continually to ensure that actions within the multiagency

network are consistent (Brannick, Prince, Prince, & Salas, 1992; Marks et al., 2001).

Establishing a shared mental model is more difficult in large teams. Being a member of a large team reduces an individual’s psychological ‘ownership’ of a problem as responsi-

bility is diffused throughout the wider team network (Kroon, ’t Hart, & van Kreveld,

1991). An individual’s sense of ownership of a problem influences how personally

responsible they feel for the potential consequences of a choice (Pierce, Kostova, & Dirks,

2001). Holding someone individually accountable for collective outcomes makes them

work harder at a team level (Kroon et al., 1991), and thus, psychological ownership

facilitates team work. In large teams, individuals have less ‘ownership’, feel less

accountable, and thus reduce their efforts as lines of communication become exponen- tially larger rather than additive (Kroon et al., 1991). When support from other team

members is lacking, individuals are less likely to process decision tasks at an individual

level (Mueller, 2012). In addition, when operating in hierarchical teams involving

command structures (such as the emergency services), these process losses associated

with increased team size are often underestimated by those in charge (Staats, Milkman, &

Fox, 2012). This could induce erroneous expectations of decision-making capabilities

from those higher up the decision-making command chain. Therefore, we hypothesized

that as team size increased, the likelihood of action would reduce due to associated reductions in psychological ownership of the decision problem, feelings of responsibility,

and diluted communications.

Superordinate goals

‘Superordinate (strategic) goals’ help guide decision-makers to generate and evaluate

different courses of action (Cannon-Bowers, Tannenbaum, Salas, & Volpe, 1995;

Dickinson & McIntyre, 1997). They are ‘goals that are urgent and compelling for all groups involved but whose attainment requires the resources and efforts of more than one

group’ (Sherif, 1962, p. 19). Strategic direction helps to guide the planning of actions to

achieve goals. It can facilitate deliberate planning, wherein principal courses of action for

mission accomplishment are formulated, contingency planning, where alternative plans

are developed to adapt to any anticipated potential events, and reactive planning which is

300 Laurence Alison et al.

in response to unanticipated changes during the incident (Cannon-Bowers et al., 1995;

Marks et al., 2001). In multiteam systems, common goals are critical for interteam

cooperation as it transforms individual units into a powerful collective (Katzenbach &

Smith, 1993). Individuals often hold competing individual and team-based goals that can make decision-making difficult (Wittenbaum, Hollingshead, & Botero, 2004). Clear

strategic direction can reduce cognitive conflict by placing boundaries on decision-

making and making collective aims salient. The setting of superordinate goals can also

help teams to adopt a collective identity, which can reduce the potential for diluted efforts

as a product of large teams and facilitate intergroup cooperation (Driskell, Salas, &

Johnston, 1999).

To establish and reach common strategic goals, a team must exchange reciprocal

information across the network. Individuals need to understand the network as a whole in terms of available resources and expertise, and also be made aware of any

potential constraints in the decision environment (Stout, Cannon-Bowers, Salas, &

Milanovich, 1999). Effective macrocognition in teams is thus dependent upon ‘action

processes’, the acts taken by members that contribute directly to goal accomplishment

(Marks et al., 2001) and help synchronize both individual and collective activities

within complex, dynamic, and unpredictable environments (Tesluk, Mathieu, Zaccaro,

& Marks, 1997). Action processes help with sense making across the team network

(Alberts & Hayes, 2006) as they involve team monitoring and the implementation of synchronized and coordinated actions (Marks et al., 2001). In addition, the commu-

nication of strategic direction must be communicated with explicit meaning and intent

across the multiteam network to achieve a shared awareness (Beck & Keyton, 2009;

Keyton & Beck, 2010). Taking the above into account, we predicted that a lack of clear

strategic direction would reduce the ability to coordinate and communicate decisions

and actions.

Method

Data were collected from a large-scale multiagency (n = 14 agencies) simulation-based training exercise of an aeroplane crash over a major city. It explored the macrocognitive

processing of a multiagency, emergency response team who would be expected to

respond to this type of incident. The training event was conducted using the ‘hydra’

system, an immersive simulated learning platform that can assist organizational training whilst facilitating research. As discussed at length by Alison, van den Heuvel et al. (2013),

simulations provide a useful platform for studying decision-making in multiteam systems

(such as the emergency services) as they provide vast amounts of varied data to explore

the social, organizational, cultural, and political factors that are inherent in such

organizational environments, whilst importantly facilitating their primary purpose in

training decision-makers by exposing them to challenging incidents. Hydra replicates a

traditional ‘tabletop’ simulation exercise by placing trainees (or participants) into small

teams split across separate ‘syndicate rooms’. Those who are running the exercise, in our case a mixture of trainers and researchers, control the simulation from the central ‘control

room’ by feeding relevant information into each syndicate room. As with a traditional

tabletop exercise, participants are split into teams and placed into syndicate rooms

according to the purpose of the simulation. Each syndicate room is equipped with a laptop

computer, a large TV screen to project the computer image, a printer, and a video/sound

recorder.

Decision inertia 301

During the exercise, syndicate rooms are fed information. Information feeds or

‘injects’ may include videos, audio clips, paper print out messages, or even role players.

Syndicate rooms can also request and receive additional information through the

‘communicator’, a chat box that links each syndicate room to the central control room. Facilitators in the control room can respond to information requests from each syndicate

room through this system, monitor decision logs, and pass information between syndicate

rooms at the request of participants. This allows simulations to be fluid, dynamic, and

responsive to the actions of the participants in the syndicate rooms, enabling an in-depth

analysis of immersive and NDM whilst maintaining a level of experimental control.

Participants Participants (n = 194) took part in a two-day simulated exercise of an aeroplane crash over a major city. Participants came from 14 different agencies, including the police

(n = 20), fire (n = 10), ambulance (n = 8), military (n = 3), local authority (n = 12), government offices (n = 15), media (n = 19), health services (n = 15), environment agency (n = 6), utilities (n = 6), science and technical advice (n = 13), and transport (n = 31), along with a multiagency gold group (n = 36). All participants were experienced real-life responders who could be required to respond to a real-life major

incident of this nature (London Emergency Services Liaison Panel, 2007). They were a mixture of strategic (‘gold’) decision-makers (who were placed in one syndicate room and

were tasked with making strategic decisions about the incident) and tactical decision-

makers from each agency (who were split into separate syndicate rooms during the

exercise, one for each agency team).

In addition to the participants in the syndicate rooms, 10 trainers and 19 subject matter

advisors were located in the control room. Advisors were experienced emergency

responders from the local resilience forum who helped design the scenario and were able

to respond to agency-specific information requests. The exercise control team was responsible for feeding information feeds to each syndicate room and responding to

communicator messages.

Procedure

Strategic ‘gold’ commanders from each agency were placed into one syndicate room, and

supporting staff from each agency were split into separate agency-specific ‘syndicate

rooms’ (Figure 1). This set-up was designed to replicate the command structure that would be in operation during a real-life major incident in the United Kingdom. Each

syndicate room received information that reflected what they would receive in a real-life

event. For example, support staff in the ambulance syndicate room received information

regarding casualties, whereas the police support staff received information on cordoning

and evacuation procedures. Agencies communicated to one another by sending logged

messages through the communicator (via the control room). They were also able to

communicate with other members of their own agency not in the syndicate room (e.g.,

those from their agency on the simulated incident ground) also via a message to the control room.

