Beruflich Dokumente
Kultur Dokumente
OPERATIONS
MANAGEMENT
ELSEVIER Journal of Operations Management 16 (1998) 21-41
Abstract
Case and industry studies have increased our understanding of time-base manufacturing and focused our attention on its
key component practices. Despite important contributions by Blackburn (1991) and Stalk and Hout (1990), we lack a clear
definition of time-based manufacturing and its relationship to Just-in-time (JIT). This study proposes a framework for
research on time-based manufacturing, reports on the development of a set of seven instruments for measuring the key
practices, and tests relationships among these practices. The instruments are valid, reliable, and generalizable across
industries and firm size. Tests of the structural model confirm Monden's (1983) notion that shop-floor employee
involvement leads to improved manufacturing practices which, in turn, lead to pull production. © 1998 Elsevier Science
B.V.
Keywords: Empirical research; Operations strategy; Measurement and methodology; Time-based manufacturing (just-in-time)
implement a set of practices designed to reduce ated with JIT is more of a side benefit than a driving
throughput time (Blackburn, 1991). A literature re- force" (Blackburn, 1991: 22). Time-based competi-
view identified seven key component practices i.e., tion and its components, including time-based manu-
shop-floor employee involvement in problem solv- facturing, focus on customers and offer a compre-
ing, reengineering set-ups, cellular manufacturing, hensive framework to build excellence in manufac-
quality improvement efforts, preventive mainte- turing.
nance, dependable suppliers, and pull production At present, little is known about the relationships
(Blackburn, 1991; Davy et al., 1992; Handfield and among the time-based manufacturing practices. Also,
Pannesi, 1995; Monden, 1983; Sakakibara et al., little is known about whether, or under what condi-
1993; Saraph et al., 1989; Stalk and Hout, 1990). tions, particular practices enhance a firm's competi-
Industry studies illustrate how these time-based prac- tive capabilities (Swamidass, 1991). To investigate
tices cut response time and enhanced competitive- these issues, instruments for measuring the extent to
ness (Blackburn, 1991; Hamilton, 1991; Lindsley et which a finn has implemented time-based manufac-
al., 1991; Merrills, 1989; Roth and Miller, 1990, turing must be developed. Empirical studies by Davy
1992). et al. (1992) and Sakakibara et al. (1993) have
Many of these time-based practices are key ele- helped to identify and measure underlying constructs
ments of Just-in-time (JIT) as defined by Monden of JIT.
(1983). Time-based manufacturing and JIT address This paper builds on germinal works by Monden
the same phenomena but with different emphases. In (1981, 1983), these empirical studies of JIT, and the
fact, Abegglen and Stalk (1985) observed that some seminal works by Blackbum (1991) and Stalk and
JIT innovators became the first time-based competi- Hout (1990). Its primary contributions are to present
tors as their emphasis on speed propelled their skills a framework for research on time-based manufactur-
in time reduction throughout the value-delivery sys- ing, offer a theoretical rationale for a set of seven
tem. Blackburn (1991) claims that JIT is the genesis practices, and report on efforts to develop instru-
of time compression. Still, JIT and time-based manu- ments to measure these practices. The study also
facturing have differences. Time-based manufactur- examines relationships among the practices which
ing is an externally focused production system that can guide implementation.
emphasizes quick response to changing customer
needs. Its primary purpose is to reduce end-to-end
time in manufacturing (Blackbum, 1991; Stalk and 2. A framework for research on time-based
Hout, 1990). Monden (1983) describes JIT as an manufacturing
internally focused production system that produces
parts on demand. It eliminates unnecessary elements Monden (1983), through his Toyota Production
in production, and its primary purpose is cost reduc- System, laid the foundation for time-based manufac-
tion. turing (Blackburn, 1991). Monden (1983) (p. 3)
According to Blackburn (1991), the transfer of claims that improvement activities by small groups
JIT concepts beyond the factory walls is often are the basis for set-up time reduction, quality im-
blocked by narrow and conventional notions of JIT provement efforts, reliable equipment, and machine
applications and results. Some firms see JIT's virtues layout. Sakakibara et al. (1993) describe 12 factors
as batch size and inventory reductions (Brown and that support the JIT construct including several cited
Mitchell, 1991; Hall, 1987; Monden, 1983) and ig- here as important time-based manufacturing prac-
nore other more important benefits such as time tices (i.e., small group problem solving, set-up time
compression and product mix flexibility (Blackbum, reduction, equipment layout, preventive mainte-
1991). The latter benefits delight customers and lead nance, JIT deliveries from suppliers, and kanban/pull
to sustainable competitive advantage. "Cycle time system support). In addition to these factors, Davy et
compression translates into faster asset turnover, in- al. (1992) identify quality improvements efforts as an
creased output and flexibility, and satisfied cus- essential component of JIT.
tomers. Thus, the diminished inventory often associ- Stalk and Hout (1990) claim that competing on
X.A. Koufteros et al. / Journal of Operations Management 16 (1998) 21-41 23
time requires a value delivery system that is two to Sakakibara et al. (1993) provide JIT measures for
three times faster and more flexible than competitors. some of the constructs cited here as components of
The path to fast and flexible factories involves set-up time-based manufacturing. Their measures are based
time reduction, product-oriented layouts, and quality on multiple responses from the same production
improvements. Blackburn (1991) identifies employee facility which includes responses from more than
involvement, set-up time reduction, product-oriented one level in the organization, e.g., plant manage-
layouts, quality improvement efforts, and supplier ment, engineering, supervision, and the workforce
relationships as essential elements of creating manu- (Sakakibara et al., 1993: 185). The reliabilities they
facturing systems that focus on time and achieve pull report are below the 0.80 or 0.90 standards suggested
production. by Nunnally (1978) for instruments used in basic or
From these efforts, seven practices that focus on applied research, respectively. Although the reliabili-
time compression were identified (see Fig. 1). The ties of these JIT instruments are problematic, they
practices measure the extent to which: provided an important stepping stone for the creation
1. Shop-floor employee involvement in problem of the time-based manufacturing instruments de-
solving: first-level employees participate in activi- scribed here.
ties to define and solve problems. Literature on JIT discusses other factors such as
2. Reengineering set-up: efforts are taken to reduce daily schedule adherence, product scheduling, and
set-up time. batch size (Davy et al., 1992; Sakakibara et al.,
3. Cellular manufacturing: units are produced in a 1993) that are not included here. These variables
product oriented layout. represent outcomes rather than causes of time-based
4. Quality improvement efforts: methods are devel- manufacturing or are captured by other factors mea-
oped and used to reduce defects and enhance sured in this study. Daily schedule adherence and
quality. product scheduling are subsumed in pull production.
5. Preventive maintenance: equipment is routinely Batch sizes is an outcome that can be reduced when
maintained on a proactive basis. set-ups are reengineered to cut time or cellular man-
6. Dependable suppliers: suppliers facilitate cus- ufacturing is implemented. The opposite is not true,
tomer needs for service quality. reducing batch size does not allow a firm to cut
7. Pull production: production is driven by demand set-up time. Reducing batch size without reducing
from the next work station and ultimately from set-up time can lead to excessive set-up costs and
the customer. delays for customers.
Reengineeringcellular
Setups
Manufacturing
Shop-Floor
Employee Quality Pull
Involvement in Improvement Efforts [ r[ Production
Problem Solving
Preventive
Maintenance
Dependable
Suppliers
Fig. 1. A framework for organizational-level research on time-based manufacturing.
24 X.A. Koufteroset al./ Journal of OperationsManagement16 (1998) 21-41
Time-based manufacturing, as illustrated in Fig. H4: Organizations with high levels of shop-floor
1, involves practices that encourage shop-floor em- employee involvement in problem solving will
ployee involvement in problem solving (Hall, 1987, achieve high levels of preventive maintenance.
