THESIS

 

 

 

D [ngADV
\ ‘N ‘35,

-. I “GIL/I "Ci .- I blate
c ‘0 2.0 U . .wersny

 

This is to certify that the
dissertation entitled

ORGANIZATIONAL CLIMATE AND PERFORMANCE: AN
EXAMINATION OF CAUSAL PRIORITY

presented by

ANTHONY S. BOYCE

has been accepted towards fulfillment
of the requirements for the

PhD. degree in Psychology

 

 

////% //Z/{/ /M/Z ‘

Major Professor’fiifinature
0/2 ' .942 ’/ (7

Date

 

MSU is an Affirmative Action/Equal Opportunity Employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 K'IPij/AolerelelRC/DateDueJndd

 

ORGANIZATIONAL CLIMATE AND PERFORMANCE: AN EXAMINATION OF
CAUSAL PRIORITY

By

Anthony S. Boyce

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Psychology

2010

ABSTRACT

ORGANIZATIONAL CLIMATE AND PERFORMANCE: AN EXAMINATION OF
CAUSAL PRIORITY

By

Anthony S. Boyce

Research and theory suggest that there is a relationship between organizational climate
and organizational performance. However, there exists both limited and conflicting
evidence as to whether climate is a cause of performance, vice-versa, or the two are
reciprocally related. The current study examines this issue of causal priority using a
longitudinal research design where data were collected on multiple occasions over a
period of six years for both organizational climate and performance. Data on
organizational climate and unit-level customer satisfaction were collected separately for
sales departments and service departments in 95 automobile dealerships. For sales
departments, unit-level sales data were also collected. Using cross-lagged panel analyses,
it was found that organizational climate exhibited causal priority over customer
satisfaction in both sales and service departments. Stable causal relationships emerged
more quickly, after one year, in service departments, while for sales departments stable
causal relationships were observed only at two-year lags. For sales departments,
organizational climate and customer satisfaction were both observed to have causal
priority over unit-level sales at two-year lag intervals, but the effects were small.
Additionally, customer service was found to fully mediate the causal relationship
between organizational climate and unit-level sales. No evidence of reciprocal

relationships, where organizational performance would have also predicted subsequent

organizational climate, were observed. The results of this study are discussed in terms of
their implications for the study of causal relationships between organizational climate and
performance, how the results compare to the results of prior research, possible directions

for future research, and potential limitations.

Copyright by
ANTHONY SCOTT BOYCE
2010

To Sarah Kaye Conklin Boyce, my best friend and wife, for her continual support,

insightful ideas, and charming wit.

ACKNOWLEDGEMENTS

I am pleased to thank Ann Marie Ryan, my Chair and Advisor. 1 was lucky to
begin working with her in 1996, as an undergraduate, and I am grateful she wanted to
continue working with me throughout my graduate career. She has always been adept at
knowing when to push me, when to offer ideas, and when to sit back and wait for me to
figure things out. It truly would have been next to impossible to write this dissertation
without her help and guidance. _

I am gratefiil to the many friends and colleagues that supported me both
intellectually and emotionally throughout this process — this list is too long to include
here, but they know who they are. I am also especially grateful to Michael Gillespie who
not only helped me to procure the data, but also spent many hours talking through ideas
with me and providing encouragement.

I am very appreciative of the insightful comments, ideas, and guidance provided
by my committee members: Steve W. J. Kozlowski, Frederick P. Morgeson, and Neal
Schmitt. I am also grateful for the guidance and training each of them provided
individually throughout my graduate studies.

Thanks as well to Denison Consulting and, especially, Daniel R. Denison for
creating and allowing me to leverage the comprehensive data set that is the core of this
dissertation.

Last, but by no means least, I am forever grateful to my brother, mother, and
father: Jason, Janis, and Larry Boyce. Without their unconditional support and help in

keeping things in the proper perspective, I would not be writing these acknowledgements.

vi

TABLE OF CONTENTS

LIST OF TABLES ........................................................................................................... viii
LIST OF FIGURES ............................................................................................................ x
Introduction ......................................................................................................................... 1
Climate ............................................................................................................................ 3
Link between Organizational Climate and Performance ................................................ 8
The Current Study ......................................................................................................... 18
Summary ....................................................................................................................... 29
Method .............................................................................................................................. 30
Sample and Procedure ................................................................................................... 30
Measures ....................................................................................................................... 34
Analytic Strategy .......................................................................................................... 40
Results ............................................................................................................................... 45
Descriptive Statistics ..................................................................................................... 45
Measurement Invariance ............................................................................................... 45
Cross-lagged Panel Analyses ........................................................................................ 50
Additional Hypotheses & Research Question 3 ............................................................ 76
Discussion ......................................................................................................................... 84
Limitations .................................................................................................................... 98
Conclusion .................................................................................................................. 1 00
Appendix ......................................................................................................................... 102
References ....................................................................................................................... l 03

vii

LIST OF TABLES

Table 1: Summary of Overall Sample Sizes for Each Department and Each Year ........ 30
Table 2: Summary of Data Collection Time Periods ...................................................... 33

Table 3: Summary of Aggregation Statistics for Each Department and Each Year ....... 39

Table 4: Descriptives Statistics and Intercorrelations — Sales Departments ................... 46
Table 5: Descriptives Statistics and Intercorrelations — Service Departments...............47
Table 6: Summary of Fit Indices for Measurement Invariance Analyses ....................... 4 9

Table 7: Summary of Fit Indices for Service Department Climate and Customer
Satisfaction One-Year Cross-Lagged Models ................................................................ 51

Table 8: Summary of Fit Indices for Service Department Climate and Customer
Satisfaction Two-Year Cross-Lagged Models ................................................................ 52

Table 9: Summary of Fit Indices for Service Department Climate and Customer
Satisfaction Three-Year Cross-Lagged Models .............................................................. 53

Table 10: Summary of Fit Indices for Sales Department Climate and Customer
Satisfaction One-Year Cross-Lagged Models ................................................................ 58

Table 11: Summary of Fit Indices for Sales Department Climate and Customer
Satisfaction Two-Year Cross-Lagged Models ................................................................ 59

Table 12: Summary of Fit Indices for Sales Department Climate and Customer
Satisfaction Three-Year Cross-Lagged Models .............................................................. 60

Table 13: Summary of Fit Indices for Sales Department Climate and Vehicle Sales
One-Year Cross-Lagged Models ..................................................................................... 6 5

Table 14: Summary of Fit Indices for Sales Department Climate and Vehicle Sales
Two-Year Cross-Lagged Models .................................................................................... 66

Table 15: Summary of Fit Indices for Sales Department Climate and Vehicle Sales
Three-Year Cross-Lagged Models .................................................................................. 67

Table 16: Summary of Fit Indices for Sales Department Customer Satisfaction and
Vehicle Sales One-Year Cross-Lagged Models .............................................................. 71

viii

Table 17: Summary of Fit Indices for Sales Department Customer Satisfaction and

Vehicle Sales Two-Year Cross-Lagged Models ............................................................. 72
Table 18: Summary of Fit Indices for Sales Department Customer Satisfaction and
Vehicle Sales One-Year Cross-Lagged Models .............................................................. 73
Table 19: Summary of Fit Indices for Sales Department Climate, Customer
Satisfaction, and Vehicle Sales Cross-Lagged Models ................................................... 77
Table 20: Summary of Results for Longitudinal Growth Models for Sales

Departments .................................................................................................................... 81
Table 20: Summary of Results for Longitudinal Growth Models for Service
Departments .................................................................................................................... 82
Table 22: Summary of Hypotheses and Research Questions .......................................... 85

Table 23: Sales Department Covariance Matrix for Measurement Invariance
Analyses .......................................................................................................................... 102

Table 24: Service Department Covariance Matrix for Measurement Invariance
Analyses .......................................................................................................................... 103

ix

LIST OF FIGURES

Figure 1: Analytic Strategy ............................................................................................. 41
Figure 2: Final Model for Service Department Climate and Customer Satisfaction ...... 56
Figure 3: Final Model for Sales Department Climate and Customer Satisfaction .......... 62
Figure 4: Final Model for Sales Department Climate and Vehicle Sales ........................ 6 9

Figure 5: Final Model for Sales Department Customer Satisfaction and Vehicle
Sales ................................................................................................................................. 74

Figure 6: Final Mediation Model for Sales Department Climate, Customer
Satisfaction, and Vehicle Sales ........................................................................................ 78

Introduction

As organizations continually face greater competition and high profit expectations
from Wall Street, it becdmes increasingly important for all functions to prove their value
to the organization. There are well known metrics for tracking the value added by finance
or marketing departments, but assessing the bottom-line contributions of the human
resources (HR) function is more difficult because the outcomes are often less tangible
(Becker, Huselid, & Ulrich, 2001; Lev, 2001; Russ-Eft & Preskill, 2005; Smith, 2003).
How, for example, does HR show the value, in terms of enhancing organizational
performance, of a climate change intervention or an employee involvement program?
When one considers the fact that HR-related expenses account annually for over $1,000
per employee, excluding the 42% of operating expenses that go directly to salaries
(Dooney & Smith, 2006), the answer to this question becomes even more pressing.
Recently researchers operating from a variety of backgrounds, including human resource
management, organizational behavior, and industrial-organizational psychology, have
begun to address this problem by attempting to link some of the outcomes of the HR
function directly to indicators of organizational performance. Some efforts at linkage
research (Wiley, 1996) have focused on linking bundles of human resource management
practices (e. g., objective selection, formal training programs) to organizational
performance (e.g., Huselid, 1995; Huselid, Jackson, & Schuler, 1997; Wright, Gardner,
Moynihan, & Allen, 2005). Other efforts have focused on HR’s responsibility for
maintaining a motivating environment by looking at how employee perceptions of the
work climate and attitudes relate to organizational performance (e.g., Borucki & Burke,

1999; Schmit & Allscheid, 1995; Schneider, Hanges, Smith, & Salvaggio, 2003). The

rationale behind both types of linkage studies is that demonstration of causal links
between HR outcomes and organizational performance can help to prove the strategic
value of the HR function. However, with a few exceptions (e. g., Koys, 2001; Kozlowski
& Farr, 1988; Ryan, Schmit, & Johnson, 1996; Schneider, White, & Paul, 1998), HR
linkage studies have utilized cross-sectional designs useful for showing that the outcomes
of HR covary with organizational performance, but not useful in establishing causal links
(Shadish, Cook, & Cambpell, 2002).

The current study contributes to the ability of HR to prove its value by
investigating the causal link between organizational climate and important indicators of
organizational performance. The current study addresses several limitations of prior
research on this topic. First, this study utilizes a longitudinal design in which climate
perceptions and indicators of organizational performance are assessed over six years.
Such a design allows for much stronger causal inferences than previous research and also
allows for investigation of reverse causality and reciprocal relationships sometimes
observed in the empirical literature (e. g., Ryan et al., 1996; Schneider et al., 1998).
Second, the sample is comprised of independently owned and operated organizations in
the same industry, carrying the exact same products, and focusing on the exact same
indicators of performance. Many previous studies have utilized samples within a single
organization, which may contribute to range restriction in climate perceptions, or samples
of organizations in different industries which creates problems in justifying common
indicators of performance. Finally, the indicators of organizational performance chosen in
the current study (i.e., customer satisfaction and unit-level sales) were chosen because of

their proximity to employee behavior and insulation from some of the external forces

(e. g., CEO scandals, downsizing) that can influence more macro-level indicators of
performance sometimes utilized in prior research (e. g., return on investment, profit, etc.).

Before describing the current study in greater detail, it is first necessary to briefly
review the concept of climate. Next, the theoretical justifications for expecting a causal
relationship between climate and organizational performance will be reviewed. Empirical
research focusing on such relationships will also be reviewed with particular attention
given to the limitations of much of the existing research for demonstrating causal links.
Finally, the current study and hypotheses will be discussed in detail.
Climate

Climate, as a concept, has its roots in Lewinian field theory, which proposes that
behavior (B) is a function of both the person (P) and the environment (E; i.e., B=f(P,E);
Lewin, 1951). In this model, individuals bring goals or desires with them to situations,
which are then perceived and interpreted in terms of these goals. These perceptions of the
environment, and their implications for an individual’s goals, then serve as guides for
subsequent action. The influence of Lewin’s field theory on the conceptualization of
climate can be clearly seen in various definitions of climate offered in the organizational

literature:

“. . .climate refers to individual attributes, namely the intervening psychological
process whereby the individual translates the interaction between perceived
organizational attributes and individual characteristics into a set of expectancies,

attitudes, behaviors, etc.” (James & Jones, 1974, p. 1110)

“[Climate] can be thought of as psychologically meaningful descriptions of
contingencies and situational influences that individuals use to apprehend order,
predict outcomes, and gauge the appropriateness of their organizational

behaviors.” (Kopelman, Brief, & Guzzo, 1990: p. 294)

“I define climate as incumbents’ perceptions of the events, practices, and
procedures and the kinds of behaviors that get rewarded, supported, and expected

in a setting.” (Schneider, 1990, p. 384).

Although semantic differences are present, common across all of these definitions is a
focus on perceptions of the environment and the implications of these perceptions for the
individual. Thus, in line with the above definitions, but in more general terms,
organizational climate is defined here as the shared perceptions employees attach to
organizational events, policies, practices, and procedures (Ostroff, 1993; Schneider &
Reichers, 1983).

In line with Lewin’s field theory, climate perceptions are thought to arise as
employees engage in sense-making behaviors aimed at understanding the implications of
these organizational features (i.e., events, policies, practices, and procedures) for the self
in terms of the types of attitudes and behaviors that are rewarded and supported by the
organization (James & James, 1989; Kopelman et al., 1990; Schneider & Reichers, 1983;
Weick, 1995). Thus, climate perceptions serve as the mediating mechanism through
which organizational features influence individuals’ behavior and attitudes. For example,

organizational features perceived by employees as encouraging collaboration, such as the

 

implementation of team building exercises, will indicate to individuals that positive
attitudes toward, and behavior consistent with, collaboration are endorsed by the
organization. Leadership and management also influence this perceptual sense-making
process by the nature of their roles as gatekeepers and framers of organizational
information and through their implementation of organizational practices, policies, and
procedures (e.g., Kozlowski & Doherty, 1989; Litwin & Stringer, 1968; McGregor,
1960).

James and Jones (1974) made a widely-embraced distinction between
psychological climate, individuals’ perceptions of the work environment, and
organizational climate, shared perceptions of the work environment aggregated across
individuals. This conceptualization of organizational climate as a higher-level isomorph
of psychological climate (Kozlowski & Klein, 2000) prompted researchers to explore the
processes through which climate perceptions came to be shared. Payne and Pugh (1976)
took a top-down structural approach and argued that shared perceptions arise directly
from the shared environments in which coworkers operate. Thus, workers within a single
unit or organization come to have shared perceptions because of shared environmental
features, while workers across different units or organizations have different shared
perceptions because of the different environmental features they experience. Other
researchers view shared climate perceptions as emerging from bottom-up processes.

The attraction-3e]ection-attrition (ASA) approach suggests that shared perceptions
emerge as a fimction of the homogeneity of persons within an organization (Schneider,
1987; Schneider, Goldstein, & Smith, 1995). From this perspective, individuals with

similar attitudes, values, goals, etc. are initially attracted to organizations, organizations

select individuals similar to those already within the organization, and dissimilar
individuals eventually attrit from the organization. Over time, this ASA process results in
a relatively homogenous group of individuals with common climate perceptions. The
symbolic interactionist approach suggests that psychological climate perceptions are
negotiated and further refined through social interaction with colleagues resulting in the
emergence of shared organizational climate perceptions that are mutually determined by
the individual and the group (Rentsch, 1990; Schneider & Reichers, 1983). It is likely
that all of these processes contribute to the emergence of a shared organizational climate
within particular units or organizations, and dissimilar organizational climates between
units or organizations.

Aside from efforts at understanding the nature and emergence of climate,
researchers have also focused attention on the appropriate content, or dimensions, of
climate. Much of this focus has centered on molar dimensions of climate thought to be
widely applicable and beneficial to all employees and organizations, although agreement
on the exact content and labels of the dimensions has varied widely (J arnes & James,
1989; Ostroff, 1993; Ostroff, Kinicki, & Tamkins, 2003; Pritchard & Karasick, 1973). On
the basis of both empirical and theoretical work on human needs, job satisfaction, and
climate (e.g., Alderfer, 1972; Campbell, Dunnette, Lawler, & Weick, 1970; Elizur, 1984;
Katz and Van Maanen, 1977), Ostroff (1993) proposed a comprehensive taxonomy
composed of three higher order facets that has gained some popularity in the literature
(e.g., Carr, Schmidt, DeShon, & Ford, 2003). Ostroff’s affective facet focuses on
participation, involvement, and social relations among workers, the cognitive facet

focuses on psychological involvement, innovation, and'development, and the

instrumental facet focuses on task involvement and work processes, such as structure.
James and James (1989) took the molar conception of climate one step further by
presenting evidence that a single hi gher-order general climate factor underlies the
dimensions of climate. These authors argued that this higher-order factor should be
expected given that climate perceptions are based on an individual’s consideration of the
implications organizational features have for the self in terms of promoting behaviors and
attitudes that are consistent with employees’ and, by extension, organizational well-
being.

Although the molar dimensions of climate have proved valuable in terms of
predicting outcomes important to organizations (e.g., Brown & Leigh, 1996; Carr et al.,
2003; Day & Bedeian, 1991; Parker etal., 2003) some researchers have argued for, and
embraced, more specific climate dimensions focused on particular referents (e. g.,
Schneider, 1975; Schneider, 1990; Zohar, 1980). On the basis of critical reviews
suggesting that the relationships between organizational climate and organizational
effectiveness were weak at best (e.g., Campbell, Dunnette, Lawler, & Weick, 1970;
Payne & Pugh, 1976), Schneider and his colleagues argued that the molar conception of
climate was too broad and inclusive to prove useful as a predictor of important
organizational outcomes (e.g., Schneider, 1975; Schneider, Parkington, & Buxton, 1980).
The alternative advocated was for a more particular conceptualization of climate with a
particular referent or strategic focus in line with an organization’s goals—that is, a
climate for something. Schneider and his colleagues argued that this conception of a
climate for something is more effective in predicting specific outcomes because it is

operationalized at a level of specificity matching many relevant criteria (Schneider, 1975;

Schneider, 1990). Some support for this notion has been found by research relating
climates for safety, service, or technical updating to correspondingly narrow individual-
and organizational-level criteria (e. g., Kozlowski & Hults, 1987; Schneider et al., 1998;
Zohar, 2000). Despite the increasing popularity of this more specific conceptualization of
climate, researchers continue to find empirical justification, in the form of predicted
relationships, with the molar conceptualization of climate adopted in the current study as
well (e.g., Ostroff, 1993; Carr et al., 2003; Parker et al., 2003).

In summary, organizational climate is defined as the shared perceptions
employees attach to organizational events, policies, practices, and procedures. These
perceptions are based at the individual-level, and over time may emerge to be shared at
the group-level, as employees engage in sense-making aimed at understanding the
implications of organizational features in terms of the behaviors and attitudes endorsed
by the organization. Thus, these shared perceptions are viewed as the mediating
mechanism between features of the organization and the collective attitudes and
behaviors of employees. Regardless of whether one adopts a molar or specific
perspective on the content of climate, climate perceptions, by definition of having
implications for collective attitudes and behaviors of employees, are expected to have
implications for organizational performance. The next section briefly reviews the process
through which climate perceptions can impact organizational performance as well as
prior empirical research attempting to establish such links.

Link between Organizational Climate and Performance
Theoretical arguments. A number of researchers have provided models linking

climate to organizational effectiveness (James & Jones, 1976; Kopelman et al., 1990;

Ostroff & Bowen, 2000; Ostroff et al., 2003). Common among all these models is the
proposition that climate influences attitudes (e. g., job satisfaction, organizational
commitment) and, both directly and indirectly through attitudes, motivation and behavior
(e. g., task-focused, organizational citizenship behavior, turnover). Employee’s behaviors
then combine to impact performance. Although almost completely ignored in the
discussion of these theories, it should be noted that some of the models do include
feedback loops flowing from organizational performance back to climate (e.g., James &
Jones, 1976).

