Critical success factors for human resource outcomes in Kaizen events: An empirical study

ARTICLE IN PRESS Int. J. Production Economics 117 (2009) 42–65

Contents lists available at ScienceDirect

Int. J. Production Economics journal homepage: www.elsevier.com/locate/ijpe

Critical success factors for human resource outcomes in Kaizen events: An empirical study Jennifer A. Farris a,, Eileen M. Van Aken b, Toni L. Doolen c, June Worley c a b c

Department of Industrial Engineering, Texas Tech University, Lubbock, TX 79409, USA Grado Department of Industrial & Systems Engineering, Virginia Polytechnic Institute & State University, Blacksburg, VA 24061, USA School of Mechanical, Industrial, & Manufacturing Engineering, Oregon State University, Corvallis, OR 97331, USA

a r t i c l e i n f o

abstract

Article history: Received 9 January 2008 Accepted 29 August 2008 Available online 17 September 2008

Kaizen events are an increasingly common organizational improvement mechanism aimed at work area transformation and employee development. While many anecdotal design prescriptions exist, there is little empirical evidence of which input and process factors are most strongly related to Kaizen event outcomes in practice. This paper uses results from a field study of 51 events in six manufacturing organizations to identify the set of input and process factors that most strongly relate to the development of employee attitudinal outcomes and problem-solving capabilities in Kaizen events. These results are used to develop guidelines for organizations and identify directions for future work. & 2008 Elsevier B.V. All rights reserved.

Keywords: Lean production Teams Productivity improvement Quality management Manufacturing companies

1. Introduction The study of improvement programs has long been a focus of the operations management (OM) and industrial engineering community (e.g., Chan et al., 2005; Dar-El, 1997; Guimaraes, 1997; Gunasekaran et al., 1994; Hales and Chakravorty, 2006; Herron and Braiden, 2006; Launonen and Kess, 2002; McIntosh et al., 2001; Van Landeghem, 2000; Vits and Gelders, 2002). Recently, lean manufacturing (Womack et al., 1990) has become a—if not the—dominant improvement paradigm, leading to a variety of studies examining this topic (e.g., Matusi, 2007; Panizzolo, 1998; Simons and Taylor, 2007; Warnecke and Huser, 1995). Within lean manufacturing, one increasingly utilized mechanism is the Kaizen event, a focused and structured continuous improvement project, using a dedicated crossfunctional team to address a targeted work area, to achieve specific goals in an accelerated timeframe (usually

 Corresponding author. Tel.: +1 806 742 3543; fax: +1 806 742 3411.

E-mail address: [email protected] (J.A. Farris). 0925-5273/$ - see front matter & 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpe.2008.08.051

1 week or shorter) (Farris et al., 2008b). In addition to potential, direct improvements in the target work area, Kaizen events are purported to serve as a ‘‘just-intime’’ training mechanism for participating employees (Drickhamer, 2004a), helping these employees develop new problem-solving capabilities and increased motivation to participate in future improvement activities. However, despite their popularity and potential benefits, Kaizen events have not been widely studied to date (Bateman, 2005; Melnyk et al., 1998). Many guidelines for Kaizen event design exist, primarily in the practitioner literature; however, these guidelines do not appear to have been tested through empirical research. Until the determinants of Kaizen event outcomes are well understood, organizations will not be able to systematically manage Kaizen events to consistently achieve positive results. This paper presents findings from a field study of 51 Kaizen events in six manufacturing organizations, where multiple regression was used to test the relationships between Kaizen event input and process factors and employee attitudinal and problem-solving capability outcomes. Findings are used to develop design guidelines for organizations using Kaizen events and to lay a foundation

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

43

for future research. Section 2 reviews the literature related to this topic, Section 3 presents the research methodology, Section 4 presents results, and Section 5 discusses study findings, limitations, and directions for future research.

needed to determine which lean system designs produce the most positive outcomes.

2. Literature review

Kaizen events appear to have originated with Toyota, who purportedly used them to train their suppliers in the 1970s (Sheridan, 1997). However, they did not become popular in the US until the 1990s (Schonberger, 2007) and do not appear in the literature until that time. Key publications from the current research literature on Kaizen events are summarized in Table 1. In addition to the limited number of studies, there is no clear agreement on which factors determine either initial outcomes or results sustainability. The methodologies used in the studies also present certain limitations. Three studies, Bateman and David (2002), and Bateman and Rich (2003), and Miller (2004), do not focus on the relationship between input and process factors and event-level outcomes, but instead on Kaizen event program-level effects. Half of the remaining studies (Farris et al., 2008b; Montabon, 2005; Patil, 2003) are based upon the analysis of a single event, while Bradley and Willett (2004) based their conclusions on interviews with participants from 12 events in a single company, Doolen et al. (2008) studied two events in a single company, and Melnyk et al. (1998) do not link their conclusions to the study of any specific events. Only Bateman (2005) studied multiple events within multiple organizations. All of the studies except Doolen et al. (2008) and Farris et al. (2008b) focus on the relationship between event characteristics and technical performance outcomes, without empirically measuring human resource outcomes. Finally, most of the studies rely heavily on qualitative data and do not include investigation of the quantitative relationships between outcomes, input factors and process factors. Doolen et al. (2008); Farris et al. (2008b) and Patil (2003) quantified certain outcomes, input factors, and process factors in their studies but, due to their small sample sizes, only very limited conclusions about quantitative relationships can be drawn. Only, Bateman (2005) empirically investigated the quantitative relationships between work area and organizational characteristics and the sustainability of event outcomes across a larger number of events. Thus, there is clearly a need for additional empirical research on Kaizen events.

2.1. Employee development within lean manufacturing Samson and Whybark (1998) issued a general charge to the OM and industrial engineering community to put more focus on ‘‘softer’’ human resource issues, and recent research suggests that many are answering this call (e.g., Jun et al., 2006; Kathuria and Partovi, 1999; Korhonen and Pirttila, 2003; Polychronakis and Syntetos, 2007; Tranfield et al., 2000). In lean manufacturing research, human resource practices, such as employee participation in continuous improvement programs, cross-functional teams, employee training, and job rotation systems, are acknowledged to form core components of a lean manufacturing program, at least in theory (e.g., Dankbaar, 1997; deTreville and Antonakis, 2006; Niepce and Molleman, 1996; Panizzolo, 1998; Shah and Ward, 2003, 2007; Warnecke and Huser, 1995). Several studies have reported a relationship between the degree of implementation of human resource practices and lean manufacturing success (Huber and Brown, 1991; Matusi, 2007; Olorunniwo and Udo, 2002; Sawhney and Chason, 2005; Schonberger, 2007). Other work has addressed the relationship between lean implementation and employee satisfaction. Findings have been mixed, with several authors suggesting a negative relationship between lean implementation and employee satisfaction (e.g., Bailey and Rose, 1988; Delbridge, 1998; Delbridge et al., 1992; Fucini and Fucini, 1990; Klein, 1989; Parker and Slaughter, 1988; Sewell and Wilkinson, 1992), while others argue that the relationship is positive (e.g., Adler 1993a, b; deTreville and Antonakis, 2006; Womack et al., 1990), null (Huber and Hyer, 1985), or ambivalent (Jackson and Mullarkey, 2000; Shafer et al., 1995). In addition, the majority of this work has been theoretical or anecdotal, rather than empirical. (Notable exceptions include Huber and Hyer, 1985; Jackson and Mullarkey, 2000; Shafer et al., 1995.) Further, potential reciprocal effects between specific lean implementation activities and employee outcomes do not appear to have been systematically investigated. Positive attitudinal outcomes from specific lean implementation activities, such as Kaizen events, could increase employee commitment to the lean program as a whole, ultimately improving the program’s success and sustainability (Adam et al., 1997; Co et al., 1998; Keating et al., 1999). Similarly, participation in Kaizen events or other problem-solving activities could help to develop operations employees’ problem-solving capabilities, which is crucial to the success of lean systems (Biazzo and Panizzolo, 2000; Brown and Mitchell, 1991; Dankbaar, 1997; Huber and Hyer, 1985; Safayeni and Purdy, 1991). Thus, the relationships between lean production implementation activities and employee outcomes are not fully understood, and additional research is

2.2. Kaizen event research literature

2.3. Kaizen events in the context of team effectiveness research A Kaizen event team represents a specific type of team—a short-duration (generally, 1 week or shorter), dedicated project team—which does not appear to have received much attention in empirical research, although teams in general have been widely studied (e.g., Campion et al, 1993; Gladstein, 1984; Hackman, 1987; Hyatt and Ruddy, 1997; Katzenbach and Smith, 1993; Kolodny and Kiggundu, 1980; Pinto et al., 1993; Sundstrom et al., 1990; Vinokur-Kaplan,1995). Furthermore, as Pagell and LePine (2002) note, there are relatively few empirical studies of

ARTICLE IN PRESS 44

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Table 1 Previous research on Kaizen events Source

Study type

Sectors

Event outcomes measured

Melnyk et al. (1998)

Empirical, observation

Various

None

Bateman and David (2002)

Empirical, observation

Automotive

Bateman and Rich (2003)

Empirical, field study of 21 organizations

Automotive

Patil (2003)

Empirical, case study in single organization

Aerospace

Bradley and Willett (2004)

Empirical, case study in single organization

Mechanical products

Miller (2004)

Various Empirical, field study in 4 organizations Automotive Empirical, field study of 40 events in 21 organizations

Bateman (2005)

Montabon (2005)

Doolen et al. (2008)

Farris et al. (2008b)

Factors related to outcomes

Short/finite duration, narrow focus, low capital investment, team-based nature, action orientation, use of specific and verifiable metrics, repetition Proposes a framework None for measuring sustainability of technical results Sustainability of Resource availability, recognition of need for change, technical results at culture that supports change, consistency of focus on organization level improvement, Kaizen program champion quality, team leader quality, management support, employee turnover, communication quality, measurement system alignment, ability to financially justify events Sustainability of Employee involvement, job security, training, employee technical results needs surveys, standard operating procedures (SOPs), follow-up reviews, time for completion of action items, strategic alignment, knowledge sharing None Problem complexity, team tool knowledge, team process knowledge, team leader experience, facilitator experience, effective internal processes, goal clarity, organizational stability, team functional heterogeneity, event scope, lack of manipulation, lack of competition, rewards for participation, flexibility in approach, action item list, historical performance data None Expectancy beliefs, task value beliefs, achievement behaviors Sustainability of technical results

5S, performance measurement, employee problemsolving systems, work area manager support, employee empowerment, full-time Kaizen event coordinator, training in new work methods, work area management involvement in Kaizen events, top management support for PI, strategic alignment Team functional heterogeneity

Empirical, case study in single organization Empirical, case study in single organization

Durable goods

None

Electronics

Empirical, case study in single organization

Large transportation equipment

Goal difficulty, management support Initial levels and sustainability of goal achievement, perceived success, and human resource impacts Goal clarity, team functional representation, team Initial goal autonomy achievement, perceived success, human resource impacts

teams in an operations management (OM) context and even fewer involving operations personnel. Likely due to the diversity of team types and contexts, there is a lack of consensus in the team literature on the exact set of variables that determines team effectiveness (Cohen and Bailey, 1997; Salas et al., 2005). However, there is some consistency in the types of factors identified (see Table 2). Many studies utilize the input-process–output framework (Guzzo and Shea, 1992; McGrath, 1964), in which process variables are assumed to at least partially mediate the relationships between inputs and outcomes. The research model for this study (Fig. 1) is based on the input-process–output framework and, particularly, the team effectiveness model by Cohen and Bailey (1997)

and the team chemistry model by Nicolini (2002). The model contains five basic factor groups:

 Kaizen event design antecedents, which are the group 



design and task design input factors describing the internal characteristics of a given event. Organizational and work area antecedents, which are the input factors describing the context of the event, e.g., the support and constraints supplied by other organizational entities. Kaizen event process factors, which describe group activities and intermediate psychosocial traits arising from team member interactions.

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

45

Table 2 Previous studies of team effectiveness Source

Type of study

Factors related to outcomes

McGrath (1964) Kolodny and Kiggundu (1980)

Theoretical Empirical

Gladstein (1984)

Empirical

Hackman (1987)

Theoretical

Pinto and Slevin (1987)

Empirical

Sundstrom et al. (1990) Susman and Dean (1992) Ancona and Caldwell (1992)

Theoretical Theoretical Empirical

Campion et al. (1993) Katzenbach and Smith (1993) Pinto et al. (1993)

Empirical Empirical Empirical

Vinokur-Kaplan (1995)

Empirical

Cohen et al. (1996)

Empirical

Belassi and Tukel (1996)

Empirical

Cohen and Bailey (1997)

Theoretical

Hyatt and Ruddy (1997)

Empirical

Janz (1999) Bailey (2000)

Empirical Empirical

Stewart and Barrick (2000) Nicolini (2002)

Empirical Theoretical

Pagell and LePine (2002)

Empirical

Lemieux-Charles et al. (2002)

Empirical

Linderman et al. (2006)

Empirical

Group composition; Group structure; Task and environment; Group processes Leadership and supervision; Organizational arrangements; Task conditions; Group characteristics; Technical skills; Group interactions Group composition; Group structure; Resource availability; Organization structure; Group processes; Task design Group design; Organizational context; Group synergy; Group processes (effort, knowledge, task strategies); Resource availability Project mission; Top management support ; Project schedule/plan; Client consultation; Personnel; Technical tasks; Client acceptance; Monitoring and feedback; Communication; Trouble-shooting Organizational context; Boundaries; Team development Integrative mechanisms; Group processes; Codification/computerization; Project risk; Goal difficulty Frequency of external communication Ambassador activities; Task coordination; Scouting; Guarding; Internal processes; Cohesion Job design; Group interdependence; Group composition; Organizational context; Group processes Group skills; Group accountability; Group commitment Superordinate goals; Physical proximity; Accessibility; Project team rules and procedures; Organizational rules and procedures; Cross-functional cooperation Group structure and composition; Task clarity; Organizational context; Training availability; Physical environment; Effort; Knowledge application; Task strategies Group task design; Encouraging supervisory behaviors; Group characteristics (composition, beliefs, processes); Employee involvement context Factors related to the project; Factors related to the project manager; Factors related to project team members; Factors related to the organization; System response variables (process variables) Task design; Group composition; Organizational context External group processes; Internal group processes; Group psychosocial traits; Environmental factors Process focus; Work group support; Goal orientation; Work group confidence; Customer orientation; Internal processes Organizational climate; Autonomy; Cooperative learning; Team development Task dependence; Training; Autonomy; Supportiveness; Rewards; Performance Feedback; Communication; Conflict; Cohesion Interdependence; Team self-leadership; Task type; Intrateam processes Project level antecedents; Business environment and organizational antecedents; Project chemistry (group processes and psychosocial traits) Team composition; Task and team structure; Technology; Trust; Novel problems to solve; Operational interdependence Organizational context; Quality improvement practices; Team norms; Process strategies; Decisionmaking process Goal difficulty; Adherence to task strategy (Six Sigma)

 Social system outcomes, which describe the human 

resource impacts of the event, e.g., problem-solving capability development and affective outcomes. Technical system outcomes, which describe the impact of the event on the technical performance of the target work area, e.g., goal achievement and perceived success.