1993; Showalter and Mulholland, 1992). Shop-floor H5: Organizations with high levels of shop-floor
employee involvement can be specified as an an- employee involvement in problem solving will have
tecedent to other time-based manufacturing practices. dependable suppliers.
This is corroborated by the employee involvement The following sections provide an explicit theoret-
and participation literature which suggests that em- ical rationale for the inclusion of each factor in a
ployee involvement drives improvement efforts time-based manufacturing research framework. Liter-
(Johnson and Rice, 1987; Locke and Schweiger, ature and logical arguments are provided that link
1979; Markus, 1983; Robey and Farrow, 1982). these practices to throughput time, competitive capa-
Shop-floor employees have a unique understanding bilities, and pull production.
of the problems present in their environment and
have insights that enable them to develop effective 2.1. Shop-floor employee involvement in problem
solutions (Badore, 1992). Hall et al. (1991) describe solving
case studies where shop-floor employees initiated
effective solutions to quality, set-up time, and pre- Employee involvement in problem solving is an
ventive maintenance problems. antecedent to other time-based changes (Johnson and
Enhanced shop-floor employee problem solving Rice, 1987; Showalter and Mulholland, 1992). In
skills facilitate the reengineering of set-ups (Monden, today's factories, work has a higher intellectual con-
1981, 1983; Shingo, 1985), the implementation of tent than it had in the past (Zuboff, 1984) and
cellular manufacturing (Pullen, 1976; Wemmerlov shop-floor employees are recognized as the actors
and Hyer, 1989), the establishment of quality im- who plan and do work. Integrating work planning
provement efforts (Deming, 1981, 1982, 1986; and work doing in employee teams is essential to
Garvin, 1983; Juran, 1978, 1981a,b; Leonard and learning through changes in set-up, layout, quality
Sasser, 1982), the initiation of effective preventative practices, preventive maintenance, and supplier rela-
maintenance programs (Bockerstette and Shell, 1993; tions. These changes allow organizations to eliminate
Schonberger, 1986), and the creation of dependable waste in the system, cut throughput time, improve
suppliers (Ansari and Modaress, 1988, 1990; Hand- competitive capabilities, and achieve pull production
field and Pannesi, 1992; Im and Lee, 1989; Lee and (Hall, 1987, 1993).
Ebrahimpour, 1984; Schmenner, 1991). With these
practices in place, time-based manufacturers seek to 2.2. Reengineering set-up
achieve pull production to cut response time to cus-
tomers and enhance competitive capabilities Shingo (1985) stresses that reducing set-up times
(Blackburn, 1991; Monden, 1981, 1983; Ohno, 1978; is essential to cutting throughput time because it
Schonberger, 1986; Shingo, 1985). This discussion increases flexibility, enabling firms to quickly switch
supports a set of hypotheses between shop-floor between products with minimum penalties. Com-
employee involvement in problem solving and five pressing set-up times permits small batch production,
of the practices. enables greater flexibility to respond to changing
HI: Organizations with high levels of shop-floor customer needs, and reduces the need for inventories
employee involvement in problem solving will have (Ohno, 1978). Reflecting on the Toyota experience,
high levels of reengineered set-ups. Monden (1981, 1983) argues that set-up time is an
H2: Organizations with high levels of shop-floor inherent component of throughput time and a deter-
employee involvement in problem solving will attain minant of shop-floor responsiveness. Set-up time
high levels of cellular manufacturing. reduction is an important element of pull production.
H3: Organizations with high levels of shop-floor H6: Organizations with high levels of reengi-
employee involvement in problem solving will have neered set-ups can achieve high levels of pull pro-
high levels of quality improvement effort. duction.
X.A. Koufteros et al. / Journal of Operations Management 16 (1998) 21-41 25
2.3. Cellular manufacturing time and induces firms to compensate for poor
equipment reliability by adding inventory. Inventory
Identifying families of products enables machines extends throughput time by clogging the factory
with different processing capabilities to be grouped floor. Firms with successful preventive maintenance
together to form a manufacturing cell (Pullen, 1976; programs involve employees in designing and per-
Wemmerlov and Hyer, 1989). Within each cell, forming maintenance activities (Bockerstette and
management attempts to keep set-up time low by Shell, 1993; Schonberger, 1986). These firms reduce
producing parts with similar size, shape, and process- unplanned downtime and achieve pull production.
ing requirements. Cellular layouts are self-contained; H9: Organizations with high levels of preventive
that is, they have all the equipment necessary to maintenance can achieve high levels of pull produc-
make a part. This reduces materials handling costs tion.
and time, slashes work-in-process inventory, cuts
throughput time, enhances quality, and increases 2.6. Dependable suppliers
flexibility (Hyer and Wemmerlov, 1984). As a result,
cellular manufacturing enables a firm to achieve pull Dependable suppliers cut throughput time, reduce
production. costs, and improve competitive capabilities (Ansari
H7: Organizations with high levels of cellular and Modaress, 1988; Giunipero, 1990; Im and Lee,
manufacturing can achieve high levels of pull pro- 1989; Lee and Ebrahimpour, 1984). On-time deliv-
duction. ery allows an organization to keep inventory low and
shorten response time to its customers. Improved
2.4. Quality improvement efforts supplier performance can help to reduce downtime
and the shortages associated with delivery delays
Deming (1982, 1986) and Juran (1988) contend (Ansari and Modaress, 1990; Blackburn, 1991 ; Clark,
that quality improvement efforts reduce throughput 1989; Im and Lee, 1989). If a supplier's parts are
time and enhance competitive capabilities. Adam defective, time-consuming rework is required or pro-
(1994) study of 187 U.S. businesses shows a strong duction is delayed until replacement parts arrive
relationship between quality improvement efforts and (Handfield and Pannesi, 1992).
operating performance. When products are rejected H10: Organizations with dependable suppliers can
for poor quality, there may be some delay while the achieve high levels of pull production.
process is adjusted and new materials are procured.
If defective units are scrapped or reworked, then 2.7. Pull production
replacement or rework time should be included in
calculating the average throughput time (Schmenner, Production is pulled by customer demand when
1992). When facilities are operating at capacity, any the firm produces to meet precisely timed customer
delay in shipping products to one customer affects requirements (Blackburn, 1991). By allowing only a
all subsequent orders. Time spent recovering from small amount of work-in-process inventory to flow
quality problems is an inherent and avoidable com- at any time, pull systems shorten the time parts stay
ponent of throughput time. To achieve pull produc- in the system by eliminating or greatly reducing
tion, firms should have successful quality improve- waiting time. Monden (1981, 1983), Ohno (1978),
ment efforts. and Shingo (1985) describe the pull system as ap-
H8: Organizations with high levels of quality plied at Toyota, and Schonberger (1986) reports on
improvement effort can achieve high levels of pull applications in the U.S.
production.