The processes whereby climate influences organizational performance operate in
parallel at both the individual- and group-levels. Additionally, within variables, there are
reciprocal relationships between the two levels. For example, as discussed previously,
psychological and organizational climates can mutually influence one another as
individuals and groups negotiate their perceptions of the environment (Schneider &
Reichers, 1983). Likewise, collective attitudes and behaviors can exert top-down
contextual influences on individual attitudes and behaviors, which themselves combine
through bottom-up emergent processes to manifest as collective attitudes and behaviors
(Kozlowski & Klein, 2000). Cross-level relationships between variables are also
proposed to exist. For example, organizational climate can impact individual attitudes
above-and-beyond the individual-level influence of psychological climate (N aumann &
Bennett, 2000). Although it is important to recognize the multi-level nature of these
processes, the conceptual focus of the current study is at the group-level. Therefore, the
individual-level, emergent, and cross-level processes will only be discussed when

necessary for understanding the primary focus of the current research.

Climate perceptions influence the attitudes and behaviors of employees (e.g., Carr
et al., 2003; Kozlowski & Hults, 1987; Ostroff, 1993), and collective attitudes and
behaviors have implications for organizational performance (e. g., Ostroff, 1992; Ryan et
al., 1996; Schneider & Bowen, 1985). From a variety of perspectives on motivation it can
be seen that, at the individual-level, climate perceptions can influence the performance of
job-related behaviors by promoting adoption of goals (Locke & Latharn, 1990), positive,
or negative, attitudes (Ajzen & Fishbein, 1980; Ajzen, 1985), or instrumentalities (e.g.,
Campbell & Pritchard, 1976; Vroom, 1964) towards the behaviors. For example, climate
perceptions related to the importance of cooperation in an organization indicate to
employees that cooperation is supported and rewarded by the organization. In such a
situation, expectancy-valence theories predict that employees are more likely to both
adopt goals related to cooperating and helping coworkers and to pursue those goals
because they believe it will lead to desired rewards. In a direct test of this notion, James,
Hartman, Stebbins, and Jones (1977) demonstrated that psychological climate was related
to employees’ instrumentalities and valences for particular behaviors. Additionally, a
meta-analysis by Parker and his colleagues (2003) also demonstrated significant
relationships between motivation and molar climate. Beyond direct effects on motivation,
climate perceptions may influence performance through an effect on job attitudes.

A number of researchers have argued that climate perceptions impact employee
motivation and behavior through job attitudes, such as job satisfaction and organizational
commitment (e.g., F riedlander & Margulies, 1969; Lawler, Hall, & Oldham, 1974;
Kopelman et al., 1990). Job satisfaction theory and research suggests that a climate

characterized by the promotion of attitudes and behaviors beneficial to employee well—

10

being is likely to result in positive affective evaluations of one’s job (i.e., job satisfaction;
Hackman & Oldham, 1976; Locke, 1976; James & James, 1989). Additionally, given that
many of the common dimensions of climate specifically focus on perceptions that the
environment supports attitudes and behaviors commonly associated with employee well-
being (e.g., involvement, cooperation, growth, role clarity, etc.; James et al., 2008) it is
not surprising that positive climate perceptions relate to job satisfaction (Carr et al., 2003;
Parker et al., 2003). Similarly, climate perceptions are likely to influence organizational
commitment as employees develop obligations of reciprocation and feelings of
identification with organizations perceived as promoting an environment conducive to
their personal welfare (e.g., Mathieu & Zajac, 1991; Meyer & Herscovitch, 2001). This is
also consistent with Blau’s (1964) conception of social exchange theory which predicts
that if employees perceive that the organization is concerned for their well-being, they
will develop an implicit obligation to reciprocate by carrying out relevant job-related
behaviors. The empirical relationships between climate and job satisfaction and climate
and commitment are well-established in the literature (e.g., Carr et al., 2003; Kozlowski
& Hults, 1987; Parker et al., 2003; Pritchard & Karasick, 1973). Job attitudes, in turn,
have implications for employee behavior as satisfied and committed employees are more
likely to engage in greater task-relevant and extra-role behavior while on the job (e.g.,
LePine, Erez & Johnson, 2002; Likert, 1961; McGregor, 1960; Mowday, Porter, &
Steers, 1982; Weiss & Cropanzano, 1996) and less likely to engage in withdrawal
behaviors (e.g. Griffeth, Horn & Gaertner, 2000; Muchinsky, 1977). Furthermore, while
not examining the mediating motivational or attitudinal mechanisms, many studies have

demonstrated relationships between climate perceptions and a number of j ob-related

11

behaviors including organizational citizenship (Schneider, Ehrhart, Mayer, Saltz, &
Niles-Jolly, 2005), continuing education (Kozlowski & Hults, 1987), safety behaviors
(Hofrnann & Stetzer, 1996; Oliver, Cheyne, Tomas, & Cox, 2002), and, more generally,
job performance (e. g., Pritchard & Karasick, 1973). Therefore, there exists ample
theoretical and empirical evidence that climate can influence the performance of j ob-
related behaviors.

Both Kopelman and his colleagues (1990) and Ostroff and Bowen (2000) propose
that the relationships between attitudes, behaviors, and performance are greater at the
group- or organizational-level than at the individual-level. For example, there are many
potential individual responses to dissatisfaction, although some are more likely than
others. A given employee may respond to dissatisfaction by being tardy or absent,
witholding information, or even working harder to improve performance. Likewise, a
satisfied employee could engage in helping coworkers, working harder, or simply
maintaining current-levels of effort. Thus, at the individual-level there are many possible
behavioral responses, with varying implications for organizational performance, to the
experience of satisfaction. However, aggregated across many different employees the net
effect, at the group-level, of satisfaction can be much greater because of the accumulation
of organizationally-desirable or undesirable behaviors (Ostroff, 1992). Similar arguments
could be made for the impact of commitment and motivation on behavior as well as for
the impact of employee behaviors on organizational performance. For example, Mathieu
and Kohler (1990) demonstrated that unit-level absenteeism has effects distinct from
individual-level absenteeism. On this topic, Ostroff and Bowen (2000) concluded, “From

a levels perspective, this suggests a bottom-up process whereby individuals’ attitudes and

12

behaviors combine to emerge into a collective effect that is greater than the simple
additive effects across individuals,” (p. 228-229).

Therefore, there is ample theoretical evidence for the supposition that
organizational climate has a causal influence on organizational performance, through an
impact on employees’ job-related attitudes and behaviors. Additionally, there are a
number of cross-sectional studies demonstrating such relationships (e. g., Borucki &
Burke, 1999; Ostroff, 1992; Schneider & Bowen, 1985; Schneider et al., 2005). Despite
theoretical evidence of a causal pathway from climate to organizational performance,
only a few studies have attempted to empirically establish this causal link.

Empirical evidence of causality. Schneider and his colleagues (1998) examined
the causal relationship of organizational climate for service and customer perceptions of
service quality. Service climate and customer service perceptions data were collected in
1990 and 1992 for over 100 bank branches. The authors examined the causal priority of
service climate by conducting a cross-lagged panel analysis. The results indicated
reciprocal causality between service climate and customer service perceptions, with no
indication that one construct was the greater cause. However, the finding of equal cross-
lags may also be indicative of spuriousness, that is the two variables may be caused by a
third, unmeasured, variable (Kenny & Harackiewicz, 1979). Additionally, the
generalizability of this finding is tempered by the study’s use of only two time periods
and units within a single organization.

Neal and Griffin (2006) examined causal links in the context of safety. These
authors assessed safety climate and individual safety motivation at two time points and

linked climate and motivation to prior and subsequent levels of accidents, at the unit-

13

level, over a five year period. Supporting theoretical arguments that climate influences
employee motivation, the researchers found a significant positive relationship between
group safety climate and safety motivation two years later. However, the researchers
failed to find a significant relationship between group-level climate and subsequent
measures of group-level accidents assessed one and three years later, although the results
were in the expected direction. The lack of significance for climate predicting accidents is
likely due to the small sample available for the group-level analyses (i.e., n=33).
Although this study did not address climate linkages to organizational performance
directly, it does demonstrate a longitudinal link between climate and employee
motivation.

Gelade and Ivery (2003) conducted a linkage study that, while not strictly
longitudinal, is also relevant to this issue of causality. In an effort to demonstrate that
correlations observed between climate and subsequent measures of organizational
performance indicators were not due to unmeasured human resource management
practices, these authors collected data from 137 geographically defined bank branch
clusters. The results revealed that relationships between molar climate and subsequent
organizational performance could not be explained by their common dependence on
human resource management practices. Although this study does help to rule out some
human resource management practices (i.e., staffing level, overtime, and professional
development) as a third variable explanation for climate-performance linkages, there are
other human resource practices that remain to be tested (e. g., employee involvement

practices, incentive compensation, etc.).

14

In summary, the empirical literature on climate-performance linkages at the
organizational- or unit-level provides sparse evidence of the causal pathway often cited in
the theoretical literature. At best, there is evidence for reciprocal causality. A few studies
have addressed the linkage between group-level employee attitudes and organizational
performance. Although employee attitudes are distinct from climate perceptions, the
theoretical position of job-related attitudes as a mediator of the climate-performance link
indicates that evidence of the attitude-performance causal pathway increases the
plausibility of the climate-performance causal link.

Indirect evidence of causality: Research on attitudes. Ryan, Schmit, and
Johnson (1996) examined the causal ordering of group-level employee satisfaction and
indicators of organizational performance across two time periods. Data were collected
from over 140 branches of a financial services organization and outcomes included
turnover, customer payment delinquency, and customer satisfaction. Turnover was
significantly predicted by employee satisfaction over time. Counterintuitively, the cross-
lagged panel analysis of employee and customer satisfaction and payment delinquency
revealed that the causal priority flowed from customer delinquency and satisfaction to
employee satisfaction. In fact, the cross-lagged relationships from employee satisfaction
to subsequent customer delinquency and satisfaction did not reach traditional significance
levels. This suggests that, at least in some circumstances, indicators of organizational
performance cause employee satisfaction. However, the results should be interpreted with
caution given the data were collected from a single organization over only two time

periods.

15

Koys (2001) conducted a similar study examining employee satisfaction,
organizational citizenship behavior, and organizational performance indicators. In this
study, data were collected at two points in time from 28 chain restaurants. Regression
analyses demonstrated that employee satisfaction was a significant predictor of
subsequent customer satisfaction, and customer satisfaction was not a significant
predictor of subsequent employee satisfaction. Additional analyses showed that
managers’ ratings of employees’ organizational citizenship behaviors were positively
related to subsequent year’s profits; again no evidence of reverse causality was observed.
Although the sample size for this study was quite small, the results indicate that causal
priority flows from employee attitudes and behaviors to customer satisfaction.

Schneider, Hanges, Smith, and Salvaggio (2003) examined job satisfaction and
macro indicators of firm financial performance (e. g., return on assets and earnings per
share) longitudinally. Job satisfaction data were collected from a relatively small sample
(n=250) of employees in 35 companies, although some companies did not participate in
some years. The data, analyzed over one-, two-, three-, and four-year time lags, showed
that indicators of financial performance had causal priority over the various aggregated
satisfaction measures. However, there was also some evidence of reciprocality.
Unfortunately, all organizations in this study did not use the same job satisfaction items
nor did a single organization necessarily use the same items over time and thus the results
should be interpreted cautiously.

On balance, the conclusions one can draw from empirical studies examining
attitude-performance links at the organizational level are very similar to those examining

climate-performance links: reciprocal causality is likely and it is unclear if either

16

climate/attitudes or organizational performance have causal priority. Despite the fact that
theory posits climate and attitudes as having causal priority, it is not unreasonable to
expect that employees base their climate perceptions (especially service climate
perceptions), at least partially, on the feedback they receive from customers as customers
themselves constitute a salient feature of the environment. For example, if the customers
are giving negative feedback then it is reasonable that employees will respond to this
feedback by perceiving a lack of service climate in their organization. Heskett et al.
(1997) referred to the relationship between employees and customers in service
organizations as a “mirror” implying that what happens for both has reciprocal influences
like those found by Schneider and his colleagues (1998) and Ryan and her colleagues
(1996). Likewise, the results of Schneider et al. (2003) are not necessarily surprising
when one considers that organizations with greater financial performance are likely to
have greater resources available to devote to human resource practices that yield greater
employee satisfaction (Wright & Gardner, 2003; Wright et al., 2005).

A note on causality. Shadish, Cook, and Campbell (2002), on the basis of John
Stuart Mill’s work on the topic, highlighted the necessary conditions that must exist to
make inferences of a causal relationship: First, the cause must be related to the effect.
Second, the cause must precede the effect in time, that is the cause must be demonstrated
to exhibit causal priority. Finally, plausible alternative causal explanations for the effect
must be ruled out. These conditions are difficult to meet for any topic, but especially
difficult when studying emergent organizational variables, like climate and organizational
performance, that are very difficult to adequately create and manipulate in experimental

lab studies. Prior research on organizational climate has been successful in demonstrating

l7

that the purported cause (i.e., climate) is related to the effect (i.e., organizational
performance; e.g., Borucki & Burke, 1999; Ostroff, 1992; Schneider & Bowen, 1985;
Schneider et al., 2005). Longitudinal, non-experimental studies, utilizing cross-lagged
panel analyses, however, have largely yielded inconclusive (i.e., conflicting) results with
respect to causal priority (e.g., Ryan et al., 1996; Schneider etal., 1998). Gelade and
Ivery’s (2003) study helped to rule out some types of human resource practices as
plausible alternative explanations, but several alternatives continue to exist. Additionally,
Schneider et al.’s (1998) analyses failed to rule out spuriousness as a potential
explanation of the observed correlations.
The Current Study

The current study contributes to the knowledge base on the causal relationship
between organizational climate and indicators of organizational performance by adding a
number of design features absent in the limited previous empirical research. First, the
current study uses data collected from multiple contexts (i.e., vehicle sales departments
and vehicle service departments). Second, the data is collected from multiple
organizations (i.e., dealerships) that carry the same products and have common
performance indicators, but nevertheless are owned and operated independently of one
another. Third, data is collected repeatedly over a period of six years allowing utilization
of cross-lagged panel analyses to examine causal priority. Fourth, a non-equivalent
control group is used to compare the organizational performance of those organizations
completing the climate measure, and subsequent action-planning process, to those

organizations that did not participate in this process over this time period.

18

As previously reviewed, it is often asserted that climate is a cause of
organizational performance (e. g., Kopelman et al., 1990; Ostroff & Bowen, 2000; Ostroff
et al., 2003). The empirical research to date has largely been inconclusive with respect to
this causation and has even provided limited evidence of possible third variable causation
or reverse causality. However, the research to date has also suffered from several
limitations that make it imprudent to hypothesize spuriousness or reverse causality. Thus,

the following hypotheses concerning causal priority are proposed:

Hypothesis la: Department-level climate perceptions will predict customer
satisfaction over time more strongly than vice-versa in both sales and service
departments.

Hypothesis 1b: Department-level climate perceptions for sales departments1 will

predict dealership vehicle sales over time more strongly than vice-versa.

Theoretical treatments of the organizational climate-performance linkage have
largely ignored aspects of reciprocality, even though some researchers did include
feedback loops suggestive of reciprocality in their graphical models (James & Jones,
1976; Ostroff & Bowen, 2000; Ostroff et al., 2003). As mentioned previously, there is
some evidence that customer feedback may represent a salient aspect of the work
environment perceived by employees, and thus may influence employees’ climate
perceptions (Schneider et al., 1998; Ryan et al., 1996). Therefore, the following

hypothesis concerning reciprocality is proposed:

 

1 Similar to Hypothesis la, a predictive relationship for sales would be expected in both vehicle sales and
service departments. However, sales data were unavailable for service departments, therefore, Hypothesis
lb focuses solely on sales departments.

19

Hypothesis 2: Department-level climate perceptions and customer satisfaction

will be reciprocally related over time.

Customer satisfaction is commonly believed to be important to an organization’s
success because satisfied customers are more likely to make repeat purchases (e. g.,
Grewal & Sharrna, 1991; Heskett, Jones, Loveman, Sasser, & Schlesinger, 1994;
LaBarbera & Mazursky, 1983; deerlund, 2002; Yi, 1990) and spread positive word-of-
mouth about the organization (e. g., Kohli & Jaworski, 1990; Maxham, 2001; Maxham &
Netemeyer, 2003; Richins, 1983; 1987; Swan & Oliver, 1989). Both repeat purchases and
positive word-of-mouth have implications for an organization’s future sales. In fact,
word-of-mouth has been shown to have a major influence on individual’s purchasing
behavior (e.g., Amdt, 1967; Grewal & Sharma, 1991; Price & Feick, 1984; Schiffman,
1971; deerlund, 2002). In the current context, word—of-mouth is likely to have an
influence on getting potential customers in the door, which should increase sales. Thus, it
is likely that customer satisfaction mediates the relationship between organizational
climate and sales. However, once a potential customer is in the dealership, other
employee behaviors that may be unrelated to customer satisfaction but influenced by
climate perceptions, such as willingness to negotiate and closing the sale, would
presumably also influence whether or not a sale is actually made. Therefore, it is likely
that customer satisfaction only partially mediates the relationship between climate

perceptions and vehicle sales.

20

Hypothesis 3: The relationship between sales department climate perceptions and

vehicle sales will be partially mediated by department-level customer satisfaction.

Employee surveys in an organization are often accompanied by a subsequent
feedback session which is then followed by an action planning process in which
organizational leaders, and sometimes employees, attempt to develop plans and
procedures for addressing opportunities for improvement highlighted by survey results
(Church & Waclawski, 1998). Such feedback and action planning processes are
sometimes asserted to motivate change by organizational development theorists (French
& Bell, 1995; Nadler, 1996; Nicholas, 1982; Solomon, 1976) and may explain one
mechanism through which organizational climate levels can be increased. While there has
been relatively little research on the effectiveness of feedback for motivating change at
the group- or organizational-levels, there is some empirical evidence that survey feedback
is an effective intervention in relation to increasing employee attitudes and perceptions
(e. g., Bowers, 1973; Brown, 1972; Ryan, Horvath, & West, 2003).

From a theoretical standpoint, the influence of feedback and action planning can
be understood as operating at the first and second stages of Lewin’s (1951) three-stage
theory of change. Lewin proposed that the first stage of organizational change involves
an “unfreezing” process where organizational members are confronted by evidence (e.g.,
survey feedback) indicating the need for change that overcomes their natural inclination
to continue operating in the present way. Lewin’s second stage represents the actual
change process. At this stage, organizational members take steps to identify what exactly

needs to be changed, develop plans for (e. g., action planning), and implement these

21

changes. As changes are implemented, employees engage in the sense-making process of
interpreting the features in terms of their implications for the types of behaviors and
attitudes supported and endorsed by the organization (i.e., employees form new climate
perceptions). Over time, as these perceptions come to be shared, through the mechanisms
discussed previously (e. g., leadership communications, social interaction, ASA), a new
organizational climate is likely to emerge—the final “refreezing” stage of Lewin’s model.
As part of the current study, the management of dealerships participating in the
climate survey process attended survey feedback and action-planning sessions facilitated
by outside consultants. Therefore, in accordance with the assumptions of theorists in the
organizational development literature and the limited empirical evidence available on the

topic, the following hypothesis is proposed:

Hypothesis 4: Climate levels will increase over the period of the study.

If climate is related to customer satisfaction and climate increases over the course
of the study, then one would expect that participating organizations should have greater
customer satisfaction at the conclusion of the study than organizations that do not
participate in the climate survey and accompanying action-planning process. Note that
climate survey data and vehicle sales data is not available for non-participating

organizations. Thus, the following hypothesis is proposed:

Hypothesis 5: Organizations participating in the climate survey and action-

planning process will have greater customer satisfaction at the conclusion of the

22

study than organizations not participating in the climate survey and accompanying

action-planning process.

Research questions. The issue of time is largely unaddressed in the literature on
climate and organizational performance. Drawing on theories of organizational change,
such as Lewin’s (1951) change model discussed above, there is recognition that it takes
time for organizational features to change, time for this change to impact organizational
climate, and time for the climate to impact individual and organizational performance
(e.g., Ostroff et al., 2003). However, there is little discussion of exactly how much time is
required for this process to unfold and what factors might influence the length of this
process.