The training event took place over 2 days. Day one of the training event simulated the

multiagency ‘response’ phase, which ran in real time from 8 AM until 10 PM and

represented two simulated days (e.g., ‘day 1 response’ and ‘day 2 response’). The second

exercise day simulated the multiagency ‘recovery’ phase, which ran in real time from 8 AM

302 Laurence Alison et al.

until 12 midday and simulated the transition to ‘day 3 recovery’ (London Emergency

Services Liaison Panel, 2007). During the simulation, participants within separate

syndicate rooms were responsible for sharing and coordinating the information they had received to relevant internal and external bodies through the communicator.

The scenario: Aeroplane collision over a major city

A general timeline was followed by the exercise control room. The control team fed

information about the overall unfolding scenario to all agencies; however, specific

timelines and information feeds were adapted during the simulation based upon the

learning requirements of the exercise and recommendations from advisors. The initial information feed (that went to all agencies) stated that two aeroplanes had crashed above

a local city airport: One crashed into a residential tower and the other crashed into a power

station (causing a large section of the city to lose power). Subsequent feeds involved

updates on traffic congestion, the failure of specific metro lines, information on toxic

plumes and reduced air quality over large and populated areas of the city, updates on

number of victims and casualties involved, and information regarding hospital generators

failing and running out of fuel. Agency-specific information feeds involved, for example,

an intelligence update that terrorism could not be ruled out (police) and that the number of available resources were being stretched due to the severity of the incident (fire and

ambulance).

Data collection

Data were collated from the electronic communication logs that were passed between the

syndicates and control room. Participants used communication logs to request additional

information about the incident from their own or other responding agencies and also to provide information when requested. They used the communicator to coordinate

decisions and actions and request information from their own agency (e.g., advisors) or

other agencies. Participants marked their communications as follows: (1) information

seeking, (2) a decision, or (3) an action. Communications were coded according to

Figure 1. A typical syndicate room used during a ‘Hydra’ simulation.

Decision inertia 303

whether they were ‘intra-agency’ (e.g., fire requesting confirmation of available resources

from control) or ‘interagency’ (e.g., police coordinating triage locations with ambulance).

Communications were coded as intra- or interagency only when they explicitly

mentioned either their own or another agency. Note that ‘intra-agency’ communications does not refer to verbal communications made within syndicate rooms, but relates to

messages concerning their agency that were sent to the control room.

Data analysis

Social network analysis

Social network analysis (SNA) usefully provides an overarching picture of the social

dynamics of team-based settings and can further be used to focus qualitative analysis. We

conducted a SNA of the communication logs to produce an overview of the social

interaction and relationships between the different agencies involved in the exercise

(Knoke & Yang, 2008). Relational SNA shows how central a group is within the network, where they are positioned within the network (which is defined by their relationship or

ties to all others), and how dense the ties are between the groups or people that make up

that network (Mizruchi, 1994).

The initial step in analysis involved inputting frequency data into the network analysis

package, UCINET (Borgatti, Everett, & Freeman, 1992). To analyse the density,

centralization, and hierarchy of the social network, two communication matrices were

produced: One that reflected the presence or absence of communication between

agencies, and another to indicate the total sum of messages between agencies (Table 1). These matrices coded communications across the entire incident, as well as specifically

within the ‘response’ (e.g., days 1 and 2) and ‘recovery’ (e.g., day 3) phases of the

simulation. In addition, a multidimensional-scale technique was used to create a visual

representation of networked communications for both the response and recovery phases

(Figures 2 and 3). This visual representation involves plotting all agencies relative to one

another so that their position within the space reflects how strongly they are socially

connected to others. Agencies that are closely located have more relative association with

each other and represent the key members of that network, whilst agencies on the fringes or periphery of the network are those agencies who communicated less with all others

(Heinz & Manikas, 1992).

Identifying the ‘critical issues’ that required interagency communication

One of the challenges of NDM research is establishing how to judge decision-making

quality (Gore et al., 2006). The SNA provided top-level frequency findings on

communications between agencies, but this does not provide us with information about communication quality. For example, some decisions may require more internal than

external collaboration and thus decentralized networks may counter-intuitively reflect

better quality coordination. To address this, we approached five subject matter experts

(SMEs; three high ranking police officers, one ex-military, and one chief fire officer) who

highlighted six ‘critical issues’ during this scenario that, according to their expertise and

experience, required a high frequency of interagency liaison (Table 2). These SMEs were

external to the exercise and were approached post-exercise (via email). They were

provided with a timeline and information feeds from the exercise in slow time and were asked to identify the ‘critical issues’, where interagency communication was required for

304 Laurence Alison et al.

T a b le

1 . F re q u e n c y o f c o m m u n ic a ti o n s b e tw

e e n (a n d w it h in ) a g e n c ie s

P o li c e 1

P o li c e 2

P o li c e 3

A m b

F ir e

M e d ia

M il it a ry

T ra n s

S T A C

L A

G o v t O ff

H e a lt h

U ti li ti e s

E n v ir o n

G o ld

C o n tr o l

E x

T o ta l

F ir e

8 5

5 7

8 6

3 2

2 2

2 2

2 2

3 5

2 1

1 0

1 6 7

T ra n s

1 0

7 6

3 4

5 9

5 4

2 2

2 2

1 6

2 1

1 0

2 3

1 5 8

H e a lt h

7 4

5 9

2 3

2 2

9 6

3 4 0

6 2

1 7

1 4

2 2

1 5 3

P o li c e 1

7 3

5 5

5 6

6 1

7 1

5 1

1 3

1 6

4 2 0

1 5 0

L A

1 2

6 6

4 4

6 4

4 4

5 3

6 6

7 5

2 7

6 1 4 2

A m b

8 6

6 9 1

1 0

2 4

0 0

0 2

0 0

3 3

8 1 3 4

U ti li ti e s

7 2

2 1

2 1

5 2

1 1

1 1

6 3

1 6

4 7

1 0 7

E n v ir o n

6 0

0 0

1 0

0 5

2 1

0 0

4 2 6

2 5

2 7

7 9

G o v t O ff

4 1

1 1

1 1

1 9

3 8

2 1

3 1

2 8

2 4

7 1

M e d ia

8 3

3 3

4 2

2 2

2 3

2 2

4 2

1 3

0 4 6

P o li c e 3

5 3

1 4

0 3

0 0

1 1

1 0

0 1

1 2

2 6

4 0

P o li c e 2

4 2 0

5 1

0 0

0 5

0 0

0 0

0 0

2 0

1 3 8

S T A C

2 2

2 2

5 2

1 1

3 2

1 2

2 1

1 2

4 3 5

M il it a ry

0 0

0 0

0 0

2 5

0 0

0 0

0 0

0 1

1 5

3 2

G o ld

6 5

3 4

5 0

1 1

0 2

1 0

0 0

0 1

2 3 1

T o ta l

1 6 0

6 9

6 3

1 3 1

1 2 4

2 9

5 5

9 9

3 0

8 6

2 1

5 9

1 1 1

4 5

1 5

8 5

1 6 5

1 3 4 7

N o te . A m b , a m b u la n c e se rv ic e ; T ra n s, tr a n sp o rt ; S T A C , sc ie n c e a n d te c h n ic a l a d v ic e c e ll ; L A , lo c a l a u th o ri ty ; G o v t o ff , g o v e rn m e n t o ffi c e ; U ti l, u ti li ti e s; E n v ir o n ,

e n v ir o n m e n t a g e n c y ; G o ld , g o ld g ro u p ; C o n t, c o n tr o l; E x t, e x te rn a l.