Environmental factors related to the external market context of the organization were not included in the research, as they were not expected to vary substantially at the event level, i.e., within organizations, during the short time horizon of the research. However, data describing each organization’s external environment were collected though an organization-level Kaizen program interview, and each organization was monitored for sudden changes in its environment. Further, due to scope constraints, this paper focuses on social system outcomes only. However, technical system outcomes are included in Fig. 1 to make the full structure of the research model clear. In accordance with many of the previous inputprocess–outcome models (e.g., Cohen and Bailey, 1997;

Hackman, 1987; Sundstrom et al., 1990; Susman and Dean, 1992), the model further hypothesizes that input factors can impact outcomes both indirectly and indirectly via group processes. Thus three types of relationships are of interest, which can be framed as three broad research hypotheses: Hypothesis H1. Input variables (Kaizen event design antecedents and organizational and work area antecedents) have a direct relationship to social system outcomes. Hypothesis H2. Kaizen event process variables have a direct relationship to social system outcomes. Hypothesis H3. Kaizen event process variables mediate the effect of input variables on social system outcomes. To identify the specific variables to be studied in this research, published accounts of Kaizen events were reviewed in addition to the studies of team effectiveness. While there has been little empirical research on Kaizen events to date (see Table 1), there have been numerous anecdotal publications, primarily in non-refereed journals, conference proceedings, industry magazines and newsletters, professional

ARTICLE IN PRESS 46

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Input Factors

Process Factors

Outcomes

Kaizen Event Design Antecedents - Goal Clarity - Goal Difficulty - Team Autonomy - Team Kaizen Experience - Team Leader Experience - Team Functional Heterogeneity Organizational and Work Area Antecedents

Social System Outcomes

Kaizen Event Process Factors - Action Orientation - Affective Commitment to Change - Intemal Processes - Tool Quality - Tool Appropriateness

- Understanding of CI - Skills - Attitude

Technical System Outcomes - Goal Achievement - Impact on Area - Overall Perceived Success

- Management Support - Event Planning Process - Work Area Routineness

Fig. 1. Initial research model.

society magazines and newsletters, websites, and newspapers. These publications typically describe the phenomenon, design suggestions, and example results. To date, the authors have reviewed over 80, primarily anecdotal, sources on Kaizen events published between 1993 and 2007. The next sections discuss how findings from the team effectiveness literature and published Kaizen event accounts were used to identify the specific variables studied in this research (Fig. 1). 2.3.1. Kaizen event design antecedents The design of a given Kaizen event can be described in terms of the design of the target task and the design of the Kaizen event team. Task design factors include both task structure characteristics, e.g., team autonomy, and task content characteristics, e.g., scope and complexity (Campion et al., 1993), while group design factors describe the characteristics of team members and the team leader (Cohen and Bailey, 1997). In the context of Kaizen events, three task design factors appear to be of particular interest: team autonomy, goal clarity, and goal difficulty. Similarly, three group design factors also appear to be of particular interest: team functional heterogeneity, team kaizen experience, and team leader experience. High team autonomy, typically including actual authority to implement changes during the event, is frequently cited as one of the key task design characteristics contributing to Kaizen event effectiveness (e.g., Adams et al., 1997; Bicheno, 2001; Bradley and Willett, 2004; Foreman and Vargas, 1999; Kumar and Harms, 2004). Conversely, lack of implementation authority (autonomy) has been suggested as one of the reasons that some continuous improvement mechanisms, such as

quality circles, failed in the US (Lawler and Mohrman, 1985, 1987). Thus, many Kaizen event practitioners suggest that the higher autonomy in Kaizen events makes these projects more successful than older improvement structures. The empirical team literature generally supports a positive relationship between team autonomy and team outcomes (Cohen and Ledford, 1994; Wall et al., 1986), although a few studies have failed to find a relationship (e.g., Janz, 1999). Most Kaizen event accounts suggest that goal clarity (i.e., specific, well-defined goals) is another key task design characteristic promoting event effectiveness (e.g., Adams et al., 1997; Bradley and Willett, 2004; Rusiniak, 1996). Generally, the accounts have further suggested the use of measurable targets or benchmarks (e.g. Bradley and Willett, 2004; Foreman and Vargas, 1999; Heard, 1997; Martin, 2004; Melnyk et al., 1998; Treece, 1993; Vasilash, 1993). However, a few Kaizen event accounts have suggested using loosely defined goals, including allowing the team to develop the goals and even identify the target work area during the event (e.g., Kumar and Harms, 2004; Wittenberg, 1994). The empirical team literature generally supports a positive relationship between goal clarity and team outcomes (e.g., Doolen et al., 2003a; Koch, 1979; Pritchard et al., 1988; Van Aken and Kleiner, 1997), although a few studies have failed to find a relationship (e.g., Gladstein, 1984). The goal-setting literature, which has typically focused on the individual level rather than the group level (Linderman et al., 2006), also lends some support for the importance of clear, specific goals, although it has generally reported that the relationship between goal specificity and performance is moderated by goal difficulty (Locke and Latham, 2002). The project

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

management literature has more consistently identified goal clarity as a key factor affecting team success (e.g., Pinto and Slevin, 1987). Kaizen event accounts acknowledge goal difficulty as an important task design characteristics, but present conflicting design recommendations. Many suggest a positive relationship between goal difficulty and Kaizen event outcomes (e.g. Bicheno, 2001; Cuscela, 1998; Kumar and Harms, 2004; LeBlanc, 1999; Minton, 1998; Tanner and Roncarti, 1994; Treece, 1993). However, others suggest that the problems targeted by Kaizen events must not be too complex and must be suitable for well-known and simple lean tools (e.g. Harvey, 2004; Sheridan, 1997), and some accounts contain both, potentially conflicting, suggestions (e.g. Bradley and Willett, 2004; Gregory, 2003; Rusiniak, 1996). Literature on goal setting has typically reported a positive relationship between goal difficulty and task performance (Locke and Latham, 1990; Locke et al., 1981; O’Leary-Kelly et al., 1994). However, some authors have suggested that this relationship might not hold at high levels of goal difficulty (Erez and Zidon, 1984; Locke and Latham, 1984, 1990). In addition, goalsetting research has generally centered on the relationship between researcher-assessed levels of goal difficulty and task performance in controlled experiments where extensive training on the task was provided prior to experimental manipulation. It is not clear that this relationship holds for other measures of goal difficulty and other contexts. For instance, Martin et al. (1999) reported a negative relationship between respondentperceived goal difficulty and task performance when goals also had high researcher-assessed goal difficulty. Earley et al. (1989) found that, when training on the specific tactics and strategies needed to achieve the goals was omitted, subjects with high researcher-assessed levels of goal difficulty had lower task performance than subjects without assigned goals. Similarly, Linderman et al. (2006) found that, while goal difficulty had a positive relationship with team performance for Six Sigma teams who closely adhered to the Six Sigma methodology, goal difficulty was negatively related to performance in teams that did not. Cross-functional heterogeneity is frequently cited as one of the group design characteristics determining event effectiveness (Bradley and Willett, 2004; Melynk et al., 1998; Montabon, 2005). The Kaizen event literature generally recommends that the team consist of a crossspectrum of the organization and, particularly, the target process (operators, managers, support personnel, etc.). A variety of more specific suggestions regarding team composition also exist, such as the inclusion of participants unfamiliar with the target process or work area (e.g., Bradley and Willett, 2004; Cuscela, 1998; David, 2000; Foreman and Vargas, 1999; LeBlanc, 1999; Martin, 2004; McNichols et al., 1999; Minton, 1998; Melnyk et al., 1998), and the inclusion of process suppliers or customers (e.g., Adams et al., 1997; Heard, 1997; Larson, 1998; McNichols et al., 1999; Melnyk et al., 1998; Tanner and Roncarti, 1994). Previous team research has found that crossfunctional teams may make better decisions (Jackson, 1992; Jehn et al., 1999; Lovelace et al., 2001; McGrath,

47

1984), but typically experience lower levels of enjoyment in working together (Ancona and Caldwell, 1992; Amason and Schweiger, 1994; Baron, 1990), suggesting a mixed effect of functional heterogeneity in terms of team outcomes. Team and team leader task experience are other group design characteristics, which may contribute to event effectiveness (e.g., Bradley and Willett, 2004). A key proposition in Bradley and Willett (2004) is that teams require at least one experienced individual, who may or may not be the team leader. Similarly, based on their study of business process reengineering (BPR) teams, Launonen and Kess (2002) suggest that it is the responsibility of the team leader to supply, either directly or indirectly, any needed problem-solving skills not present among other team members. In addition, the role of the team leader is consistently emphasized in the project management literature (e.g., Belassi and Tukel, 1996; Nicolini, 2002: Pinto and Slevin, 1987), as well as the team literature (e.g., Salas et al., 2005). Two other Kaizen event design factors ultimately not included in the research model are worthy of brief mention. The relatively short duration of Kaizen events is a task characteristics commonly mentioned as one of the defining characteristics of Kaizen events, possibly contributing to success as compared with longer-duration improvement mechanisms (e.g. Adams et al., 1997; Bicheno, 2001; Bradley and Willett, 2004; Heard, 1997; Melnyk et al., 1998). For instance, the organizational change literature has emphasized the need for short term, visible improvements to generate longer-term support for improvement programs (e.g., Keating et al., 1999; Kotter, 1995). Event duration was ultimately not included in the research due to expected limited variation across events. The final sample supported this assumption, as events ranged from 2 to 7 days, with 86% of the events lasting between 3 and 5 days. Post-hoc analyses were used to determine whether event duration was significantly related to Kaizen event outcomes, after accounting for the effects of other variables, and no significant relationships were found. In addition, many published accounts make some recommendation regarding team size, a group design characteristic. The range of recommended team sizes is relatively small, from three to five members (Rusiniak, 1996) to 14 to 15 members (Laraia 1998; Larson, 1998), and reasons for specific recommendations are generally unclear. Team size was ultimately not included in the research model due to an expected limited range of variation. The final sample supported this assumption, with team sizes ranging from three to 22, and only eight teams having more than 10 members. Post-hoc analyses were conducted to determine whether team size had a significant effect, after controlling for the other variables studied, and no significant relationships were found. 2.3.2. Organizational and work area antecedents Organizational design factors (or organizational context factors) describe the organizational environment in which the team operates. Typically, this includes structures such as reward systems, training programs, resource provision, and supervisory systems (Cohen and Bailey,

ARTICLE IN PRESS 48

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

1997). While clearly identified in theoretical models as potential determinants of team effectiveness, organizational context factors have received less attention in previous research than other, more locally controllable variables (e.g., task design, group design) (Doolen et al., 2003a; Pagell and LePine, 2002). However, some previous research has identified positive relationships between organizational context and team outcomes (e.g., Campion et al. 1993; Doolen et al., 2003a; Hyatt and Ruddy, 1997; Pinto and Slevin, 1987). Published accounts of Kaizen events frequently discuss several characteristics related to organizational context, including:

 rewards provided for Kaizen event teams after the







event (e.g., Adams et al., 1997; Foreman and Vargas, 1999; Larson, 1998; Martin, 2004; Melnyk et al., 1998; Tanner and Roncarti, 1994; Taylor and Ramsey, 1993); resources provided for the Kaizen event team during the event (e.g., Adams et al., 1997; Bicheno, 2001; Bradley and Willett, 2004; Kumar and Harms, 2004; Martin, 2004; McNichols et al., 1999; Sheridan, 1997; Tanner and Roncarti, 1994; Taylor and Ramsey, 1993); time and other resources allocated for planning the event (e.g., Sheridan, 1997; Bradley and Willett, 2004; Heard, 1997; Gregory, 2003; Foreman and Vargas, 1999); training provided to the team (e.g., Foreman and Vargas, 1999; Heard, 1997; Larson, 1998; McNichols et al., 1999; Melnyk et al., 1998; Minton, 1998; Tanner and Roncarti, 1994; Taylor and Ramsey, 1993; Treece, 1993).

This research studied the resources provided during the event (management support), as well as the total hours of support provided prior to the event (event planning process). Team-level reward mechanisms were not included, as they do not appear likely to vary much across teams within the same organization. Training was not directly studied due to the fact that many events occurring within the same organization appear to receive similar training, although training may at times be customized to the type of event, and some organizations do omit training or conduct it before the formal start of the event (Bicheno, 2001; Gregory, 2003; McNichols et al., 1999). However, data on organizational-level reward mechanisms and on training processes were collected during this research, thus enabling potential future post-hoc analyses involving these variables. In addition to support from the larger organization, Kaizen event accounts also discuss the importance of context factors in the target work area, including the stability or routineness of the target work area (e.g., Bradley and Willett, 2004; LeBlanc, 1999) and work area buy-in to the changes identified (e.g., Bicheno, 2001; Gregory, 2003; Sheridan, 1997). This research studied the effect of work area routineness. Commitment from work area employees not on the event team was not studied, as it appears to relate more to sustainability than to initial results (Bateman 2005; Patil, 2003). 2.3.3. Kaizen event process factors Group process factors describe the interactions between members of the group, and between members of

the group and outside personnel, while group psychosocial characteristics are shared knowledge, beliefs and attitudes, such as team mental models, group norms and team affect (Cohen and Bailey, 1997), which arise from team processes. Kaizen event accounts identify several process-related variables, which appear important to the study of event outcomes: action orientation, tool appropriateness, tool quality, internal processes and affective commitment to change. Team action orientation, or the extent to which the team focuses on implementation rather than analysis, is frequently cited as one of the key determinants of event success (e.g., Adams et al., 1997; Bicheno, 2001; Foreman and Vargas, 1999; Larson, 1998a; Martin, 2004; Melnyk et al., 1998; Sheridan, 1997; Smith, 2003; Tanner and Roncarti, 1994; Taylor and Ramsey, 1993; Treece, 1993; Vasilash, 1993). Published Kaizen event accounts suggest that events that emphasize ‘‘hands-on,’’ implementationfocused activities, such as moving equipment during the event to implement a future state, will be more effective than events which focus primarily on analysis or planning. However, it appears that action orientation has not yet been studied in the team or project literature. Two other process variables that appear important to the study of Kaizen event teams are the appropriateness of team tool selection and the quality of tool use (e.g., Bradley and Willett, 2004; Kumar and Harms, 2004; Minton, 1998; Tanner and Roncarti, 1994; Vasilash, 1993). Few studies have explicitly included the quality of team tool use, although a few have studied the relationship between improvement process adherence and outcomes, generally reporting a positive relationship (Douglas and Judge, 2001; Ghosh and Sobek, 2007; Handfield et al., 1999; Linderman et al., 2006). The quality of team coordination processes, or the internal process dynamics of the team, is also frequently mentioned as a contributor to event success (e.g., Bradley and Willett, 2004; Foreman and Vargas, 1999; Tanner and Roncarti, 1994; Vasilash, 1993; Wheatley, 1998). Team research has typically suggested that internal team coordination is related to attitudinal outcomes and to team perceptions of team effectiveness (e.g., Bailey, 2000; Jehn, 1995; Pinto et al., 1993; Vinokur-Kaplan, 1995), but not to objective performance (e.g., Bailey, 2000; Jehn, 1995; Vinokur-Kaplan, 1995). (One notable exception is Hyatt and Ruddy, 1997.) Finally, one group psychosocial trait explicitly mentioned is team member buy-in for the goals of the event (Bradley and Willett, 2004; McNichols et al., 1999; Melnyk et al., 1998; Taylor and Ramsey, 1993; Wheatley, 1998). Indeed, the organizational change literature suggests that affective commitment to change from participants is vital to transformation effort success (Keating et al., 1999). Team member commitment has also been emphasized in studies focused specifically on the effectiveness of lean manufacturing tools (e.g., Chan et al., 2005). 2.3.4. Social system outcomes The team literature has studied a variety of outcomes, both technical and social (e.g., Cohen and Bailey, 1997). Published Kaizen event accounts also consistently mention