Specification of this time frame is particularly important for cross-lagged panel
analyses, used in the current study, as the causal lag time period can have a large
influence on the observed results (Kenny & Harackiewicz, 1979). If the lag examined is
too short for the causal process to unfold, any causal effect observed is likely to
underestimate the true causal effect. Similarly, if the lag examined is too long, then the
observed causal effect may be underestimated because the causal impact has dissipated.
Previous researchers have observed cross-lagged effects at two-year lags (Schneider et
al., 1998) for climate and organizational performance and one-year lags (Ryan et al.,
1996) for employee satisfaction and organizational performance. However, in both cases,
the causal lag was determined by the availability of data. Given the lack of theoretical

guidance on the optimal lag for climate and organizational outcomes, the present study

23

investigated these effects at one-, two-, and three-year lags, in the context of the

following research question:

Research mestion 1: Over what lag period(s) do relationships between

organizational climate and performance emerge?

Although there is no concrete theoretical or empirical guidance on which to base a
determination of the causal lag period in the current study, speculation about the
differences in the context of the customer service between the two types of departments
suggests that the causal lag periods between climate and customer satisfaction may differ.
Contextual differences in the intangibility of the service experience and the immediacy of
feedback suggest that changes in organizational climate may impact outcomes for service
departments. more quickly than for sales departments.

The concept of intangibility was originally developed to explain the distinction
between a tangible good and an intangible service (Shostack, 1977). However, services
themselves also vary in their degree of intangibility (Schneider, 1990; Schneider &
Bowen, 1985). The intangibility of a service refers to the extent to which customer
satisfaction is ultimately based on customers’ impressions of the experience (e. g.,
customer satisfaction ratings of service quality in a retail store) versus being judged, at
least partially, on the basis of a physical (i.e., tangible) outcome (e. g., whether a vehicle
was fixed correctly; Ryan & Ployhart, 2003; Schneider, 1990).

As intangibility increases, customers must rely more heavily on the behaviors of

the service provider to form their impressions of service quality (Bowen & Schneider,

24

1985) because objective evidence is lacking. However, organizations have less control
over employees in the provision of intangible services as the intangibility makes it more
difficult to explicitly define the behaviors that employees should demonstrate. This
suggests that changes in organizational climate will take more time to influence customer
satisfaction in an intangible than in a more tangible service context because the specific
behaviors required in the former context will take more time for the organization to
recognize and reinforce and more time for the employees to discover and adopt. Whereas
in a more tangible service context, changes in organizational climate may influence
customer satisfaction more quickly because the specific behaviors supported and
expected by the organization are likely to be more obvious and apparent to both the
organization and the employees.

In the current study, the customer service experience in service departments is
likely to be more tangible in that customer satisfaction is largely inseparable from the
objective and observable outcomes of the experience (i.e., whether the vehicle was fixed
completely, correctly, and on time). In sales departments, the customer service
experience is more intangible as customer satisfaction with the experience is largely
based on customers’ impressions of whether, for example, the experience was pleasant
and whether the sales person was knowledgeable, helpful and courteous. The difference
in the tangibility of the service experience between the departments is also reflected in
the content of the customer satisfaction surveys. The service department survey focuses
on the “service visit overall” and includes content focused on the tangible outcomes of
whether the vehicle was fixed correctly and on time. The sales department survey focuses

customers’ impressions of the overall “purchase and delivery experience” and is

25

explicitly separated from customers’ satisfaction with the outcome of the service (i.e., the
vehicle) both in the wording of the questions and through the provision of a separate
satisfaction questionnaire focused on the vehicle itself. Therefore, organizational climate
may impact customer satisfaction, and vice-versa, more quickly in service departments
than in the less tangible context of sales departments.

Contextual differences influencing the proximity of customer feedback may also
impact the length of the causal lag between organizational climate and outcomes.
Proximal feedback on the success of a service encounter is likely to either reinforce the
behaviors leading to the successful encounter or discourage the behaviors leading to an
unsuccessful encounter (Herrnstein, 1970; Thorndike, 1911). Distal customer feedback,
however, makes it more difficult to systematically determine which behaviors led to a
successful service encounter and which did not. As employees attempt to identify the new
behaviors in line with a new or modified organizational climate, the proximity of
customer feedback may influence how quickly this process unfolds and employees adopt
the new behaviors thereby influencing the causal lag observed between changes in
climate and changes in customer satisfaction.

In the current study, customer feedback in service departments is likely to be
more proximal than in sales departments. Distal customer feedback is received in the
form of customer surveys on a quarterly basis in both departments. However, in service
departments, employees are likely to receive proximal feedback from dissatisfied
customers whose vehicles are not fixed on time or correctly. This proximal feedback
affords the organization and employees an immediate opportunity to learn from the

situation by determining what behaviors led to the negative outcome and also allows for

26

an opportunity to “make it up” to the customer at the moment of dissatisfaction. In sales
departments, however, employees are unlikely to receive proximal negative feedback as
cognitive dissonance theory (Festinger, 1957) suggests that someone who just spent
thousands of dollars on a new vehicle is unlikely to behave in a way that conveys
dissatisfaction in the moment, regardless of whether the customer is ultimately pleased
with the service they received. Thus, differences between the departments in the
proximity of customer feedback may result in a longer causal lag between climate and
customer satisfaction, and vice-versa, being observed for sales than for service
departments.

In line with the above speculation concerning contextual moderators of the
relationship between organizational climate and performance, the following research

question was investigated:

Research Question 2: Do differences between sales and service departments exist

in the causal lag periods between climate and customer satisfaction?

James and Jones (1976) and Ostroff and her colleagues (2000; 2003) note the
external environment is an indirect influence on organizational climate. However, the
discussion ends with this recognition and a further note that climate is rarely studied as an
outcome. One plausible environmental influence is local economic conditions. For
example, it is possible that employees in units residing in economically depressed areas
have lower climate perceptions because the depressed economy may contribute to

employee anxiety about the health of the organization. On the other hand, a booming

27

local economy may create conditions of optimism that contribute to either higher initial
levels or increases in climate perceptions over time. Although not an environmental
variable, organizational size represents an additional contextual variable that may
influence the level and change of climate perceptions. For instance, larger organizations
could have multiple subclimates (Ostroff et al., 2003) that impede the pace of climate
change efforts resulting in more gradual increases over time. The following research
questions will be examined to explore these potential economic and contextual influences

on climate levels and growth trajectories.

Research Question 3: Do differences in the initial levels and grth trajectories of
organizational climate exist across organizations (i.e., dealerships)? If so, do
local economic conditions or organizational size account for some of this

variability across organizations?

Control variables. Local economic conditions and organizational size could also
influence the organizational performance indicators of interest in this study. It is quite
probable that both variables influence the number of vehicles a dealership sells in a year.
Additionally, it is possible that organizational size impacts customer service perceptions
as well. For example, in larger organizations it may be more difficult for customers to
navigate the organization when seeking answers to questions which could negatively
impact customer satisfaction. Therefore, local economic conditions and organizational
size will be controlled for in most analyses, with the exception of those focused on

Research Question 3 where these variables are of substantive interest.

28

Summary

Many researchers and practitioners alike appear to assume that the causal
direction flows from climate to organizational performance. Unfortunately, the existing
empirical research provides only limited evidence both for and against this assumption.
The core purpose of the current study is to contribute to the systematic evaluation of this
assumption by examining causal direction in the context of a longitudinal study with a
number of design features aimed at addressing some of the limitations of prior research.
. Although no single study can definitively prove causation, or even causal priority, the
current study is an important contribution to the body of research on which any

convincing causal claims must rely.

29

Method
Sample and Procedure

Over the six years of the study, complete data were collected from a total of 95
franchise automobile dealerships selling and servicing identical products from a single
automobile manufacturer. At least some data were collected from a total of 599
dealerships over the course of the study. Unfortunately, there was substantial missing
data for over 500 of these dealerships. The missing data primarily resulted from many
dealerships not participating in the climate survey until later years, if at all. Given the
amount of missing data, it was determined that imputation and other methods of dealing
with missing data were inappropriate. Additionally, departments with less than three
respondents were dropped from the analyses due to aggregation concerns. Therefore, the
primary sample for most analyses consists of 95 sales departments and 95 service
departments for which complete climate survey and outcome data were collected across
the entire six years of the study. An additional sample of 44 sales and service departments
for which only customer satisfaction data were available will be used as a control group
for Hypothesis 5.

Climate survey data were collected at each of the four collection periods from
approximately 1,200 sales department employees and 3,000 service department
employees within the dealerships (see Table l). The average number of sales department
employees within each dealership responding to the climate survey across measurement

periods was approximately 13 with a range of three to 42 employees. The average

Table 1: Summary of Overall Sample Sizes for Each Department and Each Year

 

 

 

Year
2000 2001 2002 2004
Sales Departments 1,226 1,194 1,239 1,179
Service Departments 33,190 2,999 3,045 2184

 

30

number of service department employees within each dealership responding to the
climate survey across measurement periods was approximately 32 with a range of four to
131 employees. No information on the demographics of respondents or response rates
within the dealerships or departments was available.

Data were collected via paper-and-pencil measures in 2000 and 2001, but were
collected via a secure intemet site in 2002 and 2004. During 2000 and 2001,
representatives of an external consulting firm conducted data collection at each
dealership. All employees were requested to attend the data collection sessions during
normal working hours, but participation was voluntary. Likewise, during 2002 and 2004,
all employees were provided with instructions on how to access the web-based survey
and were requested, but not compelled, to complete the survey during working hours.

Data were collected in 2000 early in the lSt quarter (i.e., January and February).
For subsequent years, data were collected towards the end of the 4th quarter. This resulted
in some differences in the amount of time between data collection periods, with the time
lapse between 2000 and 2001 being approximately 2 years, between 2001 and 2002 being
1 year, and between 2002 and 2004, again, being approximately 2 years. In order to
account for these differences, and allow for alignment between time periods in which
data were collected for the other substantive variables in this study (i.e., customer
satisfaction and vehicle sales), climate data collected in 2000 were considered Time 1,
2001 as Time 3, 2002 as Time 4, and 2004 as Time 6. Climate data were unavailable for
Time 2 and Time 5. Table 2 summarizes this information for each variable.

Survey results were first reported back to the management of each dealership by

consulting firm representatives. The survey feedback sessions included formal

31

presentation of the results, normative comparisons to other organizations and dealerships,
and action-planning procedures. The action-planning focused on identifying opportunity
areas, brainstorming potential reasons for the results, and developing short- and long-
term plans for addressing the underlying issues reflected in the survey results.
Management had responsibility for making the results available to non-management
employees.

Customer satisfaction surveys were mailed to all customers purchasing or having
a vehicle serviced. Surveys were mailed to customers by, and returned to, an independent
consulting firm. On average, 41 surveys were available, on a quarterly basis, for each
dealership’s sales department and 85 surveys were available for each service department.
The response rates were approximately 50% for sales departments and 35% for service
departments. In order to more closely align the time periods in which customer
satisfaction and climate data were collected, the customer satisfaction data from
approximately two quarters preceding and succeeding collection of the climate survey
data were averaged to construct this variable for analyses at each time period. Therefore,
the data for Time 3, for example, was composed of customer satisfaction data from the
third and fourth quarters of 2001 and the first and second quarters of 2002. For Time 1,
however, only data from the two quarters succeeding (i.e., first and second quarters of
2000) climate data collection were available (see Table 2).

The number of new vehicle sales for each dealership was available on a quarterly
basis. On average dealerships included in this study sold approximately 93 vehicles
(Median = 77), each quarter. No vehicle sales information was available for 2002. Similar

to the procedure used for customer satisfaction, the sales data from the two quarters

32

 

NO ._.0 Hmoom NO #0 Hvoom . . . . NOIWFO Hroom . . mm_mw
v0 .mO “vooN «.0 .m0 ”moom NO _.O .mooN v0 m0 .Foom V0 .mO ”oooN NO F0 .oooN m_oEm>
NO #0 nmoom NO .FO H“VooN NO #0 HmooN NO ._.0 HNooN NO .FO U_.ooN NO ._.0 .oooN cozoflwzmw
v0 .mO HwooN v0 .mO Hmoom v0 .mO HNoom v0 .mO Hroom v0 .mG HoooN . .wEoumao

v0 u.VooN <\Z v.0 HNooN v0 H_.ooN (\2 v0 Hooow 05.9.2.0

0 mg» m. we: 4 as: m 9:: N 9:: F 95
moored 95... 53260 San ho meE3m ”N cam...

33

preceding and succeeding collection of the climate survey data were averaged to
construct this variable for analyses at each time period. For Times 1 and 4, only sales data
from the two quarters approximately succeeding climate data collection (i.e., first and
second quarters of 2000 and 2003, respectively) were available. For Time 3, only data
from the two quarters approximately preceding climate data collection (i.e., third and
fourth quarters of 2001) were available (see Table 2). The distribution of sales across
dealerships exhibited a negative skew, so sales data were logarithmically transformed
prior to analyses to account for this skew.
Measures

Climate survey. The proprietary climate survey used in this study consists of 60
items assessed using a 5-point likert-type scale ranging from strongly disagree to strongly
agree. This instrument was designed to measure the key aspects of Denison’s (1990)
model of effective organizational cultural values which has theoretical roots in the human
relations movement (e.g., McGregor, 1960), Schein’s (1985; 1992) culture theory, and
the Competing Values Framework (e.g., Cameron & Quinn, 1999; Quinn & Rohrbaugh,
1983). Drawing on each of these theories and research streams as well as on his own
extensive quantitative and qualitative research, Denison proposed that effective cultures
are characterized, at more visible levels, by values and practices focusing on employee
involvement, internal consistency, adaptability, and a clear mission (e.g., Denison, 1990;
Denison & Mishra, 1995). While some theorists have argued that the alignment, or fit,
between an organization’s culture and its environment is necessary for organizational
effectiveness (e.g., Cameron & Quinn, 1999; Perrow, 1970; Quinn & Rohrbaugh, 1983),

Denison argues that effective organizations have all of these cultural values and that the

34

balancing and simultaneous pursuit of the competing demands these values represent is
the key to organizational effectiveness (Dension, 1990; Denison & Mishra, 1995).
Recognizing that cultural values, and the deeper-level assumptions on which they
are based, are difficult to assess quantitatively in organizations (Ashkanasy, Broadfoot, &
Falkus, 2000; Schein, 1990; 2000), Denison developed the perception-based measure
used in the current study to assess the climate-level manifestations of these assumptions
and values. This method of measurement is also consistent with the notion that
employees’ perceptions of organizational features, and the organizational features
themselves, are important mediators of the impact culture can have on organizational
performance (Kopelman et al., 1990; Ostroff et al., 2003). Denison’s measure is
organized around employees’ climate perceptions reflecting the four cultural values
identified in Denison’s model. The employee involvement dimension assesses employee’s
perception of the work environment as encouraging empowerment, team-based
cooperation, and individual learning and development (e. g., “Decisions are usually made
at the level where the best information is available”). The internal consistency dimension
measures employees’ perceptions of organizational features as promoting a clear set of
espoused values, agreement on these values, and the individual and inter-departmental
coordination that should arise from this common and agreed upon set of values (e. g.,
There is good alignment of goals across levels”). The adaptability scale contains items
focused on assessing employees’ perceptions that the work environment is oriented
toward Ieaming from its competitors and customers and has practices and procedures that
promote flexible and adaptive responses at both the organizational- and employee-level

(e. g., “Customer comments and recommendations often lead to changes”). The mission

35

dimension assesses employees’ perceptions that the organization has a clearly articulated
strategic direction that provides context for action and goals against which progress can
be tracked (e. g., “There is a long-term purpose and direction”).

Although Denison’s measurement dimensions are based in culture theory, the
dimensions exhibit overlap with some existing dimensions and taxonomies of molar
climate (e.g., James & James, 1989; Ostroff, 1993). For example, Ostroff (1993)
developed a comprehensive taxonomy that included three higher-order facets. Ostroff’s
affective facet focuses on involvement and social relations among workers, elements of
which are reflected in both Denison’s employee involvement and internal consistency
dimensions. The cognitive facet, focusing on growth, innovation, autonomy, and intrinsic
rewards, is partially reflected in Denison’s dimensions of employee involvement and
adaptability. However, Ostroff’s instrumental facet and Denison’s mission dimension
each seem to be unique to the particular models.

Denison and his colleagues (2006) assessed the factor structure of this measure on
the basis of over 35,000 employees from 160 different organizations. As expected the
four factor model provided good fit to the data (RMSEA=.048, CFI=.98), but the latent
factors had an average intercorrelation .90. Consistent with prior research looking at
molar climate dimensions, this degree of intercorrelation at the latent-level suggests that a
higher-order general climate factor underlies responses to this measure (e. g., James &
James, 1989; Gelade & Ivery, 2003; Parker, 1999). James and James (1989) argued that
the emergence of a general climate factor should be expected given that climate
perceptions are based on an individual’s consideration of the implications organizational

features have for the self. A confirmatory factor analysis using all of the available data,

36

across all time periods, from the current study demonstrated that a single hi gher-order
factor fit the data (RMSEA=.04, SRMR=.03, CFI=.92, TLI=.91) equally as well as the
four factor model (RMSEA=.04, SRMR=.03, CFI=.92, TLI=.91). Therefore, for the
purposes of this study, the Denison climate measure will be interpreted and analyzed as a
single molar indicator of a climate for effectiveness.

In order to justify aggregating individuals’ perceptions of climate to the
organizational-level, it is first necessary to show that a minimum degree of consensus

exists among group members (Bliese, 2000). Statistical justification of this consensus

relies on demonstration of adequate within-group agreement (ngU))’ interrater reliability

(ICCl), and group mean reliability (ICC2). The rwgo) values were computed using

equations for multiple-item scales provided by James, Demaree, and, Wolf (1984). When

using the rng) as an index of within-group agreement, it is necessary to specify a null

random response distribution against which the observed distribution of ratings is
compared. Although most researchers tend to use only a uniform null distribution, many
researchers have argued that other plausible null distributions should be used for
comparison as well (e.g., Bliese, 2000; James et al., 1984; Kozlowski & Hattrup, 1992).

Evidence of a slight negative skew was observed in the current dataset when examining

individuals’ climate response frequencies, therefore rwgo) estimates were computed

utilizing both a uniform and a slightly negatively skewed null distribution. The average

rng) values, across years, observed for sales departments were .97 using a uniform and

.88 using a slightly skewed null distribution, which represent the upper and lower limits

37

of the actual rng) values (Kozlowski & Hults, 1987; see Table 3). For service

departments, the average rwgo) values were .98 using a uniform and .92 using a slightly

skewed null distribution (see Table 3). All of these values are above the .70 cutoff
commonly referred to (e. g., Schneider et al., 2003), but rarely cited (Lance, Butts, &
Michels, 2006), in the literature.

Interrater reliability was examined using equations provided by Bliese (2000) for
ICCl. The average ICCl, across years, was .18 for sales and .14 for service departments
(see Table 3). These values are above the median ICCl of .12 observed in the
organizational literature by James (1982) and within the range (.05 to .20) reported by
Bliese (2000). The reliability of the group means was also examined using equations
provided by Bliese (2000) for ICC2. The average ICC2, across years, was .74 for sales
and .84 for service departments (see Table 3).

In summary, the within-group agreement, interrater reliability, and group mean
reliability observed in the current data provides ample justification for aggregation.
Therefore, individuals’ climate perceptions will be averaged to obtain organizational
climate values for each dealership.

Customer satisfaction. Mean customer responses to a single customer
satisfaction survey item were available on a quarterly basis for each sales department:
“Based on your overall purchase/lease and delivery experience, how satisfied are you
with XYZ Dealership.” Similar mean customer responses to a single item were available
for each service department: “Based on this service visit overall, how satisfied are you

with XYZ Dealership?” Customers made ratings on a 4-point likert-type rating scale

38

S. B. 8.
3. cm. S.
moo. 62 News.

vooN

F

8:39.56 .832... 29:82 >2.ng

Season .3: 85:5 F
S. 8. 2. 8. 8. S. 2. S. 8. E. 2. mm. 8. 3525800228
8. E. S. a. 8. E. 2. E. 3. E. 2. 8. 3. acmsemaoomgmm
em}. 82 60. New}. Fame. 8962 New} ma} moo. So. New} Fem}

NooN room ooom
cmo>

 

Lmo> comm ucm EoEtmaoo zoom .9 83¢.sz cozmmflmme. 055:0 Co meEsm um cam...