Equipment that has not been properly maintained Several objectives were defined for these time-
causes unplanned downtime that increases waiting based manufacturing instruments. An effective in-
26 X.A. Koufteros et a l . / Journal of Operations Management 16 (1998) 21-41
Reliability of the remaining items comprising each 2500 organizations with more than 100 employees
scale was evaluated using Cronbach's alpha (Cron- that are engaged in discrete part manufacturing. SME
bach, 1951). Items were eliminated if Cronbach's mailed a survey and cover letter to the senior manu-
alpha was at least 0.80 and the content of the scale facturing executive at these firms. A notification card
was not significantly altered. Four items were re- was sent two weeks prior to mailing the question-
moved from their scales without significantly chang- naire. Out of 253 responses received, 244 were
ing alpha. usable resulting in a response rate of 9.9%. Response
rates of 10% or less are typical for organizational-
3.4. Modifications after pilot study level studies involving senior management.
The usable responses are from fabricated metal
Before the large-sample administration, the scales products (35%), industrial and commercial machin-
were examined with respect to the research objec- ery (30%), electrical equipment and components
tives and the overall pilot study results. Attention (12%), transportation equipment (15%), and miscel-
lbcused on the constructs with reliabilities near the laneous (8%). The respondent's position are: presi-
minimum standard of 0.80, i.e., cellular manufactur- dents (12%), vice-president (31%), directors (11%),
ing, reengineering set-up, and preventive mainte- managers (19%), and miscellaneous (27%). Seventy
nance with alphas of 0.81, 0.80, and 0.84 respec- percent of the responses are from firms with less
tively. In the cellular manufacturing scale, two dis- than 500 employees, and firms with more than 1000
tinct types of items were identified--items that mea- employees account for only 15% of the sample.
sure whether manufacturing cells exist and items that These sample statistics are not significantly different
measure whether the resources in the cells are flexi- from the corresponding population parameters which
ble. Flexibility is an outcome of cellular manufactur- implies no difference in characteristics between re-
ing, not a practice; thus, questions related to flexibil- spondents and non-respondents.
ity were dropped. Also, firms without cells may not
understand what was intended by the term 'manufac- 3.6. Large-scale research methods: measurement
turing cell'; so, the remaining items were dropped. model
Eight new items were developed that measure the
To develop good measures of these seven time-
extent to which the firm engages in cellular manufac-
based practices, it was necessary to purify items,
turing practices. These replacement items do not use
perform exploratory factor analysis, check reliability,
the words 'manufacturing cell'.
test for convergent and discriminant validity, and
To improve reliability of the reengineering set-up
assess predictive validity.
scale, three items were reworded to improve clarity.
One item compounded two activities (i.e., reconfig- 3.6.1. Purification
uring equipment and redesigning tools). This item Items were purified, as recommended by Churchill
was replaced by two simpler questions (RS16 and (1979), because substantial questionnaire changes
RS 17). Two new items were added that asked about were made in the pilot study. Items were eliminated
the use of special tools (RSI8) and training (RS19). before further large sample analysis if their corrected
To improve the reliability of the preventive mainte- item-total correlation was below 0.60.
nance scale, two items were added (PM7, PM8).
After the pilot study revisions, the instrument con- 3.6.2. Factor analysis
sisted of 45 items (see Table 1) measuring seven To examine construct validity, an exploratory fac-
manufacturing practices. tor analysis was conducted using principal compo-
nents and oblimin as the means of extraction and
3.5. Survey methods and sample characteristics method of rotation, respectively. Only the items
referring to six of the practices were included be-
The Society of Manufacturing Engineers (SME) cause shop-floor employee involvement is consid-
co-sponsored a large-scale survey to test and refine ered to be an antecedent (Monden, 1983) to the other
the instrument. From its membership, SME selected constructs and was factor analyzed separately. Items
28 X.A. Koufteros et al. / Journal of Operations Management 16 (1998) 21-41
with factor loadings below 0.60 or with cross-load- ally, 1978). This rule suggests that only factors that
ings above 0.30 were eliminated. explain more variance than the average amount ex-
The number of factors extracted in this research is plained by one of the original items should be re-
based on Kaiser's eigenvalues greater than l (Nunn- tained. If the factor explains more variance than an
Table 1
Corrected-item total correlations for time-based manufacturing in large-scale survey (n = 244)
Items Corrected-item
total correlation
Shop-floor employee involvement in problem solving
PSI. Shop-floor employees are involved in problem solving efforts 0.71
PS3. Shop-floor employees are involved in suggestion programs 0.63
PS4. Shop-floor employees are involved in designing processes 0.67
and tools that focus on improvement
PS5. Shop-floor employees are involved in improvement efforts 0.81
PS6. Shop-floor employees are involved in problem solving teams 0.76
Reengineering set-ups
RS 11. We have been working towards improving set-up times 0.59
RS13. Standard set-ups are developed for new processes 0.40
RS14. Employees work on set-up improvement 0.73
RS 15. Tools for set-up are conveniently located 0.51
RS16. Employees redesign or reconfigure equipment to shorten set-up time 0.61
RS 17. Employees redesign jigs or fixtures to shorten set-up time 0.71
RS 18. We use special tools to shorten set-up 0.65
RS 19. Our employees are trained to reduce set-up time 0.73
Cellular manufacturing
CM1. Products with design or processing similarities are produced together 0.47
CM2. Products that share similar design or processing requirements 0.69
are grouped into families of products
CM3. Products are classified into groups with similar processing requirements 0.66
CM4. Products are classified into groups with similar routing requirements 0.67
CM5. A coding classification is used to group parts into families 0.49
CM6. Our factory layout groups different machines together to produce families of products 0.45
CM7. Equipment is grouped to produce families of products 0.69
CM8. Families of products determine our factory layout 0.62
Preventive maintenance
PM1. There is a separate shift, or part of a shift, reserved for 0.58
preventive maintenance activities
PM3. We emphasize good preventive maintenance 0.78
PM5. Records of routine maintenance are kept 0.63
PM6. We do preventive maintenance 0.84
PM7. We do preventive maintenance during non-productive time 0.61
PM8. We maintain our equipment regularly 0.80
X.A. Koufteros et al. / Journal of Operations Management 16 (1998) 21-41 29
Table 1 (continued)
Corrected-item total correlations for time-based manufacturing in large-scale survey (n = 244)
Items Corrected-item
Dependable suppliers
DS 1. We receive parts from suppliers on time 0.60
DS4. We receive the correct number of parts from suppliers 0.60
DS7. We receive the correct type of parts from suppliers 0.76
DS9. We receive parts from suppliers that meet our specifications 0.75
DSI0. Our suppliers accommodate our needs 0.73
DS11. We receive high quality parts from suppliers 0.73
Pull production
PPI. We do not produce unless there is a demand in the next station 0.45
PP3. Production is 'pulled' by the shipment of finished goods 0.64
PP5. Production at stations is "pulled' by the current demand of the next stations 0.69
PP6. We use a 'pull' production system 0.77
PP8. We use kanban to pull production 0.48
original item, then some degree of condensation is Discriminant validity was also tested by compar-
achieved (DeVellis, 1991). To achieve a stable factor ing the average variance extracted with the squared
structure, Tinsley and Tinsley (1987) suggest that the correlation between constructs. Fomell and Larcker
ratio of respondents to items should be at least 5 or (1981) suggest that discriminant validity exists if the
10 to 1. Comrey (1988) states that a sample size of items share more common variance with their re-
200 is normally adequate for factor analysis involv- spective construct than any variance that construct
ing fewer than 40 items. shares with other constructs. Therefore, the average
variance extracted for a construct should be substan-
3.6.3. Convergent and discriminant validity tially higher than the squared correlation between
In addition to factor analysis which tests for that construct and all the other constructs.
convergent validity, convergent validity was evalu-
ated by testing the lowest within group correlation to
3.6.4. Reliability
determine if it is significant at P < 0.01. Discrimi-
The reliability of all scales was examined using
nant validity was assessed at the item-level using a
Cronbach's alpha and the average variance extracted.
single-method, multiple-trait approach (Campbell and
Average variance extracted (Fornell and Larcker,
Fiske, 1959). The lowest correlation for a particular
1981) measures the amount of variance for the speci-
item and any other item within the factor was com-
fied indicators (items) that is accounted for by the
pared to correlations of that item and all items
latent construct. The average variance extracted is a
outside of the factor. If the former correlation was
complementary measure to the construct reliability
less than the latter, then a violation occurred (i.e., a
value. Bagozzi and Yi (1988) suggest that the aver-
violation occurred when the within factor correlation
age variance extracted for a construct should exceed
was less than the between factor correlation).
0.50.