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

49

1998; Minton, 1998; Smith, 2003; Tanner and Roncarti, 1994; Watson, 2002). Many published accounts of Kaizen events further suggest that the visibility of changes made during the events, along with the direct employee participation, create employee enthusiasm for and buy-in to the lean manufacturing program (Bradley and Willett, 2004; Butterworth, 2001; Heard, 1997; Melnyk et al., 1998; Watson, 2002; Tanner and Roncarti, 1994). Thus, published accounts suggest that Kaizen events can be intrinsically motivating, impacting employee attitudes by creating the desire for participation in future improvement activities (Bicheno, 2001; David, 2000; Hasek, 2000; Heard, 1997; Kumar and Harms, 2004; McNichols et al., 1999; Rusiniak, 1996; Tanner and Roncarti, 1994; Wittenberg, 1994).

the importance of both technical and social system outcomes, but have rarely attempted to measure social system outcomes. In fact, the development of participating employee attitudes, knowledge and skills is often a formal objective of Kaizen events (Sheridan, 1997; Melnyk et al., 1998; Laraia et al., 1999), although these effects have not been validated empirically. In alignment with the knowledge, skills, and attitude framework from the industrial and organizational psychology literature (Muchinsky, 2000), as well as published Kaizen event accounts, this research aimed to study two outcomes related to employee problem-solving capability development, understanding of continuous improvement and skills, and one outcome measuring employee attitude toward Kaizen events. Published Kaizen event accounts suggest that Kaizen events can serve as a ‘‘just-in-time’’ training mechanism (Drickhamer, 2004a), impacting employee knowledge and skills related to continuous improvement. Team members typically receive training on the improvement tools needed to address their goals as a formal part of the event and then are empowered to immediately apply those tools, with a trained facilitator available to provide coaching and to guide the team through the solution process (Bicheno, 2001; Foreman and Vargas, 1999; Heard, 1997; Melnyk et al., 1998; Minton, 1998; Perry, 1995). Thus, it is suggested that participating employees develop new knowledge, skills, and abilities that they may apply to subsequent problem-solving tasks, as well as to the current Kaizen event (e.g., Adams et al., 1997; Bradley and Willett, 2004; David, 2000; Drickhamer, 2004b; Foreman and Vargas, 1999; Jusko, 2004; Laraia et al, 1999; Martin, 2004; McNichols et al., 1999; Melnyk et al.,

3. Methodology 3.1. Sample selection This research used a multi-site cross-sectional field study of six manufacturing organizations (Table 3). Organizations were non-randomly selected, due to the need for access to data from multiple events, as well as certain organization-level data, requiring top management buy-in and longer-term commitment to the study. However, several selection criteria (Table 4) were applied to increase the reliability and validity of study results (Yin, 1994; Eisenhardt, 1989). Within each organization, Kaizen events were randomly selected during a 9 month study period (October 2005–June 2006). Three organizations agreed to provide data for all events conducted during the

Table 3 Characteristics of study organizations

Org. description

SIC code Public/private Year founded No. employees First Kaizen event Event rate during research Percent of org. experiencing events Major processes targeted

Percent of events in manufacturing areas No. events sampled (retained) Percent of events sampled during study period Target sampling percentage

Org. A Secondary wood products manufacturer

Org. B Electronic motor manufacturer

Org. C Secondary wood products manufacturer

Org. D Manufacturer of large transportation equipment

Org. E Specialty equipment manufacturer

Org. F Steel component manufacturer

2434 Public 1946 560 1998 2–3 per month 100

3621 Public 1985 700 2000 1 per month 90

3731 Public 1939 18,000 1998 5–6 per month 85

3843 Private 1964 950 2000 6–8 per month 100

3443 Private 1913 3500 1995 1 per month 20

Operations

Operations, sales and marketing, customer service and technical support, product design, production planning and inventory control, process design 75% manufacturing

2434 Public 1946 500 1992 2 per month Data not available Operations

Engineering and related activities

All areas of organization

Manufacturing, order entry, accounts receivable, distribution, vendors, engineering product development

70% nonmanufacturing 4(4) 13

Data not available 12(11) 33

80–85% manufacturing 6(6) 24

25–33

50

60

Almost 100% manufacturing 15(15) 100

8(8) 56

Almost 100% manufacturing 11 (7) 100

100

100

100

ARTICLE IN PRESS 50

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Table 4 Selection criteria for study participants Criterion

Description

Purpose

Organization type

All participating organizations manufacture products of some type. However, organizations in different industries were recruited to increase the generalizability of results All participating organizations had been conducting Kaizen events for at least 1 year prior to the start of the study

To provide baseline similarities in focus, fundamental processes and metrics used to measure performance

Kaizen event experience

Systematic use of Kaizen events

Kaizen event frequency

All participating organizations use Kaizen events as part of a formal organizational improvement strategy, rather than as a ‘‘single use’’ change mechanism All participating organizations conduct Kaizen events relatively frequently, i.e., at least one event per month on average

To minimize learning curve effects from organizations just starting to implement Kaizen events and thus to increase the reliability, i.e., representative nature, of the internal sample To indicate senior management commitment to the Kaizen event program, and systematic and strategic planning of Kaizen events To enable an adequate sample size of Kaizen events to be drawn from each organization within a reasonable length study period

study period, enabling a census approach; the other three organizations requested a lower data collection frequency, requiring a systematic sampling procedure (Scheaffer, et al., 1996). For most organizations, the actual sampling frequency was somewhat lower than the target sampling frequency due to non-response (see Table 3). Comparison of known characteristics (event type, facilitator, etc.) of responding events versus non-responding events did not clearly indicate any differences. Organizations were initially identified through prior contact in university–industry partnerships, conferences and word of mouth. The selection criteria were then applied to determine which organizations should be invited to participate. As incentive to participate, organizations were provided with a description of the study objectives and the benefits to the organization. Each organization was provided with a short feedback report for each event, as well as a summary of findings within and across participating organizations (provided after study conclusion). Seven organizations initially agreed to participate. However, one organization withdrew after providing data for only one Kaizen event which was not included in the analysis due to insufficient withinorganization sample size. The seventh organization withdrew due to shifting internal priorities, as well as the turnover of a key individual, a general manager, who had strongly supported the research. The authors did not receive any negative feedback on the data collection methods as a reason for withdrawal. The withdrawal of

the organization does not appear to bias the sample due to the fact that the withdrawing organization was from the same industry as two of the other participating organizations (secondary wood products) and sample size was sufficient without including events from this organization. Data were initially collected on 56 events. However, five events were ultimately excluded due to incomplete data, leaving the final sample size of 51. Thus, only a relatively small percentage of data was ultimately excluded due to partial response (8.9% of original sample), leaving a sufficient sample size for regression modeling. Comparisons of the excluded events with the included events on their known characteristics (e.g., duration, size, types of goals, type of target area) did not reveal any systematic differences, even within Organization C, which contributed four out of five excluded events, suggesting a lack of bias.

3.2. Instruments and measures Five data collection instruments were used in the research. All instruments were developed in accordance with commonly accepted principles for questionnaire and interview script design (e.g., Dillman, 2000), based on previously existing instruments where possible, and revised based on pilot research, which included feedback on usability and initial analysis of Cronbach’s a for survey scales. Table 5 summarizes the administration sequence and content of the data collection instruments. For each organization, the mid-level manager responsible for coordinating the Kaizen event program (Kaizen event program coordinator) served as the contact person responsible for working with the authors to coordinate data collection. For each event studied, data were directly collected by the facilitator of the event. The use of internal organizational personnel increased the naturalism of the data collected and enabled the collection of data from multiple concurrent events within multiple organizations. A standard data collection protocol was used to ensure consistency across facilitators and organizations. In addition, standard training was provided to facilitators in how to collect the data, and data collection instruments were designed to be as stand-alone as possible, such that the facilitator’s interaction with the team during data collection was minimal. All questionnaires, i.e., kickoff questionnaire, team activities log, report out questionnaire and event information sheet, included a cover page with instructions, description of the study motivation, and contact information for the researchers and their institutional review boards. The kickoff questionnaire and report out questionnaires were distributed to the team by the event facilitator, who also read a short standard script with brief instructions, which were repeated in the questionnaire cover letter, and then left the room for a prespecified period of time. An envelope was available to collect the surveys of the team members who wished to participate. The sealed envelope was then collected by the facilitator and returned to the researchers via mail. The team activities log included an instruction page and

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

51

Table 5 Summary of data collection instruments Instrument

Frequency

Timing

Respondents

Method

Length

Measures

Kaizen program interview

Once per organization

Commencement of study participation

Kaizen event coordinator

Semi-structured telephone or inperson interview

38 items

Kickoff questionnaire

Once per event Once per event

All Kaizen event team members One Kaizen event team members

Group administered questionnaire Self-administered questionnaire

19 closeended items

Team activities log

Report out questionnaire

Once per event

End of kickoff meeting at beginning of event Ongoing during event, begins immediately following kickoff meeting End of report out meeting at end of event

Organizational characteristics, approach to conducting events, perceived benefits from Kaizen event programa Goal clarity, goal difficulty, affective commitment to change, team kaizen experience High-level description of event activities broken down by days and by half hour incrementsb

All Kaizen event team members

Group administered questionnaire

Event information questionnaire

Once per event

Initial request sent immediately following notice of end of event

Kaizen event facilitator

Self-administered questionnaire or telephone interview

One page template

37 closedended items, 2 open-ended items 15 items, primarily factual and open-ended

Attitude, understanding of CI, skillsc, team autonomy, management support, action orientation, internal processes Team functional heterogeneity, team leader experience, event planning process, work area routineness, tool appropriateness, tool quality

a These interviews were not used to draw inferences about individual Kaizen events, but to collect data on organizational characteristics (see Table 3) and to better understand the organizational context of the events studied. b Data from the team activities log were not directly used in the analyses presented in this paper, but were used to help the research team better understand the context and activities of each event. c Note that the understanding of CI and skills scales were ultimately combined in the kaizen capabilities scale.

two examples of completed logs. At the end of the event, a team member volunteer returned the completed activities log to the facilitator in a sealed envelope. The team facilitator was asked to complete the event information questionnaire electronically and return it to the research team within 1 week of receiving it. Reminder emails were sent if the event information questionnaire was not received within 2 weeks of the initial request. In cases where the event information questionnaire was still not received 4 weeks after the reminders, a follow-up via phone was made to collect the data using an alternative method (phone interview). As the data collected via the event information questionnaire are largely factual, this mixed method approach appears warranted and unlikely to create undue bias in the data. In addition to the data collected through the five instruments, a copy of the actual report out presentation created by each team was collected. The report out presentations were not directly used in the analyses reported in this paper; however, they were reviewed by the researchers to verify the accuracy of data collected for certain variables, e.g., team functional heterogeneity, team goals, to provide more detailed descriptions of events for post-hoc analyses, and to create feedback reports for participating organizations. Most variables were operationalized as multi-item scales, which, where possible, were based on previously existing instruments. Several of the variables measured through perceptual scales, e.g., affective commitment to change, are inherently perceptual in nature and cannot be accurately measured through other means. Other variables measured through perceptual scales, e.g., management support and action orientation, are more objective in

nature but difficult to measure through other means or to normalize across events. Finally, some variables, i.e., team functional heterogeneity, team kaizen experience, team leadership experience, event planning process, were measured through factual (objective) data. The potential for bias due to the use of perceptual measures was addressed by using multiple types of measures (perceptual and nonperceptual), multiple data sources (facilitator and team members), multiple respondents per team, and previously tested instruments, as well by cross-validating data, conducting tests of reliability and validity, and analyzing the consistency of responses for individuals within the same team.

3.3. Construct validity Exploratory factor analysis was used to evaluate the construct validity of the multi-item scales. All factor analyses were conducted using principal components extraction with oblique (direct quartimin) rotation, to allow correlation between factors hypothesized to be related (Fabrigar et al., 1999, Finch, 2006). The established heuristic of extracting all factors with eigenvalues greater than 1.0 (Field, 2005; Johnson, 1998) was used to determine the number of factors in each analysis. In the factor analyses, individual items were considered to have loaded onto a given factor when the primary loading was 0.500 or greater and all cross-loadings were less than 0.300 (Kline, 1994). The initial data set contained 347 kickoff questionnaires and 305 report out questionnaires. Prior to performing further analyses, individual-level data were

ARTICLE IN PRESS 52

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

screened for evidence of systematic response bias and for missing data at the item level. This screening resulted in the removal of a small proportion of individual-level data (o1% for the kickoff questionnaire and 8% for the report out questionnaire). Three factor analyses were performed: kickoff questionnaire measures, report out questionnaire input and process measures, and report out questionnaire outcome measures. Tables 6–8 present the results of the factor analyses. For each item label in the tables, the letters refer to the scale name, while the number refers to the item ID. For instance, GC1 refers to the first item in the original goal clarity scale (see the Appendix A for a full list of items by scale). In cases where all the items in a given scale failed to load onto a single factor, or items from different scales loaded onto a single factor, content

Table 6 Pattern matrix for factor analysis of kickoff questionnaire scalesa

Table 8 Pattern matrix for factor analysis of report out questionnaire outcome scalesa Component

UCI1 SK3 SK1 UCI4 AT1 UCI3 UCI2 AT3 IMA4 IMA2 IMA3 IMA1 SK2 AT4 AT2 SK4

1

2

3

0.829 0.804 0.780 0.767 0.759 0.755 0.713 0.707 0.571 0.118 0.031 0.018 0.445 0.004 0.205 0.105

0.000 0.207 0.032 0.107 0.078 0.067 0.040 0.164 0.299 0.835 0.808 0.741 0.489 0.001 0.009 0.173

0.031 0.028 0.094 0.012 0.014 0.010 0.172 0.032 0.055 0.082 0.218 0.272 0.161 0.848 0.712 0.607

Component a

ACC6 ACC1 ACC5 ACC4 ACC3 ACC2 GDF2 GDF1 GDF3 GDF4 GC3 GC1 GC4 GC2 a

1

2

3

0.774 0.770 0.756 0.722 0.715 0.711 0.060 0.043 0.003 0.122 0.013 0.024 0.098 0.001

0.128 0.019 0.079 0.019 0.086 0.081 0.862 0.843 0.747 0.723 0.046 0.030 0.026 0.007

0.022 0.061 0.055 0.154 0.077 0.130 0.059 0.026 0.074 0.025 0.881 0.875 0.808 0.754

Bold text indicates that the item loads onto the given component.