39

ranging from “Not At All Satisfied” to “Completely Satisfied.” Customer satisfaction,
like vehicle sales, will be treated as a descriptive outcome variable rather than an
aggregated construct that represents the shared perceptions of customers.

Control variables. The number of employees within each dealership was not
available. However, all employees were provided with an opportunity, and encouraged,
to fill-out the climate survey at each administration. Therefore, department size was
approximated by averaging the number of respondents from each department across all
four climate survey data collection points. Unemployment rates for the statistical
metropolitan areas in which each dealership is located were used as indicators of local
economic conditions. This information was obtained from US. Bureau of Labor Statistics
website (http://bls.gov).

Analytic Strategy

Figure 1 depicts the overall analytic strategy used to examine Hypotheses 1a, 1b,
2, 3, and Research Questions 1 and 2. The analyses progressed through four stages: (1)
evaluation of measurement invariance over time and across departments; (2) evaluation
of the full cross-lagged reciprocal model; (3) evaluation of simpler competing models;
and (4) evaluation of the consistency of the models across time. Stages 2 through 4 were
repeated separately for each department, and, for the sales department, for each outcome
(i.e., customer satisfaction and vehicle sales). This analytic strategy imposes and
compares increasingly strict assumptions about the underlying relationships of the
variables depicted in the model and allowed for examination of the substantive

hypotheses as well as differences in the hypothesized relationships over time.

40

Figure 1: Analytic Strategy

 

 
     

Allow selected parameters
to vary across time and/or
groups.

Stage 1. Evaluate measurement Invariance
across tlme and departments.

Acceptable
Fit?

 

 

 

 

 

 

Model modification and
respeciflcatiort

     

Acceptable

Stage 2. Evaluate full cross-lagged model. Fit?

 

 

 

 

 

 

Sig. Worse
Fit?

   

Stage 3. Compare competing models. Select simpler model.

 

   

 

Retain comparison model.

 

 

L

 

t

 

 

 

 

 

Sig. Constraln selected
Stage 4. Zest consistency of model across Differences? No , parameters to be identical
me. across time.
Yes

 

Estimate parameters
independently at each time
point.

 

 

 

 

A

 

 

Final Model

Valid inferences about the similarity or differences of structural relationships are
contingent upon evidence that the constructs of interest have the same underlying
meaning and are measured consistently across time and between groups. Thus, the first
stage in the analytic process (Stage 1) involved investigation of the invariance of the
climate measure across departments and time according the procedures outlined by

Vandenberg and Lance (2000).

41

Cross-lagged panel analyses began with the full cross-lagged reciprocal model
(Stage 2). The key features of this model are: (1) cross-lagged paths from climate to
subsequent outcomes (i.e., satisfaction or sales); (2) cross-lagged paths from outcomes to
climate; (3) autoregressive paths within constructs; and (4) residual correlations between
climate and outcomes within each measurement occasion. The cross-lagged paths
estimate the hypothesized relationships. The autoregressive lags control for prior levels of
the variable of interest when estimating the impact of the cross-lagged variable. The
residual correlations were included on the basis of evidence, provided by Anderson and
Williams (1992), that failure to account for these correlations can lead to biased estimates
of the cross-lagged effects. At this stage, all structural paths were allowed to vary across
time periods and departments. The impact of wave-skipping autoregressive lags (e. g.,
Time 1 to Time 3, Time 2 to Time 4, etc.) were also investigated at this stage on the basis
of findings in previous cross-lagged research that these paths often improve model fit
(e.g., Hays, Marshall, Wang, & Sherbourne, 1994; Madon, Willard, Guyll, Trudeau, &
Spoth, 2006).

Stage 3 models compared competing, more parsimonious, models to the full
cross-lagged reciprocal model by constraining selected cross-lagged paths to zero. The
first model examined the alternative that climate influences outcomes over time, but
outcomes have no direct effect on climate, by constraining the cross-lagged paths from
outcomes to climate to be zero. The second model examined the converse where
outcomes influence climate over time, but climate has no direct effect on the outcomes.
The final model in this stage, an autoregressive null model, examined the alternative that

no direct causal relationships exist among the variables by constraining all cross-lagged

42

paths to zero.

Assessment of the consistency of the models over time was examined at Stage 4.
The first set of models at this stage constrained the autoregressive lags within a construct
to be equal over time (e. g., all one year autoregressive lags within a construct were
constrained to be equal). This model examined whether, for example, customer
satisfaction consistently influenced subsequent measures of customer satisfaction to the
same degree, or whether these relationships changed over the period of the study. Next,
consistency of the cross-lagged paths was assessed by constraining corresponding cross-
lags to be equal over time. This model examined whether, for example, the impact of
climate on subsequent customer satisfaction increased, decreased, or remained stable over
time.

Stages 2 through 4 were repeated for the different time lags over which the cross-
lagged climate—outcome relationships may emerge (i.e., 1 year, 2 year, and 3 year lags) in
order to investigate Research Questions 1 and 2. As noted previously, specification of the
appropriate causal lag time period can have a large influence on the results of cross-
lagged panel analyses. If the causal lag examined is too short for the causal process to
unfold, any causal effect observed is likely to underestimate the true causal effect.
Similarly, if the causal lag examined is too long, then the observed causal. effect may be
underestimated because the causal impact has dissipated.

Stages 2 through 4 were slightly modified for the examination of Hypothesis 3
(i.e., customer satisfaction partially mediates the relationship between climate and vehicle
sales) to account for the addition of a third set of variables. Stage 2 examined the

hypothesized partial mediation model. Stage 3 examined a full mediation model and

43

more restricted models constraining selected cross-lagged effects to zero. Stage 4, again,
examined consistency of the resulting model across time.

The best-fitting model resulting from each stage served as the initial comparison
model for the subsequent stage of analyses. On the basis of Hu and Bentler’s (1998;
1999) recommendations, model fit was assessed using the following criteria: SRMR S .08
and (RMSEA S .06 or CFI Z .95). Chi-square statistics are also reported, and used for
model comparisons, but were not considered in assessment of overall model fit.
Comparisons between nested models were assessed with the chi-square difference test
(Bentler & Bonett, 1980). All measurement and structural models were estimated using
Amos 7.0.

Hypothesis 4 was examined by means of a dependent groups t-test comparing
Time 1 climate perceptions to Time 6 climate perceptions. Hypothesis 5 was examined
with repeated-measures AN OVA with participation in the climate survey process as the
between-subjects factor. Research Question 3 was examined via longitudinal growth

modeling analyses with time nested within departments using HLM 6.04.

44

Results

Descriptive Statistics

Tables 4 and 5 contain means, standard deviations, reliabilities, and
intercorrelations for all variables. An increasing trend is evident for both mean climate
perceptions and satisfaction over time and across departments, though not for vehicle
sales. For sales departments, the pattern of correlations between climate perceptions and
satisfaction with sales were generally in the hypothesized direction (i.e., positive).
However, the correlations with vehicle sales were both slight and largely not significant.
For service departments, the pattern of correlations between climate and satisfaction were
generally in the hypothesized direction and significant. Unemployment figures were
largely unrelated to both sales and satisfaction, across both departments, indicating that
unemployment is not a necessary control variable for subsequent analyses. Department
size was consistently positively related to vehicle sales indicating its importance as a
control variable for subsequent analyses involving vehicle sales.
Measurement Invariance

Measurement invariance (Stage 1) of the climate survey across time and
departments was examined at the department-level using item parcels defined by
Denison’s four dimensions (i.e., Involvement, Consistency, Adaptability, Mission; see
Appendix for item parcel covariance matrices) in accordance with the procedures
outlined by Vandenberg and Lance (2000), with one exception. Some authors have noted
that the omnibus test suggested by Vandenberg and Lance as the first step in establishing

invariance can lead to erroneous conclusions (Byme, 1998; Raju, Laffitte, & Byme,

45

._.o.v0 0m 0585ch Em 3:00-22. 5 0:250:00 .mova .0 0:85:90 9m 0.00 5 22.9250 .082

 

.R. on. ow. “V. 3... oo. 3.- oo.- vor #0.- mo.- nor Nor oo. oo.- oo.- 3.- oo.- oo. oo.- 2.- oo.- voo mo.v ooEF EoE>o_m.Em:D .MN
-- 3. Vs. ox. on. No. oo.- oo.- nor Nwr oo.- oo.- of- vo. 3.- nor oo.- For 00. no.- mo.- mor who om.m mmEF 0208.33.55: .NN
- no. No. 8. Nov 9..- o_..- 9.. 0N.- 3r 9.. NN.- no. 9.- No.- oo.- mor No. oo. no. mo. woo woo voEF 305303895 .FN

- 3. on. No.- 3.- oo.- oo.- 9.. oo.- Nor _.N.- oo. 9..- —.o.- oo.- oo. or. no. or. oo. mod 36 moEC. E0E>2QE®¢D .ON

- on. No. oo.- not oo.- 5...- E..- of- NN.- mo. of- vor oo.- 5. N. or. 5. mo. #50 56 No.5... EmE>o_an:D .ow

- oo. oo. oo. No.- mwr oo.- nor NN.- oo.- 5.- oo.- oo.- _.o. oN. no. G... Fo. who who PoEE. EmE>oEEmcD .ww

- or. NN. oN. NN. NN. NN. 3.- 3.- opr mo. co.- oo.- oo.- oo.- oo. owr ofio oon 0N5 EoEtmqmo NF
- oo. oo. No. No. 3. NN. t. oN. 3. oN. _.N. ov. Fo. vN. :. oNo no... ooEE. 86w .9.

- no. no. 3. no. N. t. or. or. oN. ON. oo. oo. 9.. oo. cod 5... no.5... 92mm .9.

- no. no. No. 2.. op. or. oo. 0N. FN. oo. No. 2. oo. woo oo.? VoEFQOw .3.

.- oo. oo. NN. 3. 1N. or. on. on. no. 0o. 3. no. omo nmé moEFmew .9.

I no. or. or. _.N. NF. 2... QN. oo. 5.- 3. mo. Poo Poé NoE_._.mm_mw .N—.

I NN. 3. NN. 3.. X... mN. o... 5. or. no. Nmo no.— 005:. mmfim .5.

- 5‘. mV. NM. NN. 9.. of _.N. NN. 9.. mod Foam 09:? cozompmzmw mmfiw .o_.

.. 00. 00. «0. 3. 00. 2. 00. a. 00.0 00.0 most 00000308 8.8 .0
-- B. E. 0... 00. 0m. 00. 00. :0 050 «05; 02805000 8.8 .0

t a... 00. «0.- 0.- «a. «N. 00.0 0k...” 005: cofiewfim 8am A

t 00. 0.. N0. 00. 9. 30 E0 Nos: 0000000000 8.00 .0

.- 3. 2. 2. t. 00.0 :0 Postcesaaammoaw .0

$00 3. 8.. 00. 3.0 40.0 00.5 995.0 .0

000.0 00. S. 8.0 5.0 «as: 2025 .0

80.0 00. 3.0 0&0 nee: 20200 .N

80.0 00.0 50.0 2.5: 2055 .F

 

NNwNoNowoFtoevaFmFNr:o_.owhomvm N PowcmmS.

 

 

mocmEtmaoo 02mm - 00002208090. new mesmzmum 909.0024 ”v 2004

46

._.o.v0 00 0:005:90 0.0 00:02-23 0. 0.20.00.02.00 .002. 00 0:005:90 0.0 0.00 0. 05.00.0200 .0002

 

E. 00. 0... .0. .0. 00.- 3. 3.- 00.- 0..- v..- 0..- 00.- 00. 00. 00. 3.0 00... 005.023.630.05 .0.
-- .0. 0... 0.. 00. 00.- 00. 0..- 00.- 0..- 00. 00.- 00.- 0..- 00. 00. 0.0 00.0 005.200.520.005 .0.
I .0. 00. 00. 0..- 00. 3.- 00.- ...- 00. 00. 00.- ...- 0.. 00. .00 00.0 «05.00005203500: .0.

.- 3. 0.. .0: 3. 00.- .0.- 00.- 0.. .0. 0.. .0.- 00. 00. 00.0 00.0 005.500.020.000: .0.

.. 00. 00.- 00. 00.- 00.- 0..- 0.. 00. .0. 00. .0. 00. K0 .0... 005.200.50.305 .0.

-- 00. 00.- 0..- 0..- 0..- 3. 00. 00. 00. 00. 00. 0.0 0.0 05.500.00.305 .0.

.- 00.- .0.- 00. .0. 00.- 0..- t. 00. 00. 3. 00.0. 00.0 02000200000 ...

-- 00. 00. 0.... 00. 00. 0.. 00. .0. t. 0.0 .00 00...: 02.00.0000 02200 .0.

-- .0. 0... 00. 00. 00. .0. 00. 0.. 0.0 00.0 00...: 02.00.0000 02200 .0

.- 00. 0... .0. 00. 00. 00. .0. 0.0 00.0 00...: 02.00.0000 02200 .0

.- 00. 00. 0.. .0. .0. .0. 0.0 00.0 00....» 0200.200 02200 .0

-- 0.. 0.. 0.. 0.. 00. 0.0 00.0 00:... 02.00.0000 02200 .0

.- 3. ... 00. 0.. 0.0 00.0 .05.. 0200.200 02200 .0

.000 0... 00. 00. 00.0 .00 00:... 0.02.0 .0

.000 0... 0... 00.0 00.0 00.20 0.02.0 .0

.000 0... 00.0 00.0 00...: 0.0.2.0 .0

.000 00.0 .00 .00... 0.0.2.0 ..

 

0.0.3.29330 m n o 0 v 0 N.owcm0_2
0EOEtma0o 00.20w - 000.020.200.005 0:0 0000.005 02.9.0000 .0 0.000

47

2002). Therefore, this step was excluded from interpretation, but, for completeness, the
results of the omnibus test (Model 0) are included, along with the results of the other
models, in Table 6. Correlations among measurement errors of the same observed
variables across all measurement occasions were included in the models because repeated
measures of the same variable generally results in correlated measurement errors (Bollen,
1989; Kessler & Greenberg, 1981).

Examination of equivalence of factor patterns (i.e., configural invariance) with
factor loadings freely estimated across both time and departments indicated that the
model fit the data well (Model 1). The second set of measurement models assessed metric
equivalence by constraining the factor loadings to be equal across departments (Model

2a), time (Model 2b), and both time and departments (Model 2c). Neither Models 2a,

2 2
Xdifl(12,N = 190) = 12.87, ns, nor 2b, x diff (18, N = 190) = 18.1], ns, fit the data

significantly worse than the configural model. Furthermore, Model 2c did not fit the data

2 ' 2
significantly worse than 2a, x difl (9, N = 190) = 11.74, ns, nor 2b, x diff (3, N = 190) =

6.50, ns, indicating that the latent constructs are measured similarly across both
departments and time. Next, scalar invariance was examined by constraining intercept
terms to be equal across departments (Model 3a) and time (Model 3b). The results of the
scalar invariance analyses indicated a significant reduction in fit of the models to the data

2
across departments, x diff (16, N = 190) = 123.60, p < .01. Similarly, the failure of

Model 3b to converge suggests that scalar invariance is not present across time in the
responses to the climate survey across both departments and time and that no further tests

of more restrictive forms of measurement invariance are justified.

48

0902.00 00 00:0“...

..o.va .0. 0:005:90 0.0 00:02-28 0. 00:_0> 0.0000220 .mo.v0 .0 0:005:90 0.0 0.00 c. 000_0> 05:00-20 .0002

 

 

 

I I I oo.. - - I - 00E: - 00:252.. 550m no

or awe-MN: 0N 582 no: madam oo. no. oo. . 02.08.5000 - 00.5_5>:_ 550m mm

m omo 0N .0022

o 0.5. 3 0N 72005. oo.. NNKMN oo. oo. No. 0E_._. w EcmEtmaoo - 00:952.. 0305. 0N

or _._..o_. _. .0005. mm: ns.omN oo. oo. No. 0E_._. - 00:352.. 050—2 0N

N: hoNF _. 5.00.2 oo: omfiVN oo. oo. No. flcwEtmamo .- 00:952.. 050.2 0N
I we: NoNNN oo. oo. No. 0:5. w ficmEtmamo - 00cmtw>5 5.05550 _.

I I - oNP umfimh vo. oo. oo. .00:- 0:n_:EO o

8cw_..m>c_ wCGEQthmw—Z ”—- 01mg”
.5 < 0525 _0Uo_2 .b 052% <ww_2m_ E0 12mm .2022

-20 < 000.5950

.20

 

0002004. 09.0.52: 0205050005. .0. 000.05 :0. .0 Ewe—20w .0 0.00..-

49

Although full measurement invariance was not supported, the existence of metric
invariance provides sufficient justification for proceeding with the structural analyses of
interest in the present study (Bollen, 1989; Cheung & Rensvold, 1998). Consequently, all
subsequent analyses included longitudinal equality constraints on factor loadings.
Cross-lagged Panel Analyses

Customer satisfaction.

Service departments. Hypothesis 1a, proposing that climate perceptions would
predict customer satisfaction over time more strongly than vice-versa, and Hypothesis 2,
proposing a reciprocal relationship, were examined according the staged approach
discussed above. Summaries of model fit and comparisons for all one-, two-, and three-
year lag models are presented in Tables 7, 8, and 9, respectively.

The initial full cross-lagged reciprocal model for one-year lags (Model 1) did not
satisfy all of the criteria for acceptable fit of the model to the data (i.e., SRMR > .08).
However, an alternative model (Model 1a) including wave-skipping autoregressive lags
(e. g., Time 1 to Time 3, Time 2 to Time 4, etc.) did result in acceptable fit, and fit

significantly better than the initial model without the wave-skipping autoregressive lags,

2
x difi” (6, N = 95) = 38.29, p < .01. As would be expected, a similar pattern of results was

2
observed for both the two-, x djfl (6, N = 95) = 39.10, p < .01, and three-year lag

2
models, x diff (6, N = 95) = 38.86, p < .01. Therefore, Model 1a was retained as the

comparison model for subsequent stages of analyses for each of the different lag periods.

The results for the one-year lag models diverged from the two- and three-year lag models

50

.5.v0 .0 .0005090 0.0 00._0..-0_0n 0. 00:.0> 05:00.00 .mo.v.. .0 0005090 0.0 0.0.. 0. 00:_0> 05:00.00 who...

 

 

 

 

 

 

N N. 0 08.2 8. 00.08 0... 00. 0... 0252.8 0.5.2.. 000.0090 .080 0
0 3.0. N .0022 0.: 3.03 0... 0... B. 000.. 020020222 .030 0
0.0... 000.04. .0005. .0 500.0.0000 .00 .. .0 0m0.m
m 3.2 N .822 N0. 8N5 8. 00. 3. 08.080 o... .32 9.08.0292 ..
0 .00. m. .822 0... 3.00N 0... 00. 0... .00 0.05.0 A- .00 00.2030 - 880.0020 0
0 «N... 2 08.2 at ENNN 0... 00. 0... 2:0 .00 052030 A- 0.05.0 - 880.0090 N
0.000.). m0..000.00 050.000 .0 0W0.w
-- .. .. . . . . 003 0

m... cm 3N mo mm 3 0200050550. 00.00.0505; 0..>> 00.0090 =0“. .
0 0N0... 0. 08.2 N... 2.00N S. 00. N0 00.0090 .3. .
.0005. 003040090 =0“. 0.0:_0>w ”N 0m0.m
.0 0 98.0 08.2 .0 0.0.00 «0.0.20 .00 0.200 08.2
-_..0 4 8003050 -_..0

 

0.000.). 003040090 .00 >000 00..00.0..0m 5.00.000 000 0.0.5.0 .00Et0000 8.00m .0. 000.00. ..u. .0 >50..an K 0.00...

51

.500 .0 5000.090 0.0 00:0..-0_00 0. 00:.0> 05:00.00 .mo.V0 .0 5000.090 0.0 0.00 0. 00:.0> 05:00.00 MED).