Linear structural equation modeling (LISREL)
methodology (Bagozzi and Phillips, 1982) was used
to test discriminant validity between pairs of con- 3.6.5. Predictive validity
structs. These tests were run with the correlation To assess predictive validity, the seven time-based
between the latent variables fixed at 1.0 and with manufacturing scales were correlated with competi-
correlation between the latent variables freed to as- tive capabilities. Literature on manufacturing strat-
sume any value. Large X 2 differences between the egy (Roth and Miller, 1990; Swamidass and Newell,
fixed and freed solutions provide evidence of dis- 1987), capabilities (Miller et al., 1992), and perfor-
criminant validity. mance (Maskell, 1991) identifies six key competitive
30 X.A. Koufleros et al. / Journal of Operations Management 16 (1998) 21-41
capabilities used by time-based competitors to im- days of work-in-process inventory, and days in fin-
prove organizational performance (i.e., competitive ished goods inventory. The other competitive capa-
pricing, premium pricing, value to customer quality, bilities were assessed using 7-point Likert scales that
dependable deliveries, product innovation, and compared the firm's capabilities with the average
throughput time) (Koufteros, 1995). capability in the industry. The scales are: 1 = Much
Throughput time was measured by aggregating Below, 2 = Moderately Below, 3 = Slightly Below,
the days of raw material inventory, value added time, 4 = About Average, 5 = Slightly Above, 6 =
Table 2
Exploratory factor analysis for time-basedmanufacturingitems after large-samplepurification(n = 244)
Item Preventive Dependable Cellulars Pull Quality Reengineering
maintenance suppliers manufacturing production improvement
efforts set-ups
PM6 0.94
PM8 0.89
PM3 0.81
PM5 0.72
PM7 0.68
DS9 0.85
DS7 0.85
DS 11 0.82
DS10 0.79
DS4 0.72
DS1 0.71
CM2 0.87
CM4 0.83
CM3 0.82
CM7 0.69
CM8 0.62
PP3 0.89
PP6 0.87
PP5 0.79
QII 0.86
QI16 0.83
QI8 0.80
QI5 0.79
RS 17 -0.86
RS 14 -0.75
RS19 -0.73
RS16 - 0.72
RS18 -0.71
Eigenvalue 8.08 3.50 2.46 2.00 1.86 1.42
% of variance 28.8 12.5 8.8 7.1 6.6 5.1
Cumulative 28.8 41.3 50.1 57.2 63.9 69.0
% of variance
C M 2 C M 3 C M 4 C M 7 C M 8 RS14 R S I 6 RSI7 RSI8 RSI9 PM3 PM5 P M 6 P M 7 P M 8 PP3 PP5 PP6 QII QI5 QI8 Q I 1 6 DSI DS4 DS7 DS9 D S 1 0 D S I I PSI PS3 PS4 PS5 PS6
CM2 1.00
CM3 0.65 1.00
CM4 0.62 0.61 1+00
CM7 0.54 0.51 0+50 1.00
CM8 0.49 0.38 0.43 0.61 1.00
RSI4 0.31 0.29 0.24 0.34 0.27 1.00 .~
RSI6 0.23 0.24 0+32 0.32 0.33 0,54 1.00 .~
RS17 0.21 0.26 0.20 0.34 0.27 0.62 0.63 1.00
RS18 0.28 0.36 0.27 0.41 0+35 0.56 0.45 0.56 1.00 ~.~
RS19 0.19 0.24 0.23 0.32 0.31 0.64 0.52 0.63 0.60 1.00 e~
PM3 0.17 0.25 0.13 0.24 0.19 0.47 0.35 0.40 0.39 0.49 1.00
PM5 0.12 0.21 0.12 0.19 0.14 0.33 0.24 0.28 0.29 0.38 0.54 1.00 ~,~
PM6 0.10 0.18 0.15 0.24 0.12 0.34 0.34 0.33 0.32 0.39 0.78 0.62 1.00
PM7 0.19 0.24 0.18 0.23 0.25 0.24 0.26 0.34 0.23 0.30 0.52 0.44 0.59 1.00
PM8 0.14 0.21 0.14 0.26 0.21 0.40 0.33 0.36 0.42 0.47 0.76 0.63 0.82 0.54 1.00
PP3 0.13 0.08 0.09 0.24 0.18 0.20 0.17 0.25 0.20 0.19 0.22 0.14 0.16 0.32 0.14 1.00 ".t
PP5 0.23 0.11 0.22 0.27 0.32 0.31 0.29 0.28 0.20 0+29 0.24 0.26 0.19 0.35 0.24 0.58 1.00
PP6 0.23 0.13 0.20 0.30 0.31 0.27 0.24 0.29 0.24 0.26 0.29 0.23 0.25 0.39 0.23 0.76 0.69 1.00 ~,~
QII 0.18 0.21 0.19 0.14 0.19 0.25 0,26 0.22 0.16 0.31 0.28 0.17 0.17 0.16 0.14 0.09 0.20 0.20 1.00
QI5 0.10 0.13 0.18 0.19 0.20 0.35 0.36 0.32 0.30 0.39 0.29 0.22 0.27 0.22 0.26 0.18 0.27 0.27 0.57 1.00 "~
QI8 0,14 .014 0.08 0.23 0.21 0.38 0.32 0.32 0.36 0.42 0.32 0.26 0.32 0.22 0.29 0.16 0.19 0.18 0.55 0.58 1.00
QI16 0.20 0.15 0.20 0.22 0.18 0.39 0.32 0.32 0.35 0.40 0,29 0,29 0.30 0.22 0.26 0.10 0.26 0.22 0.55 0.63 0.70 1.00 ~.
DSI 0.06 0.06 0.01 0.07 0.07 0.13 0.06 0.14 0,14 0.19 0,17 0.24 0.13 0,12 0.17 0.10 0,14 0.14 0.05 0.08 0.03 0.08 1.00
DS4 0.06 0.14 0.00 0.04 0.01 0.06 0.05 0.03 0.10 0.06 0.16 0.27 0.14 0.07 0.14 0.12 0.09 0.08 0.04 0.03 0.01 0.02 0.48 1.00 ~.
DS7 0.06 0.21 0.08 0.13 0.05 0.17 0.09 0.11 0.22 0.20 0.14 0.29 0.20 0.15 0.20 0.09 0.10 0.12 0.04 0.12 0.11 0.05 0.49 0.58 1.00
DS9 0.05 0.20 0.08 0.06 0.07 0.14 0.12 0.15 0.16 0.23 0.14 0.23 0.16 0.17 0.19 0.11 0.11 0.14 0.03 0.09 0.03 0.03 0.47 0.49 0.70 1.00
DSI0 0.14 0.19 0.14 0.11 0.11 0.11 0.11 0.14 0.13 0.22 0.18 0.28 0.24 0.19 0.27 0.01 0.04 0.04 0.10 0.12 0.13 0.09 0.54 0.46 0.64 0.62 1.00
DSI1 0.13 0.23 0.08 0.16 0.11 0.13 0.10 0.16 0.24 0.28 0.19 0.23 0.20 0.15 0.21 0.05 0.08 0.08 0.03 0.16 0.13 0.11 0.49 0.45 0.62 0.71 0.64 1.00
PSI 0.22 0.30 0.13 0.26 0.26 0.57 0.42 0.45 0.43 0.50 0.27 0.22 0.12 0.10 0.21 0.13 0.13 0.19 0.35 0.30 0.29 0.29 0.13 0.01 0.14 0.21 0.13 0.19 1.130
PS3 0.27 0.19 0.12 0.22 0.27 0.48 0.38 0.45 0.42 0.51 0.33 0.29 0.25 0.24 0.31 0.17 0.28 0.29 0.31 0.30 0.40 0.39 0.08 0.03 0.10 0.08 0.09 0.12 0.53 1.00
PS4 0.16 0.22 0.17 0.24 0.26 0.52 0.49 0.48 0.46 0.52 0.32 0.21 0.21 0.12 0.27 0.13 0.20 0.20 0.29 0.36 0.33 0.33 0.09 0.05 0.13 0.08 0.10 0.14 0.55 0.46 1.00
PS5 0.16 0.32 0.20 0.28 0.28 0+60 0.50 0.50 0.43 0.55 0.42 0.36 0.29 0.19 0.35 0.21 0.22 0.27 0.39 0.38 0.40 0.37 0.08 0.01 0.16 0.12 0.16 0.11 0.73 0.55 0.66 1.00
PS6 0.07 0.15 0.13 0.18 0.21 0.49 0.49 0.47 0.38 0.51 0.36 0.31 0.25 0.20 0.27 0.23 0.27 0.29 0.51 0.51 0.52 0.50 0.06 0.13'0 0.11 0.12 0.11 0.13 0 . 5 9 0 . 6 1 0 . 6 0 0 . 7 1 1.00
C M 2 C M 3 C M 4 C M 7 C M 8 R S I 4 RS16 RS17 RS18 R S I 9 P M 3 P M 5 P M 6 PM7 PM8 PP3 PP5 PP6 QI1 QI5 QI8 QII6 DSI D S 4 DS7 DS9 D S I 0 D S I I PS1 PS3 PS4 PS5 PS6
Mean 3.6 3.71 3.63 3.58 3.45 3.27 2.97 3.01 3.28 3.02 3.35 3.78 3.49 3.32 3.53 3.38 3.14 3.12 2.45 2.15 3.09 2.80 3.60 3.74 3.93 3.93 3,88 3.80 3.60 3.46 2.96 3.62 3.29
S.D. 1.09 0.95 1.05 1.07 1.17 0.91 0.96 1.02 0.95 1.03 1.04 1.06 I.II 1.06 1.06 1.33 1.16 1.32 1.28 1.19 1.35 1.44 0.75 0.72 0.67 0.68 0.69 0.72 0.89 1.10 1.02 0.96 1.10
Number of 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
violations
excluding
PS items
Correlations above 0.13 are significant at 0.05, Correlations above 0.17 are significant at 0.01.