Table 7 Pattern matrix for factor analysis of report out questionnaire input and process scalesa Component

IP3 IP2 IP4 IP5 IP1 AO4 AO2 AO1 MS3 MS2 MS5 MS1 TA3 TA2 AO3 TA1 MS4 TA4 a

1

2

3

4

0.882 0.872 0.850 0.617 0.591 0.423 0.090 0.223 0.025 0.021 0.115 0.160 0.037 0.132 0.001 0.054 0.153 0.025

0.096 0.051 0.094 0.075 0.158 0.066 0.861 0.733 0.046 0.033 0.148 0.009 0.054 0.139 0.301 0.052 0.181 0.003

0.152 0.041 0.015 0.102 0.060 0.119 0.062 0.169 0.891 0.839 0.655 0.396 0.014 0.013 0.235 0.276 0.280 0.266

0.055 0.004 0.109 0.085 0.201 0.028 0.036 0.059 0.029 0.003 0.121 0.350 0.830 0.794 0.763 0.508 0.481 0.479

Bold text indicates that the item loads onto the given component.

Bold text indicates that the item loads onto the given component.

analysis was used to verify the loading patterns and propose interpretations for the resulting factors. Confirmatory factor analysis was also used to verify the final scales (Farris, 2008a). As shown in Table 6, in the kickoff questionnaire factor analysis, three underlying factors were extracted, and these factors exactly matched the content of the three multi-item scales as originally designed. Thus, the original constructions of all three scales (goal clarity, goal difficulty, and affective commitment to change) were used in further analyses. As Table 7 shows, the factor analysis of the report out questionnaire input and process factors measures generally supported the construction of the scales but suggested some revisions to the original constructions. In the factor analysis, all five of the internal processes items (IP1–IP5) loaded onto a single factor (component one), supporting the initial construction of this scale. Meanwhile, three of the five management support items (MS3, MS2, and MS5) loaded onto a single factor (component two), three of the four team autonomy items (TA1, TA2, and TA3) loaded onto a single factor (component four), and two of the four action orientation items (AO1, AO2) loaded onto a single factor (component two), suggesting the need to remove the other items from these scales (see the Appendix A). As Table 8 shows, the loading pattern produced by the factor analysis of report out questionnaire outcome measures was the most complex in the study, but also generally supported the scale construction. The report out questionnaire originally contained three scales designed to measure the human resource outcomes of Kaizen events:

 understanding of continuous improvement (UCI), designed to measure the increase in team members’ knowledge of the philosophy of continuous improvement;

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

 skills (SK), designed to measure the increase in team 

AT1, AT3, and IMA4). Although the emergent factor consisted of items from different scales, it appeared to contain a consistent focus on employee problem-solving capability development. To differentiate it from the previous knowledge (UCI) and skills (SK) scales, a new label, kaizen capabilities (KC), was developed. The only item that did not load cleanly onto any factor was SK2, which was therefore removed from further analyses. Although examination of the items in the final attitude and kaizen capabilities scales (Table 9) supports the content validity of the revised scales, the scales should be tested in future research, and it is possible that further revisions may be warranted. The factor analysis also included a fourth scale from the report out questionnaire, impact on area (IMA), designed to measure event impact on the target work area, a technical system outcome. A third factor (component three), including three out of four IMA items, also emerged from the factor analysis.

members’ problem-solving skills; and attitude (AT), designed to measure impact on team members’ motivation.

However, the factor analysis suggested only two underlying dimensions. All the items related to affective response toward Kaizen event participation (AT2, AT4, and SK4) loaded onto a single factor (component three). This revised scale measures employee affect for Kaizen event activities and thus the original attitude (AT) name was retained. A second factor (component one) also emerged, consisting of items related to growth in employee problem-solving capabilities, including understanding of continuous improvement, problem-solving skills, motivation to examine daily work, and ability to work in problem-solving teams (UCI1, UCI2, UCI3, UCI4, SK1, SK3,

3.4. Reliability

Table 9 Final construction of attitude and kaizen capabilities scales Kaizen capabilities (KC)



      

Following the factor analyses, Cronbach’s a (Cronbach, 1951) was used to evaluate the reliability of the final, revised scales. The a values (see Table 10) were evaluated against the commonly applied thresholds of 0.70 for established scales (Nunnally, 1978) and 0.60 for newly developed scales (deVellis, 1991). All scales except one had a values greater than 0.70 and most scales (six out of nine) had a values of 0.80 or greater. The final action orientation scale had an a of 0.64. Although this a value is acceptable for newly developed scales, particularly given the small number of items in this scale (two), this value suggests that the action orientation measure could benefit from further refinement in future research.

Attitude (AT)

 Overall, this Kaizen event increased our team members’ knowledge of what continuous improvement is. (UCI1) In general, this Kaizen event increased our team members’ knowledge of how continuous improvement can be applied. (UC2) Overall, this Kaizen event increased our team members’ knowledge of the need for continuous improvement. (UCI3) In general, this Kaizen event increased our team members’ knowledge of our role in continuous improvement. (UCI4) Most of our team members can communicate new ideas about improvements as a result of participation in this Kaizen event. (SK1) Most of our team members gained new skills as a result of participation in this Kaizen event. (SK3) In general, this Kaizen event motivated the members of our team to perform better. (AT1) Overall, this Kaizen event increased our team members’ interest in our work. (AT3) Overall, this Kaizen event helped people in this area work together to improve performance. (IMA4)

 Most of our





53

team members liked being part of this Kaizen event. (AT2) Most members of our team would like to be part of Kaizen events in the future. (AT4) In general, our Kaizen event team members are comfortable working with others to identify improvements in this work area. (SK4)

3.5. Team-level properties The theoretical unit of analysis in this research was the team. To support inferences at the team level, all study measures used the team as the referent. However, for the measures of team perceptions, questionnaires were administered to all members within a team to increase the reliability of the data collected (Chan, 1998). Thus, it was necessary to verify that the actual data collected demonstrated evidence of shared, team-level properties.

Table 10 Summary of psychometric properties of multi-item scales Measure

Low. prim. load.

Cron. a

ICC(1) (p)

Avg.rwg

Percentage rwg 40.70

Goal clarity (GC) Goal difficulty (GDF) Affective commitment to change (ACC) Internal processes (IP) Management support (MS) Team autonomy (TA) Action orientation (AO) Kaizen capabilities (KC) Attitude (AT)

0.754 0.723 0.711 0.591 0.655 0.508 0.733 0.571 0.607

0.86 0.81 0.86 0.86 0.80 0.79 0.64 0.93 0.74

0.060 (0.046) 0.216 (o0.001) 0.173 (o0.001) 0.326 (o0.001) 0.125 (0.003) 0.185 (o0.001) 0.373 (o0.001) 0.121 (o0.001) 0.300 (o0.001)

0.88 0.83 0.94 0.97 0.88 0.88 0.76 0.98 0.90

94 88 96 100 90 98 75 100 96

ARTICLE IN PRESS 54

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Interrater agreement (rwg) (James et al. 1984, 1993) and the intraclass correlation coefficient (1), ICC(1), were used to determine whether the data collected on individual team members via the kickoff and report out questionnaires demonstrated evidence of shared team-level properties, and therefore could justifiably be aggregated. Interrater agreement, rwg, was evaluated using the commonly applied threshold of 0.70, which is interpreted as strong within-group agreement (George, 1990; Klein and Kozlowski, 2000; Van Mierlo et al., 2006). For each study variable, both the average rwg and the proportion of teams with rwg greater than 0.70 were analyzed. As Table 10 indicates, all kickoff and report out survey variables had an average rwg value greater than 0.70, and further had at least 75% of individual team rwg values greater than 0.70. In addition, for seven out of nine teams, 90% or more of individual team rwg values were greater than 0.70. Thus, an examination of rwg supports the existence of teamlevel properties for all kickoff and report out questionnaire variables. In addition, the range of rwg within each team was examined to determine whether any teams displayed low rwg on more than 50% of the perceptual measures, which might suggest that the team be removed from the research, as aggregation of individual-level data to the team level might not be appropriate in such a case. However, no such cases were found. ICC(1) was calculated using the Bartko (1976) formulation, where MSB is the between-team mean square, MSW is the within-team mean square and k is the average team size from analysis of variance (ANOVA). As the data set contained two levels of nesting (teams within organizations), a nested ANOVA was used to control for error due to correlation between teams within a given organization (Jetten et al., 2002). A significant ANOVA with team as the main effect indicates that ICC(1) is significant (Bliese, 2000; Klein and Kozlowski, 2000), supporting the presence of shared team-level properties for the given measure. In addition, the magnitude of ICC(1) was evaluated using the commonly applied thresholds of 0.200, indicating strong team-level properties (Molleman, 2005), and 0.100, indicating weaker team-level properties (James, 1982; Molleman, 2005; Schneider et al., 1998). As shown in Table 10, except for one variable, all kickoff and report out questionnaire variables had significant ICC(1) values that were greater than 0.100. In addition, four out of nine variables had ICC(1) values greater than 0.200. For one variable, goal clarity, the ICC(1) value was lower than the threshold of 0.100 but still significant. This, combined with the rwg values, led the authors to conclude that the inference of team-level properties was justified for goal clarity on the balance of the evidence. ICCð1Þ ¼

MSB  MSW MSB þ ½ðk  1Þ  MSW

(1)

4. Results Regression modeling was used to identify which event input and process factors were most strongly related to each of the two human resource outcomes (attitude and kaizen capabilities). Multiple regression was used to test

direct relationships (H1–H2), while mediation analysis (Judd and Kenny, 1981; Baron and Kenny, 1986; Kenny et al., 1998) was used to test indirect relationships (H3). Due to the nested structure of the data, it could not be assumed that the data for teams within a given organization were uncorrelated (Kenny and Judd, 1986), a fundamental and necessary assumption of multiple regression using ordinary least squares (OLS) estimation (Neter et al., 1996). Correlations between observations can result in biased standard error estimates and, therefore, spurious statistical test results, although estimates of the regression parameters themselves remain asymptotically unbiased (Hox, 1994; Lawal, 2003). Therefore, generalized estimating equations (GEE) (Liang and Zeger, 1986), executed in SAS 9.1.3 using PROC GENMOD, were used to account for correlation between teams within the same organization. GEE is a quasilikelihood estimation procedure, which can be used with regression models where the response variable comes from a distribution within the exponential family (Horton and Lipsitz, 1999). The GEE procedure uses a correlation matrix to model the association between clustered observations in the data set, e.g., repeated measures data, individuals within groups, etc., when developing parameter and standard error estimates, often resulting in differences in the estimates for parameter standard errors for GEE versus OLS. However, since the parameter estimates are asymptotically unbiased for both OLS and GEE, GEE models usually produce similar parameter estimates to OLS regression models. GEE has seen recent, growing use in the life and social sciences (Hardin and Hilbe, 2003), e.g., Pickering and Kisangani (2005), Szinovacz and Davey (2001), Whitford and Yates (2003), where hierarchically nested data are common. For each regression in this research, the GEE model used a Gaussian link function, assuming a normally distributed dependent variable, and an exchangeable correlation matrix, assuming equal correlation between all observations within a given cluster, i.e., teams within a given organization. Other options for modeling the correlation matrix include autoregressive and similar time-based correlation structures, none of which appeared to be strictly applicable to the current data set, especially given differences in timing, i.e. the interval between events, across organizations. Other approaches for modeling nested data include SEM (e.g., Koufteros and Marcoulides, 2006) and hierarchical linear modeling, HLM (e.g., Raudenbush and Byrk, 2002), both of which require larger sample sizes, e.g., 50–100 observations per level. Due to the exploratory nature of this research, a relatively large number of candidate predictors were tested, and the sample size was fairly small at the organization level. Sections 4.1 and 4.2 describe the regressions used to identify direct and indirect predictors, respectively. Prior to building the regression models, the distributions of all study variables were examined to determine whether any variables demonstrated severe departures from normality. In the cases of three variables, team leader experience, team kaizen experience, and event planning process, a log transformation was used to reduce the variance and absolute distance to outlying observations. The variance

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

inflation factor (VIF) was also calculated for all 14 input and process variables to determine whether multicollinearity appeared to be problematic in the current data set. A single VIF greater than 10.00 (Neter et al., 1996) or an average VIF greater than 3.00 is generally taken to indicate a multi-collinearity problem. In the current research, the maximum VIF was 2.91, while the average VIF was 1.83. Thus, multi-collinearity was deemed not to be problematic. The VIF values were also calculated for the variables in each final regression model, and these values were also well below the suggested thresholds.