 

 

 

 

 

 

0.000..00< .02 302.0000 0.0.5... 000.0020 .0:0m. 0
0 00.0. e .0022 .0. 0.00N 0... 00. N0 80.. 02.08.0035,. .080 0
02.: 0020.0. .0005. .0 2000.0.0000 .00 .. .0 0005
m «0.0 m .0005.
0 N00 N .822 N0. 8.N..N 8. 00. E. 000-0020 o... .32 0208.00.92 ..
m 00.0 0. .0005. 0: 00. SN oo. 00. E. 200 0.00.:0 A- .00 505.000 - 0000070020 0
m. 00.0 0: .0005. at. 00.000. 00. 00. mo. 5.00 .00 50.0.0:0 A- 0.02:0 - 0000070020 N
0.000.243.5950 0500.00 .0 000.0
I I I . . . . 000.. 0
0... 00 cam 00 mm 00 0200200554. 00.000505; 0..>> 00.0020 ..:n. _.
0 2.0.0 0: .0005. N0. Sufism mo. 3. 0.. . 00.0020 ._:n. F
.0005. 0000000020 .5“. 0.0:_0>m “N 000.0
.0 0 05:00 .0005. .0 05:00 .0me0 .00 15.00 .0005.
-20 < 8202.00 -20

 

0.0005. 003000020 .00>.03... 02.00.0000 5.00.0:0 000 0.02:0 .00....5000 00.200 .0. 000.00. E .0 250.050 “m 0.00 0.

52

.Fo.v0 .0 .0000.00.0 05 00..0..-0.00 0. 00:.0> 05:00-00 .mo.v0 .0 .50....00.0 05 0.00 0. 00:_0> 05:00-00 ..m...02

 

 

 

 

 

 

 

208.2,. .oz 22.28 25.2.. 32.0020 .0000 0
0 00.8 30002 a. 0.000 00. 00. NF. 000. 02000090550000 0
00.. .. 00204. .0005. .0 20.0.0000 .00 0. .0 000.0
0 .00 0. .0002 00. 00.000 00. 8? :. 002-0020 8:52 02.000.00.220
N 0.0 0. 0022 00. 00.3.0 00. 00. :. 2:0 22....0 A-000.00.20.00-0082-080 m
N 00.0 0. .0002 00. 00.000 00. 00. 8. 200 00.00.2000 42052000002090 0
0.0005. 000000.00 050.000 Hm 0010.0
I I I . . . . 000.. 0
at 0» 00m mo 00 00 0>.000.00.0.:< 00.00.00-053 0..>> 00.0020 __:0 F
0 00.2.. 0:000: 00. 00.2.0 3. B. 0.. 02080.2.
.0005. 0000000020 ..:n. 0.0:.0>m. 0000.0
.00 0.000 0022 .0 0.0000 50.20 0.0 0200 0022
-200 002.8500 -00

0.0022

0000000020 50>-00.0.. 02.00.0000 50.0.0:0 000 0.00...0 .000..500o 00.200 .0. 000.00. ..... .0 252050 .0 0.00...

53

for subsequent stages of analyses, so the results of all one-year lag models will be
discussed next followed by a discussion of the two- and three-year lag models.

Models 2, 3, and 4 compared the full cross-lagged models from Stage 2 analyses
to the alternative, more parsimonious Stage 3 models. Model 2, which constrained the
one-year cross-lagged effects of customer satisfaction on climate to be zero, fit the data

2
well and did not fit significantly worse than Model 1a, x difl (3, N = 95) = 4.23, ns.

Model 3, which constrained the one-year cross-lagged effects of climate on customer

satisfaction to be zero, did not provide acceptable fit, and fit the data significantly worse

2
than Model 1a, x diff (3, N = 95) = 19.54, p < .01. Similarly, Model 4, an autoregressive

null model with no cross-lagged effects, did not result in acceptable fit. Additionally, the

2
autoregressive null model fit the data significantly worse than Model 2, x difi” (3, N = 95)

= 19.87, p < .01 . These results indicate that, for service departments, there is a direct
effect of climate on subsequent customer satisfaction for one-year lags, but there is no
direct, reciprocal, effect of customer satisfaction on subsequent climate.

Analyses were conducted next to examine the consistency of the one-year lag
model across time (Stage 4). Model 5, which constrained the autoregressive lags to be
equal over corresponding time periods, provided acceptable fit to the data, and did not fit

2
the data significantly worse than Model 2, x diff (9, N = 95) = 15.4], ns. Imposing

additional constraints of equal cross-lags across the different time periods (Model 6) also

2
provided acceptable fit and did not fit the data significantly worse than Model 5, x difl

(2, N = 95) = 1.23, ns. This indicates that the influence of climate on customer

54

satisfaction is stable across the one-year lags. Model 6, including both equal
autoregressive and equal cross-lags, was retained as the final one-year lag model for
service departments.

Examination of two- and three-year lag models indicated that no cross-lagged
effects existed at these longer lag periods. For both sets of models, the autoregressive null
model (Model 4) did not fit the data significantly worse than models containing cross-
lagged effects. Furthermore, none of the models for two- and three-year lags met all of
the criteria for acceptable model fit. Thus, one-year lags appear to be the optimal time lag
available in the current study for examination of the longitudinal effects of climate and
customer satisfaction in service departments.

Overall, the one-year lag model with equal cross-lagged effects from climate to
customer satisfaction and equal autoregressive lags (Model 6) provided the best fit to the
data for service departments. Final parameter estimates for this model are provided in
Figure 2. The cross-lagged paths from climate to customer satisfaction were stable over
time and significant—standardized estimates ranged from .16 to .20 (note that the
coefficients were constrained to be equal in the unstandardized solution, but
standardization leads to slightly different estimates). There was no evidence of reciprocal
relationships from customer satisfaction to climate. Therefore, for service departments,
the results support the hypothesis that climate has causal priority over customer
satisfaction (Hypothesis 1a), but fail to support the hypothesis of reciprocal relationships

over time (Hypothesis 2).

55

.oSwm 05 88a 3880 soon 38. 880888 3838.0.6 EB

1608 8082388 .3820 3m ._o.vQ an .5833 98 8:82:09 5 mean; do!“ 5 ESE—Emu 8m Eon

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

         

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E 328/ 385588 5 88833 8868858: 5:» .565 2a 356508 58 Bfiwumucsm $.th
o 28:. m 08:. v 08:. m BE... N 08F _ 06:.
A83 A83 ASE
A33 . NS. =-
2:. a cn— "
+ F 4 <
cowofimuam 183. c 828$QO Sum.» cozofimtum Sum», 880$QO Sum u. aouofimcmm $3..» coroflmumm
588.30 mm. . 88830 «.3. . 58930 mam: 38930 a: 682,050 _ . 88230
so: «8.- A83 we. A83 on. 689 E.
«mg.» .03. «Mg.» m2. 33.» V3.
33..» 28: 33. u 33..» 38:
3M. .5 Sm. N3: .5
$23 at;
can. mm»:

cowofimumm 38880 88 038:0 808.3an 83.8w 8m ESE 38m ”N 035

56

Sales departments. The same staged approach was used to examine Hypotheses
1a and 2 for sales departments. Tables 10, 11, and 12, respectively, contain summaries of
model fit and comparisons for the one-, two-, and three-year lag models. Similar to the
results for the service departments, the addition of wave-skipping autoregressive lags
(Model 1a) resulted in significantly increased model fit for all three sets of models. Also
similar to the service department models, the remaining models for the different lags
diverged. However, for sales departments, it was the results of the subsequent two-year
lag models that diverged while the results of the one- and three-year lag models were
similar. Therefore, the results of the two-year lag sales department models will be
considered next followed by a discussion of the one- and three-year lag models.

The results for the two-year lag sales department models for climate and customer
satisfaction followed a similar pattern to the results of the one-year lag models for service
departments. Model 2, estimating only the cross-lagged effects of climate on customer

satisfaction, fit the data well, did not fit significantly worse than the full cross-lagged

2
Model 1a, x difi” (3, N = 95) = 1.72, ns, and fit the data significantly better than the

2
autoregressive null model (Model 4), x my (3, N = 95) = 10.08, p < .05. Model 3, which

estimated only the cross-lagged effects of customer satisfaction on climate, fit the data

2
significantly worse than Model 1a, x difl (3, N = 95) = 10.08, p < .05 . Thus, Model 2

was retained as the best fitting model at this stage of analyses.

57

._.o.v0 .0 2005090 00 02.3.0.2. a. 003.9 20:00-...0 .mo.v0 .0 E00500? 0.0 0.00 c. 00:.0> 90:00-...0 ..m...02

 

 

 

 

290232 82 00.528 25.3. 02.320 0:00 0

m 2.8 v 0022 x: 2.30 oo. 3. S. 03 02820055.. 0.00 m
00...? 000..0< .0005. .0 5:20.300 .00.. ”v mama
o 8.0 3 082 NS 3.30 8. mm. 8. 302-390 2. .52 2080203 e
m a: 2 .0022 at 3.200 mo. 8. 3. 2:0 205:0 A- .00 02290 - 882-390 0
0 E0 3 08.2 at 00.80 .8. 8. 8. 2:0 .00 0592.0 A- 905:0 - 882-390 0
0.0033500 0.09000 ”m mama
. . . . 0002

.- I I 0: mm mum mo mm 3 0200050093. 00.00.6003 5.2. 00.0090 =3“. 0F
0 02.2. S 08.2 08 00.30 B. S. 8. 02-390 :3 F

 

.0005. Bag-000.0 =0“. 0.020%. ”N mafim

 

.0 < 20:00 08.2 .6 92:00 50.2”. 0.0 0.200
-20 a 82.8500 -20
0.000.). 00302-00000 000.4000 00:00.0..0m 000.2030 000 0.0:...0 .coEt0m0n. 00.0w .0. 000.00. 0.“. .0 E0855 ”or 0.00b

58

.300 .0 E00530 0.0 00:80.00 0. 00:_0> 0.0000250 .0000 .0 2.00.5090 0.0 0.00 c. 00:_0> 0.0000250 mtbz

cos-00:00 0.5.2.. 000.0020 .005 m

 

 

 

 

 

 

 

N 8.0 N .085. .8 8.0NN 8. 8. 8.
m 3.8 N 822 08 8.08 8. 8. 8. 80.. 020.00.00.05... .800 0
0..—FF. 000.04. .0022 .0 5:903:00 .00... “v manuw
0 8.3 N 822 N0. 8.08 8. 8. 8. 82-0.8.0 2.. 82 020.00.00.05... 0
m 8.2 0. 08.2 at 5N8 8. 8. 8. .80 0.05:0 A- .00 8.90:0 - 0082-0090 0
m N: 2 085. at 03.8 8. 8. 8. 2:0 .00 02980 A- 205:0 - 0082-0090 N
0000300000500 0.00500 ”m 0.0430
-- .- -- . . . . 000.. 0
0: 0.- NNN mo mm 00 0280.00.05... 00.00.05.003 5.>> 00.0020 =0”. F
0 8.3.. 0. 822 N8 .58 8.. B. NF. 02-0020 :3. F
.0005. 0091040080 =3“. 0.0:.0>m ”N $0.0
.0 < 0.200 .0022 .0 0.230 5020 :0 0.200

-20 < 000009000

.20

0.000.). 004802-0020 .00>.0>..-.. 00.80.0000 050.000 000 0.0.0.6 0.0.5.8000 00.0w .0. 000.00. g... .0 E0885 “3 0.00...

59

.5.v0 00 00005090 000 00..05-0.00 c. 00:.0> 000000-20 .mo.va 00 «0005000 000 0.00 c. 00:.0> 000000-20 MROZ

 

 

 

 

0.08.02 02 02588 55.3. 80.0020 08m. 0

0 2.8 0 082 8 0:8 8. 8. 8. 83 020028.50... 080 0
0E...- 0020< .0005. 00 5:20.050 000.5 “0 0m0uw
v 80 E 08.2 N? 8.0.8 8. 8. 8. 08.0090 8. .02 02009800... 0
N 5.0 3 08.2 08 8.08 8. 8. 8. >00 005.0 A- .00 02980 - 0082-0020 0
N 03 3 0.8.2 08 8.08 8. 8. 8. >00 .00 05280 A- 005.0 - 880.0090 N
0.0002g00E00 00000.00 ”m 0m0um
. . . . 0000

-- - .- 0: 00 gm 00 mm 00 02000600003 050050-903 55> 00.0090 =0“. 0r
0 8.8 3 .822 x: 8N8 B. 8. 00 8.0020 .3“. r

 

.0022 00800-0090 =0“. 0000.0>m N 0&6

 

 

500 0.0000 .0005.
-500 000009000

“.0 886$ (mm—2m EU mfimm

-30

0.000.). 00300-00000 -.00>-00E.r 0000000000 00.0000 000 000E_.0E0Et010-0O 00.0w .0». 000.00. 5“. 00 b0EEzm ”NF 0.00 -.-

6O

Examination of the consistency of the two-year lag model across time (Stage 4), revealed
that, in contrast to the results for service departments, the autoregressive lags within

construct were not stable over time as Model 5 fit the data significantly worse than Model

2
2, x diff (9, N = 95) = 32.21, p < .01. However, comparison of Model 6 to Model 2,

demonstrated that the cross-lagged effects of climate on customer satisfaction were stable

2
over time, x difi” (2, N = 95) = .64, ns. These results indicate that, for sales departments,

there is a stable direct effect of climate on subsequent customer satisfaction for two-year
lags, but there is no direct, reciprocal, effect of customer satisfaction on subsequent
climate.

Examination of one- and three-year lag sales department models indicated that no
cross-lagged effects existed at these lag periods. For both sets of models, the
autoregressive null model (Model 4) did not fit the data significantly worse than models
containing cross-lagged effects. Furthermore, the models for one- and three-year lags
generally failed to meet all of the criteria for acceptable model fit. In contrast to the
results observed for service departments, the two-year lag period appears to be the
optimal time lag available in the current study for examination of the longitudinal effects
of climate and customer satisfaction in sales departments.

Overall, for sales departments, the two-year lag model with equal cross-lagged
effects from climate to customer satisfaction (Model 6) provided the best fit to the data.
Final parameter estimates for this model are provided in Figure 3. The cross-lagged paths

from climate to customer satisfaction were stable over time and significant—standardized

61

0.59.. 05 Eot 002.50 0000 0>0z 0080:0000 000000306 000

 

 

 

 

 

 

 

 

 

   

 

 

 

 

  

.0008 0008050008 500.0 00”— ..o.V0 0 00005.80 000 8:00.200 0. 00:.0> .moVa 00 0002-000?- 08 0.00
E 00:.0> 8000000000 5 0208000 0080000000000 .005 03000 000 080.0508 500 0080000005 MEG?
0 as: 0 2:; v 25.8 m as: N 2:: . 2:0
38.. «N08 38..
«V~N» . co _
. 8. 8.-
08 _ 0
+ + + e
00000-0000w «Nam... 00000-0000m «03..» 000000003 «000”», 000000000m «3N.» 00000-0003 «3.0-» 000000003
00500000 n:- 0058000 5% 00800000 NS. 00.09000 00.09000 0089000

 

 

 

 

 

 

\

   

 

         
 

 

 

 

 

 

 

 

 

 

 

 
 

 

 

 

:80 .8. 08.0 08. 28.. 8... .80.. 8 ..
08.. N2. 08.. EN 08.. 80
6.3 00:: 0.8.. 08.. o 2:
3.. . .0 EN. 00.0. 0 .0
0 8.0 0.2..
80. SN.

000000000m 0083000 000 808:0 000§00Q 00.0w 00.“ .0002 Rim ”m 03E

62

estimates ranged from .13 to .20 (note that the coefficients were constrained to be equal
in the unstandardized solution, but standardization leads to slightly different estimates).
There was no evidence of reciprocal relationships from customer satisfaction

to climate. Therefore, for sales departments, the results support the hypothesis that
climate has causal priority over customer satisfaction (Hypothesis 1a), but fail to support
the hypothesis of reciprocal relationships over time (Hypothesis 2).

Summary. The results for both service and sales departments were similar in
indicating that climate perceptions impact subsequent customer satisfaction but customer
satisfaction does not impact subsequent climate perceptions, and, therefore, climate has
causal priority over customer satisfaction. Additionally, for each department, the
magnitudes of the longitudinal cross-lagged effects were equal over time indicating that
changes in climate perceptions have a stable impact on subsequent customer satisfaction.
However, despite the consistency of the effect within each department, results suggest
that the causal process unfolds more rapidly in the service departments than in the sales
departments.

Vehicle sales. Hypothesis 1b proposed that climate perceptions would predict
vehicle sales over time more strongly than vice-versa. The same staged analytic process
described above in the context of customer satisfaction was used to investigate this
hypothesis. Department size was related to the number of vehicle sales and, therefore,
was used as a control variable in all analyses. Across all three sets of models, the Stage 2
models with wave-skipping autoregressive lags (Model 1a) again provided significantly
better fit than the models without these additional autoregressive lags (Model 1). Similar

to the results of the sales department models examining climate and customer

63

satisfaction, the results of the two-year lag model diverged from the similar results of the
one- and three-year lag models, and will be discussed first. Summaries of model fit and
comparisons for all one-, two-, and three-year lag models are presented in Tables 13, 14,
and 15, respectively.

The results for the two-year lag sales department models for climate and vehicle
sales followed the same pattern as the results of the two-year lag models for customer
satisfaction. Model 2, with cross-lagged effects from climate to vehicle sales, fit the data

well, did not fit the data significantly worse than the full cross-lagged model (Model la),

2
x difl (3, N = 95) = 2.92, ns, and fit the data significantly better than the autoregressive

2
null model (Model 4), x difl (3, N = 95) = 8.10, p < .05 . Alternative Model 3, with only

cross-lagged effects from vehicle sales to climate, fit the data significantly worse than

2
Model 1a, x difl (3, N = 95) = 8.10, p < .05 . Therefore, consistent with the sales

department results for the two-year lag model of climate and customer satisfaction,
Model 2 was the best fitting model.

Similar results to sales department analyses with customer satisfaction were also
obtained for the consistency of the two-year lag model over time: Model 5 (equal

2
autoregressive lags) fit the data worse than Model 2, x diff (9, N = 95) = 37.19, p < .01,

2
and Model 6 (equal cross-lagged effects) did not, x difl (3, N = 95) = 1.72, ns. These

results indicate that there is a stable direct effect of climate on subsequent vehicle sales

for two-year lags, but there is no reciprocal effect of vehicle sales on subsequent climate.

64

...0.v0 .0 E00530 00 02.8.0.0: :_ 00:.0> 90:00-...0 00.x. .0 .:00_.._:0_0 0.0 0.00 :_ 00:_0> 90:00-...0 M502

 

 

 

 

 

 

 

0.000.390 .02 90:00:00 55.3. 000.0090 .0:0m. 0
m 00.00 c 08.2 80 8.000 8. 5. S. 000.. 9.0020205... 0:00 0
0E... 0090< .0005. .0 >0:0.0_0:00 .00 .r ”0 000.0
0 3.0 0.. .0005. 0.0. mo. 50 00. n0. 00. 000.0090 0:. __:z 0>_000._09o.:< v
m 00.0 0F .0022 00. 0.0.000 00. 0.0. 3. 0:0 0.0:...0 A- 00.00 0.0.:0> - 00000.0090 m
0 0:0 0. 08.2 .0. 00.80 8. 3. 8. 2:0 0000 002.5 A- 205:0 - 8000.080 0
0.000.). 9:09:00 0.0%E00 um 0%.0
-- .. -- . . . . 000.. 0
.0. .0 000. 00 0.0 00 0200909054. 0:.000.m0>0>> :..>> 00.0090 ._:u. F
0 00.00 0.. .0005. N0. 00.000 00. 00. 00. 00.0090 .3”. F
.0005. $000090 .3”. 0.0:.0>.n.. “N 000.0
00 4 90:00 .0005. 00 90:00 (mm—2”. _u.0 0.210

-_:0 0 :00000500

.....0

0.0005. 00001000090 .00>-0:O 00.0w 0.0....0> 0:0 0.09:0 .:0:000d0o 00.0w .0. 0000:. ..u. .0 0:05:50 “m. 0.00 ._.