k~
Table 4
Descriptive statistics, correlations, reliabilities, average variance extracted, X 2 test for discriminant validity, and squared correlations for time-based manufacturing scales ..,..
I
X.A. Koufteros et al./ Journal of Operations Management 16 (1998) 21-41 33
Moderately Above, and 7 = Much Above (Koufteros, practices are shown in Table 1. Items with CITC
1995). below 0.60 were eliminated. Overall, 12 items were
The reliabilities (alpha) of the individual scales eliminated.
are: competitive pricing (0.94), premium pricing
(0.93), product innovation (0.86), value to customer 4.1. Factor analysis
quality (0.90), and customer delivery service (0.95).
Because throughput time was a single item scale, Factor analysis was conducted on 28 items, ex-
reliability cannot be calculated. A composite mea- cluding the items for shop-floor employee involve-
sure of these six competitive capabilities was ob- ment in problem solving. Without specifying the
tained by standardizing the scales and summing them. number, six factors with eigenvalues greater than 1
This 24-item composite measure had a reliability of emerged (see Table 2). These six factors explained
0.90. 69.0% of the variance. Only loadings greater than
0.30 are shown. All items loaded on their theoretical
3.7. Large-scale research methods: structural model construct and had loadings greater than 0.60. Ac-
cording to Comrey (1988), when item groupings are
LISREL (Joreskog and Sorbom, 1986) was used identified prior to factoring, a result that is consistent
to test the structural relationships identified in Fig. 1. with these groupings provides evidence of factorial
In LISREL, shop-floor employee involvement in validity. A factor analysis of the five items measur-
problem solving was treated as an exogenous vari- ing shop-floor employee involvement revealed a sin-
able and the other variables were endogenous. This gle factor that explained 68.0% of the variance.
is consistent with the idea of Monden (1983) that These results are also given in Table 2.
improvement activities by small groups is an an-
tecedent to changing manufacturing practices. To 4.2. Convergent and discriminant validi~
assess the overall fit of the model to the data, the
goodness of fit index (GFI) was calculated (Byrne, Tests for convergent and discriminant validity
1989). In addition, three widely used incremental fit were conducted at the item-level using single method,
indices were calculated, the comparative fit index multiple trait (Table 3). The lowest correlation within
(CFI) (Bentler, 1990), the incremental fit index (ICI), a factor was found in Table 3. These correlations for
and the normed-fit-index (NFI) (Bentler and Bonnet, each factor are high and are significant at P < 0.01
1980). These indices range from 0.0 (no fit) to 1.0 indicating good convergence. For discriminant valid-
(perfect fit). Values between 0.80 and 0.89 represent ity at the item level, there were zero violations out of
a good fit while values of 0.90 or better represent a a total of 648 potential violations. This provides
very good fit. initial evidence of discriminant validity.
If the model fits the data adequately, the magni- The LISREL methodology was used to test for
tude of the gamma coefficients (from exogenous to discriminant validity between the factors. X 2 differ-
endogenous) and beta coefficients (from endogenous ences for all 21 pairs of latent variable are greater
to endogenous) can be examined for statistical signif- than 40.53 and are shown in Table 4. For 21 compar-
icance. The test for significance compares the esti- isons, the X 2 value for any pair must be greater than
mated parameter to its standard error and has a or equal to 9.1 for significance at P < 0.05 (Cohen
t-distribution (Marsch and Hocevar, 1985). A t-sta- and Cohen, 1983: 167). In all cases, the X 2 differ-
tistic greater than 1.96 is significant at P < 0.05 and ences are several times greater than the critical value
greater than 2.33 is significant at P < 0.01. which implies that the differences are significant.
A second test for discriminant validity at the
factor level compares the average variance extracted
4. Results for the measurement model for a factor to the squared correlations between that
factor and all other factors. In all cases except for
The corrected-item total correlations (CITCs) for shop-floor employee involvement in problem solving
all 45 items for the seven time-based manufacturing versus reengineering set-ups, the average variance
34 X.A. Koufteros et al. / Journal of Operations Management 16 (1998) 21-41
extracted is at least twice the squared correlation. For seven time-based manufacturing scales were corre-
this one exception, the squared correlation is 0.52 lated with a composite measure of competitive capa-
and the average variance extracted for these factors bilities. This measure was determined by standardiz-
are 0.60 and 0.58, respectively. Even in this case, the ing the scales and summing them with equal weights
items share more variance within a factor than that (reverse scale for throughput time). The correlations
factor shares with another factor. Overall, the evi- between each of the time-based manufacturing mea-
dence indicates that discriminant validity exists sures and the composite measure of competitive
among the constructs. capabilities are significant at P < 0.01. The correla-
tions for shop-floor employee involvement in prob-
4.3. Reliability lem solving, reengineering set-up, cellular manufac-
turing, quality improvement efforts, preventive main-
Table 4 contains the means, standard deviations, tenance, dependable suppliers, and pull production
correlations between the seven manufacturing prac- are 0.30, 0.33, 0.23, 0.23, 0.36, 0.30, and 0.18,
tice scales, and Cronbach's alpha which measures respectively. A composite scale (equally weighted)
scale reliability. All the scales have alphas that ex- of the 33 items measuring time-based manufacturing
ceed 0.80 with the lowest reliability being 0.85 for practices has a correlation of 0.41 with the composite
cellular manufacturing. In addition, the average vari- measure of competitive capabilities ( P < 0.01).
ance extracted for all factors exceeds the 0.50 crite- To consider the impact of time-based manufactur-
rion suggested by Bagozzi and Yi (1988). ing practices on throughput time, the composite scale
for time-based manufacturing practices was corre-
4.4• Predictive validity lated with throughput time (scale not reversed). The
negative correlation, R = - 0.31 ( P < 0.01), implies
A measure has predictive validity when it is that organizations that have attained high levels of
appropriately related to associated constructs. In this these manufacturing practices have low throughput
case, organizations engaging in time-based manufac- time. To investigate which time-based manufacturing
turing practices should increase their competitive practices may have a significant impact on reducing
capabilities and reduce their throughput time. The throughput time, each of the practices was correlated
I
4:.