Decision inertia 305

success (or ‘good’ decision-making). An issue was highlighted as ‘critical’ if four of the five

SMEs identified and agreed that it was. SMEs also identified which agencies should be

involved; only five of 14 agencies were identified as requiring involvement with these

critical issues (police, fire and rescue, ambulance, military, and local authority) which the

authors suspect is due to their need to be involved in the dynamic response phase of the

incident. Each ‘critical issue’ was also coded for the prevalence of our three barriers: (1) non-time-bounded problem, (2) large number of agencies involved, (3) and/or lacked

Figure 3. MA network during the response phase of incident management.

Figure 2. The interaction between communication content and audience on communication frequency.

306 Laurence Alison et al.

clear strategic direction. Qualitative analyses of communications about these critical

issues explored how ‘barriers’ and interagency discussions interacted. Quantitative

analyses of to whom these communications were made was also conducted to establish

Table 2. The six critical issues identified by the subject matter experts

Critical issue

(agencies involved) Definition Optimal response

Resource

coordination

(police, fire,

ambulance, local

authority,

military)

The timely resolution of any major

incident requires agencies to

continually monitor and implement

intra-agency resources whilst sharing

this information and receiving

information from other agencies

Sharing of resource information

should occur continually, but it is

especially important during the

initial response phase

Casualty triage

(police, fire,

ambulance)

Participants had to consider how to

coordinate the recovery of casualties

and bodies during the response phase

A significant amount of coordination

regarding the issue should occur at

the tactical (bronze) level on scene.

Strategic decisions and actions with

regard to the distribution of

resources by police, fire, and

ambulance at ‘gold’ level is also

required

Communication

channels (police,

fire, ambulance,

local authority)

During the initial response phase, the

communication (‘airwave’) channels

were disrupted. The police received a

direct request from a tactical officer on

whether or not they should implement

the ‘Mobile Telecommunication

Privileged Access Scheme (MTPAS)’ – A UK procedure for prioritizing mobile

telephone networks for emergency

services

Collaboratively decide to invoke the

scheme and provide interagency

rationale for the decision

Reception centre

location (police,

local authority)

This is essential for ensuring the welfare

of those civilians affected by the

incident

There needs to be ongoing review of

the needs of civilians. This should be

primarily led by the local authority

as it feeds into the later recovery

phase when attempting to restore

normality

Emergency

evacuation

(police, fire,

ambulance, local

authority)

All agencies received information that a

smoke plume caused by the aeroplane

crash might potentially be toxic and

thus hazardous to responders and

civilians in affected area

There needs to be discussion

amongst blue light responders and

the local authority, with advice

from science and technical advisors

(STAC) or the airline investigation

branch. A joint decision must then

be made on whether or not to

evacuate both responders and

civilians from the affected areas

Handover of

primacy (police,

local authority)

At some point, the police must hand

over the primary management of the

incident to the local authority

This indicates that the response

phase is over the recovery phase

has begun. Thus, the timing of this

decision is crucial

Decision inertia 307

whether the response was ‘interoperable’. It was classed as ‘interoperable’ if more than

50% (i.e., the majority) of logged communications about this issue were ‘interagency’.

Results

Descriptive analyses of inter- and intra-agency communication frequency

A repeated measures ANOVA found that across the incident as a whole agencies made

significantly more intra-agency communications (M = 68.42, SD = 41.52) than intera- gency communications (M = 33.00, SD = 20.13), F(1, 11) = 13.581, p = .004. There was a significant difference in the number of communications made depending upon their content, F(2, 22) = 6.246, p = .007. Significantly more messages discussed information seeking (M = 55.58, SD = 35.82) than decisions (M = 30.00, SD = 22.68) or actions (M = 16.17, SD = 28.229). There was a significant interaction between communication content, audience, and frequency, F(1.23, 13.58) = 6.871, p = .016, as information- seeking communications were especially intra-agency (Figure 2).

Social network analysis UNICET was used to calculate ‘network density’ and how embedded or isolated each

‘node’ (e.g., agency) was within the wider network (Hanneman & Riddle, 2005). The

average density value was 3.45, which represented an average density tie of 59.5%. This

meant that, across the entire incident, 59.5% of all possible reciprocal communications

between agencies occurred with an average number of messages between agency pairs of

3.5 messages. This indicated that 40.5% of possible communications between agencies

were not made. Low density occurred because agencies focused on contacting individuals

within their own organizations instead of the network (Table 1). The highest number of interagency communications (n = 16; transport and utilities) was less than a quarter of the number of intra-agency communications (n = 86; fire and rescue). Network density was significantly higher during the response phase of the incident (density = 2.95) compared to the recovery phase (density = 0.49), t(0.61) = 2.61, p < .05, as 75.4% of all possible communications were made during response compared to only 27.1% during

recovery (Table 3). This pattern was also illustrated in the visual representations of the

network, as the response phase had more (strong) ties between pairs of agencies

(Figure 3) than the recovery phase (Figure 4). Centralization measures how well connected a node is with the rest of the network. In-

degree centrality reflects the number of communications leading into that node, whereas

out-degree centrality reflects the number of communications going out from that node

(Hanneman & Riddle, 2005). High in-degree centralization indicates that a small number

of agencies are being consulted by all others, whereas high out-degree centralization

indicates that a small group of agencies are sending information out (Scott, 2001). The

Table 3. Average density values and percentage of present ties across network for entire incident and

per incident phase

Density Density: % of ties

Entire incident 3.45 59.5

Response phase 2.95 75.4

Recovery phase 0.49 27.1

308 Laurence Alison et al.

multiagency network was found to be decentralized as there were relatively low levels of

in- and out-degree centralization, although in-degree asymmetry was almost twice as high

as out-degree (28.5% and 15.7%, respectively, see Table 4). The network was more

centralized during the initial response phase (in-degree = 25.49% and out- degree = 17.12%) than the recovery phase (in-degree = 10.82% and out- degree = 6.25%). Table 5 presents each agency’s itemized centrality for both phases of the incident and across the entire incident. In addition, the SNA found no evidence of a

hierarchy within the network at any phase of the incident or overall (Table 4). This meant

that no single agency commanded the incident as a key player (albeit this finding would be

expected as a product of the UK Gold Command structure).

Interoperability during six ‘critical issues’

Critical issues were coded for the presence or absence of our three hypothesized ‘barriers’ to multiagency decision-making, namely whether the issue was ongoing in duration (e.g.,

lacked a deadline), involved three or more agencies, and/or lacked a clear superordinate

goal (e.g., lacked strategic direction from gold; Table 6). Although there was no significant

correlation between the number of barriers present and the frequency of interagency

communications, there was an interesting trend in the data. It was found that barrier

frequency was negatively correlated to interagency communication frequency

(r = �.542, p = .266); as the number of barriers increased, communication with external agencies decreased. Likewise, a nonsignificant positive correlation was found between

Figure 4. MA network during the recovery phase of incident management.

Table 4. Hierarchy and centralization degrees for the entire incident and per incident phase

Hierarchy Centralization – In-degree (%) Centralization – Out-degree (%)

Entire incident 0 28.57 15.78

Response phase 0 25.49 17.12

Recovery phase 0 10.82 6.25

Decision inertia 309

barrier frequency and intra-agency communication (r = .691, p = .128); as the number of barriers increased, the number of within-agency communication also increased. This

suggests that when decision-making is more complex (e.g., increased barriers),

communication with other agencies decreases and internal-agency communication

increases. It is possible that the lack of statistical significance in this trend was due to the

lack of power generated by our single (albeit large) naturalistic data set. Further research

in more experimental and repeatable settings may help to clarify these effects.