65

...o.va .0 .c00...:0.0 90 02.3.0.0: :_ 00:.0> 90:00-30 .00.v0 .0 .:00_...:0_0 90 0.00 :_ 00:.0> 90:00-...0 ..m...02

 

 

 

 

 

 

 

0 00.0 0 .082 00. 3.000 B. 00. 00. 02528 55.3. 000.0020 .0300 0
0 050 0 08.). 000 9.000 00. .0. B. 000.. 02.000.00.93 .0000 0
09.... 00904. .0005. .0 >0:0.0.0:00 .00 0.0 000.0
0 2.0 0 .08.). .0. 00. .00 00.. .0. 00. 000.080 o... .32 02.000.00.22 v
0 2.0 0. .082 .0. 9.000 00. .0. 3. 2:0 0.0:...0 A- 00.00 00...? - 8000.080 0
0 00.0 0. .08.). .0. 00.000 3. 00. 00. 2:0 00.00 22.05 A- 0.05.6 - 8000.080 0
0.0005. 00000900 90300 .0 000.0
-- .. -- . . . . 000.. 0

.0. 00 000 no 00 00 0>.009090.:< 0:.00_x0-0>0>> 55> 00.0090 .3... F
0 .00.. 0. .0005. .0. 00.0.0 00. 00. 00. 00.0090 .3“. .
.0005. 000030090 ..:n. 0.0:.0>m UN 0000.0
.0 0 20:00 .0022 .0 0.0000 «00.20 .00 0.200

-_:0 < 50000900

-20

0.0005. 000030090 .00>.0.s._. 00.00 0.0.:0> 0:0 0.0:...0 .:0E...000o 00.00 .0. 000.05 .E .0 b05050 “0.. 0.000

66

..0.vq .0 9000900 90 02.3.0.3 :. 00:.0> 90:00-90 .mo.v0 .0 90890.0 90 0.00 :. 00:.0> 90:00-90 ..m....02

 

 

 

 

 

0.08.02 .oz 002.008 0.5.3. 000.0090 .0000 0

0 00.00 0 .0005. 000 2.000 00. .0. .0. 000.. 020.00.00.92 .0000 0
09. .- 0090.4 .0005. .0 >0:0.0_0:00 .00 .F “v 00040
0 00.0 0. 000.2 .0. 00. .00 00. .0. 00. 000.0020 05 __:z 0>_000.00.os< 0
0 0... 0. .08.). 00. 00. .00 00. .0. 00. .00 0.09.0 A- 00.00 0.0.005 - 0080.080 0
0 00.0 0. .0022 00. 00.000 00. .0. 0o. .00 00.00 00:.0> A- 0.0:...0 - 80.00.0020 0
0.000.). 0:..00900 900900 .0 000.0 .
. . . . 000:

- - - 00. 00 000 00 .0 00 020090905... 09:020-902. 5.)) 00.0090 .50 0.
0 00.0 0. .0022 00. 00.000 00. 00. 0.. 00.0090 =0“. .

 

 

.0005. 0000000090 .50 0.0:_0>w ”N 000.0

 

 

.00 0.0000 .085. .0

-90 0 :00000900

 

0.0300 $05.0. EU m2m0
-20

0.000.). 0000000090 .00>-09:... 00.00 0.0.:0> 0:0 0.09:0 909.0000 00.00 .0. 000.09 ..n. .0 b099:0 ”m. 0.00 .r

67

Examination of one- and three-year lag models of climate and vehicle sales
indicated that no cross-lagged effects existed at these lag periods. For both sets of
models, the autoregressive null model (Model 4) did not fit the data significantly worse
than models containing cross-lagged effects. These results are consistent with the results
of the sales department models for climate and customer satisfaction, and provide
additional evidence that the climate to organizational outcomes causal process takes
approximately two years to unfold for sales departments.

The two-year lag model with equal cross-lagged effects from climate to sales
(Model 6) provided the best fit to the data. Final parameter estimates for this model are
provided in Figure 4. The cross-lagged paths from climate to vehicle sales were small,
but stable over time and significant—standardized estimates were approximately .03
(note that the coefficients were constrained to be equal in the unstandardized solution, but
standardization leads to slightly different estimates). There was no evidence of reciprocal
relationships from vehicle sales to climate. The results provide only tentative support for
the hypothesis that climate has causal priority over vehicle sales (Hypothesis 1b), due to
the small effect size.

Customer satisfaction and vehicle sales. Hypothesis 3 proposed that customer
satisfaction would partially mediate the relationship between climate perceptions and
vehicle sales. The sales department models described above were used to inform the
construction of the mediation models estimated to examine this hypothesis. That is, the
previous results indicated that the mediation models did not require the inclusion of

cross-lagged effects of customer satisfaction on climate and vehicle sales on climate.

68

.050... 0... 80... 000.90 000.. 0.6.. 90.080000 0.00.0? .0008 000 00:00:00.0
.3008 9090000009 5.00.0 .0". ..c.V0 .0 900.0007. 0.0 02.8.0.2. 0. 00:.0> .mo.V0 .0 0.500.000 0.0 0.0..
0. 00:.0> 0000500000 0. 00.08000 00300000809. 0..? 952.0 0.0 0.00.0880 £00 0000000080 MED?

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 20.... 0 20.. 0 20F m 20.. N 20F _ 20..
0.0 . .0
. :03
.500 .00: 000. 00..
000. . 00. .
+ + l+ ¢
00.0w «0.0... 00.00 «0.0.. .. 00.00 x. 00... 00.00 80.. .0 . 00.00 80.0 .0 00.00
20.00., .00. 220$ 00.. 0.0.03. 000. 00:.0> a... 20.00., 0.0. 220$
.0000 .00.- .0000 000. :00.-. 000.- .0000 000.
0.00.. 0.0. 0.00.. .00 0.00.. 0.0.
.0000 225 0 :0... :03 22..

00.. ._ 0... .00. .0

«.3... $0..

.00 .00.

00.00 0.0.0.0> 000 0.0800 .00§00Q 00.00 00.. .0003. .00.... ”v 050....

69

Prior results also indicated that it was unnecessary to examine one- or three-year cross-
lagged effects of climate on customer satisfaction or vehicle sales. In order to further
inform construction of the mediation models, an additional set of models examining the
relationship between customer satisfaction and vehicle sales were also examined and are
discussed next.

The same staged approach used to examine the bivariate relationships involving
climate was used to examine the longitudinal relationships between customer satisfaction
and vehicle sales. Summaries of model fit and comparisons for all one-, two-, and three-
year lag models are presented in Tables 16, 17, and 18, respectively. For all three models,
the inclusion of wave-skipping autoregressive lags (Model la) significantly improved
model fit. Similar to both prior sets of sales department models, two-year lags appeared
to be the optimal time period for longitudinal relationships to emerge between customer
satisfaction and vehicle sales. For both the one- and three-year lag models, the
autoregressive null model (Model 2) was the best fitting model, indicating no cross-
lagged relationships were present at these lag periods, while for the two-year lag analyses
Model 6, containing equal cross-lagged effects of customer satisfaction on vehicle sales
and no reciprocal effects, provided the best fit to the data.

Similar to the results for climate and vehicle sales, the cross-lagged effects of
customer satisfaction on vehicle sales were small, but stable over time and significant—
standardized estimates ranged from .02 to .04 (note that the coefficients were constrained
to be equal in the unstandardized solution, but standardization leads to slightly different

estimates). Final parameter estimates for Model 6 are provided in Figure 5. Therefore,

70

.Fo.va .0 6005090 00 00:00.28 0. 00:_0> 20:00-20 .093 .0 .002..ch 0.0 200 c. 00:_0> 20:00-30 ”5.02

 

0.08:0? .02 002.008 5505. 002.820 0000 0

3. am...“ 0 .0002 5 ONCE. 9. mm. 2.. 0mm. 02.000.00.056. .mzcw m
005 000.0... 800.2 .0 55.0.0000 .00 ._. ”0 $05

 

0. .00. 0. 000.2 .0 00.00 2. .0. 00. 000-0020 00. =02 020020222 0.

0 .0... 0. .0022 0.. 00.0 3. .0. 00. 0:0 .00 080.000 A- 00.00 202$ - 8000.085 0

0 0.0 0. 000.2 .0 0.00 2. .0. 0.. 0:0 00.00 202.5 A- .00 0050.000 - 8000-00.90 N
0.000%..030500 0.00E00 ”m 0d0.m

 

 

 

 

-- .. I . . . . 000.
.m 8 mm m. 3 no 0200059050. 00.00.6020; 55> 00.0090 =0”. 0.
m 8&0 0.. _000_2 0.. 00.02 or. 00. .0. 000000.5 =0”. F
_000_2 003000020 =0“. 0.0:_0>m. ”N $05
.0 0 0.00.5 .0005. .0 0.005 <mm§m EU 1.21m
-30 < 08000500 -20
0.0005.

0030400000 .00>-000 00.0w 00:.0> 0:0 :0_.00.0_.0m 0.0.0.000 0.00.0080 00.0w .0. 000.05 .E .0 6050.5 ”0 _. 030...

71

.S.v0 .0 0.005090 0.0 00__0..-0_00 c. 00:.0> 0.0:00-....0 .mo.va .0 .c0o...:m.0 0.0 200 c. 00:.0> 0.0:00-20 mtoz

 

 

 

 

 

 

 

 

 

0 .0... N .08.). 0.. .0. .0 0.. .0. 00. 8.50:8 55.5 000.0090 .0300 0
E 2.00 N .08.). .0 00. .3 0.. 00. 0.. 000.. 0208.00.92 .0300 0
0...:- mwo..0< 500—2 .0 5:306:00 .00... ”v 0am

v 3.0. N .085. .0 00.00 3. .0. 00. 000.0020 as .32 0203.00.03 v

v 3.3 0. .08.). 0.. 0.00 2. .0. .0. .20 .00 550.020 A- 00.00 0.0.0.5 - 00002-0020 0

v 0: 0. .08.). 00 00.00 0.. .0. 00. .20 00.00 00:.0> A- .00 .3905 - 8000.080 0

2000.2 m:..00E00 0.00:.00 “m 04.0.0

- - - . . . . 000.. 0
mm mm 00 3 mm .o 0>.000.m0.0.:< 00.0020-90.5 5.>> m0_-000.0 =0“. F

0 00.00 0. .085. .v 8.00. 3. v0. 0.. 00.0020 .3. .

_00o_>_ 0030-70020 =3... 0.0:_0>m ”N 0d0.m
.0 q 90:00 .082 .0 0.080 4.005... E0 0.200
-20 < 80.0058 -20

0.08.2

 

00000-70080 .00 >03 .- 020m 0_0.c0> 0:0 02.00.0000 .0E0.0:0 2.0.5.0000 00.0w .0. 000.05 ...-._ .0 b00550 E. 050..

72

..o.va .0 .000...0m.0 0.0 02.00.23 0. 000_0> 20000.00 .mo.v0 .0 .000...0m.0 0.0 0.00 0. 000_0> 20000-...0 ..m..02

 

 

 

 

 

 

 

 

 

008.2502 002.028 2.5.2.. 000.0020 .030 0
... 00.00 30.8.2 .0 0.00. 0.. 00. ... 000.. 0200200553080 0
0...... 000.0< _000_2 .0 500.0.0000 .00 .- .v 0m0.w
0 .00 0. .080. ... 00.00 ... .0. 0o. ...-00.0020 05.02 0208.00.22 ..
0 0.... 0. .082 S. 00.0 ... .0. .0. 200.00.005.05 A0200 o.o...0>-00000_-0090 0
0 0.... 0. .0022 3 00.00 ... .0. 00. 2:0 00.00 0.0.29 A- 000.050.000-00000085 0
00002300200 2000.00 .0 0m0.m
-- .- .. . . . . 000.. 0
.0 0. .0 . . .0 .0 020020290... 00.00.20-903 55> 0070020 =0“. .
0 00.0.. 0. .082 0.. ....0. .... 00. 0..! 00.08.90... .
.0005. 00300-0020 :00. 0.00_0>m ”N 0.9.0
.00 20:00 .0022 ... 0.0300 «00.2.. ..0 0.200
-...0< 80.09.50 -...0
0.08.2

00300-0020 .00>-00...-_. 00_0m 0_0...0> 000 02.00.0000 020.000 .00200000 0200 .0. 000.00. ..u. .0 b02200 “w. 050.

73

..0008 808800008 $.00? 00...

0.00.... 0... 80... 000.80 0000 30.. 008080.00 0.00_.0> 30:00 000 60:00:00.0

.SVQ .0 602.0%? 0.0 00:02-22. 0. 00:.0> .mo.va .0 .00050w0 0.0 0.00

 

 

 

 

 

 

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S. 00:.0> 000050200 E 00.08000 000.05.000.00: 5.? 03000 0.0 0.5.2....000 £00 005080005 ..M.QZ
0 25. 0 2a.. .. as... m ”a... N ”a; . on...
8: ..
. «..m y
3.... a»: :0. 0....
. 3:.
00. _ .
+ + +
00.0w «Mg.» 020m «mm. .. 00.0w 90m... 8.0m «...... 0 00.0m «a: .0 00.0m
0.2.; 5.. 22.3 .3... 20...; $00. 20...“; .3... 2%.; .3... 22...;
«N§§u Sum: Soc. ME. coo. moo. coo. as. Coo.-. ow _ .- «MNQ.» NmN.
:2... .8. :2... ..8. 28.. .0... :2... 08.
0280.003 «3.0.», :0_.00..0_.0m «Va... 00:00.0..0m «:0... 00000.000m «RN.» 02.000003 «0%.... .» :0_.00..0_.0m
M0»: 580.000 mu». . 680.000 0%... 0080.000 .3: 0080.030 .30. 550.000
. . . .
MNN.» . 4
00.. 0.8.. an... E3
5. 9...»:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

00.0w 2030.» 0:0 02.00.36 580.000 308.003. 00.0w .0.. .0002 .05.... um 05m...

 

 

 

 

 

 

 

 

 

 

 

w...

74

consistent with the other sales department models, these results indicate that it is
unnecessary to examine one- or three-year lag mediation models and it is unnecessary to
estimate cross-lagged effects of vehicle sales on customer satisfaction.

The initial partial mediation model (Stage 2) contained two-year cross-lagged
effects of climate on customer satisfaction, climate on vehicle sales, and customer
satisfaction on vehicle sales. Consistent with all previous models, the inclusion of wave-
skipping autoregressive lags (Model 1a) significantly improved model fit (see Table 19
for a summary of fit indices and model comparisons for all mediation models). The full

mediation model (Model 2) fit the data well, did not fit the data significantly worse than

2
Model 1a, x dlfir (3, N = 95) = 6.68, ns, and fit the data significantly better than the

2
autoregressive null model (Model 4), x difl (7, N = 95) = 19.56, p < .01. Alternative

Model 3a, which constrained the cross-lagged effects of customer satisfaction on vehicle

2
sales to zero, fit the data significantly worse than Model 1a, x any (7, N = 95) = 17.14, p

< .05 . Likewise, alternative Model 3b, which constrained the cross-lagged effects of

climate on customer satisfaction to zero, also fit the data significantly worse than Model

2
la, x dzfl (6, N = 95) = 15.80, p < .05. As expected on the basis of prior sales

department analyses, constraining corresponding autoregressive lags to be equal (Model

2
5) resulted in significantly worse model fit, x difl (16, N = 95) = 65.65, p < .01, and

2
constraining cross-lagged effects to be equal (Model 6) did not, x difl (5, N = 95) = 5.47,

75

ns. These results indicate that the longitudinal relationship between climate and vehicle
sales is fully mediated by customer satisfaction.

The full mediation model with equal cross-lagged effects from climate to
customer satisfaction and customer satisfaction to vehicle sales (Model 6) provided the
best fit to the data. Final parameter estimates for this model are provided in Figure 6. The
cross-lagged paths from climate to customer satisfaction were slightly smaller than those
obtained in the bivariate climate-customer satisfaction model—standardized estimates
ranged from .12 to .19 (compared to .13 to .20 for the bivariate model). The cross-lagged
paths from customer satisfaction to sales were consistent with the results for the bivariate
customer satisfaction-vehicle sales model—standardized estimates were approximately
.03. The results do not support the partial mediation hypothesis (Hypothesis 3) as
customer satisfaction fully mediated the relationship between climate and vehicle sales.
Additional Hypotheses & Research Question 3

Hypothesis 4 states that climate perceptions will be more positive at the
conclusion of the study than at the beginning. Dependent groups t-tests comparing

perceptions at Time 1 and Time 6 separately for each department were significant
(tsaleS=-3.45, df=94, p < .01; tService='3-72a df=94, p < .01). Thus, Hypothesis 4 was
supported. This indicates that the survey feedback and action planning process may have
contributed to enhanced climate perceptions over time. However, it is important to note

that other factors (e. g., concordant changes in staff or policy) may have also led to more

positive perceptions over time.

76

...o.v0 .0 .000...00.0 20 00:00-28 0. 00:.0> 20000.00 .mo.v0 .0 000.0090 20 0.00 0. 00:.0> 20000.00 ..m..02

oo.-0.0000 0.5.2.. 000.0020 .000m. 0

 

 

 

 

 

 

 

m .00 N .0005. 0N0 0.0.0 .0. .0. .0.

0. 00.00 N .0005. 000 00.000 00. 00. 00. 0000 0>.002020.0< .000m. 0

0E...- 0020< .0005. .0 >000.0.0000 .00 .-
. 00.0. m .0022 000 00.8.. B. 00. 00. 000.0020 o... .02 020090992 ..
m N...0 N .0005. 0N0 00.000 .0. 00. 00. ..00 00.00 0.0.00> A- ..00 0.00.000 00
0 .06.. N .0005. .Nm 0.. ..00 .0. 00. .o. 200 ..00 00.00.0000 A- 0.0.0.5 mm
m 00.0 0. .0005. Na mm. ..0 .0. .0. .0. 0000.005. =0... N
0.000.). 00009000 200.000
.. - .- . . . . 000... 0>.002020.0<
80 00 000 B .0 .0 2.2.0.20; 5.2. 8.0.8.2 .0000 0.
0. .0... 0. .0002 000 0.000 00. 00. 0.. 0000.005. .0200 F
.0005. 000.073-0020 =0“. 0.00.0.6
.0 0 20000 .0005. .0 20000 <mm§m .00 1.2mm

-....0 < 000000.000

-...0 ,
0.0022

000000-0020 00.0w 0.0.00> 000 02.00.0000 0020.000 .0.0.0__0 00.000000 00.0w .0. 000.00. ..n. .0 000.000 .0. 0.00 .-

77

0.33. 0... 8000 00.300 003 0>00 000.00.00.00 00. 03000030800 00.0020-g0? 000 02.0.0300 005300.00 0:50.003
6.00.03 .0....00 000003.00 ..0008 608200008 $000.0 .0”. ..o.V0 .0 8000.007. 000 00000-200 0. 00:.0> 00.x. .0
0.000.000 0.0 0.00 0. 00:.0> 00.000.00.00 0. 00.08.00 000.000.000.000 0..? 03000 0.0 0.00.0808 0.00 005000003m mag/N

 

 

o 00:5.
00.0m «00.. ,
0_UEQ> .Nn.

 

 

  

 

 

00..00..0..0m
0060.000

 

 

 

m 25,—-
00.0m «an. 0 -
0.030> 00. .

 

 

0 0.5-.-
00.0m «.3...»
00:.0> .Rm.

 

 

 

32.. 08.

 

 

 

 

 

 

 

 

 

 

 

 

«0.0. _ 0000000000 «0....»- 00..00-.0..0m «000.»
{00. 5.00.000 000. 580.000 000.
«mm-Q.» MVN. «am-0.» no...
.23 2 000..
m2. 0..-EU nnN.

00.0w 0.0005 000 0000000000 880.000 6.08:0 .88500Q 00.0w .0.. .0002 0000.002 .00.". ”0 0.00.".

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

m 00...- N 0E...- . 0:0-.-
00.0m «.0. 0 . 00.0w «000 0 7 00.0m
0.0005 a. . 0.00.0> 000 . 0.0...0>

5.0.. 08.
00:00.0..0m «NaN . 00.80.0000 «000 0 0000000000
0200.000 ..0... , 0200.000 .000. 0050.000

 

 

     
  

 

 

 

«...-0.» 0:.

«0%.»

 

 

 

 

.30..