4.1. Identification of direct predictors of outcomes Due to the exploratory nature of this research, no established hierarchy of predictor variable importance existed. Therefore, an exploratory regression (e.g. Olorunniwo and Udo, 2002) with a variable selection procedure was used to build the models. A backward selection procedure was used, because it is less likely to result in the exclusion of important variables than a forward or stepwise selection procedure (Neter et al., 1996; Field, 2005). At each step in the selection procedure, if the p-value for one or more variables was greater than a ¼ 0.10/k, where k is the number of parameters in the model (i.e., the number of predictor variables plus one), the variable with the largest p-value was removed. This procedure was repeated until all remaining variables were significant at the a ¼ 0.10/k level. After the initial model was developed, auxiliary search methods, e.g., automated OLS-based searches using Cp, R2, and adjusted R2, were used to test the robustness of the final solution and to identify additional models for consideration. A candidate model was considered to be viable if all variables had p-values less than 0.05 and most had p-values less than 0.10/k. A conservative approach, i.e., using small p-values, was used due to the fact that an overall test for the significance of the regression cannot be accurately calculated for quasi-likelihood estimation methods such as GEE in standard statistical software (Hardin and Hilbe, 2003; SAS Institute Inc., 2006). Following each regression, R2, adjusted R2, and residual analysis were used to examine model fit. Potential departures from linearity and normality were evaluated through the examination of residual plots and partial regression plots and through Wald–Wolfowitz runs tests, all of which have been recommended for use with GEE models (Chang, 2000; Hardin and Hilbe, 2003). The Wald–Wolfowitz test is a nonparametric sign test which analyzes the degree to which regression residuals follows a random distribution (Field, 2005). The residual plots, partial regression plots and Wald–Wolfowitz test did not indicate any apparent departures from linearity or normality for the final models. In addition, all standardized residual values for both models were less than 3.0 and all but a few (three for the attitude model and two for the kaizen capabilities model) were less than 2.0. Table 11 shows the final regression models for the two outcome variables: attitude and kaizen capabilities. As

55

Table 11 Regression analysis for direct predictors of attitude (AT) and kaizen capabilities (KC) Variable

Model 1 (y ¼ AT) _

b

p

Model 2 (y ¼ KC) _

b

Intercept 0.467 0.447 0.219 Team functional heterogeneity 0.547 0.012 Management support 0.250 0.013 Internal processes 0.694 o0.001 0.465 Goal difficulty 0.119 Team autonomy 0.234 Team kaizen experience 0.398 Team leader experience 0.195 Work area routineness 0.094 Affective commitment to change 0.222 R2 0.605 0.706 R2a 0.580 0.658 a r 0.019 0.071

p 0.687

o0.001 0.032 0.014 o0.001 0.020 0.002 0.049

a This value is the intraclass correlation coefficient, which measures the extent of correlation between model residuals from teams within the same organization.

shown, three variables were significant predictors of employee attitude toward events:

 internal processes (b ¼ 0.694, po0.001);  management support (b ¼ 0.250, p ¼ 0.013); and  team functional heterogeneity (b ¼ 0.547, p ¼ 0.012). The attitude model R2 was 0.61 and the adjusted R2 was 0.58, indicating that the model explains approximately 60% of the variance in the outcome. The threeand two-way interactions of the predictors in the final model were also tested to determine whether there were any significant interaction effects; however, none were found. As shown in Table 11, seven variables were significant predictors of kaizen capabilities:

      

internal processes (b ¼ 0.465, po0.001); team autonomy (b ¼ 0.234, p ¼ 0.014); affective commitment to change (b ¼ 0.222, p ¼ 0.049); goal difficulty (b ¼ 0.119, p ¼ 0.032); work area routineness (b ¼ 0.094, p ¼ 0.002); team kaizen experience (b ¼ 0.398, po0.001); and team leader experience (b ¼ 0.195, p ¼ 0.020).

The kaizen capabilities model R2 was 0.71 and the adjusted R2 was 0.66, indicating that the model explained approximately 70% of the variance in the outcome. Due to the larger number of predictors in the final model, the two-way and higher-order interactions were not tested for significance.

 Thus, overall, H1 and H2 were partially supported in that each social system outcome was significantly predicted by at least one input factor and at least one process factor.

ARTICLE IN PRESS 56

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

4.2. Mediation analysis to identify indirect predictors of outcomes Mediation analysis was used to test indirect relationships. A mediation model hypothesizes that a variable (x) indirectly influences a second variable (y) by acting though a third variable (z), which is called the mediating variable or mediator (Baron and Kenny, 1986). This mediating variable is typically a process variable, whereas the first variable, x, can be an input variable or another process variable. Several methods of testing mediation models are currently available (MacKinnon et al., 1995). Although the approaches differ somewhat in the estimation methods and the statistical tests used, they all involve jointly or separately testing the regression paths involving x, y, and z. In the current analysis, a modified version of Judd and Kenny (1981) and Baron and Kenny’s (1986) classic approach, as presented in Kenny et al. (1998) and Kenny (2006), was used. This approach includes testing three regression coefficients in two equations (see steps 1 and 2 below). Since each regression coefficient is tested separately, a Bonferroni correction (a ¼ 0.05/3 ¼ 0.0167) is needed to account for the increased chance of type-I error (Kenny, 2006). All significant process variables from the regression models for director predictors (see Table 11) were tested as potential mediators:

Table 12 Mediation analysis results for attitudea Step l (y0 ¼ IP, separate regression)

a

p

Goal clarity Goal difficulty Team autonomy Team functional heterogeneity Team kaizen experience Team leader experience Management support Event planning process Work area routineness

0.571 0.031 0.408 0.088 0.117 0.004 0.336 0.212 o0.001

o0.0001* 0.7342 0.0001* 0.7745 0.4650 0.9767 0.0036* 0.0759 0.9892

Step 2 (y0 ¼ AT, separate regression) b Internal processes Goal clarity Internal processes Team autonomy Internal processes Management support





Table 12 shows the mediation analysis results for attitude, while Table 13 shows the mediation analysis results for kaizen capabilities. The mediation analysis suggests the following:

c0

p

0.147

0.2762

0.185

0.0988

0.725 o0.0001* 0.701

o0.0001*

0.678 o0.0001* 0.285 0.0068

Step 3 (y0 ¼ IP, simultaneous regression)

a0

p

Goal clarity Team autonomy Management support Calculated effect of goal clarity on attitudeb

0.445 0.154 0.096 0.414

o0.0001* 0.2176 0.3984

a

 Step 1: The mediating process variable (z) was separately regressed on each of the nine input variables (x) and the resulting coefficient (a) was tested for significance. Step 2: If a significant relationship was demonstrated in step one, the outcome variable (y) was regressed on both the input variable (x) and the mediating process variable (z), and the resulting regression coefficients were tested for significance. A significant relationship between the mediating process variable (z) and the outcome (y) (coefficient b) suggests a mediation relationship. The relationship between the outcome variable (y) and the input variable (x) (coefficient c0 ) can be either significant (partial mediation) or nonsignificant (full mediation). Step 3: After the two preceding steps were accomplished for all nine input variables, the mediating process variable (z) was simultaneously regressed on all the input variables (xi) significant in step one. This step was performed to confirm whether each input variable (xi) was a significant unique predictor of the mediator (z), after controlling for the other input variables.

p

b



An asterisk indicates a significant relationship. Based on a  b; a closer estimate is a  bIP in Table 11, or 0.396.

kaizen capabilities. Although three input variables were significant in the first step of the mediation analysis, subsequent steps failed to provide clear support for mediation effects. Thus, H3 was partially supported with one significant mediation effect for each outcome variable.

5. Conclusions 5.1. Discussion of findings Following the conclusion of the study, findings were reported to the participating organizations to evaluate the validity of study conclusions and to allow the organizations to benefit from the results. In general, participating organizations found the study results convincing and used study feedback to evaluate the effectiveness of their current practices and to identify potential changes. As mentioned, the study provided partial support for the three broad research hypotheses. Specific findings of the study are as follows:

 Internal processes and goal clarity were the strongest  Internal processes is a significant mediator of the effect



of goal clarity on both attitude and kaizen capabilities (see Tables 12 and 13), accounting for an estimated 57% of the direct effect of internal processes. Affective commitment to change is not a significant mediator of the effect of any of the input variables on

predictors of both kaizen capabilities and attitude.

 Management support and team functional heteroge

neity were significant predictors of attitude, but not kaizen capabilities. Team autonomy, affective commitment to change, goal difficulty, work area routineness, team kaizen

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

57

Table 13 Mediation analysis results for Kaizen capabilitiesa Step l (y0 ¼ z., separate regression)

z ¼ IP A

p

z ¼ ACC a

p

Goal clarity Goal difficulty Team autonomy Team functional heterogeneity Team kaizen experience Team leader experience Management support Event planning process Work area routineness

0.571 0.031 0.408 0.088 0.117 0.004 0.336 0.212 o.001

o.0001* 0.7342 0.0001* 0.7745 0.4650 0.9767 0.0036* 0.0759 0.9892

0.663 0.168 0.350 0.269 0.051 0.005 0.338 0.119 0.032

o.0001* 0.0556 0.0013* 0.3913 0.7641 0.9646 0.0043* 0.3377 0.4917

Step 2 (y0 ¼ KC, separate regression)

b

p

Mediator variable (z) Goal clarity Mediator variable (z) Team autonomy Mediator variable (z) Management support

0.610

o.0001*

c0

0.227 0.500

o.0001*

0.614

o.0001*

0.382 0.250

p

c0

b

p

0.254

0.1317

0.230

0.0489

0.334

0.0139*

0.0770 o.0001* 0.0142*

p

0.357

0.0424

0.495

o.0001*

0.338

0.0059*

Step 3 (y0 ¼ z., simultaneous regression)

a0

p

b

p

Goal clarity Team Autonomy Management Support Calculated effect of Goal clarity on Kaizen capabilitiesb

0.445 0.154 0.096 0.348

o.0001* 0.2176 0.3984

0.603 0.046 0.066 N/a

o.0001* 0.6858 0.5178

a b



An asterisk indicates a significant relationship. Based on a  b; a closer estimate is a  bIP in Table 11, or 0.266.

experience and team leader experience were significant predictors of kaizen capabilities but not attitude. Some variables proposed to affect Kaizen event outcomes, i.e., action orientation, tool quality, tool appropriateness, and event planning process, showed no significant relationship to either outcome in this study.

Internal processes was by far the strongest predictor of both attitude (b ¼ 0.694) and kaizen capabilities (b ¼ 0.465), suggesting that maintaining positive group dynamics may be the most important factor for generating employee motivation to participate in future improvement activities and for developing employee problemsolving capabilities. This result is aligned with previous empirical team research, which has consistently reported a positive relationship between internal processes and human resource outcomes (e.g., Bailey, 2000; Jehn, 1995; Pinto et al., 1993; Vinokur-Kaplan, 1995). Acting indirectly through internal processes, goal clarity is also one of the strongest predictors of attitude and kaizen capabilities, accounting for approximately 60% of the direct effect of internal processes. This finding is aligned with several previous studies in team effectiveness reporting a positive relationship between goal clarity and effectiveness (e.g., Doolen et al, 2003a; Koch, 1979; Pritchard et al., 1988; Van Aken and Kleiner, 1997). This finding also supports the majority of published Kaizen event accounts, which have suggested using clear and

specific goals (e.g., Adams et al., 1997; Bradley and Willett, 2004; Rusiniak, 1996), and contradicts the few that have suggested using more loosely defined goals (Kumar and Harms, 2004; Wittenberg, 1994). The finding that management support had a significant positive relationship to attitude, but not to kaizen capabilities, partially supports the consistent emphasis on management support in published Kaizen event accounts (e.g., Adams et al., 1997; Bateman, 2005; Bicheno, 2001; Bradley and Willett, 2004; Kumar and Harms, 2004; Martin, 2004; McNichols et al., 1999; Patil, 2003; Sheridan, 1997; Tanner and Roncarti, 1994; Taylor and Ramsey, 1993). In addition, it agrees with previous team studies which have reported a positive relationship between management support and team effectiveness (e.g., Campion et al. 1993; Doolen et al., 2003a; Hyatt and Ruddy, 1997; Pinto and Slevin, 1987). The finding that team functional heterogeneity had a significant negative relationship to attitude, but not to kaizen capabilities, is aligned with previous team research suggesting that cross-functional teams may experience lower levels of enjoyment in working together (Ancona and Caldwell, 1992; Amason and Schweiger, 1994; Baron, 1990). However, the mediation analysis showed no significant relationship between team functional heterogeneity and internal processes, suggesting that more diverse teams were still able to develop effective communication processes (Azzi, 1993; Earley and Mosakowski, 2000). This appears to agree with team research indicating that cross-functional teams may experience high

ARTICLE IN PRESS 58

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

quality decision-making processes (e.g., Jackson, 1992; Jehn et al., 1999; Lovelace et al., 2001; McGrath, 1984). The finding that team autonomy had a significant positive relationship to kaizen capabilities, but not to attitude, partially supports the assertions in the published Kaizen event accounts that autonomy promotes positive outcomes (e.g., Adams et al., 1997; Bicheno, 2001; Bradley and Willett, 2004; Foreman and Vargas, 1999; Kumar and Harms, 2004). Some empirical team studies have similarly reported a positive relationship between autonomy and outcomes (Cohen and Ledford, 1994; Wall et al., 1986). The finding that affective commitment to change had a significant positive relationship to kaizen capabilities, but not to attitude, partially supports the emphasis in the published Kaizen event accounts on creating team buy-in for the improvements targeted by the event (Bradley and Willett, 2004; McNichols et al., 1999; Melnyk et al., 1998; Taylor and Ramsey, 1993; Wheatley, 1998). In addition, this finding is aligned with the organizational change literature, which suggests a relationship between participant commitment and change initiative effectiveness (Chan et al., 2005; Keating et al., 1999). The finding that goal difficulty had a significant positive relationship to kaizen capabilities, but not to attitude, suggests that, from the standpoint of employee learning, the more challenging the goals, the more positive the skill development results, which would support the Kaizen event accounts that advocate using highly challenging goals (e.g. Bicheno, 2001; Bradley and Willett, 2004; Cuscela, 1998; Gregory, 2003; Kumar and Harms, 2004; LeBlanc, 1999; Minton, 1998; Rusiniak, 1996; Tanner and Roncarti, 1994; Treece, 1993). However, findings related to technical system outcomes revealed negative relationships between goal difficulty and technical success outcomes (Farris et al., 2008a), suggesting a trade-off between technical and skill development outcomes. Meanwhile, the finding that work area routineness had a significant positive relationship to kaizen capabilities, but not to attitude, supports the suggestions in published Kaizen event accounts that less complex work areas provide favorable ‘‘learning laboratories’’ for skill development (LeBlanc, 1999; Bradley and Willett, 2004), and aligns with findings in the organizational learning literature (Vits and Gelders, 2002). This effect may be due to the fact that more predictable work areas allow more complete implementation of lean practices. The finding that team kaizen experience had a significant negative relationship to kaizen capabilities, but not to attitude, is consistent with the well-known learning curve effect (Wright, 1936), which predicts decreasing incremental gains in learning from each successive attempt at a given application. The finding that team leader experience had a significant negative relationship to kaizen capabilities, but not to attitude, is unexpected and warrants further investigation. A manager in one of the participating organizations suggested that, based on his experience, teams with more seasoned leaders tended to skip or hurry through steps in the Kaizen event process, under the paradigm that they ‘‘already know what they are doing.’’ In this study, although data were collected on the activities of the team

through the team activities log, there was no formal measure of the extent to which teams adhered to the formal Kaizen event process. Recent literature has indicated a link between process adherence and team and organizational outcomes (Douglas and Judge, 2001; Ghosh and Sobek, 2007; Handfield et al., 1999; Linderman et al., 2006), although some published Kaizen event guidelines propose a less rigid adherence to the organization’s Kaizen event format (e.g., Bradley and Willett, 2004), including even omitting training in some cases (e.g., Bicheno, 2001; Gregory, 2003; McNichols et al., 1999). Teams with high team leader experience might also be more likely to jump to solutions similar to those used in previous events that the leader has conducted, thus limiting team creativity and team member participation in designing solutions. Or, more experienced leaders might tend to adopt a more directive leadership style, allowing individual team members less role in shaping the solution. Such directive leadership has been shown to be negatively related to team effectiveness in at least one study (Durham et al., 1997). Finally, it is possible that in teams with less experienced leaders the Kaizen event facilitator, who is typically highly experienced in problem-solving tools and applications, may tend to become more involved in coaching the team through the solution process, providing on-the-floor training, etc., thus contributing to greater learning. Future research on the relationship between team leader experience, team process adherence, team creativity, leadership style, facilitator involvement, and kaizen capabilities is clearly needed. Finally, some variables hypothesized to affect outcomes, i.e., action orientation, tool quality, tool appropriateness, and event planning process, showed no significant relationship to either outcome in this study. This finding does not prove that these variables are unimportant to the development of attitude and kaizen capabilities, although it does suggest that they might be less important than other variables. It should also be noted that these variables may be related to technical system outcomes or to social system outcomes not studied in this research. Although admittedly preliminary, the findings of this research can be further used to develop initial guidelines for organizations (to be tested in future research), including the following:

 To obtain maximum impact on both employee attitude



and problem-solving ability development, these results suggest that organizations should seek to maintain a high level of positive internal team dynamics, through the use of structured mechanisms (e.g., team ground rules, ice breakers, training, charters) and facilitator coaching. A key structural mechanism is the development of explicit goal statements in advance of the event. To further develop employee motivation to participate in events, these results suggest that organizations must maintain strong and visible management support during events. Meanwhile, organizations should recognize that cross-functional heterogeneity may have a negative impact on team member affect toward events, and potentially compensate for this effect through

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65





other variables, e.g., internal processes, management support. To further increase team problem-solving capability development, these results suggest that teams should be allowed a high degree of autonomy and that the organization should use mechanisms designed to increase participant buy-in for the event (e.g., clearly describing the reasons for the event, demonstrating how the event will positively impact both individual and organizational interests). Organizations should also recognize that the use of difficult goals could increase team learning, although other findings suggests that goal difficulty may compromise technical system results (Farris et al., 2008a). Finally, these results suggest that organizations should recognize that more complex work area may contribute to a lower degree of team learning, and potentially compensate for this effect through other mechanisms, e.g., autonomy, goal clarity, etc. Organizations should also recognize that greater team and team leader experience do not necessarily result in greater team learning, and should consider compensating for this effect through other mechanisms or through using less experienced personnel.

5.2. Limitations and future research To the authors’ knowledge, this study was the first to empirically investigate the quantitative relationships between input and process variables and initial human resource outcomes in Kaizen events, using multiple events from multiple organizations. Additional research is needed to confirm and extend these findings. Other study limitations include:

 Although data were collected from 51 Kaizen events



and over 300 participants, the sample represented only six organizations. Furthermore, the selection criteria deliberately resulted in the selection of manufacturing organizations that were relatively experienced in the strategic use of Kaizen events. Thus, it is possible that findings might not generalize to markedly different organizational contexts. However, the six organizations spanned a variety of industries and approximately 16% of the events were in non-manufacturing work areas. Events were selected using a random sampling process designed to capture the mixture of events occurring naturally within the participating organizations; the researchers did not attempt to control the percentages of specific event types sampled. Most of the events in this research (71%) were general processes improve-





59

ment events, often using the ‘‘standard work’’ methodology. A smaller number of events were of other types: total productive maintenance (TPM) (14%), single minute exchange of die (SMED) (8%), 5S (6%) and value stream mapping (VSM) (2%). (Note, these percentages total 101% due to rounding.) In addition, most events included implementation of the solution during the event (75%), targeted manufacturing processes or work areas (84%), and were highly successful with 100% goal achievement (69%). Thus, it is possible that findings might not generalize to events of markedly different types. This research only included event-level variables. It is possible that some organizational-level effects also contribute to event-level outcomes. While GEE accounts for correlation in regression residuals between teams within the same organization, it also assumes uniformity of regression slopes across organizations. Hierarchical linear modeling (HLM) has the capability to model differences in effects of both organization and event-level variables across organizations. However, as noted, a larger data set are needed to construct models of any size (Raudenbush and Byrk, 2002). Finally, this research only investigated the determinants of initial event outcomes, while sustainability of both human resource and technical outcomes is of great interest, and many organizations appear to struggle to sustain initial gains from Kaizen events (e.g., Bateman and David, 2002; Bateman 2005; Laraia et al, 1999).

In summary, this research has systematically investigated the empirical relationships between input and process variables and human resource outcomes using a sample of 51 events from six organizations. In closing, the authors reiterate the call from Melnyk et al. (1998) regarding the need for additional research on this topic. If the proliferation of books, websites, conference proceedings and non-refereed articles is any indication, organizational interest in Kaizen events only appears to have increased since the late 1990s, and little scholarly research has been conducted to date. It is the responsibility of the operations management and industrial engineering research community to help organizations better understand this phenomenon and how it contributes to longer-term organizational success. Appendix A See Table A1 for summary of study variables and measures.

60

Table A1 Summary of study variables and measures Scale

Instrument

Itemsa

Sources

Goal clarity (GC)

Kickoff questionnaire

Van Aken and Kleiner 6-point Likert (1997) and Wilson et type al. (1998)

Team average for scale

Goal difficulty (GDF)

Kickoff questionnaire

6-point Likert Ivancevich and McMahon (1977) and type Hart et al. (1989)

Team average for scale

Affective commitment to change (ACC)

Kickoff questionnaire

Our team has clearly defined goals. (GC1) The performance targets our team must achieve to fulfill our goals are clear. (GC2) Our goals clearly define what is expected of our team. (GC3) Our entire team understands our goals. (GC4) Our team’s improvement goals are difficult. (GDF1) Meeting our team’s improvement goals will be tough. (GDF2) It will take a lot of skill to achieve our team’s improvement goals. (GDF3) It will be hard to improve this work area enough to achieve team’s goals. (GDF4) In general, members of our team believe in the value of this Kaizen event. (ACC1) Most of our team members think that this Kaizen event is a good strategy for this work area. (ACC2) In general, members of our team think that it is a mistake to hold this Kaizen event. (REVERSE CODED IN FACTOR ANALYSIS) (ACC3) Most of our team members that this Kaizen event will serve an important purpose. (ACC4) Most of our team members think that things will be better with this Kaizen event. (ACC5) In general, members of our team believe that this Kaizen event is needed. (ACC6) Our team communicated openly. (IP1) Our team valued each member’s unique contributions. (IP2) Our team respected each others’ opinions. (IP3) Our team respected each others’ feelings. (IP4) Our team valued the diversity in our team members. (IP5) Our team had enough contact with management to get our work done. (MS1)* Our team had enough materials and supplies to get our work done. (MS2) Our team had enough equipment to get our work done. (MS3) Our team had enough help from our facilitator to get our work done. (MS4)* Our team had enough help from others in our organization to get our work done. (MS5) Our team had a lot of freedom in determining what changes to make to this work area. (TA1) Our team had a lot of freedom in determining how to improve this work area. (TA2) Our team was free to make changes to the work area as soon as we thought of them. (TA3) Our team had a lot of freedom in determining how we spent our time during the event. (TA4)* Our team spent as much time as possible in the work area. (AO1) Our team spent very little time in our meeting room. (AO2) Our team tried out changes to the work area right after we thought of them. (AO3)* Our team spent a lot of time discussing ideas before trying them out in the work area. (REVERSE CODED IN FACTOR ANALYSIS) (AO4)* Overall, this Kaizen event increased our team members’ knowledge of what continuous improvement is. (UCI1)* In general, this Kaizen event increased our team members’ knowledge of how continuous improvement can be applied. (UCI2)* Overall, this Kaizen event increased our team members’ knowledge of the need for continuous improvement. (UCI3)* In general, this Kaizen event increased our team members’ knowledge of our role in continuous improvement. (UCI4)* Most of our team members can communicate new ideas about improvements as a result of participation in this Kaizen event. (SK1)* Most of our Kaizen event team members are able to measure the impact of changes made to this work area. (SK2)* Most of our team members gained new skills as a result of participation in this Kaizen event. (SK3)* In general, our Kaizen event team members are comfortable working with others to identify improvements in this work area. (SK4)*

Herscovitch and Meyer (2002)

6-point Likert type

Team average for scale

Hyatt and Ruddy (1997)

6-point Likert type

Team average for scale

Doolen et al. (2003a)

6-point Likert type

Team average for scale

Kirkman and Rosen (1999), Groesbeck (2001), and Hayes (1994) Original to this researchc

6-point Likert type

Team average for scale

6-point Likert type

Team average for scale

Original to this researchd

6-point Likert type

N/ab

Original to this researchc

6-point Likert type

N/a

Report out questionnaire

Team autonomy (TA)

Report out questionnaire

Action orientation Report out (AO) questionnaire

Understanding of CI (UCI)

Report out questionnaire

Skills (SK)

Report out questionnaire

ARTICLE IN PRESS

Management support (MS)

Value

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Internal processes Report out (IP) questionnaire

Meas. scale

Attitude (AT)

Report out questionnaire

Team kaizen experience (TOTKE) Team functional heterogeneity (FUNHET)

Kickoff questionnaire Event information questionnaire

In general, this Kaizen event motivated the members of our team to perform better. (AT1)* Most of our team members liked being part of this Kaizen event. (AT2) Overall, this Kaizen event increased our team members’ interest in our work. (AT3)* Most members of our team would like to be part of Kaizen events in the future. (AT4) Not including this event, how many Kaizen events total have you participated in?

Original to this researchd

6-point Likert type

Team average for scale

N/a

Continuous

Team average plus one

Continuous

H index value

Continuous

Unadjusted response value

6-point Likert type

Average for four items

6-point Likert type

Average rating for all tools listed

Please fill-in the number of team members in each job category. (choices: operator, technician, engineer, Shannon (1948) and manager, supervisor, other) Teachman (1980) Calculated as follows, where pi is the proportion of team members from each functional category P H¼ pi ðlogð1=pi ÞÞ i

Tool appropriateness (TAPP) Tool quality (TQUAL)

Event information questionnaire Event information questionnaire Event information questionnaire

Event planning process (HRSPLN)

a

Including this event, how many Kaizen events total has the team leader led or co-led in the past three years?

N/a

The work the target work area does is routine. (WAC1) The target work area produces the same product (SKU) most of the time. (WAC2) A given product (SKU) requires the same processing steps each time it is produced. (WAC3) Most of the products (SKUs) produced in the work area follow a very similar production process. (WAC4)

Perrow (1967), Duncan (1972), Miltenberg (1995), Withey et al. (1983), and Gibson and Vermeulen (2003) (Respondents first listed the problem-solving tools used by the team). For each tool, please rate the team’s N/a use of the tool on appropriateness of using this tool to address the team’s goals (Uses the same tool list above.) For each tool, please rate the quality of the team’s use of this tool.

N/a

6-point Likert type

Average rating for all tools listed

How many hours total did you and others spend planning and doing other pre-work for this event?

N/a

Continuous

Unadjusted response value

An asterisk indicates that the item was removed from its original scale following the factor analysis (see Section 3.3). The understanding of CI and skills variables were ultimately replaced by the kaizen capabilities variable (see Section 3.3), the value of which is the team average for the scale. c The items in this scale were newly developed, i.e., original to this research stream; however, the concept of team action-orientation was informed by the concepts of personal initiative (Frese et al., 1997) and individual-level action orientation (Kuhl, 1992). d The items in this scale were newly developed, i.e., original to this research stream; however, pilot study items and statistics were reported in Doolen et al., 2003b and Doolen et al., 2008. b

ARTICLE IN PRESS

Event information questionnaire Event information questionnaire

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Team leader experience (LDREXP) Work area routineness (WAC)

61

ARTICLE IN PRESS 62

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

References Adam, E.E., Corbett, L.M., Flores, B.E., Harrison, N.J., Lee, T.S., Rho, B.H., Ribera, J., Samson, D., Westbrook, R., 1997. An international study of quality improvement approach and firm performance. International Journal of Operations and Production Management 17 (9), 842–873. Adams, M., Schroer, B.J., Stewart, S.K., 1997. QuickstepTM process improvement: Time-compression as a management strategy. Engineering Management Journal 9 (2), 21–32. Adler, P.S., 1993a. The new ‘‘learning bureaucracy’’: New United Motor Manufacturing Inc. In: Staw, B.M., Cummings, L.L. (Eds.), Research in Organizational Behavior, vol. 15. JAI Press, Greenwich, CT, pp. 111–194. Adler, P.S., 1993b. Time-and-motion regained. Harvard Business Review 71 (1), 97–108. Amason, A.C., Schweiger, D.M., 1994. Resolving the paradox of conflict, strategic decision making and organizational performance. International Journal of Conflict Management 5, 239–253. Ancona, D., Caldwell, D.F., 1992. Demography and design: Predictors of new product team performance. Organization Science 3, 321–341. Azzi, A.E., 1993. Implicit and category-based allocations of decisionmaking power in majority–minority relations. Journal of Experimental Psychology 29, 203–228. Bailey, D.E., 2000. Modeling work group effectiveness in high-technology manufacturing environments. IIE Transactions 32 (4), 361–368. Bailey, J.E., Rose, R.L., 1988. Maybe Winnebago just wasn’t ready for big city bosses. Wall Street Journal 17 (17 October), 1. Baron, R.A., 1990. Countering the effects of destructive criticism: The relative efficacy of four interventions. Journal of Applied Psychology 75, 235–245. Baron, R.M., Kenny, D.A., 1986. The moderator–mediator variable distinction in social psychological research: Conceptual, strategic and statistical considerations. Journal of Personality and Social Psychology 51, 1173–1182. Bartko, J.J., 1976. On various intraclass correlation reliability coefficients. Psychological Bulletin 83 (5), 762–765. Bateman, N., 2005. Sustainability: The elusive element in process improvement. International Journal of Operations and Production Management 25 (3), 261–276. Bateman, N., David, A., 2002. Process improvement programmes: A model for assessing sustainability. International Journal of Operations and Production Management 22 (5), 515–526. Bateman, N., Rich, N., 2003. Companies’ perceptions of inhibitors and enables for process improvement activities. International Journal of Operations and Production Management 23 (2), 185–199. Belassi, W., Tukel, O.I., 1996. A new framework for determining critical success/failure factors in projects. International Journal of Project Management 14 (3), 141–151. Biazzo, S., Panizzolo, R., 2000. The assessment of work organization in lean production: The relevance of the worker’s perspective. Integrated Manufacturing Systems 11 (1), 6–15. Bicheno, J., 2001. Kaizen and Kaikaku. In: Taylor, D., Brunt, D. (Eds.), Manufacturing Operations and Supply Chain Management: The LEAN approach. Thomson Learning, London, pp. 175–184. Bliese, P.D., 2000. Within-group agreement, non-independence, and reliability: Implications for data aggregation and analysis. In: Klein, K.J., Kozlowski, S.W.J. (Eds.), Multilevel Theory, Research and Methods in Organizations. Jossey-Bass, San Francisco, pp. 349–381. Bradley, J.R., Willett, J., 2004. Cornell students participate in Lord Corporation’s Kaizen projects. Interfaces 34 (6), 451–459. Brown, K.A., Mitchell, T.R., 1991. A comparison of just-in-time and batch manufacturing: The role of performance obstacles. Academy of Management Journal 34 (4), 906–917. Butterworth, C., 2001. From value stream mapping to shop floor improvement: A case study of Kaikaku. In: Taylor, D., Brunt, D. (Eds.), Manufacturing Operations and Supply Chain Management: The LEAN Approach. Thomson Learning, London, pp. 185–193. Campion, M.A., Medsker, G.J., Higgs, A.C., 1993. Relations between work group characteristics and effectiveness: Implications for designing effective work groups. Personnel Psychology 46, 823–850. Chan, D., 1998. Functional relationships among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology 83 (2), 234–246. Chan, F.T.S., Lau, H.C.W., Ip, R.W.L., Chan, H.K., Kong, S., 2005. Implementation of total productive maintenance: A case study. International Journal of Production Economics 95, 71–94. Chang, Y., 2000. Residual analysis of the generalized linear models for longitudinal data. Statistics in Medicine 19 (10), 1277–1293.