Qualitative analyses of critical issues characterized by where the intra-agency communications:

Resource coordination and casualty triage

Table 6 shows how the two ‘critical issues’ that had all three barriers present were the

only two critical issues were communications were predominantly intra-agency focussed

(i.e., not intereoperable). Both ‘resource coordination’ and ‘casualty triage’ critical issues

had ongoing duration (e.g., lacked a deadline for a final decision and action), involved

more than two agencies, and, following analysis of the communications sent from the ‘gold’ strategic syndicate, had no superordinate goals. The only reference made to

‘resources’ in the gold syndicate was unclear: ‘Police gold: Scene management and

resourcing at scene’, and communications on ‘casualty triage’ focused on information

seeking rather than setting necessary direction: ‘Fatalities. What is the scale of the incident

we are dealing with?’; ‘Bodies left on scene at present? Those at hospital being dealt with’.

Thus, both issues were considered by the SMEs to lack strategic direction.

Analyses of communication content at the tactical level supported our earlier finding

that agencies focused on intra-agency matters and that these communications were

Table 5. Itemized in-degree and out-degree centrality for each agency for the entire incident and per

incident phase

Entire incident Response Recovery

Incoming Outgoing Incoming Outgoing Incoming Outgoing

Police 1 10.60 8.82 5.29 6.53 2.40 2.29

Local Authority 5.73 8.29 1.79 6.88 1.60 1.47

Police 2 4.60 2.24 2.86 2.12 0.67 0.12

Police 3 4.20 2.35 3.00 1.82 0.67 0.53

Fire 8.27 9.82 2.43 9.06 0.73 0.77

Ambulance 8.73 7.88 2.50 7.88 0.06 0.06

Transport 6.60 9.29 2.86 9.00 0.47 0.29

Health 3.93 9.00 1.29 6.88 0.87 1.88

Utilities 7.40 6.29 2.86 4.53 1.07 1.77

Environment 3.00 4.65 1.29 2.82 0.53 1.71

Government Office 1.40 4.18 1.29 3.82 0.13 0.41

Military 3.67 1.88 1.93 1.77 0.33 0.12

Science and Technical

Advice Cell

2.00 2.06 1.86 2.00 0.07 0.06

Media 1.93 2.71 1.71 2.59 0.20 0.12

Gold Group (Strategic

Coordinating Group - SCG)

3.40 1.82 3.21 1.71 0.40 0.12

External 11.00 N/A 8.67 N/A 2.20 N/A

Control 5.67 N/A 4.87 N/A 0.87 N/A

310 Laurence Alison et al.

T a b le

6 . T h e n u m b e r o f b a rr ie rs

a n d fr e q u e n c y o f c o m m u n ic a ti o n s fo r e a c h c ri ti c a l is su e

N o n -t im e -

b o u n d e d (Y /N

)?

M o re

th a n

2 a g e n c ie s (Y /N

)?

P o o r st ra te g ic

d ir e c ti o n (Y /N

)?

T o ta l n u m b e r

o f b a rr ie rs

F re q u e n c y o f

c o m m u n ic a ti o n s

D id in te ro p e ra b il it y

o c c u r (%

o f in te ra g e n c y

c o m m u n ic a ti o n s) ?

In tr a -a g e n c y

In te ra g e n c y

R e so u rc e

c o o rd in a ti o n

Y Y

Y 3

3 4

8 N o (1 9 )

C a su a lt y tr ia g e

Y Y

Y 3

1 4

1 0

N o (4 2 )

C o m m u n ic a ti o n

c h a n n e l fa il u re

N Y

Y 2

4 4

Y e s (5 0 )

R e c e p ti o n c e n tr e

lo c a ti o n

Y N

N 1

9 1 3

Y e s (5 9 )

E m e rg e n c y

e v a c u a ti o n

N Y

Y 2

8 1 2

Y e s (6 0 )

P ri m a c y h a n d o v e r

Y N

N 1

4 1 4

Y e s (7 8 )

Decision inertia 311

predominantly concerned with information seeking (Figure 5). For example, the

ambulance service focused on information seeking about their own resources: ‘How is

ambulance coping with core business? i.e. availability of resources?’; as did the fire and

rescue service: ‘All bulk foam units are currently in use. What reserve stocks of foam do we have and do we need to order more?’ For the casualty triage issue, it was also found that

agencies were primarily concerned with information about internal resources: ‘Request

information from ICR regarding our casualty figures from all sites’. Despite this, the SMEs

recommended that communication between agencies should be interagency focused to

reach a successful resolution. As was found, an intra-agency focus on internal resources

reduced the ability of agencies to coordinate resources on scene. Likewise, agencies

discussed casualty triage by gathering information from one another, yet failed to

coordinate their efforts on scene to facilitate action. Both issues were characterized by the three identified barriers, and both issues induced a failure to communicate between

agencies and execute action.

Qualitative analyses of critical issues characterized by interagency communication: Communication

channel failure, reception centre location, emergency evacuation, and primacy handover

Discussions about the ‘communication channel failure’, ‘reception centre location’,

‘emergency evacuation’, and ‘primacy handover’ were predominantly interagency (i.e., >50% of communications were interagency) and tended to focus less on information seeking and more on action execution (Figure 5). The only exception was the

‘communication channel failure’ issue (Table 6), as this issue was minimally discussed

(n = 8) and only just reached the 50% interagency communication margin. In terms of the barriers that were present for these issues, there were mixed findings. As mentioned, the

frequency of interagency communications was negatively, albeit non-significantly,

associated with the number of barriers. It is likely that this non-significant trend was in

part due to the mixed findings when the number of barriers dropped below 3. Both team size and duration of issue are barriers to interoperability that are difficult to

influence from a solutions perspective. As such, we focused on the setting of

0

5

10

15

20

25

30

C om

m un

ic at

io n

fr eq

ue nc

y

Critical issue

Information seeking

Decision

Action

Figure 5. The frequency of communications made for each critical issue broken down by communi-

cation content.

312 Laurence Alison et al.

superordinate goals from ‘gold’ decision-makers for our analysis. Although gold

acknowledged communication channel failures as an issue that needed to be resolved:

‘Mobile phone network may not be working. Not sure how to communicate to

community’, they did not outline a strategic goal. Likewise, the emergency evacuation issue lacked clear strategic direction as commanders focused on information seeking over

setting goals: ‘Need to have update re smoke/direction and intensity etc’. This lack of

superordinate strategy further impacted tactical discussions as responders focused on

reacting to the impact of the issue rather than generating tactics for resolution or to

prevent further escalation or harm: ‘boroughs need to be mindful of the lack of

communication methods due to the loss of power’; ‘Contact AAIB for specialist advice

regarding hazards contained within smoke plume’. Thus, although there was a high

proportion of interagency communication, they did not induce action. As indicated in Figure 5, the content of communications for these two issues was fairly evenly dispersed

between information seeking, decision-making, and action execution. Although they

were not dominated by information seeking as was found for the two intra-agency issues,

the proportion of communications was less focused on action execution than their

interagency counterpart issues (e.g., reception centre location, primacy handover).

Fundamentally, a lack of strategic direction diluted the focus of discussions on action

execution.

In contrast, the ‘reception centre location’ and ‘primacy handover’ issues did have clear strategic direction. For the reception centre location issue, gold communicated a

clear goal: ‘Agree to active humanitarian centre’ and further specified who should take

charge: ‘LA [local authority] to organise humanitarian centre’. This meant that the local

authority was able to take charge of the tactical response and coordinate action with other

agencies: ‘LA understands police have been posed question about flowers and other

tributes and are identifying suitable areas. Expect contact soon on this’. This was similar

for the primacy handover issue: ‘LA have the lead. Key messages, major tragedy. All

agencies have worked together; now return to normality’. Interestingly, although this message was not communicated until the very end of the response phase of the incident,

the LA (in accordance with their role) had already begun contingency planning for taking

primacy: ‘Recovery group meeting convened at 13.00. Rationale: To initiate process and

an impact assessment in the early stages and to horizon scan for the next 48 hours’.