36 X.A. Koufteros et al. / Journal of Operations Management 16 (1998) 21-41
with throughput time. Quality improvement efforts because they work with the systems daily and are
(R = - 0.36), preventive maintenance (R = - 0.18), responsible for their success. Shop-floor employee
and pull production (R = - 0 . 2 3 ) are correlated with involvement also helps to build a network of depend-
throughput time ( P < 0.01). Reengineering set-ups able suppliers. It creates an environment where hori-
(R = - 0 . 1 3 ) is correlated with throughput time ( P zontal, across the value chain, communications are
< 0.05). more frequent and more effective. This advantage
may not be restricted by artificial barriers such as the
4.5. Generalizability o f the scales boundary between organizations. As a result, infor-
mation flows smoothly between suppliers and cus-
For scales to achieve wide spread application, tomers.
they must be generalizable across different groups of Three time-based practices, reengineering set-ups
respondents. This is an attractive feature because (/3 = 0.22, t = 1.74, P < 0.10), cellular manufactur-
researchers may use such scales in different contexts. ing (/3 = 0.22, t = 2.49, P < 0.01), and preventive
To assess the generalizability of the scales across maintenance (/3 = 0.18, t = 2.09, P < 0.05), have
industries, reliability (Cronbach's alpha) was calcu- statistically significant relationships with pull pro-
lated for the four industries that had an adequate duction, hypotheses H6, H7, and H9, respectively.
number of responses. With only one exception, the Creating pull systems seems to require equipment
seven scales had alphas above 0.80 (see Table 5). that can be quickly switched from one product to the
The reliability for the quality improvement scale in next and cellular layouts than can respond with
SIC 37 was 0.76. In addition, reliability was assessed minimum delay. The equipment must be reliable
by finn size. For small firms (sales below US$100 which helps to explain the importance of preventive
million) and larger firms, the reliabilities were all maintenance. It was expected that quality improve-
above 0.80. These reliabilities provide preliminary ment efforts would be related to pull production, but
evidence that the scales can be used across industries the beta coefficient was not significant. Also, de-
and across finn size. pendable suppliers does not appear to be directly
related to pull production. The absence of statisti-
cally significant beta coefficients may be caused by
5. Results for the structural model differences between industries and may require in-
dustry-specific studies.
The model shown in Fig. 1 was investigated using
LISREL. GFI = 0.82, CFI = 0.92, IFI = 0.92, and
NFI = 0.83 which indicate a good fit. Ten hypothe-
ses described earlier in the paper were tested. The 6. Discussion
results indicate that shop-floor employee involve-
ment in problem solving is related to reengineering Based on the assumption that the measurement of
set-ups, cellular manufacturing, quality improvement manufacturing phenomena must be handled differ-
efforts, preventive maintenance, and dependable sup- ently across industries, manufacturing research has
pliers (see Fig. 2). These relationships, which are tended to be industry or finn specific. This has
described in Hypotheses HI, H2, H3, H4, and H5, inhibited the development of generalizable theories
have significant gamma coefficients 0.89, 0.50, 0.60, in manufacturing. While structural relationships be-
0.63, and 0.15 with t-statistics of 10.98, 5.38, 6.65, tween constructs may differ among industries, the
7.86, and 2.98, respectively. All are significant at measurement models may not. Measurement models
P < 0.01. that permit large-scale, cross-industry research are
Designing, implementing, and executing pro- essential for developing generalizable theories of
grams to reengineer set-ups, create cellular manufac- manufacturing.
turing, initiate quality improvement efforts, and im- In disciplines like manufacturing where an empir-
plement preventive maintenance systems seem to ical research tradition is emerging, standardized in-
require the participation of shop-floor employees struments may be the cause rather than the conse-
X.A. Koufteros et al. / Journal of Operations Management 16 (1998) 21-41 37
quence of research progress. Developing standard- then supplier dependability may not significantly
ized instruments requires researchers to carefully impact throughput time.
define a variable so it can be accurately measured
and to design items to ensure complete coverage of
the phenomenon. As a result, standardized instru- 7. Conclusions
ments improve theory development, facilitate the
testing of substantive hypotheses, and enhance the This paper provides a research framework that
additivity of research efforts. builds on the Toyota Production System of Monden
The instruments developed here enable manufac- (1983) as well as the vision of time-based competi-
turing research that is broad in scale and, thus, opens tion of Blackburn (1991) and Stalk and Hout (1990).
up promising new avenues for research. Tentative It describes an evolution from JIT and its internal
results indicate that the instruments have good relia- focus on inventory and cost reduction to time-based
bility across industries and across firm size, suggest- manufacturing and its external focus on time and
ing that cross-industry studies and comparisons be- customers. This framework provides a foundation for
tween industries may be possible. Further research is research in time-based manufacturing and is part of
needed to determine whether the measurement mod- the research framework for time-based competition.
els are invariant across industries. Additional work in product development, marketing
These instrument development efforts are a first- and distribution, and competitive capabilities are
step in testing the impact of time-based manufactur- needed to complete the latter.
ing practices on organizational performance. Predic- A theoretical rationale is provided for a set of
tive validity tests indicate that firms with high levels time-based manufacturing practices, and instruments
of time-based manufacturing practices may have high are developed to operationalize the seven dimension
competitive capabilities and low throughput times. so they are applicable to industries that manufacture
Statistically significant correlations between the indi- discrete parts. The final instruments, listed in Ap-
vidual manufacturing practices and the composite pendix A, satisfy the objectives outlined earlier in
measure of competitive capabilities provide a start- the paper. The instruments are short and easy to use.
ing point for establishing a network of constructs to Each scale has six or fewer items and the total
assess time-based competition. Price charged, value number of items across all the scales is only 33. The
to customer quality, and customer delivery service content domain of the constructs has been adequately
are clearly influenced by a manufacturing system covered because care was taken during item genera-
that is both fast and flexible. Such a system will also tion. The items are short and easy to understand with
allow product innovations to be adopted quickly and each item having fewer than 15 words. The factor
passed onto the customer. structure is simple and has high loadings, and the
When the composite measure of manufacturing scales have both discriminant and convergent valid-
practices is correlated with throughput time, the re- ity. Investigation and analysis have demonstrated
sults are statistically significant. The individual prac- that the instruments exceed generally accepted valid-
tices that seem to have the greatest impact on ity and reliability standards for basic research. The
throughput time are reengineering set-ups, quality instruments can be used separately to investigate the
improvement effort, preventive maintenance, and pull antecedents and consequences of specific manufac-
production. Cellular manufacturing and dependable turing practices or together to assess the conse-
suppliers are not significantly correlated with quences of implementing time-based manufacturing.
throughput time, but there may be industry differ- The results obtained here need to be cross-validated
ences that explain these non-significant correlations. in confirmatory studies.