0.08:0

78

Hypothesis 5 states that customer satisfaction for departments participating in the
survey and accompanying feedback and action planning process would be greater at the
conclusion of the study than non-participating departments. This hypothesis was
examined using repeated-measures AN OVA with time as a within-subjects factor and
participation in the climate survey as a between-subjects factor. Note that there were no

significant differences between the participating and non-participating departments in

customer satisfaction observed at Time 1 (tSales='-74a df=137, ns; tsewice=-.81, df=94,

ns). Analyses demonstrated no interaction between participation and time (FSales

(1,133)=1.86, ns; F Service (5,133)=1.44, ns), although the main effect of time was

significant for both sets of departments (FSales (1,133)=15.49, p < .01; FService

(5,133)=15.75, p < .01) indicating that customer satisfaction was increasing for both
groups over time. Thus Hypothesis 5 was not supported, suggesting that factors other
than participation in the survey process were responsible for increases in customer
satisfaction during the study.

Research Question 3 examined the possible influence of local economic
conditions (i.e., local unemployment rates) and organizational size on initial levels and
changes in department-level climate perceptions over time. Longitudinal growth models
were constructed with time nested within departments to examine these relationships
separately for both sales and service departments. The results of these analyses are

presented in Tables 20 and 21.

79

Initial null models, containing no predictors, and unconditional linear and
quadratic growth trend models were constructed and compared prior to the substantive
analyses. Initial null models indicated that 33.1% (sales departments) and 43.0% (service
departments) of the total variance in climate perceptions were between departments. The
linear trends were significant predictors of within-department changes in climate
perceptions for both sales and service departments (,8 = .063, p < .01, and fl = .045, p <
.01, respectively). Comparison of the linear trends models and the null models indicated
that the linear trends accounted for 16% (sales departments) and 12% (service
departments) of the Level-l within-department variance. The quadratic trends were not
significant predictors (,8 = .002, ns, and ,6 = .000, ns, for sales and service departments,
respectively), and were excluded from subsequent analyses.

The unconditional linear growth models indicated that there was significant
between-department variance in initial climate levels (i.e., intercepts) for both
departments, but only sales departments had significant between-department variance in
changes over time (i.e., slopes). For sales departments, the correlation between initial
climate levels and changes over time was -.571, indicating that sales departments with
lower initial climate levels tend to have greater increases in these perceptions over time
than departments with higher initial levels. Conditional linear growth models, including
unemployment as a Level-l predictor, did not account for significant within-department
variance for either sales or service departments (,6 = .003, ns, and ,8 = -.006, ns,
respectively).

Intercepts-as-outcomes and slopes-as-outcomes models were constructed to

examine whether unemployment or department size predicted between-department

80

Table 20: Summary of Results for Longitudinal Growth Models
for Sales Departments
Sales Departments

 

DV: Climate Level-2 Level-2
Level-1 Intercept Slope
Estimate Var. Var. Var.
Null Model .100 .050 -
ICC( 1) .331
Unconditional Linear Growth Model .084 .078 .006
Intercept 3.60
Linear Trend .063
2
R within-group '16
Unconditional Quadratic Growth Model .085 .059 .002
Quadratic Trend .002
2
R within-group '00
Conditional Linear Growth Model .084 .078 .006
Level-1
Unemployment .003
within-group '00
lntercepts-as-Outcomes Model 084 1’80 -005
Level-2
Unemployment .032
Department Size -.003
2
R between-group (intercept) '00
Slopes-as-Outcomes Model 084 -032 005
Le veI-2 - Slope
Unemployment -017
Department Size 000
2

R beMeen-groqp (slope) -00

2
NOTE: All R estimates are computed in comparison to immediately

preceding step; values in italics are marginally significant p<.10, values
in bold are significant at p<.05, values in bold-italics are significant at
p<.01.

1
All Beta estimates are at entry.
2 . . . .
No Significant level-2 variance to estimate.

3
No level-2 variance to predict.

81

Table 21: Summary of Results for Longitudinal Growth Models
for Service Departments

Service Departments

 

 

DV: Climate Level-2 Level-2
Level-1 Intercept Slope
Estimate Var. Var. Var.
Null Model .051 .038 -
ICC( 1) .430
Unconditional Linear Growth Model .045 .028 .001
Intercept 3.36
Linear Trend .045
2
R within-group '12
Unconditional Quadratic Growth Model .044 .034 .000
Quadratic Trend .000
2
R within-group '02
Conditional Linear Growth Model .048 .039
Level-1
Unemployment -.006
2
R within-group '00
2
lntercepts-as-Outcomes Model -047 ~039 --
Level-2
Unemployment .044
Department Size .002
2
R between-group (intercept) '00
2

Slopes-as-Outcomes Model - --
Level-2 - Slope

3
Unemployment -
3
Department Size --
2

R between-grouflslope)

2
NOTE: All R estimates are computed in comparison to immediately

preceding step; values in italics are marginally significant p<.10, values
in bold are significant at p<.05, values in bold-italics are significant at
p<.01.

1
All Beta estimates are at entry.
2
No significant level-2 variance to estimate.

3
No level-2 variance to predict.

82

differences in initial climate levels, for both sales and service departments, and changes
in climate perceptions over time, for sales departments only. For both sales and service
departments, neither unemployment (,6 = .032, ns, and ,6 = .044, ns, respectively) nor
department size (6 = -.003, ns, and ,8 = .002, ns, respectively) significantly predicted
between-department differences in initial climate levels. Similarly, neither unemployment
(fl = .017, ns) nor department size (,8 = .000, ns) significantly accounted for the between-
department differences in changes in climate perceptions over time observed for sales
departments. Thus, while the analyses showed some between-department differences in
initial climate levels and changes over time, none of these differences were explained by

unemployment or department size.

83

Discussion

Consistent with much of the existing theory (e. g., Bowen & Ostroff, 2004; James
& Jones, 1976; Kopelman et al., 1990; Ostroff et al., 2003) and research (e.g., Ryan et al.,
1996; Schneider et al., 1993; Schneider et al., 2003), organizational climate and
organizational performance, operationalized as customer satisfaction and sales quantity,
were significantly related. Furthermore, by utilizing longitudinal data and cross-lagged
panel analyses, it was demonstrated that organizational climate had causal priority over
both types of performance indicators, and that, for sales departments, customer
satisfaction fully-mediated the relationship between climate and sales. However, it should
be noted that the size of effect for prediction of sales was quite small making causal
claims tentative. The observed causal lag periods differed between the sales and service
departments. Specifically, the results indicated that for service departments,
organizational climate predicted customer satisfaction one year later, while for sales
departments organizational climate predicted customer satisfaction and sales two years
later.

Contrary to expectations, no evidence of reciprocal relationships between
organizational climate and performance were observed. Additionally, while
organizational climate was, on average, more positive at the end of the study than the
beginning, there was no evidence that departments participating in the climate survey
process had greater customer satisfaction scores at the conclusion of the study than
departments that chose not to participate. Exploratory analyses demonstrated that while
there were between-department differences in initial climate levels and changes over

time, none of these differences were explained by unemployment or department size.

84

Table 22 summarizes the results of the formal hypothesis tests and the outcomes of the

research questions. The remainder of this manuscript will provide an integrated

discussion of the implications of this research for the study of linkages between climate

and organizational performance, how the current results compare to the results of similar

studies, and possible directions for future research, followed by a discussion of potential

limitations of the current study.

Table 22: Summary of Hypotheses and Research Questions

 

Hypothesis

Result

 

Hypothesis la: Department-level climate perceptions will predict
customer satisfaction over time more strongly than vice-verse.

Hypothesis lb: Department-level climate perceptions for sales
departments will predict dealership vehicle sales over time more
strongly than vice-versa.

Hypothesis 2: Department-level climate perceptions and customer
satisfaction will be reciprocally related over time.

Hypothesis 3: The relationship between sales department climate
perceptions and vehicle sales will be partially mediated by
customer satisfaction.

Hypothesis 4: Climate levels will increase over the period of the
study.

Hypothesis 5: Organizations participating in the climate survey
and action-planning process will have greater customer
satisfaction at the conclusion of the study than organizations not
participating in the climate survey and accompanying action—
planning process.

Research Question 1: Over what lag period(s) do relationships
between organizational climate and performance emerge?

Research Question 2: Do differences exist in the causal lag periods
between climate and customer satisfaction?

Research Question 3: Do differences in the initial levels and
growth trajectories of organizational climate exist across
organizations? If so, do local economic conditions or
organizational size account for some of this variability across
organizations?

Supported

Supported

Not Supported

Not Supported — Full
mediation observed

Supported

Not Supported

Service Departments — One-
year lag periods only

Sales Departments — Two-
year lag periods only

Yes

Differences do exist, but were
not predicted by economic
conditions or organizational
size

 

85

Organizational researchers from many different perspectives propose models of
organizational performance that link human resource practices to organizational
performance through the mediating mechanisms of climate, motivation, employee
attitudes, and behavior (e.g., Applebaum, Bailey, Berg, & Kalleberg, 2000; Becker,
Huselid, Pickus, & Spratt, 1997; James & Jones, 1976; Kopelman et al., 1990; Ostroff et
al., 2003). Assumed, explicitly or implicitly, in all of these models is a causal chain
whereby human resource practices lead to climate perceptions, which lead to employee
motivation and attitudes to behaviors, which aggregate to result in organizational
performance. Unfortunately, despite repeated calls for research that would allow for
investigation of the causal ordering of these constructs (e. g., Paauwe, 2009; Wright &
Haggerty, 2005), the vast majority of prior research examining these links has used cross-
sectional or limited longitudinal designs that provide little basis for inferring causation or
exploring reciprocal relationships.

Three conditions must exist to make inferences of a causal relationship: the cause
must be related to the effect, the cause must precede the effect in time (i.e., causal
priority), and plausible alternative causal explanations for the effect must be ruled out. In
the context of climate, previous research has demonstrated that organizational climate
and performance are related (e. g., Borucki & Burke, 1999; Ostroff, 1992; Schneider et
al., 2005), but has been inconclusive on issues of causal priority (e. g., Schneider et al.,
1998) and has sometimes failed to rule out spuriousness, or third variable causation, as an
alternative explanation of the observed relationships. The current study attempted to fill
this gap by utilizing design features and analyses that strengthen causal inferences and

allow for investigation of reverse causation and reciprocal relationships.

86

This study demonstrated that organizational climate has causal priority over two
indicators of organizational performance by demonstrating that organizational climate
predicts subsequent measures of customer satisfaction and vehicle sales over time.
Contrary to some previous research, there was no evidence of reverse causation or
reciprocal relationships. The use of a longitudinal panel design, allowing for control of
previous levels of organizational climate and organizational performance in estimating
the longitudinal relationships, strengthens causal inferences by showing that changes in
organizational climate predict subsequent changes in organizational performance thereby
demonstrating that organizational climate has causal priority over the indicators of
organizational performance examined in this study. The replication of these relationships
over multiple time periods also provides stronger evidence of causal relationships than
previous studies examining only one longitudinal lag between two time periods. While it
was not possible to control for many potential third variables that could account for the
relationships observed, the finding of no significant reverse causation or reciprocal
influences reduces spuriousness as a plausible explanation (Kenny & Harackiewicz,
1979). Additionally, the use of a sample of independently owned and operated
organizations carrying identical products and services and the replication of this finding
across two different organizational contexts (i.e., sales and service departments), at least
for customer satisfaction, enhances the generalizability of these findings. Thus, this study
provides the strongest evidence to date that organizational climate can causally impact
organizational performance.

As noted previously, Ostroff and her colleagues (2003), among others (e.g., James

& Jones, 1976; Kopelman et al., 1990), propose that organizational climate contributes to

87

organizational performance by promoting performance of behaviors consistent with the
organization’s goals. The results of the current study corroborate this theory. However,
many of these models also include a reciprocal path whereby organizational performance
also influences organizational climate, which is not corroborated by the current results.
This reciprocal influence is thought, by some, to be a result of higher performing
organizations having greater resources to devote to human resource practices that
enhance the well-being of employees (e. g., Wright & Gardner, 2003; Wright et al., 2005).
Another explanation, particularly relevant to the outcome of customer satisfaction, is that
customers themselves represent a salient feature of the environment for front-line
employees and as such climate may include an appraisal of the degree to which an
organization’s policies, procedures, etc. promote the well-being of customers (Burke,
Borucki, & Hurley, 1992; Heskett et al., 1997). As mentioned previously, prior research
also provides some evidence of this reciprocal relationship between climate and
organizational performance (Schneider et al., 1998). While it is not entirely clear why
such a reciprocal relationship was not observed in the current study, there are a number of
possible explanations.

Schneider and his colleagues found a reciprocal relationship between
organizational climate for service and customer satisfaction in the context of bank branch
employees providing face-to-face transactional services to customers. The lack of finding
a reciprocal relationship in the current study could be due to the utilization of a molar
organizational climate measure, which is potentially less susceptible to the influences of

customer satisfaction. Alternatively, differences in the nature of the employee-customer

88

service encounter may limit employees’ opportunity to directly observe customer
dissatisfaction, which may reduce the reciprocal influence.

In the context of directly interacting with a customer to provide services, as in the
Schneider study, it is likely that there are immediate visual and/or verbal cues of
customer satisfaction, especially dissatisfaction, that increase the salience of customer
satisfaction as a feature of the work environment. However, in the context of selling
someone a car, a dissatisfied potential customer may simply not choose to purchase the
car from that particular dealership without the sales person even being aware that the
customer was dissatisfied with their service. Additionally, as noted previously, cognitive
dissonance theory (Festinger, 1957) suggests that someone who just spent thousands of
dollars on a new vehicle is unlikely to behave in a way that conveys dissatisfaction in the
moment, regardless of whether the customer is ultimately pleased with the service they
received. Therefore, in sales departments, the observation of customer satisfaction is
likely limited to satisfaction, with dissatisfaction only being communicated later, and
indirectly, through the customer feedback survey process. In the service departments
examined in the current study, the majority of employees work behind the scenes and
have no direct interaction with the customer, which would, likewise, limit their ability to
directly observe customer satisfaction or dissatisfaction. However, it should be noted that
service employees do get proximal indirect feedback of customer dissatisfaction if the
customer’s problem with the vehicle was not fixed correctly. In both of these contexts,
the limited direct observation of customer dissatisfaction may decrease the extent to
which customer satisfaction is perceived as a salient feature of the work environment.

Additionally, in the absence of direct observation of customer dissatisfaction, it may be

89

that any reciprocal influences must flow through a longer causal chain before being
perceived by employees (e. g., customer satisfaction results are reviewed by management,
policies and procedures are revised to address problem areas, which then become a
salient part of the environment for new employees) which is likely to limit the size of the
effect (Shrout & Bolger, 2002). Similar arguments may be made for why Ryan and her
colleagues (1996) observed evidence of a causal relationship between customer
satisfaction and employee satisfaction in their examination of customer contact
employees in a financial services organization. Future research should explore these
alternatives by examining the reciprocal relationship of climate and customer satisfaction
longitudinally across multiple types of employee-customer service contexts.

There is also a possible methodological explanation for why no reciprocation was
found in the currently study. The overall sample size, while larger than many cross-
lagged panel studies conducted at the organizational level of analysis, was smaller than is
generally recognized as desirable for conducting structural equation model analyses (e.g.,
Kline, 1998; Muthén & Muthén, 2002). The major impact of a small sample size is to
reduce the power to detect small true effects. Thus, one possible reason that the
hypothesized reciprocal effects between organizational performance and climate were not
observed is because there was too little power in the analyses to detect small reciprocal
effects. Interestingly, the reciprocal effects for both customer satisfaction and sales
predicting climate were also highly variable with some lags approaching statistical
significance and other lags not, for both departments. Future research with greater sample

sizes is necessary to fully explore this possibility.

90

It was hypothesized that customer satisfaction would only partially mediate the
relationship between climate and vehicle sales, but full mediation was found. While this
is consistent with theory that suggests that customer satisfaction results in positive word-
of-mouth and repeat purchases in the firture (e. g., Heskett et al., 1994; Maxham &
Netemeyer, 2003), it appears inconsistent with the notion that other behaviors influenced
by climate perceptions, but unrelated to customer satisfaction, play a role in determining
organizational financial outcomes. For example, a positive climate is expected to promote
a number of behaviors that are not directly related to customer satisfaction, such as
willingness to negotiate with difficult customers, attending sales technique training, or
network building, and these behaviors are expected to promote sales. However, it may be
that these behaviors do positively influence organizational financial outcomes, only not
by increasing sales but by increasing per vehicle profit. For instance, when a sales person
leverages new sales techniques that result in a higher profit for the dealership (e.g.,
focusing on the monthly payment instead of the overall vehicle price, presenting optional
products/services as “included” in the sales price, “bumping” interest rates above bank
quotes, etc.). Future research in similar contexts is necessary to examine the
generalizability of this finding and test the alternative proposed. Future research
investigating this finding in other contexts would also be informative. For example, it is
possible that stronger relationships between climate and sales would be observed in
contexts where employees are not paid on commission, but are still encouraged to
increase sales (e. g., hourly/salaried sales associates at companies like Best Buy or

AutoZone).

91

The effect sizes in the current study, despite being significant, were relatively
modest in the case of organizational climate predicting customer satisfaction, and were
quite small when predicting sales. However, when considering the size of these effects, it
is important to remember that prior levels of the performance indicators, and, hence, the
influence of prior climate levels, has already been controlled for. Additionally, the effect
sizes must be considered in the context of the distal nature of climate in relation to the
performance indicators. Shrout and Bolger (2002) note that as the relationship between
variables becomes more distal, the effect size decreases due to the number of links in the
causal chain, competing causes, and other random factors. In the case of climate, theory
postulates that motivation, attitudes, and behaviors mediate its relationship with
indicators of performance. It seems likely that estimates of these theoretically more
proximal potential determinants of customer satisfaction and sales would yield stronger
effect sizes, although cross-sectional studies examining the relationships of some of these
additional variables with customer satisfaction have yielded effect sizes roughly
equivalent in magnitude to the cross-sectional correlations observed in the current study
(e.g., Brown & Lam, 2008; Harter, Schmidt, & Hayes, 2002). Finally, it is important to
point out that even small effects can have large practical implications (Abelson, 1985;
Lipsey, 1990; Rosenthal, Rosnow, & Rubin, 2000).

As noted, the effect sizes for prediction of sales were extremely small, even with
the more proximal variable of customer satisfaction as the predictor, which showed an
average concurrent partial correlation (controlling for department size) of .24. It is
difficult to argue that a standardized effect of approximately .03 for predicting sales from

organizational climate is practically significant — this effect size translates to, for an

92

average dealership, an increase in annual vehicle sales of just over five vehicles (about
$7,500 in gross profit, or less than .2% of average annual gross profitz; NADA, 2007).
For the larger dealerships (i.e., those selling 400 or more vehicles annually), this would
represent an negligible increase, but for smaller dealerships (i.e., those selling
approximately 100 vehicles annually) this could represent a practically significant
increase in sales. Additionaly, it is important to note that prior sales and department size
accounted for approximately 90% of the total variance in subsequent sales figures,
leaving little variance left to account for. It is possible that other indicators of financial
performance, such as per vehicle profit or dealership return on investment, would have
shown stronger effects. Although other researchers predicting financial indicators of
organizational performance with human resource practices, commitment, and employee
satisfaction have also observed large reductions in the size of relationships, generally to
the point of non—significance, once controlling for prior financial performance (Guest,
Michie, Conway, & Sheehan, 2003; Koys, 2001; Wright et al., 2005). The concurrent and
lagged correlations in these studies were of similar size to that observed for customer
satisfaction and organizational performance in the current study.

Wright and his colleagues (2005) suggest a number of reasons why relationships
between various HR outcomes (i.e., climate, job satisfaction, motivation, etc.) and
financial indicators of performance may disappear when controlling for prior financial
performance. First, they suggest that such results may be indicative of a reciprocal
relationship where firms that perform well financially, invest more heavily in HR

practices, and this results in further increased financial performance. Second, they

 

2 . . . .
These estimates do not, however, account for indirect profits that may accrue in subsequent years due, for
example, to vehicle service visits or repeat purchases.

93

propose these types of findings may indicate that such relationships are spurious, and
instead both HR outcomes and financial performance are caused by third variables. While
the cross-lagged panel analyses provided no evidence of reciprocality in the current
study, the lagged correlations with previous and subsequent sales for both customer
satisfaction and climate are quite similar to those found by Wright and his colleagues.
Given the similarity of the lagged correlations, the small effect sizes observed for
predicting sales in the cross-lagged panel analyses, and the relatively low power for
detecting small reciprocal effects in the current study, it is not possible to conclusively
rule out these alternative explanations for vehicle sales.