Communications predominantly focused on action execution rather than redundant

information seeking (Figure 5). Thus, issues that were guided by strategy from gold were

dealt with through increased interagency communications (as recommended by the

SMEs) that specifically focused on the useful execution of actions.

Discussion

NDM can be used in a novel way to explore the intricate cognitive and social processes to

understand failures to act. Data were collected from an immersive simulated training

exercise of the emergency response to a major disaster. The network as a whole was decentralized and communications tended to be intra-agency-focused and dominated by

information seeking. SMEs identified six critical issues that required interagency

communication for a successful resolution (e.g., decisions needed to be coordinated

jointly between agencies). If critical issues were non-time-bounded, involved more than

two agencies, and lacked clear strategic direction, then communications predominantly

involved redundant intra-agency information seeking, as opposed to the useful

Decision inertia 313

interagency coordination of decisions and actions (as recommended by the SMEs). The

decision-making process was delayed due to decision inertia. However, when one or more

of these barriers was missing, interagency communications increased and discussions

were more focused on decisions and actions. Specifically, when issues had clear strategic direction, they were associated with action execution. We argue that the relatively under

explored notion of ‘decision inertia’ and the contexts that provoke it warrant further

attention, as it is a salient problem in real-world decision-making. Given the repeated

criticism of the emergency services for failures to act rather than for errors of judgment,

we suggest that specific attention be directed at this interesting phenomenon of what

Anderson (2003) more broadly describes as ‘The Psychology of Doing Nothing’.

Limitations

It is difficult to dispute the academic integrity of the NDM movement in extending our

understanding of real-world choice, yet there is an inherent methodological problem in

deriving conclusions from such uncontrolled, complex, and dynamic contexts. By

researching ‘natural’ environments where the researcher cannot manipulate rigorously

defined independent variables, nor quantify objective dependent outcomes, the

objectivity of research conclusions is reduced. There is a risk that researchers’ interests

could bias the interpretation of the data, creating the potential for unreliable or invalid findings (Lipshitz et al., 2001; McAndrew & Gore, 2013), as researchers continually

return to and attempt to explain the data, thereby risking post-hoc rationalization and

assumed significance.Likewise, it is possible that the ability to immerse oneself in the data,

with the freedom for methodological flexibility, is a strength for NDM research that

primarily desires pragmatic solutions derived from the data to help practitioners.

It has been suggested that a useful way to improve objectivity may be through the

use of simulations, whereby researchers can maintain a level of control over the scenario

and data recordings (Alison, van den Heuvel et al., 2013). Yet this may come at the cost of reduced stress and cognitive demand on the decision-maker compared to real life. The

behaviour of participants may also be influenced by the simulated environment as, for

example, participants focus more on information seeking over action implementation as

this is perceived as the main purpose of the communicator. Alternatively, the traditional

laboratory-based decision-making environments may improve objectivity of conclusions,

but this comes at the cost of artificially clean decision-making environments that assume

pure rationality of thought (Thwaites & Williams, 2006). It is noteworthy that research

on emergency decision-making is notoriously complex and might involve many variables, such as strategic group-level presentation (Van Leeuwen, 2007), intricate

power relations (Nadler & Halabi, 2006), perceived dangerousness of emergencies

(Fischer et al., 2011), and a whole spectrum of non-verbal factors, such as the body

language and facial expression of decision-makers, and a number of serious ethical

considerations would also arise (Binik, Mah, & Kiesler, 1999). Furthermore, we should

bear in mind that even the currently available virtual reality systems offer only crude

approximations of natural experiences and only half-credible computer-generated

responses (Rovira, Swapp, Spanlang, & Slater, 2009). To deal with (but not to completely overcome) this challenge, we adopted a complementary qualitative method

that could bring other kinds of subtle insights, which are sometimes out of scope for

experimental methods (Bryman, 2007), to light.

Thus, we suggest that rather than judge research according to how ecologically ‘valid’

or ‘objective’ it is, it should be judged according to the ‘credibility’ and ‘transferability’ of

314 Laurence Alison et al.

the conclusions drawn from the context-rich environment (Mischler, 1990). Kuhn (1962)

highlights how the definition of ‘science’ is a culturally defined phenomenon according to

the benchmarks set by the scientists of the day; thus, NDM may reflect a paradigm shift

from a top-down experimental focus to a more bottom-up focus on contextualized enquiry. Overall, although we acknowledge that there are methodological concerns for

NDM and simulation-based research, we argue that the unique advantage of context and

grounded meaning to conclusions is an unparalleled advantage of this research domain.

Three barriers to macrocognition in multiteam systems: Non-time-bounded choice,

large team size, and a lack of superordinate direction

We found that when a decision task was non-time-bounded, involved multiple agencies, and lacked clear strategic direction, then interagency action execution failed and decision

inertia occurred through the redundant deliberation of intra-agency information.

Interestingly, the only two issues where the majority of communications were intra-

agency and focused on information seeking were also the only two issues that were

characterized by all three decision barriers. It is acknowledged that whilst intra-agency

resource monitoring is useful during a major incident response (National Police

Improvement Agency, 2009), interagency coordination and information sharing is

equally critical, and further that communications on decisions and actions will inevitably be less than those on information seeking; as the latter informs the former. To minimize

these effects, we focussed on the six ‘critical issues’ that SMEs identified as requiring

interagency communication.

The effectiveness of a multiteam system is dependent upon the sharing of task

interdependence across the team network (De Dreu, 2007), and thus, an overt focus on

intra-agency information can distract decision-makers from collective goals and

behaviours (Comfort, 2007; De Dreu, 2007). By solely communicating agency-specific

information, there is the risk of an ‘information starvation’ effect within the multiteam system, where important information is not shared with the wider team, a prevalent

problem during emergency response (Netten, Bruinsma, van Someren, & de Hoog, 2006).

Although a decentralized approach to multiteam systems may usefully allow experts to

focus on their own specialisms (House et al., 2014), if this occurs in the absence of

interagency coordination of efforts, then this means that key decisions and actions may be

missed or redundantly duplicated (Brannick et al., 1992). Interagency communication is

essential to facilitate a shared understanding of the requirements of the dynamic decision

environment. One barrier that we predicted would reduce decisions and actions was a lack of

deadlines or specific time demands. Individuals avoid difficult choices when possible (van

den Heuvel et al., 2012) and so when critical issues were non-time-bounded, we found

that teams failed to communicate interagency decisions and actions. We acknowledge, as

with most NDM research, that we cannot conclude a causal relationship between time

demands and decision outcomes, but existing literature may help explain it. It is possible

that a lack of deadlines induced inaction as teams tried to avoid potential anticipated

negative emotions, such as anticipated regret (Anderson, 2003). Individuals may have utilized more selective search strategies when evaluating options because they felt they

had the time to do so. Indeed, when decision-makers are placed under time pressure, then

they tend to use time-saving processes such as satisficing to reach a choice (Simon, 1956);

yet when deadlines are absent, individuals use more selective alternative-based

processing to continually deliberate on potential courses of action to find the ‘best’

Decision inertia 315

option (Parker & Schrift, 2011). We suggest that when a critical issue was non-time-

bounded, teams redundantly sought further information about the choice from within

their own agency, rather than committing to a joint choice with other agencies.