Cohen, S.G., Bailey, D.E., 1997. What makes teams work: Group effectiveness research from the shop floor to the executive suite. Journal of Management 23 (3), 239–290. Cohen, S.G., Ledford, G.E., 1994. The effectiveness of self-managing teams: A quasi-experiment. Human Relations 47 (1), 13–43. Cohen, S.G., Ledford, G.E., Spreitzer, G.M., 1996. A predictive model of self-managing work team effectiveness. Human Relations 49 (5), 643–676. Co, H.C., Patuwo, B.E., Hu, M.Y., 1998. The human factor in advanced manufacturing technology adoption, an empirical analysis. International Journal of Operations and Production Management 18 (1), 87–106. Cronbach, L., 1951. Coefficient alpha and the internal structure of tests. Psychiatrika 16, 297–334. Cuscela, K.N., 1998. Kaizen blitz attacks work processes at Dana Corp. IIE Solutions 30 (4), 29–31. Dankbaar, B., 1997. Lean production: Denial, confirmation or extension of sociotechnical systems design? Human Relations 50 (5), 567–583. Dar-El, E.M., 1997. What we really need is TPQM!. International Journal of Production Economics 52, 5–13. David, I., 2000. Drilled in Kaizen. Professional Engineering 13 (9), 30–31. Delbridge, R., 1998. Life on the Line in Contemporary Manufacturing. Oxford University Press, Oxford. Delbridge, R., Turnbull, P., Wilkinson, B., 1992. Pushing back the frontiers: Management control and work intensification under JIT/TQM factory regimes. New Technology, Work and Employment 7 (2), 97–106. deTreville, S., Antonakis, J., 2006. Could lean production job design be intrinsically motivating? Contextual, configurational, and levels-ofanalysis issues. Journal of Operations Management 24 (2), 99–123. deVellis, R.F., 1991. Scale Development: Theory and Application. Sage Publishing, Newbury Park, CA. Dillman, D., 2000. Mail and Internet Surveys: The Tailored Design Method. Wiley, New York. Doolen, T.L., Hacker, M.E., Van Aken, E.M., 2003a. The impact of organizational context on work team effectiveness: A study of production teams. IEEE Transactions on Engineering Management 50 (3), 285–296. Doolen, T.L., Worley, J., Van Aken, E.M., Farris, J., 2003b. Development of an assessment approach for kaizen events. In: 2003 Industrial Engineering and Research Conference, Portland, OR. Doolen, T.L., Van Aken, E.M., Farris, J., Worley, J., Huwe, J., 2008. A field study of kaizen events and organizational performance. International Journal of Productivity and Performance Management, in press. Douglas, T.J., Judge, W.Q., 2001. Total quality management implementation and competitive advantage: The role of structural control and exploitation. Academy of Management Journal 44 (1), 158–169. Drickhamer, D., 2004a. Just-in-time training. Industry Week 253 (7), 69. Drickhamer, D., 2004b. Braced for the future. Industry Week 253 (10), 51–52. Duncan, G.M., 1972. Characteristics of organizational environments and perceived environmental uncertainty. Administrative Science Quarterly 17, 313–327. Durham, C.C., Knight, D., Locke, E.A., 1997. Effects of leader role, team-set goal difficulty, efficacy, and tactics on team effectiveness. Organizational Behavior and Human Decision Processes 72 (2), 203–231. Earley, P.C., Mosakowski, E., 2000. Creating hybrid team cultures: An empirical test of transnational team functioning. Academy of Management Journal 43 (1), 26–49. Earley, P.C., Connolley, T., Ekegren, G., 1989. Goals, strategy development and task performance: Some limits on the efficacy of goal setting. Journal of Applied Psychology 74 (1), 24–33. Eisenhardt, K.M., 1989. Building theories from case studies. Academy of Management Review 14, 532–550. Erez, M., Zidon, I., 1984. Effect of goal acceptance on the relationship of goal difficulty to performance. Journal of Applied Psychology 69, 69–78. Fabrigar, L.R., Wegener, D.T., MacCallum, R.C., Strahan, E.J., 1999. Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods 4 (3), 272–299. Farris, J., Van Aken, E.M., Doolen, T., Worley, J., 2008a. Identifying the determinants of Kaizen event success: An empirical investigation. Working Paper. Farris, J., Van Aken, E.M., Doolen, T.L., Worley, J., 2008b. Learning from less successful Kaizen events: A case study. Engineering Management Journal 20 (3), 10–20. Field, A., 2005. Discovering Statistics Using SPSS, second ed. Sage Publications Ltd., London.

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Finch, H., 2006. Comparison of the performance of varimax and promax rotations: Factor structure recovery for dichotomous items. Journal of Educational Measurement 43 (1), 39–52. Foreman, C.R., Vargas, D.H., 1999. Affecting the value chain through supplier Kaizen. Hospital Materiel Management Quarterly 20 (3), 21–27. Frese, M., Fay, D., Hilburger, T., Leng, K., Tag, A., 1997. The concept of personal initiative: Operationalization, reliability and validity in two German samples. Journal of Occupational and Organizational Psychology 70, 139–161. Fucini, J.J., Fucini, S., 1990. Working for the Japanese: Inside Mazda’s American Plant. The Free Press, New York. George, J.M., 1990. Personality, affect, and behavior in groups. Journal of Applied Psychology 75, 107–116. Ghosh, M., Sobek, D., 2007. Fostering second-order problem-solving in healthcare. In: 2007 Industrial Engineering and Research Conference, Nashville, TN. Gladstein, D.L., 1984. Groups in context: A model of task group effectiveness. Administrative Science Quarterly 29 (4), 499–517. Gibson, C., Vermeulen, F., 2003. A healthy divide: Subgroups as a stimulus for team learning behavior. Administrative Science Quarterly 48 (2), 202–239. Gregory, A., 2003. Haring for change. Works Management 56 (5), 18–21. Groesbeck, R., 2001. An empirical study of group stewardship and learning: Implications for work group effectiveness. Unpublished Doctoral Dissertation, Grado Department of Industrial and Systems Engineering, Virginia Tech, Blacksburg, VA. Guimaraes, T., 1997. Empirically testing the antecedents of BPR success. International Journal of Production Economics 50, 199–210. Gunasekaran, A., Korukondab, A.R., Virtanerf, I., Yli-Olli, P., 1994. Improving productivity and quality in manufacturing organizations. International Journal of Production Economics 36, 169–183. Guzzo, R.A., Shea, G.P., 1992. Group performance and intergroup relations in organizations. In: Dunnette, M.D., Hough, L.M. (Eds.), Handbook of Industrial and Organizational Psychology, vol. 3. Consulting Psychologists Press, Palo Alto, CA, pp. 269–313. Hackman, J.R., 1987. The design of work groups. In: Lorsch, J.W. (Ed.), The Handbook of Organizational Behavior. Prentice Hall, Englewood Cliffs, NJ, pp. 315–342. Hales, D.N., Chakravorty, S.S., 2006. Implementation of Deming’s style of quality management: An action research study in a plastics company. International Journal of Production Economics 103, 131–148. Handfield, R., Jayaram, J., Ghosh, S., 1999. An empirical examination of quality tool deployment patterns and their impact on performance. International Journal of Production Research 37 (6), 1403–1426. Hardin, J.W., Hilbe, J.M., 2003. Generalized Estimating Equations. Chapman and Hall/CRC, Boca Raton, FL. Hart, S.H., Moncrief, W.C., Parasuraman, J., 1989. An empirical investigation of salespeople’s performance, effort and selling method during a sales contest. Journal of the Academy of Marketing Science 17 (1), 29–39. Harvey, J., 2004. Match the change vehicle and method to the job. Quality Progress 37 (1), 41–48. Hasek, G., 2000. Extraordinary extrusions. Industry Week 249 (17), 79–80. Hayes, B.E., 1994. How to measure empowerment. Quality Progress 27 (2), 41–46. Heard, E., 1997. Rapid-fire Improvement with Short-cycle Kaizen. American Production and Inventory Control Society, Washington, DC. Herron, C., Braiden, P.M., 2006. A methodology for developing sustainable quantifiable productivity improvement in manufacturing companies. International Journal of Production Economics 104, 143–153. Herscovitch, L., Meyer, J.P., 2002. Commitment to organizational change: Extension of a three-component model. Journal of Applied Psychology 87 (3), 474–487. Horton, N.J., Lipsitz, S.R., 1999. Review of software to fit generalized estimating equation regression models. The American Statistician 53 (2), 160–169. Hox, J.J., 1994. Hierarchical regression models for interviewer and respondent effects. Sociological Methods and Research 17 (1), 300–316. Huber, V.L., Brown, K.A., 1991. Human resource issues in cellular manufacturing: A sociotechnical analysis. Journal of Operations Management 10 (1), 138–159. Huber, V.L., Hyer, N.L., 1985. The human factor in cellular manufacturing. Journal of Operations Management 5 (2), 213–228. Hyatt, D.E., Ruddy, T.M., 1997. An examination of the relationship between work group characteristics and performance: Once more into the breach. Personnel Psychology 50 (3), 553–585.

63

Ivancevich, J.M., McMahon, J.T., 1977. A study of task-goal attributes, higher order need strength, and performance. Academy of Management Journal 20 (4), 552–563. Jackson, S., 1992. Team composition in organizations. In: Worchel, S., Wood, W., Simpson, J. (Eds.), Group Process and Productivity. Sage, London, pp. 1–12. Jackson, P.R., Mullarkey, S., 2000. Lean production teams and health in garment manufacture. Journal of Occupational Health Psychology 5, 231–245. James, L.R., 1982. Aggregation bias in estimates of perceptual agreement. Journal of Applied Psychology 69 (2), 219–229. James, L.R., Demaree, R.G., Wolf, G., 1984. Estimating within-group interrater reliability with and without response bias. Journal of Applied Psychology 69, 3–32. James, L.R., Demaree, R.G., Wolf, G., 1993. rwg: An assessment of withingroup interrater agreement. Journal of Applied Psychology 78 (2), 306–309. Janz, B.D., 1999. Self-directed teams in IS: Correlates for improved systems development work outcomes. Information and Management 35, 171–192. Jehn, K., 1995. A multimethod examination of the benefits and detriments of intragroup conflict. Administrative Science Quarterly 40 (2), 256–282. Jehn, K., Northcraft, G., Neale, M., 1999. Why differences make a difference: A field study of diversity, conflict, and performance in workgroups. Administrative Science Quarterly 44 (4), 741–763. Jetten, J., Duck, J., Terry, D.J., O’Brien, A., 2002. Being attuned to intergroup differences in mergers: The role of aligned leaders for low-status groups. Personality and Social Psychology Bulletin 28 (9), 1194–1201. Johnson, D.E., 1998. Applied Multivariate Methods for Data Analysis. Brooks/Cole Publishing Company, Pacific Grove, CA. Judd, C.M., Kenny, D.A., 1981. Process analysis: Estimating mediation in treatment evaluations. Evaluation Review 5, 602–619. Jun, M., Cai, S., Shin, H., 2006. TQM practice in Maquiladora: Antecedents of employee satisfaction and loyalty. Journal of Operations Management 24, 791–812. Jusko, J., 2004. Lonely at the top. Industry Week 253 (10), 58–59. Kathuria, R., Partovi, F.Y., 1999. Work force management practices for manufacturing flexibility. Journal of Operations Management 18, 21–39. Katzenbach, J.R., Smith, D.K., 1993. The Wisdom of Teams: Creating the High Performance Organization. Harvard Business School Press, Boston. Keating, E.K., Oliva, R., Repenning, N.P., Rockart, S., Sterman, J.D., 1999. Overcoming the improvement paradox. European Management Journal 17 (2), 120–134. Kenny, D.A., 2006. Mediation. White paper. Available URL: /http:// davidakenny.net/cm/mediate.htmS, November 7, 2006. Kenny, D.A., Judd, C.M., 1986. Consequences of violating the independence assumptions in analysis of variance. Psychological Bulletin 99 (3), 422–431. Kenny, D.A., Kashy, D.A., Bolger, N., 1998. Data analysis in social psychology. In: Gilbert, D., Fiske, S.T., Lindzey, G. (Eds.), The Handbook of Social Psychology, vol. 1. McGraw-Hill, Boston, pp. 233–265. Kirkman, B.L., Rosen, B., 1999. Beyond self-management: Antecedents and consequences of team empowerment. Academy of Management Journal 42 (1), 58–74. Klein, J.K., 1989. The human costs of manufacturing reform. Harvard Business Review 67 (2), 60–66. Klein, K.J., Kozlowski, S.W.J., 2000. From micro to meso: Critical steps in conceptualizing and conducting multilevel research. Organizational Research Methods 3 (3), 211–236. Kline, P., 1994. An Easy Guide to Factor Analysis. Routledge, London. Koch, J.L., 1979. Effects of goal specificity and performance feedback to work groups on peer leadership, performance, and attitudes. Human Relations 33, 819–840. Kolodny, H.F., Kiggundu, M.N., 1980. Towards the development of a systems model in woodlands mechanical harvesting. Human Relations 33, 623–645. Korhonen, K., Pirttila, T., 2003. Cross-functional decision-making in improving inventory management decision procedures. International Journal of Production Economics 81–82, 195–203. Kotter, J., 1995. Leading change: Why transformation efforts fail. Harvard Business Review 73 (2), 59–67. Koufteros, X., Marcoulides, G.A., 2006. Product development practices and performance: A structural equation modeling-based multigroup analysis. International Journal of Production Economics 103, 286–307.