For example, in industries where small lots of fabri- Tests of hypotheses H1 to H5 indicate that shop-
cated parts or subassemblies are frequently delivered floor employee involvement in problem solving is a
by suppliers, missed shipments create delays that catalyst for reengineering set-ups, cellular manufac-
immediately impact throughput time. If suppliers turing, quality improvement efforts, preventive main-
deliver large quantities of raw materials infrequently, tenance, and dependable suppliers. These relation-
38 X.A. Koufleros et al. / Journal of Operations Management 16 (1998) 21-41
ships may be independent of the industry involved performance. Much additional research is needed to
because several industries are included in the study. understand time-based manufacturing and its impli-
Test of hypotheses H6 to H10 indicate that three cation on organizational effectiveness.
time-based manufacturing practices (reengineering
set-ups, cellular manufacturing, and preventive main-
tenance) have a significant relationship with pull Acknowledgements
production. The lack of relationships between quality
improvement efforts and pull production and be- The authors wish to thank the Information Sys-
tween dependable suppliers and pull production may tems and Operations Management Department's
be caused by industry differences. For example, Academic Challenge Grant Committee at The Uni-
some factor of production available in one industry, versity of Toledo for its generous support of this
such as technology, may create uniformly high qual- research.
ity within the industry. In this case, there may not be
enough variation in quality to show a relationship or
Appendix A. Time-Based Manufacturing Prac-
correlation with pull production.
tices
While the primary contribution of this paper is to
develop valid and reliable instruments to measure the A.1. Shop-floor employee involvement in problem
extent to which time-based manufacturing practices solving
have been achieved, some management insights are
also available. Involving shop-floor employees may PS1 Shop-floor employees are involved in
be essential to successfully implement new manufac- problem solving efforts
turing practices. Shop-floor employees have substan-
tial experience with the processes and associated PS3 Shop-floor employees are involved in sug-
problems and often provide important insights for gestion programs
problem solving. Pull production may be facilitated PS4 Shop-floor employees are involved in de-
by several time-based practices rather than a parame- signing processes and tools that focus on
ter that management can set directly. This would improvement
imply that achieving small lot sizes and responding PS5 Shop-floor employees are involved in im-
quickly to customer demands are the result of funda- provement efforts
mental changes to the manufacturing system and not PS6 Shop-floor employees are involved in
something that can be achieved by high-level policy problem solving teams
changes alone.
The instruments developed in this study as well as A.2. Reengineering set-up
the structural analysis present opportunities for addi-
RS 14 Employees work on set-up improvement
tional research. The interactions between the time-
based practices, illustrated in Fig. 1, may not fully RS16 Employees redesign or reconfigure equip-
describe key relationships. For example, reengineer- ment to shorten set-up time
ing set-ups to reduce time may facilitate cellular RS17 Employees redesign jigs or fixtures to
manufacturing. Quality improvement efforts may im- shorten set-up time
pact the need for preventive maintenance. More in- RSI8 We use special tools to shorten set-up
vestigation of the structural relationships is needed. RS19 Our employees are trained to reduce set-up
It would also be appropriate to complete a confirma- time
tory study for instrument development, perform in-
dustry studies that examine issues related to the A.3. Cellular manufacturing
structural model, examine relationships between
time-based manufacturing and other time-based prac- CM2 Products that share similar design or pro-
tices such as product development, and test relation- cessing requirements are grouped into fam-
ships between time-based manufacturing and firm ilies of products
X.A. Koufieros et al. / Journal of Operations Management 16 (1998) 21-41 39
CM3 Products are classified into groups with PP5 Production at stations is ' p u l l e d ' by the
similar processing requirements current d e m a n d of the next stations
CM4 Products are classified into groups with PP6 W e use a ' p u l l ' production system
similar routing requirements
CM7 E q u i p m e n t is grouped to produce families
of products References
CM8 Families of products determine our factory Abegglen, J., Stalk, G., 1985. Kaisha, The Japanese Corporation.
layout Basic Books, New York.
Adam, E.E., 1994. Alternative quality improvement practices and
organization performance. J. Oper. Manage. 12, 27-44.
A.4. Quality improvement efforts Ansari, A., Modaress, B., 1988. JIT purchasing as a quality and
productivity center. Int. J. Prod. Res. 26 (1), 19-26.
W e use fishbone type diagrams to identify Ansari, A., Modaress, B., 1990. Just-in-Time Purchasing. The
QI1
Free Press, New York.
causes of quality problems Badore, N.L., 1992. Involvement and empowerment: the modern
QI5 W e use design of experiments (i.e., Taguchi paradigm for management success. In: Heim, J.A., Compton,
methods) W.D. (Eds.), Manufacturing Systems: Foundation of World-
QI8 Our employees use quality control charts class Practice. National Academy Press, Washington, DC.
(e.g., SPC charts) Bagozzi, R.P., Phillips, L.W., 1982. Representing and testing
organizational theories: a holistic construct. Admin. Sci. Q.
QI16 W e c o n d u c t process capability studies 27, 459-489.
Bagozzi, R.P., Yi, Y., 1988. On the evaluation of structural
A.5. Preventive maintenance equation models. Acad. of Marketing Sci. 16 (1), 74-94.
Bayus, B., 1994. Are product life cycles really getting shorter. J.
Prod. Inno. Manage. 11,300-308.
PM3 W e emphasize good preventive mainte- Bentler, P.M., 1990. Comparative fit indexes in structural models.
nance Psychol. Bull. 107, 238-246.
PM5 Records of routine m a i n t e n a n c e are kept Bentler, P.M., Bonnet, D.G., 1980. Significance tests and good-
ness of fit in analysis of covariance structures. Psychol. Bull.
PM6 W e do preventive m a i n t e n a n c e 88, 588-606.
PM7 W e do preventive m a i n t e n a n c e during Blackburn, J., 1991. Time-Based Competition. Business One Ir-
n o n - p r o d u c t i v e time win, Homewood, IL.
PM8 W e m a i n t a i n our e q u i p m e n t regularly Bockerstette, J., Shell, R., 1993. Time-based Manufacturing. Mc-
Graw-Hill, Norcross, GA.
Brown, K., Mitchell, T., 1991. A comparison of Just-in-time and
A.6. Dependable suppliers batch manufacturing: the role of performance obstacles. Acad.
Manage. J. 34, 906-917.
Byrne, B.M., 1989. A Primer of LISREL: Basic Applications and
DS1 W e receive parts from suppliers on time Programming for Confirmatory Factor Analytic models.
DS4 W e receive the correct n u m b e r of parts Springer-Verlag, New York.
from suppliers Campbell, D.T., Fiske, D.W., 1959. Convergent and discriminant
DS7 W e receive the correct type of parts from validation by the multitrait-multimethod matrix. Psychol. Bull.
suppliers 56 (1), 81-105.
Churchill, G.A., 1979. A paradigm for developing better measures
DS9 W e receive parts from suppliers that meet
of marketing constructs. J. of Marketing Res. 16, 64-73.
our specifications Clark, K., 1989. Project scope and project performance: the
DS10 Our suppliers a c c o m m o d a t e our needs effects of parts strategy and supplier involvement on product
DS11 W e receive high quality parts from suppli- development. Manage. Sci. 35, 1247-1263.
ers Cohen, J., Cohen, P., 1983. Applied Multiple Regression/Correla-
tion Analysis of the Behavior Science, 2nd edn. Lawrence,
Erlbaum Assoc., Hillsdale, NJ.
A. 7. Pull production Comrey, A.L., 1988. Factor ana!ytic methods of scale develop-
ment in personality and clinical psychology. J. of Consulting
Clin. Psychol. 56, 754-761.
PP3 Production is ' p u l l e d ' by the shipment of Cronbach, L.J., 1951. Coefficient alpha and the internal structure
finished goods of tests. Psychometrika 16, 297-334.
40 X.A. Koufteros et al./ Journal of Operations Management 16 (1998) 21-41
Davy, J., White, R., Merritt, N., Gritzmacher, K., 1992. A deriva- Juran, J., 1978. Japanese and western quality: a contrast in
tion of the underlying constructs of Just-in-time management methods and results. Manage. Rev. 67 (11), 27-45.
system. Acad. Manage. J. 35,653-670. Juran, J., 1981a. Product quality--a prescription for the west: Part
Deming, W.E., 1981. Improvement of quality and productivity I. Manage. Rev. 70 (6), 8-14.
through action by management. Natl. Productivity Rev. 1 (l), Juran, J., 198 lb. Product quality--a prescription for the west: Part
12-22. II. Manage. Rev. 70 (7), 57-61.
Deming, W.E., 1982. Quality, Productivity, and Competitive Posi- Juran., J., 1988. Quality Control Handbook, 4th edn. McGraw-Hill,
tion. MIT Center for Advanced Engineering, Cambridge, MA. New York.
Deming, W.E., 1986. Out of the Crisis. MIT Center for Advanced Koufteros, X., 1995. Time-based competition: developing a nomo-
Engineering, Cambridge, MA. logical network of constructs and instrument development.