The involvement of more than two agencies also reduced communication on decisions and actions. It is possible that an increase in team size reduced the ‘psychological

ownership’ over decision outcomes due to diluted feelings of responsibility and

accountability (Kroon et al., 1991; Pierce et al., 2001). Agencies must keep one another

informed of their actions in multiteam systems to orchestrate the smooth sequencing of

activity (Healey, Hodgkinson, & Teo, 2009), yet when responsibility is diluted, roles and

responsibilities are confused (Bearman et al., 2010). Role confusion can permeate both

intra- and interagency action to reduce the overall effectiveness of the multiagency team

(Quarantelli, 1985). An increase in team size may further decrease the level of trust throughout the team network. Trust is reflected by a willingness of the trustor to take risks

based upon advice received from the trustee (Mayer et al., 1995) and within multiteam

networks the building of trust is based upon reciprocity (Rusman et al., 2010). When

team size was larger, agencies may have focused on information concerning their own

team due to poor trust within the wider multiteam network. Fundamentally, any increase

in interagency team size means that individuals can ‘hide in the crowd’ and avoid taking

responsibility for action (Kroon et al., 1991).

A further barrier associated with reduced interagency action was when there was a lack of clear superordinate direction (LePine et al., 2008). Interestingly, not only was a

lack of strategic direction associated with intra-agency information seeking, but when

strategic direction was present, this appeared to increase interagency communications on

‘actions’, suggesting a bidirectional relationship. In other words, a lack of strategy reduced

action, whilst strategic direction facilitated action. Unclear goals at a strategic level may

have reduced cooperation as decision-makers were distracted by their own intra-agency

goals (de Bruijn, 2006). Intra- and interagency goals tend to compete (Sonnenwald &

Pierce, 2000) and so if strategic goals are unclear, then intra-agency goals are more salient. It was found that when strategic goals were set, discussions became focused on facilitating

action, which was judged by the SMEs as necessary. Effective communications are

essential for facilitating action in multiteam systems where different backgrounds and

cultures must work together towards an effective resolution (Bearman et al., 2010) and so

strategic goals must be set with clear meaning and intent (Keyton & Beck, 2010). This is

achieved by establishing reciprocal relationships to develop a shared understanding and

team cognition (Cooke, Gorman, & Rowe, 2009). Thus, when strategic direction is lacking

either completely or due to implicit assumptions, then teamwork failed to be interdependent and goal-focused (Salas, Cook, & Rosen, 2008); yet when strategy is

clear, it can facilitate the implementation of action in multiteam networks.

Conclusion and Implications

Our research has both methodological and applied implications. Firstly, we demonstrated

how NDM can be used to research failures to make decisions. Whereas traditional

laboratory-based decision experiments treat the absence of choice as a null effect, NDM’s emphasis on context and real-world choice allows us to explore the dynamic and

interrelating processing associated with decision inertia. In addition, our research

provides a worked example of how NDM allows for both practitioners and academics can

combine training and research to derive mutually beneficial conclusions. Secondly, we

found that the barriers of non-time-bounded choice, large team size, and a failure to set

316 Laurence Alison et al.

clear strategic direction can inhibit multiteam system decision-making, a finding that

requires further investigation. For example, experts can adapt to time pressure efficiently

using satisficing strategies (Klein, 2008), and so it would be interesting to explore whether

experts are more willing to take action despite the absence of deadlines or whether they will continue to avoid choice if the opportunity arises. It may also be possible to conduct

cognitive task analysis (CTA) style interviews in order to unpack how these variables

influence the cognitive processing of emergency workers in real life (Crandall et al.,

2006) or to compare differences in decision-making styles between ‘critical’ decisions and

those that are less ‘high-stake’. As strategic goals facilitated interagency communications

of decisions and actions, this highlights the importance of superordinate goals during

emergency response command and control, creating clear implications for effective

training of practitioners.

References

Alberts, D. S., & Hayes, R. E. (2006). Understanding command and control. CCRP Publication

Series. Washington, DC: CCRP.

Alison, L., Doran, B., Long, M., Power, N., & Humphrey, A. (2013). The effects of subjective time

pressure and individual differences on hypotheses generation and action prioritization in police

investigations. Journal of Experimental Psychology: Applied, 19(1), 83–93. doi:10.1037/ a0032148

Alison, L., van den Heuvel, C., Waring, S., Power, N., Long, A., O’Hara, T., & Crego, J. (2013).

Immersive simulated learning environments for researching critical incidents: A knowledge

synthesis of the literature and experiences of studying high-risk strategic decision making.

Journal of Cognitive Engineering and Decision Making, 7, 255–272. doi:10.1177/ 1555343412468113

Ancona, D. G., & Chong, C. (1996). Entrainment: Pace, cycle, and rhythm in organizational behavior.

In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (pp. 251–284). Greenwich, CT: JAI Press.

Anderson, C. J. (2003). The psychology of doing nothing: Forms of decision avoidance result from

reason and emotion. Psychological Bulletin, 129(1), 139–167. doi:10.1037/0033-2909.129. 1.139

Ask, K., & Granhag, P. A. (2007). Motivational bias in criminal investigators’ judgements of witness

reliability. Journal of Applied Psychology, 37, 561–591. doi:10.1111/j.1559-1816.2007.00175.x Bearman, C., Paletz, S. B. F., Orasanu, J., & Thomas, M. J. W. (2010). The breakdown of coordinated

decision making in distributed systems. Human Factors, 52(2), 173–188. doi:10.1177/ 0018720810372104

Beck, S. J., & Keyton, J. (2009). Perceiving strategic meeting interaction. Small Group Research, 40,

223–246. doi:10.1177/1046496408330084 Binik, Y. M., Mah, M., & Kiesler, S. (1999). Ethical issues in conducting sex research on the Internet.

The Journal of Sex Research, 36(1), 82–90. doi:10.1080/00224499909551971 Borgatti, S. P., Everett, M. G., & Freeman, L. C. (1992). Ucinet IV: Network analysis software.

Connections, 15, 12–15. Brannick, M. T., Prince, A., Prince, C., & Salas, E. (1992). Team coordination: Measuring an elusive

construct. Paper presented at the meeting of the Human Factors Society, Atlanta, GA.

Brooks, M. E. (2011). Management indecision. Management Decision, 49, 683–693. doi:10.1108/ 00251741111130788

Bryman, A. (2007). Barriers to integrating quantitative and qualitative research. Journal of Mixed

Methods Research, 1(1), 8–22. doi:10.1177/2345678906290531 Byrnes, P. D. (2011). Why haven’t I changed that? Therapeutic inertia in general practice.

Australian Family Physician, 40(1/2), 24–28.

Decision inertia 317

Cannon-Bowers, J. A., Tannenbaum, S. I., Salas, E., & Volpe, C. E. (1995). Defining Competencies and

Establishing Team Training Requirements. In R. A. Guzzo & E. Salas (Eds.), Team effectiveness

and decision making in organization (pp. 333–380). San Francisco, CA: Jossey-Bass. Chen, R., Sharman, R., Rao, R., & Upadhyaya, S. J. (2008). Coordination in emergency response

management. Communications of the ACM, 51(5), 66–73. doi:10.1145/1342327.1342340 Comfort, L. K. (2007, December). Crisis Management in Hindsight: Cognition, Communication,

Coordination, and Control. Public Administration Review, 67(1), 189–197. doi:10.1111/ j.1540-6210.2007.00827.x

Cooke, N. J., Gorman, J. C., & Rowe, L. J. (2009). An ecological perspective on team cognition. In E.

Salas, J. Goodwin & C. S. Burke (Eds.), Team effectiveness in complex organisations: Cross-

disciplinary perspectives and approaches (pp. 157–182). Hove, UK: SIOP Organizational Frontiers Series. Taylor & Francis.