ARTICLE IN PRESS 64

J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Kuhl, J., 1992. A theory of self-regulation: Action versus state orientation, self-discrimination, and some applications. Applied Psychology 41 (2), 97–129. Kumar, S., Harms, R., 2004. Improving business processes for increased operational efficiency: A case study. Journal of Manufacturing Technology Management 15 (7), 662–674. Laraia, A., 1998. Change it NOW. Manufacturing Engineering 121 (4), 152. Laraia, A.C., Moody, P.E., Hall, R.W., 1999. The Kaizen Blitz: Accelerating Breakthroughs in Productivity and Performance. The Association for Manufacturing Excellence, New York. Larson, M., 1998. Lantech’s Kaizen diary: Monday through Friday. Quality 37 (6), 40. Launonen, M., Kess, P., 2002. Team roles in business process reengineering. International Journal of Production Economics 77, 205–218. Lawal, B., 2003. Categorical Data Analysis with SAS and SPSS Applications. Lawrence Erlbaum Associates Inc., London. Lawler, E.E., Mohrman, S.A., 1985. Quality circles after the fad. Harvard Business Review 85 (1), 64–71. Lawler, E.E., Mohrman, S.A., 1987. Quality circles: After the honeymoon. Organizational Dynamics 19, 15–26. LeBlanc, G., 1999. Kaizen at Hill-Rom. Center for Quality of Management Journal 8 (2), 49–53. Lemieux-Charles, L., Murray, M., Baker, G.R., Barnsley, J., Tasa, K., Ibrahim, S.A., 2002. The effects of quality improvement practices on team effectiveness: A mediational model. Journal of Organizational Behavior 23, 533–553. Liang, K.Y., Zeger, S.L., 1986. Longitudinal data analysis using generalized linear models. Biometrika 73, 13–22. Linderman, M., Schroeder, R.G., Choo, A.S., 2006. Six Sigma: The role of goals in improvement teams. Journal of Operations Management 24, 779–790. Locke, E.A., Latham, G.P., 1984. Goal Setting: A Motivational Technique That Works. Prentice-Hall, Englewood Cliffs, NJ. Locke, E.A., Latham, G.P., 1990. A Theory of Goal Setting and Task Performance. Prentice-Hall, Englewood Cliffs, NJ. Locke, E.A., Latham, G.P., 2002. Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist 57 (9), 705–717. Locke, E.A., Shaw, K.N., Saari, L.M., Latham, G.P., 1981. Goal setting and task performance: 1969–1980. Psychological Bulletin 90 (1), 125–152. Lovelace, D., Shapiro, D.L., Weingart, L.R., 2001. Maximizing crossfunctional new product teams’ innovativeness and constraint adherence: A conflict communications perspective. Academy of Management Journal 44 (4), 779–793. MacKinnon, D.P., West, G., Dwyer, J.H., 1995. A simulation study of mediated effect measures. Multivariate Behavioral Research 30 (1), 41–62. Martin, K., 2004. Kaizen events: Achieving dramatic improvement through focused attention and team-based solutions. Society for Health Systems Newsletter August 2004, 6–7. Martin, B.A., Snell, A.F., Callahan, C.M., 1999. An examination of individual differences in the relation of subjective goal difficulty to performance in a goal-setting model. Human Performance 12 (2), 115–135. Matusi, Y., 2007. An empirical analysis of just-in-time production in Japanese manufacturing companies. International Journal of Production Economics 108, 153–164. McGrath, J.E., 1964. Social Psychology: A Brief Introduction. Hold Rinehart and Winston Inc., New York. McGrath, J.E., 1984. Groups: Interaction and Performance. Prentice-Hall, Englewood Cliffs, NJ. McIntosh, R.I., Culley, S.J., Mileham, A.R., Owen, G.W., 2001. Changeover improvement: A maintenance perspective. International Journal of Production Economics 73, 153–163. McNichols, T., Hassinger, R., Bapst, G.W., 1999. Quick and continuous improvement through Kaizen blitz. Hospital Materiel Management Quarterly 20 (4), 1–7. Melnyk, S.A., Calantone, R.J., Montabon, F.L., Smith, R.T., 1998. Short-term action in pursuit of long-term improvements: Introducing Kaizen events. Production and Inventory Management Journal 39 (4), 69–76. Miltenberg, J., 1995. Manufacturing Strategy. Productivity Press, Portland, OR. Miller, B.D., 2004. Achievement behaviors within kaizen blitz from an expectancy-value perspective: An empirical study. Unpublished Doctoral Dissertation, Department of Industrial Engineering, University of Houston, Houston, TX.

Minton, E., 1998. Profile: Luke Faulstick—‘Baron of blitz’ has boundless vision of continuous improvement. Industrial Management 40 (1), 14–21. Molleman, E., 2005. The multilevel nature of team-based work research. Team Performance Management 11 (3–4), 113–124. Montabon, F., 2005. Using Kaizen events for back office processes: The recruitment of frontline supervisor co-ops. Total Quality Management and Business Excellence 16 (10), 1139–1147. Muchinsky, P.M., 2000. Psychology Applied to Work, sixth ed. Wadsworth/Thomson Learning, Belmont, CA. Neter, J., Kutner, M.H., Nachtsheim, C.J., Wasserman, W., 1996. Applied Linear Statistical Models, fourth ed. McGraw-Hill Companies Inc., New York. Nicolini, D., 2002. In search of ‘project chemistry’. Construction Management and Economics 20 (2), 167–177. Niepce, W., Molleman, E., 1996. A case study: Characteristics of work organization in lean production and sociotechnical systems. International Journal of Operations and Production Management 16 (2), 77–90. Nunnally, J.D., 1978. Psychometric Theory, second ed. McGraw-Hill, New York. Olorunniwo, F., Udo, G., 2002. The impact of management and employees on cellular manufacturing implementation. International Journal of Production Economics 76, 27–38. O’Leary-Kelly, A.M., Martocchio, W., Frink, D.D., 1994. A review of the influence of group goals on group performance. Academy of Management Journal 37 (5), 1258–1301. Pagell, M., LePine, J.A., 2002. Multiple case studies of team effectiveness in manufacturing organizations. Journal of Operations Management 20, 619–639. Panizzolo, R., 1998. Applying the lessons learned from 27 lean manufacturers. The relevance of relationships management. International Journal of Production Economics 55, 223–240. Parker, M., Slaughter, J., 1988. Choosing Sides: Unions and the Team Concept. South End Press, Boston, MA. Patil, H., 2003. A standard framework for sustaining Kaizen events. Unpublished Master’s Thesis, Department of Industrial and Manufacturing Engineering, Wichita, KS. Perrow, C., 1967. A framework for the comparative analysis of organizations. American Sociological Review 32, 194–208. Perry, L., 1995. Effective facilitators—A key element in successful continuous improvement processes. Training for Quality 3 (4), 9–14. Pickering, J., Kisangani, E.F., 2005. Democracy and diversionary military intervention: Reassessing regime type and the diversionary hypothesis. International Studies Quarterly 49, 23–43. Pinto, J.K., Slevin, D.P., 1987. Critical factors in successful project implementation. IEEE Transactions on Engineering Management 34 (1), 22–27. Pinto, M.B., Pinto, J.K., Prescott, J.E., 1993. Antecedents and consequences of project team cross-functional cooperation. Management Science 39 (10), 1281–1297. Polychronakis, Y.E., Syntetos, A.A., 2007. ‘Soft’ supplier management related issues: An empirical investigation. International Journal of Production Economics 106, 431–449. Pritchard, R.D., Jones, S., Roth, P., Stuebing, I.C., Ekeherg, S., 1988. Effects of group feedback, goal setting, and incentives on organizational productivity. Journal of Applied Psychology 73 (2), 337–358. Raudenbush, A.S., Byrk, S.W., 2002. Hierarchical Linear Models: Applications and Data Analysis Methods, second ed. Sage, Newbury Park, CA. Rusiniak, S., 1996. Maximizing your IE value. IIE Solutions 28 (6), 12–16. Safayeni, F., Purdy, L., 1991. A behavioral case study of just-in-time implementation. Journal of Operations Management 10 (2), 213–228. Samson, D., Whybark, D.C., 1998. Tackling the ever so important ‘soft’ issues in operations management. Journal of Operations Management 17, 3–7. SAS Institute Inc., 2006. Generalized Linear Models Theory. SAS/STAT User’s Guide, SAS OnlineDocs 9.1.3. Available URL: /http://support. sas.com/onlinedoc/913/docMainpage.jspS, November 7, 2006. Sawhney, R., Chason, S., 2005. Human behavior based exploratory model for successful implementation of lean enterprise in industry. Performance Improvement Quarterly 18 (2), 76–96. Schonberger, R.J., 2007. Japanese production management: An evolution—With mixed success. Journal of Operations Management 25, 403–419. Scheaffer, R.L., Mendenhall, W.III., Ott, R.L., 1996. Elementary Survey Sampling, fifth ed. Wadsworth Publishing Company, Belmont, CA. Schneider, B., White, S., Paul, M.C., 1998. Linking service climate and customer perceptions of service quality: Test of a causal model,. Journal of Applied Psychology 83 (2), 150–163.

ARTICLE IN PRESS J.A. Farris et al. / Int. J. Production Economics 117 (2009) 42–65

Sewell, G., Wilkinson, B., 1992. Someone to watch over me: Surveillance, discipline and the just in time labour process. Sociology 26 (2), 271–289. Salas, E., Sims, D.E., Burke, C.S., 2005. Is there a ‘‘big five’’ in teamwork? Small Group Research 36 (5), 555–599. Shafer, S.M., Tepper, B.J., Meredith, J.R., Marshall, R., 1995. Comparing the effects of cellular and functional manufacturing on employees perceptions and attitudes. Journal of Operations Management 12, 63–74. Shah, R., Ward, P.T., 2003. Lean manufacturing: Context, practice bundles, and performance. Journal of Operations Management 21 (2), 129–149. Shah, R., Ward, P.T., 2007. Defining and developing measures of lean production. Journal of Operations Management 25, 785–805. Shannon, C., 1948. A mathematical theory of communications. Bell Systems Technical Journal 27, 397–423, 623–656. Sheridan, J.H., 1997. Kaizen blitz. Industry Week 246 (16), 18–27. Simons, D., Taylor, D., 2007. Lean thinking in the UK red meat industry: A systems and contingency approach. International Journal of Production Economics 106, 70–81. Smith, B., 2003. Lean and Six Sigma—A one–two punch. Quality Progress 36 (4), 37–41. Stewart, G.L., Barrick, M.R., 2000. Team structure and performance: Assessing the mediating role of intrateam process and the moderating role of task type. Academy of Management Journal 43 (2), 135–148. Sundstrom, E., DeMeuse, K.P., Futrell, D., 1990. Work teams: Applications and effectiveness. American Psychologist 45 (2), 120–133. Susman, G.I., Dean, J.W., 1992. Development of a model for predicting design for manufacturability effectiveness. In: Susman, G.I. (Ed.), Integrating Design and Manufacturing for Competitive Advantage. Oxford University Press, New York, pp. 207–227. Szinovacz, M.E., Davey, A., 2001. Retirement effects on parent–adult child contacts. The Gerontologist 41 (2), 191–200. Tanner, C., Roncarti, R., 1994. Kaizen leads to breakthroughs in responsiveness—And the Shingo Prize—At Critikon. National Productivity Review 13 (4), 517–531. Taylor, D.L., Ramsey, R.K., 1993. Empowering employees to ‘just do it’. Training and Development 47 (5), 71–76. Teachman, J.D., 1980. Analysis of population diversity. Sociological Methods and Research 8 (3), 341–362. Tranfield, D., Smith, S., Foster, M., Wilson, S., Parry, I., 2000. Strategies for managing the teamworking agenda: Developing a methodology for team-based organization. International Journal of Production Economics 65, 33–42.

65

Treece, J.B., 1993. Improving the soul of an old machine. Business Week 3342, 134. Van Aken, E.M., Kleiner, B.M., 1997. Determinants of effectiveness for cross-functional organizational design teams. Quality Management Journal 4 (2), 51–79. Van Landeghem, R., 2000. Experiences with a concurrent engineering self-assessment tool. International Journal of Production Economics 64 (1–3), 295–309. Van Mierlo, H., Vermunt, J.K., Rutte, C.G., 2006. Composing group-level constructs from individual-level survey data. Working Paper. Available URL: /http://www.spitswww.uvt.nl/vermunt/mierlo2006.pdfhttp:// spitswww.uvt.nl/vermunt/mierlo2006.pdfS, October 4, 2006. Vasilash, G.S., 1993. Walking the talk of Kaizen at Freudenberg-NOK. Production 105 (12), 66–71. Vinokur-Kaplan, G.S., 1995. Treatment teams that work (and those that don’t): An application of Hackman’s group effectiveness model to interdisciplinary teams in psychiatric hospitals. The Journal of Applied Behavioral Science 31 (3), 303–327. Vits, J., Gelders, L., 2002. Performance improvement theory. International Journal of Production Economics 77, 285–298. Wall, T.D., Kemp, N.J., Jackson, P.R., Clegg, C.W., 1986. Outcomes of autonomous workgroups: A long-term field experiment. Academy of Management Review 29 (2), 280–304. Warnecke, H.J., Huser, M., 1995. Lean production. International Journal of Production Economics 41, 37–43. Watson, L., 2002. Striving for continuous improvement with fewer resources? Try kaizen. Marshall Star, 28 November 2002, vol. 1, 4 pp. Wheatley, B., 1998. Innovation in ISO registration. CMA 72 (5), 23. Whitford, A.B., Yates, J., 2003. Policy signals and executive governance: Presidential rhetoric in the war on drugs. The Journal of Politics 65 (4), 995–1012. Wilson, L., Van Aken, E.M., Frazier, D., 1998. Achieving high performance work systems through policy deployment: a case application. In: Proceedings of the 1998 International Conference on Work Teams. Withey, M., Daft, R.L., Cooper, W.H., 1983. Measures of Perrow’s work unit technology: An empirical assessment and a new scale. Academy of Management Journal 26 (1), 45–63. Wittenberg, G., 1994. Kaizen—The many ways of getting better. Assembly Automation 14 (4), 12+ (6pp.). Womack, J.P., Jones, D.T., Roos, D., 1990. The Machine that Changed the World. Harper Perennial, New York. Wright, R., 1936. Factors affecting the cost of airplanes. Journal of Aeronautical Sciences 3, 122–128. Yin, R.K., 1994. Case Study Research: Design and Methods. Sage Publications, Inc., Thousand Oaks, CA.