DeVellis, R.F., 1991. Scale Development: Theory and Applica- PhD Dissertation, Toledo OH.
tion. Sage Publications, Newbury Park, CA. Lawrence, P.R., Dyer, D. 1983. Renewing the American Industry.
Doll, W.J., Vonderembse, M.A., 1991. The evolution of manufac- Free Press, London.
turing systems: towards the post-industrial enterprise. Lee, S., Ebrahimpour, M., 1984. Just-in-time production system:
OMEGA: Int. J. Manage. Sci. 19 (5), 401-411. some requirements for implementation. Int. J. Oper. Prod.
Fornell, C., Larcker, D.F., 1981. Evaluating structural equation Manage. 4 (4), 3-15.
models with unobservable variables and measurement error. J. Leonard, F., Sasser, W., 1982. The incline of quality. Harvard
Marketing Res. 18 (2), 39-50. Business Rev. 60 (5), 163-171.
Garvin, D., 1983. Quality on the line. Harvard Business Rev. 61 Lindsley, W., Blackburn, J., Elrod, T., 1991. Time and product
(5), 65-75. variety competition in the book industry. J. Oper. Manage. 10
Gatewood, R.D., Field, H.S., 1994. Employee Selection. Dryden (3), 344-362.
Press, Fort Worth, TX. Locke, E.A., Schweiger, D.M., 1979. Participation in decision
Giunipero, L., 1990. Motivating and monitoring JIT supplier making: one more look. In: Staw, B. (Ed.), Research in
performance. J. of Purchasing Mater. Manage. 26 (3), 19-24. Organizational Behavior. JAI Press, Greenwich, CT, pp. 265-
Hall, R., 1987. Attaining manufacturing excellence. Dow Jones- 339.
Irwin, Homewood, IL. Manufacturing Studies Board, 1986. Towards a New Era in U.S.
Hall, R., 1993. The Soul of the Enterprise. Harper Business, New Manufacturing: The Need for a National Vision. National
York. Academy Press, Washington, DC.
Hall, R.W., Johnson, H.T., Turney, P.B.B., 1991. Measuring Up: Markus, M.L., 1983. Power, politics, and MIS implementation.
Charting Pathways to Manufacturing Excellence. Business Commun. ACM 26 (6), 430-444.
One Irwin, Homewood, IL. Marsch, H.W., Hocevar, D., 1985. Application of confirmatory
Hamilton, S., 1991. New-product development and manufacturing factor analysis of the study of self-concept: first and higher
competitiveness: a Hewlett-Packard perspective. In: Black- order factor models and their invariance across groups. Psy-
burn, J. (Ed.), Time-based Competition. Business One Irwin, chol. Bull. 97, 562-582.
Homewood, IL, pp. 191-208. Maskell, B.H., 1991. Performance Measures for World Class
Handfield, R.B., Pannesi, R.T., 1992. An empirical study of Manufacturing. Productivity Press, Cambridge, MA.
delivery speed and reliability. Int. J. Oper. Prod. Manage. 12 Merrills, R., 1989. How Northern Telecom competes on time.
(2), 58-72. Harvard Business Rev. 67 (4), 108-114.
Handfield, R.B., Pannesi, R.T., 1995. Antecedents of leadtime Miller, J., De Meyer, A., Nakane, J., 1992. Benchmarking Global
competitiveness in make-to-order manufacturing firms. Int. J. Manufacturing. Business One Irwin, Homewood, IL.
Prod. Res. 41 (4), 511-537. Monden, Y., 1981. How Toyota shortened supply lot production
Hayes, R., Wheelwright, S., Clark, K., 1988. Dynamic Manufac- lead time, waiting time and conveyance time. Industrial Engi-
turing. The Free Press, New York. neering, September: 22-30.
Huber, G., 1984. The nature and design of post-industrial organi- Monden, Y., 1983. Toyota Production System: A Practical Ap-
zations. Manage. Sci. 30, 928-951. proach to Production Management. Industrial Engineers and
Hyer, N., Wemmerlov, U., 1984. Group technology and produc- Management Press, Norcross, GA.
tivity. Harvard Business Rev. 62 (4), 140-149. Naisbitt, J., 1982. Megatrends. Warner Books, New York.
Im, J., Lee, S., 1989. Implementation of Just-in-time system in Nunnally, J.C., 1978. Psychometric Theory. McGraw-Hill, New
U.S. manufacturing firms. Int. J. Oper. Prod. Manage. 9 (1), York.
5-14. Ohno, T., 1978. Toyota Seissu Hoshiki (Toyota Production Sys-
Johnson, B., Rice R., 1987. Managing Organizational Innovation: tem). Diamond.
The Evolution from Word Processing to Office Information Pullen, R., 1976. A survey of cellular manufacturing cells. The
Systems. Columbia Univ. Press, New York. Prod. Eng. 55 (9), 451-454.
Joreskog, K.G., Sorbom, D., 1986. LISREL VI: Analysis of Robey, D., Farrow, D., 1982. User involvement in information
Linear Structural Relationships by Maximum Likelihood, In- systems development: a conflict model and empirical test.
strumental Variables, and Least Squares Methods. Scientific Manage. Sci. 28 (l), 73-85.
Software, Moorsville, IN. Roth, A.V., Miller, J.G., 1990. Manufacturing strategy, manufac-
X.A. Koufieros et al. / Journal of Operations Management 16 (1998) 21-41 41
turing strength, managerial success, and economic outcomes. Strategies for Service Organizations. Business Horizons,
In: Ettlie, J., Burstein, M., Fiegenbaum, A. (Eds.), Manufac- July-August: 82-87.
turing Strategy. Kluwer Academic Publisher, Norwell, MA, Skinner, W., 1985. Manufacturing: The Formidable Competitive
pp. 97-108. Weapon. Wiley, New York.
Roth, A.V., Miller, J.G., 1992. Success factors in manufacturing. Stalk, G., Hout, T., 1990. Competing Against Time. The Free
Business Horizons 4, 73-81. Press, New York.
Sakakibara, S., Flynn, B., Schroeder, R., 1993. A Just-in-Time Swamidass, P., 1991. Empirical science: new frontier in opera-
management framework and measurement instrument. Prod. tions management research. Acad. Manage. Rev. 16, 793-813.
Oper. Manage. 2, 177-194. Swamidass, P., Newell, W., 1987. Manufacturing strategy, envi-
Saraph, J., Benson, G., Schroeder, R., 1989. An instrument for ronmental uncertainty and performance: a path analytic model.
measuring the critical factors in quality management. Decision Manage. Sci. 33, 509-524.
Sci. 20, 810-829. Tinsley, H.E.A., Tinsley, D.J., t987. Use of factor analysis in
Schmenner. R., 1991. Speed and productivity. In: Blackburn, J. counseling psychology research. J. of Counseling Psychol. 34.
(Ed.), Time-based Competition. Business One Irwin, Home- 414-424.
wood, 1L, pp. 102-118 Toffier, A., 1970. Future Shock. Random House, New York.
Schmenner, R., 1992. So you want to lower costs? Business Weiss, D.J., 1970. Factor analysis in counseling research. J. of
Horizons, July-August: 24-28. Counseling Psychol. 17, 477-485.
Schonberger, R., 1986. World-class Manufacturing: The Lessons Wemmerlov, U., Hyer, N., 1989. Cellular manufacturing in the
of Simplicity Applied. The Free Press, New York. U.S. industry: a survey of users. Int. J. Prod. Res. 27, 1511-
Shingo, S.. 1985. A Revolution in Manufacturing: The SMEAD 1530.
System. Productivity, Cambridge, MA. Zuboff, S., 1984. In the Age of the Smart Machine: The Future of
Showalter, M., Mulholtand, J., 1992. Continuous Improvement Work and Power. Basic Books, New York.