Crandall, B., Klein, G., & Hoffman, R. R. (2006). Working minds: A practitioner’s guide to cognitive

task analysis. Cambridge, MA: MIT Press.

Cristancho, S. M., Vanstone, M., Lingard, L., LeBel, M. E., & Ott, M. (2013). When surgeons face

intraoperative challenges: A naturalistic model of surgical decision making. The American

Journal of Surgery, 205, 156–162. doi:10.1016/j.amjsurg.2012.10.005 Crotty, S. K., & Thompson, L. (2009). When your heart isn’t smart: How different types of regret

change decisions and profits. International Journal of Conflict Management, 20, 315–339. doi:10.1108/10444060910991048

de Bruijn, H. (2006). One Fight, One Team: The 9/11 Commission report on Intelligence,

Fragmentation, and Information. Public Administration, 84, 267–287. doi:10.1111/j.1467- 9299.2006.00002.x

De Dreu, C. K. W. (2007). Cooperative outcome interdependence, task reflexivity and team

effectiveness: A motivated information processing approach. Journal of Applied Psychology,

92, 628–638. doi:10.1037/0021-9010.92.3.628 Dickinson, T. L., & McIntyre, R. M. (1997). A conceptual framework for teamwork measurement. In

M. T. Brannick, E. Salas & C. Prince (Eds.), Team performance assessment and measurement

(1st ed., pp. 19–44). Hillsdale, NJ: Lawrence Erlbaum. Driskell, J. E., Salas, E., & Johnston, J. (1999). Does stress lead to a loss of team perspective? Group

Dynamics: Theory, Research and Practice, 3, 291–302. doi:10.7771/2327-2937.1008 Eyre, M., Alison, L., Crego, J., & McLean, C. (2008). Decision inertia: The impact of organisations on

critical incident decision-making. In L. J. Alison & J. Crego (Eds.), Policing critical incidents:

Leadership and critical incident management (pp. 201–239). Devon, UK: Willan. Fiore, S. M., Rosen, M. A., Smith-Jentsch, K. A., Salas, E., Letsky, M., & Warner, N. (2010). Toward an

understanding of macrocognition in teams: Predicting processes in complex collaborative

contexts. Human Factors, 52, 203–224. doi:10.1177/0018720810369807 Fischer, P., Krueger, J. I., Greitemeyer, T., Vogrincic, C., Kastenm€uller, A., Frey, D., . . . Kainbacher,

M. (2011). The bystander-effect: A meta-analytic review on bystander intervention in dangerous

and non-dangerous emergencies. Psychological Bulletin, 137, 517–537. doi:10.1037/a0023304 Gore, J., Banks, A., Millward, L., & Kyraikidou, O. (2006). Naturalistic decision making: Reviewing

pragmatic science. Organization Studies, 27, 925–942. doi:10.1177/0170840606065701 Hanneman, R. A., & Riddle, M. (2005). Introduction to social network methods. Riverside, CA:

University of California.

Healey, M. P., Hodgkinson, G. P., & Teo, S. (2009, June). Responding effectively to Civil

Emergencies: The Role of Transactive Memory in the Performance of Multiteam Systems.

Proceedings of NDM9, the 9th International Conference on Naturalistic Decision Making.

London, UK.

Heinz, J. P., & Manikas, P. M. (1992). Networks among elites in a local criminal justice system. Law &

Society Review, 26, 831–861. doi:10.2307/3053820 House, A., Power, N., & Alison, L. (2014). A systematic review of the potential hurdles of

interoperability to the emergency services in major incidents: Recommendations for solutions

318 Laurence Alison et al.

and alternatives. Cognition, Technology and Work, 16(3), 319–335. doi:10.1007/s10111-013- 0259-6

Jenkins, D. P., Stanton, N. A., Salmon, P. A., Walker, G. H., & Rafferty, L. (2010). Using the

decision-ladder to add a formative element to naturalistic decision-making research.

International Journal of Human-Computer Interaction, 26(2–3), 132–146. doi:10.1080/ 10447310903498700

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decisions under risk.

Econometrika, 47, 263–291. doi:10.2307/1914185 Katzenbach, J. R., & Smith, D. K. (1993). The wisdom of teams: Creating the high-performance

organization. Boston, MA: Harvard Business School.

Keyton, J., & Beck, S. J. (2010). Perspective: Examining communication as macrocognition in STS.

Human Factors, 52, 335–339. doi:10.1177/0018720810371338 Klein, G. (1997). The recognition-primed decision (RPD) model: Looking back, looking forward. In

C. E. Zsambok & G. Klein (Eds.), Naturalistic decision making (pp. 285–292). Mahwah, NJ: Lawrence Erlbaum.

Klein, G. (2008). Naturalistic decision making. Human Factors, 50, 456–460. doi:10.1518/ 001872008X288385

Klein, G., & Militello, L. (2004). The knowledge audit as a method for cognitive task analysis. In H.

Montgomery, R. Lipshitz & B. Brehmer (Eds.), How professionals make decisions (pp. 335– 342). Mahwah, NJ: Lawrence Erlbaum.

Knoke, D., & Yang, S. (2008). Social network analysis (2nd ed.). Thousand Oaks, CA: Sage.

Kroon, M. B. R., ’t Hart, P., & van Kreveld, D. (1991). Managing group decision making processes:

Individual versus collective accountability and groupthink. The International Journal of

Conflict Management, 2, 91–115. doi:10.1108/eb022695 Kuhn, T. (1962). The structure of scientific revolutions. London, UK: The University of Chicago

Press.

Kurokawa, K. (2012). The official report of The Fukushima Nuclear Accident Independent

Investigation Commission: Executive summary. Tokyo, Japan: The National Diet of Japan

Fukushima Nuclear Accident Independent Investigation Commission.

LePine, J., Piccolo, R. F., Jackson, C. L., Mathieu, J. E., & Saul, J. R. (2008). A Meta-Analysis of

Teamwork Processes: Tests of a Multidimensional model and relationships with team

effectiveness criteria. Personnel Psychology, 61, 273–307. doi:10.1111/j.1744-6570.2008. 00114.x

Lipshitz, R. (1993). Converging themes in the study of decision making in realistic settings. In G. A.

Klein, J. Orasanu, R. Calderwood & C. E. Zsambok (Eds.), Decision making in action: Models

and methods (pp. 103–137). Norwood, NJ: Ablex. Lipshitz, R., Klein, G., Orasanu, J., & Salas, E. (2001). Focus article: Taking stock of naturalistic

decision making. Journal of Behavioral Decision Making, 14, 331–352. doi:10.1002/bdm.381 London Emergency Services Liaison Panel. (2007). Major incident procedure manual (7th ed.).

London, UK: TSO.

Marks, M. A., Mathieu, J. E., & Zaccaro, S. J. (2001). A Temporally based framework and taxonomy

of team processes. Academy of Management Review, 26, 356–376. doi:10.5465/ AMR.2001.4845785

Mathieu, J. F., Marks, M. A., & Zaccaro, S. J. (2001). Multi-team systems. In N. Anderson, D. Ones, H.

K. Sinangil & C. Viswesvaran (Eds.), International handbook of work and organizational

psychology (pp. 289–313). London, UK: Sage. Mayer, R. C., Davis, J. H., & Schoorman, D. F. (1995). An integrative model of organizational trust.

Academy of Management Review, 20, 709–734. doi:10.5465/AMR.1995.9508080335 McAndrew, C., & Gore, J. (2013). Understanding preferences in experience-based choice: A study of

cognition in the ‘wild’. Journal of Cognitive Engineering and Decision Making, 7(2), 179–197. doi:10.1177/1555343412463922

Mischler, E. G. (1990). Validation in inquiry-guided research: The role of exemplars in narrative

studies. Harvard Educational Review, 60, 415–441.

Decision inertia 319

Mizruchi, M. S. (1994). Social network analysis: Recent achievements and current controversies.

Acta Sociologica, 37, 329–343. doi:10.1177/000169939403700403 Mueller, J. (2012). Why individuals in larger teams perform worse. Organizational Behavior and

Human Decision Processes, 117(1), 111–124. doi:10.1016/j.obhdp.2011.08.004 Nadler, A., & Halabi, S. (2006). Intergroup helping as status relations: Effects of status stability,

identification, and type of help on receptivity to high-status group’s help. Journal of Personality

and Social Psychology, 91(1), 97–110. doi:10.1037/0022-3514.91.1.97 National Police Improvement Agency. (2009). Guidance on multi-agency interoperability.

Bedfordshire, UK: National Policing Improvement Agency.

Netten, N., Bruinsma, G. W. J., van Someren, M., & de Hoog, R. (2006). Task-adaptive information

distribution for dynamic collaborative emergency response. International Journal of

Intelligent Control & Systems, 1, 238–247. Orasanu, J., & Connolly, T. (1993). The reinvention of decision making. In G. A. Klein, J. Orasanu, R.

Calderwood & C. E. Zsambok (Eds.), Decision making in action: Models and methods

(pp. 3–20). Norwood, NJ: Ablex. Parker, J. R., & Schrift, R. Y. (2011). Rejectable choice sets: How seemingly irrelevant no choice

options affect consumer decision processes. Journal of Marketing Research, 48, 840–854. doi:10.1509/jmkr.48.5.840

Pierce, J. L., Kostova, T., & Dirks, K. (2001). Toward a theory of psychological ownership

in organizations. Academy of Management Review, 26, 298–310. doi:10.5465/amr.2001. 4378028

Quarantelli, E. L. (1985). Organizational behavior in disasters and implications for disaster

planning. Newark, DE: Disaster Research Center, University of Delaware.

Rovira, A., Swapp, D., Spanlang, B., & Slater, M. (2009). The use of virtual reality in the study of

people’s responses to violent incidents. Frontiers in Behavioural Neuroscience, 3(59), 1–10. doi:10.3389/neuro.08.059.2009

Rusman, E., van Bruggen, J., Sloep, P., & Koper, R. (2010). Fostering trust in virtual project teams:

Towards a design framework grounded in TrustWorthiness Antecedents (TWAN) schema.

International Journal of Human-Computer Studies, 68, 834–850. doi:10.1016/j.ijhcs.2010. 07.003

Salas, E., Cook, N. J., & Rosen, M. A. (2008). On teams, teamwork, and team performance.

Discoveries and developments. Human Factors, 50, 540–547. doi:10.1518/001872008X288457 Schneider, S. L., & Shanteau, J. (2003). Emerging perspectives on judgment and decision research.

Cambridge, UK: Cambridge University Press.

Sch€on, D. (1987). Educating the reflective practitioner. San Francisco, CA: Jossey-Bass. Schubert, C. C., Denmark, K., Crandall, B., Grome, A., & Pappas, J. (2012). Characterizing novice-

expert differences in macrocognition: An exploratory study of cognitive work in the

emergency department. Annals of Emergency Medicine, 61(1), 96–109. doi:10.1016/ j.annemergmed.2012.08.034

Scott, W. R. (2001). Organizations: Rational, national, and open systems. Englewood Cliffs, NJ:

Prentice Hall.

Sherif, M. (1962). Intergroup relations and leadership. New York, NY: Wiley.

Simon, H. A. (1956). Rational choice and the structure of the environment. Psychological Review,

63, 129–138. doi:10.1037/h0042769 Sonnenwald, D. H., & Pierce, L. G. (2000). Information behaviour in dynamic group work contexts:

Interwoven situational awareness, dense social networks and contested collaboration in

command and control. Information processing and Management, 36, 461–479. doi:10.1016/ S0306-4573(99)00039-4

Staats, B. R., Milkman, K. L., & Fox, C. R. (2012). The team scaling fallacy: Underestimating the

declining efficiency of larger teams. Organizational Behavior & Human Decision Processes,

118(2), 132–142. doi:10.1016/j.obhdp.2012.03.002 Stanton, N. A., Wong, W., Gore, J., Sevdalis, N., & Strub, M. (2011). Critical thinking. Theoretical

Issues in Ergonomics Science, 12, 204–209. doi:10.1080/1464536X.2011.564479

320 Laurence Alison et al.

Stout, R. J., Cannon-Bowers, J. A., Salas, E., & Milanovich, D. M. (1999). Planning, shared mental

models, and coordinated performance: An empirical link is established. Human Factors, 41, 61– 71. doi:10.1518/001872099779577273

Tesluk, P., Mathieu, J. E., Zaccaro, S. J., & Marks, M. A. (1997). Task and aggregation issues in the

analysis and assessment of team performance. In M. T. Brannick, E. Salas & C. Prince (Eds.), Team

performance and measurement: Theory, methods, and applications (pp. 197–224). Mahwah, NJ: Lawrence Erlbaum.

Thwaites, E., & Williams, C. (2006). Service recovery: A naturalistic decision-making approach.

Managing Service Quality, 16, 641–653. doi:10.1108/09604520610711945 van de Ven, A. H., Delbecq, A. L., & Koenig, R. (1976). Determinants of coordination modes within

organizations. American Sociological Review, 41, 322–338. doi:10.2307/2094477 van den Heuvel, C., Alison, L., & Crego, J. (2012). How uncertainty and accountability can derail

strategic ‘save life’ decisions in counter-terrorism simulations: A descriptive model of choice

deferral and omission bias. Journal of Behavioral Decision Making, 25(2), 165–187. doi:10.1002/bdm.723

van den Heuvel, C., Alison, L., & Power, N. (2014). Coping with uncertainty: Police strategies for

resilient decision-making and action implementation. Cognition, Technology & Work, 16, 25– 45. doi:10.1007/s10111-012-0241-8

Van Leeuwen, E. (2007). Restoring identity through outgroup helping: Beliefs about international

aid following the December 2004 tsunami. European Journal of Social Psychology, 37, 661– 671. doi:10.1002/ejsp.389

Von Neumann, I., & Morgenstern, O. (1947). Theory of games and economic behavior. Princeton,

NJ: Princeton University Press.

Waring, S. K., Alison, L., Cunningham, S., & Whitfield, K. C. (2013). The impact of accountability on

motivational goals and the quality of advice provided in crisis negotiations. Psychology, Public

Policy and Law, 19(2), 137–150. doi:10.1037/a0030629 Wittenbaum, G. M., Hollingshead, A. B., & Botero, I. C. (2004). From cooperative to motivated

information sharing in groups: Moving beyond the hidden profile paradigm. Communication

Monographs, 71, 286–310. doi:10.1080/0363452042000299894 Yates, J. F. (2003). Decision management. San Francisco, CA: Jossey-Bass.

Young, D. L., Goodie, A. S., Hall, D. B., & Wu, E. (2012). Decision making under time pressure in a

prospect theory framework. Organizational Behavior and Human Decision Processes, 118,

179–188. doi:10.1016/j.obhdp.2012.03.005 Zeelenberg, M., & van Dijk, E. (1997). A reverse sunk cost effect in risky decision making: Sometimes

we have too much invested to gamble. Journal of Economic Psychology, 18, 677–691. doi:10.1016/S0167-4870(97)00029-9

Received 27 February 2014; revised version received 14 January 2015

Decision inertia 321

Copyright of Journal of Occupational & Organizational Psychology is the property of Wiley- Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.