Existing theory linking HR outcomes to organizational performance is relatively
silent on issues of time, with the exception of noting that it takes time for changes to
unfold and ultimately have a measurable impact on performance (e. g., Ostroff et al.,
2003). Previous research using cross-lagged panel designs has generally been limited by
the availability of data to examination of only a single causal lag period between two
points in time (e.g., Ryan et al., 1996; Schneider et al., 1998). The current research
included an investigation of one-, two-, and three-year lag periods between organizational
climate and performance, and demonstrated that the causal lag period differed between
service departments, where stable relationships only emerged at one-year lag periods, and
sales departments, where stable relationships only emerged at two-year lag periods.
Cross-lagged panel analyses revealed that models containing no cross-lagged
relationships between organizational climate and performance fit the data better for the
other lag periods, even though some small cross-lagged relationships were evident in

models in which these relationships were estimated. This suggests that the causal

94

influence either had begun to dissipate, in the case of longer periods, or had not yet had
time to fully emerge. In line with the speculation presented previously, this implies that
contextual differences may moderate how long it takes for causal relationships between
organizational climate and performance to emerge.

In addition to the contextual differences cited previously (i.e., intangibility and
proximity of customer feedback), other differences may also impact the length of causal
lag periods between organizational climate and different indicators of performance. For
example, in the context of customer satisfaction and retail contexts, pay structure (e.g.,
commission-based vs. hourly or salary) may influence how long it takes for relevant
changes in climate to have an impact on customer satisfaction. Regardless of how
positive the overall organizational climate and how much the organization promotes
customer service, commission-based employees are motivated to make sales which may
motivate behaviors that sacrifice customer service, such as using high-pressure sales
tactics. The amount of time it takes for climate to influence organizational performance
may depend on the indicator of performance examined. For example, consistent with
Shrout and Bolger’s (2002) notions about effect sizes and proximity between cause and
effect, organizational performance outcomes that are theoretically more distal to
employee behavior, such as return on investment or profit, may take longer to be
impacted by climate than more proximal outcomes such as customer satisfaction or sales.
Future theOretical work needs to more fully explore and define the temporal aspects of
these relationships in order to promote more focused empirical investigations.

The current study also contributes to the debate over whether the molar

conceptualization of climate is too broad to be useful in predicting organizational

95

 

outcomes. Some researchers argue for abandoning research on molar conceptions of
climate based on the contention that it is too broad and inclusive to predict organizational
outcomes, — which the current results dispute -— and advocate, instead, that climate should
be assessed with a more strategic focus around a criterion of interest, such as innovation
or customer service (e. g., Schneider, 1975; Schneider et al., 1980; Schneider, 1990). This
is intuitively appealing as it is consistent with an underlying premise of Aj zen and
Fishbein’s (1975) work on attitudes — that is, the predictor and criterion variables should
be operationalized at the same level of specificity. However, defining climate in these
more narrow terms may result in ignoring an aspect of many conceptualizations of
climate. Some authors contend that climate is not just about simple perceptions of the
work environment, but also includes an appraisal of the degree to which the work
environment is personally beneficial or detrimental to the organizational well-being of
the individual (James & James, 1989; James et al., 2008).

In pursuing the more narrow conceptualizations of climate, it is easy to see, for
example, how an organization’s policies, procedures, etc. could promote a focus on
customer service, and thereby a climate for service, but at the same time fail to include
features employees perceive as promoting their own well-being. Consider the cliche' that
the “customer is always right,” which could reasonably be assumed to be a policy that
promotes climate for customer service. If this is coupled with management practices that
limit employee involvement and autonomy, then it is likely that despite the climate for
service, employees are less likely to perceive the environment as promoting their

organizational well-being.

96

The current results, by demonstrating that organizational performance indicators
are predicted by molar organizational climate, suggest the possibility that ignoring the
“well-being” aspect of the definition of climate may come at a cost in terms of reducing
the prediction of organizational outcomes if climate is more narrowly defined.
Interestingly, the average concurrent correlations of organizational climate and customer
satisfaction observed in the current study (i.e., Fsmice = .22 and holes = .21) are similar to
the concurrent correlations with customer satisfaction observed for service climate by
Borucki & Burke (1999; f = .26, averaged across two time periods) and are actually
higher than that reported by Schneider and his colleagues (2005; r = .15, ns.) and
Sowinski and his colleagues (2008; r = .13, ns.). This is not to imply that research using
the more narrow conception of climate should be abandoned. Rather, it is meant to
highlight the need for researchers to investigate both conceptions of organizational
climate simultaneously, preferably using longitudinal designs. For example, as implied
by the discussion above, it could be that molar climate acts as a moderator of the
relationship between climate for service and customer satisfaction. Additionally, current
definitions of molar climate tend to be amorphous and not very well defined, so future
research should continue efforts to define the conceptual space of molar climate more
thoroughly.

More broadly, this study contributed to the literature on the linkages between HR
outcomes and organizational performance by demonstrating that organizational climate
has causal priority over indicators of organizational performance. However, there are
many other links in the overall causal chain that were not examined by the current study

and have not been adequately examined by prior studies. Future research needs to more

97

fully examine both the proposed antecedents of climate (e. g., organizational policies,
procedures, leadership practices, culture, strategy, etc.), additional outcomes of climate
(e. g., motivation, behaviors, job attitudes, other indicators of organizational performance,
etc.), and potential moderators at each stage (e. g., contextual factors, climate strength,
climate configurations, etc.). While numerous studies have investigated these causal
pathways and moderators, many of them have failed to include the design features
necessary to truly untangle the causal pathways implied by the various models (Paauwe
& Boselie, 2005; Wright & Haggerty, 2005). Research also needs to be conducted that
more fully explores the multi-level nature of these relationships by using designs that
include assessments of top-down effects of organizational features on individual’s
climate perceptions, motivation, behaviors, and attitudes and investigates the bottom-up
processes whereby these individual-level constructs interact and combine to yield
organizational-level performance. By fully explicating these multi-level causal chains,
organizational researchers and practitioners can begin to identify the most impactful
interventions for increasing organizational performance.
Limitations

As with all research, there exist a number of limitations that should be considered
when interpreting the results of the current study. First, the research was conducted with
archival data and, therefore, the researcher could not influence the availability of data and
the measures used. This limitation manifests itself most obviously in the asymmetry of
the time periods over which data were collected. While efforts were made to align the
time periods in analyzing the data, it is possible that if the data had been aligned at the

outset, the results may have been somewhat different. However, if this was a major

98

 

problem, then it is unlikely that the results observed for customer satisfaction would have
been so consistent across the two departments. The presence of missing sales data for
2002, and the resulting requirement that the analyses for the affected time periods were
composed of only the last and first two quarters of the preceding and following period
also may have introduced some degree of error into the analyses involving vehicle sales.
However, there were no obvious differences between the correlations for these time
periods and the periods for which there was complete data.

A common limitation of organizational-level research is limited sample sizes. As
mentioned previously, this is a limitation of the current study as well. The largest
problem with smaller sample sizes is low power to detect significant effects, but in the
context of structural equation modeling it can also result in over— or under-estimates of
standard errors (Muthén & Muthén, 2002). While the lower power for detecting
significant effects does not impact the findings regarding causal priority, it is possible
that with larger sample sizes reciprocal effects similar to those found by other researchers
(e.g., Schneider et al., 1998) would have been observed in this study. Due to the effects of
a small sample size on standard errors, an alternative explanation for the prediction of
vehicle sales is that the finding of a small significant effect is due to under-estimated
standard errors for the sales variable. The replication of the findings for customer
satisfaction across both departments and the larger size of this effect suggest that
misestimated standard errors are a less plausible explanation for this outcome.

Third, it is important to highlight that the causal claims are based on non-
experimental data, which reduces the strength of causal inferences. Although the

collection of longitudinal data and the use of cross-lagged panel analyses strengthens

99

claims of causal priority, other longitudinal designs including, at least, quasi-
experimental features should be conducted to corroborate the current findings. The
current study included one quasi-experimental design feature, a non-equivalent control
group, and hypothesized that this control group would have lower customer satisfaction
at the conclusion of the study than the group participating in the climate survey and
action planning process. The lack of support for this hypothesis makes it difficult to
conclude that the survey and action planning processes were responsible for the
increasing climate levels observed for the participating departments, and may be seen as
limiting the strength of causal inferences made on the basis of the other analyses.
However, since nothing other than customer satisfaction level was known about the
control group in this study, it is impossible to know whether self-selection factors (i.e.,
choosing not to participate in the climate survey) or other unmeasured factors resulted in
the observed increases in customer satisfaction, and perhaps unobserved increases in
climate levels, over the period of the study. This highlights the need for higher quality
quasi-experimental design features (e. g., equivalent control groups, switching
replications, etc.) in future studies, especially those investigating the effectiveness of
interventions for changing climate.
Conclusion

In summary, this study contributed to research and theory on the causal
relationship between organizational climate and performance in several ways. First,
although no single study can definitely determine causality, the results of the current
investigation provide evidence that organizational climate has causal priority over some

indicators of organizational performance (i.e., customer satisfaction and sales). Given the

100

 

theoretical placement of organizational climate as a mediator between HR practices and
organizational performance, the establishment of this causal link is critical for further
demonstrating the value of the HR function in contributing to organizational
performance. Second, the results demonstrate the value of continuing to research and
assess molar organizational climate, suggesting it is too soon to wholesale reject
examinations of molar climate in favor of the more narrow conceptualizations advocated
by some researchers (e.g., Schneider, 1975; Schneider, 1990). Finally, by demonstrating
that the causal impact of climate emerged, and disappeared, over the examination of
different time lags, this study highlights the importance of considering time in future
longitudinal investigations of these relationships.

Future research on climate and the other outcomes of the HR function (e.g.,
selection, training, etc.) needs to continue to make the case that HR is a strategic
department with a critical role in the effectiveness of the organization. While platitudes
such as “our people are our most important asset” suggest that organizations value HR,
only by proving that the outcomes of the HR function have important impacts on
organizational performance will HR truly come to be viewed as a strategic partner critical

to the success of organizations.

10]

 

Appendix
Covariance Matrices for Measurement Invariance Analyses

 

 

 

oow. mo... wow. mo... ooo. owo. _.oo. Nwo. Poo. ooo. woo. woo. owo. owo. oeo. owo. omEE. :23: .9.
mo... mow. ooF. mow. nwo. owo. owo. owo. wno. ooo. ooo. Noo. owo. hmo. FNo. owo. omE_._.b___nEamD< .o_.
wo... oow. oor. oow. owo. owo. woo. wwo. Nno. Noo. ooo. ooo. o_.o. owo. owo. owo. omE_._.>o:9m_mcoo .3.
mo... mor. oow. oo... oo. owo. Noo. owo. ooo. ooo. Noo. ooo. o_.o. NNo. oro. NNo. ooE_.::oEm>_o>c_ .ow
ooo. hwo. owo. Foo. mop. oo... Fww. _.w_.. ooo. ooo. woo. woo. Eo. ooo. owo. ..oo. meE. co_mm__2 .Nr
owo. owo. owo. owo. oo_.. oor. om... our. wwo. owo. ..wo. nwo. ooo. nwo. ooo. Nwo. on_._.b___nEamu<.:
Foo. owo. ..oo. Noo. ewe. oNF. on... oov. ooo. ooo. ooo. oo. oo. owo. owo. owo. me_._.>o:9m_mcoO .o_.
Nwo. owo. wwo. owo. _.w_.. our. 02. oov. ooo. moo. ooo. ooo. woo. ..oo. ooo. Noo. me_._.EmEm>_o>c_ .o
Foo. wno. Nno. ooo. ooo. Vwo. ooo. ooo. tN. oow. ot. mt. ooo. owo. moo. wno. omEE. co_mm=2 .o
ooo. ooo. Noo. ooo. ooo. owo. ooo. moo. oo_.. oo... ..wv. _.o_.. moo. noo. ooo. ooo. ooEE. b=ESQNO< N
Boo. ooo. ooo. woo. woo. Fwo. ooo. ooo. ot. ..w—a No... mo... woo. ooo. ooo. ooo. omEF >oc2m_mcoo .o
woo. woo. ooo. ooo. voo. hwo. voo. ooo. om... _.o_.. No... ht. who. woo. woo. owo. mmEr—LcmEmZoE. .o
owo. oo. oo. o_.o. who. ooo. woo. woo. ooo. moo. woo. owo. vow. tr. oo... om... rmEE. co_wm=2 .w
oNo. NNo. owo. NNo. ooo. Nwo. owo. Foo. ouo. noo. ooo. ..oo. 5... nor. ..oo. ooo. ..mEP b___nflamb<.o
9.0. _.No. owo. owo. owo. ooo. owo. ooo. noo. ooo. moo. woo. mo... 50. No... No... FmE_._.>ocoum_mcoo .N
oNo. owo. oNo. NNo. ..oo. Nwo. owo. Noo. who. ooo. ooo. who. om—x ooo. No... o3. ..oEFEmEoZoE. .F
or o_. w—. or NF 3. o_. o o n o o w o N F

 

momzmcw. 8:99;. 20693822 .8 x392 oocmtm>oo EoEtmaoo mofio ”mm 2an

102

 

 

oo_.. 2.... o:. w_._.. ooo. ooo. owo. _.wo. moo. ooo. ooo. woo. ooo. ooo. woo. ooo.
_._._.. ooo. ooo. oo_.. ooo. woo. moo. ooo. wwo. owo. mwo. owo. moo. _.oo. ooo. omo.
o_._.. ooo. 3.... oor. moo. _.oo. owo. ooo. owo. wwo. ooo. ooo. Noo. _.oo. ooo. ooo.
win oo_.. oo... o3. wwo. ooo. wwo. owo. owo. mwo. Foo. Noo. ooo. ooo. ooo. moo.
ooo. ooo. moo. wwo. NF... _.oo. ooo. ooo. ooo. owo. owo. owo. wwo. ooo. ooo. owo.
ooo. woo. _.oo. ooo. Foo. Foo. omo. moo. owo. moo. ooo. ooo. woo. owo. Noo. ooo.
owo. moo. owo. wwo. ooo. omo. ooo. ooo. ooo. ooo. ooo. wwo. Noo. omo. Noo. ooo.
_.wo. ooo. ooo. owo. ooo. woo. ooo. Noo. owo. ooo. owo. owo. ooo. _.oo. ooo. ooo.
moo. wwo. owo. owo. ooo. owo. ooo. owo. oo... omo. _.oo. ooo. Nwo. moo. ooo. wwo.
ooo. owo. wwo. mwo. owo. moo. ooo. ooo. omo. vmo. omo. mmo. ooo. ooo. woo. woo.
ooo. mwo. ooo. Poo. owo. ooo. ooo. owo. _.oo. omo. mmo. ooo. woo. ooo. woo. moo.
woo. owo. ooo. Noo. owo. ooo. _.wo. owo. ooo. mmo. ooo. _.oo. owo. ooo. ooo. wwo.
ooo. woo. woo. ooo. wwo. woo. woo. ooo. Nwo. ooo. woo. owo. ooo. moo. ooo. omo.
ooo. oo. _.oo. ooo. ooo. omo. owo. Foo. moo. ooo. ooo. ooo. moo. ooo. ooo. woo.
woo. ooo. ooo. ooo. ooo. woo. woo. ooo. ooo. woo. woo. ooo. ooo. ooo. ooo. _.mo.
ooo. omo. ooo. moo. owo. ooo. ooo. ooo. _.wo. woo. moo. wwo. omo. woo. _.mo. ooo.
o_. o_. w_. or N_. E. o_. o o m o o w o N _.

o 9.5. 576:2 .

m we: £5982 .
o 95... 5533.80 .
o oE_._. EoEo>_o>c_ .
w 95... co_mm__2 .

e 9:: £5382 .
w 95% 5563.50 .
w oEP EoEm>_o>c_ .
m we; :23: .

m 95 $3983. .
o 95... 533550 .
m mEF .ceEm>_o>c_ .
F 95m :23: .

F as; b___neamu< .

F 95... 555.800 .
F 92:. E95295 .

FNCOVLOCO
FFv-Pt-F

FNCOVIDCDINCDQO
‘—

 

momzmc< 85252.. EmEoSmmos. .8 5.65. mocmtm>oo EoEtmqwo 85.3 new 033.

103

References

Abelson, R. P. (1985). A variance explanation paradox: When a little is a lot.
Psychological Bulletin, 97(1), 1 29-1 33.

Alderfer, C. P. (1972). Existence, relatedness and growth: Human needs in
organizational settings. New York: Free Press.

Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In J. Kuhl &
J. Beckman (Eds), Action control: From cognition to behavior (pp. 11-39).
Heidelberg: Springer.

Aj zen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behavior.
Englewood Cliffs, NJ: Prentice-Hall.

Anderson, S. E., & Williams, S. J. (1992). Assmnptions about unmeasured variables with
studies of reciprocal relationships: The case of employee attitudes. Journal of
Applied Psychology, 77(5), 63 8-650.

Appelbaum, E., Bailey, T., Berg, P. and Kalleberg, A. (2000). Manufacturing advantage:
Why high performance work systems pay ofl. Ithaca, NY: ILR Press.

Amdt, J. (1967). Role of product-related conversations in the diffusion of a new product.
Journal of Marketing Research, 4, 291-295.

Ashkanasy, N. M., Broadfoot, L. E., & Falkus, S. (2000). Questionnaire measures of
organizational culture. In N. M. Ashkanasy, C. P. M. Wilderom & M. F. Peterson
(Eds), Handbook of organizational culture & climate (pp. 131-146). Thousand
Oaks, CA: Sage.

Becker, B.E., Huselid, M.A., Pickus, P.S., & Spratt, M. HR as a source of shareholder
value: Research and recommendations. Human Resource Management Journal,
31(1), Spring, 1997, 39-47.

Becker, B. E., Huselid, M. A., & Ulrich, D. (2001). The HR scorecard: Linking people,
strategy, and performance. Boston: Harvard Business School Press.

Bentler, P., & Bonett, D. (1980). Significance tests and goodness of fit in the analysis of
structures. Psychological Bulletin, 88(3), 588-606.

Blau, P. M. (1964). Exchange and power in social life. New York: John Wiley & Sons.

Bliese, P. D. (1998). Group size, ICC values, and group-level correlations: A simulation.
Organizational Research Methods, 1(4), 355-373.

Bliese, P. D. (2000). Within-group agreement, non-independence, and reliability:

Implications for data aggregation and analysis. In K. J. Klein & S. W. J.
Kozlowski (Eds), Multilevel theory, research, and methods in organizations:

104

Foundations, extensions, and new directions (pp. 349-381). San Francisco, CA:
Jossey-Bass.

Bollen, K. (1989). Structural equations with latent variables. Oxford England: John
Wiley & Sons.

Borucki, C. C., & Burke, M. J. (1999). An examination of service-related antecedents to
retail store performance. Journal of Organizational Behavior, 20(6), 943-962.

Brown, S., & Lam, S. (2008). A meta-analysis of relationships linking employee
satisfaction to customer responses. Journal of Retailing, 84(3), 243-255.

Bowers, D. G. (1973). OD techniques and their results in 23 organizations: The Michigan
ICL study. Journal of Applied Behavioral Science, 9(1), 21-43.

Brown, L. D. (1972). "Research action": Organizational feedback, understanding, and
change. Journal of Applied Behavioral Science, 8(6), 697-711.

Brown, S. P., & Leigh, T. W. (1996). A new look at psychological climate and its
relationship to job involvement, effort, and performance. Journal of Applied -
Psychology, 81(4), 358-368.

 

Burke, M., Borucki, C., & Hurley, A. (1992). Reconceptualizing psychological climate in
a retail service environment: A multiple-stakeholder perspective. Journal of
Applied Psychology, 77(5), 717-729.

Byme, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS:
Basic concepts, applications, and programming. Mahwah, NJ: Erlbaum.

Cameron, K. S., & Quinn, R. E. (1999). Diagnosing and changing organizational
culture. Reading, PA: Addison-Wesley.

Campbell, D. T., & Pritchard, R. (1976). Motivation theory in industrial and
organizational psychology. In M. Dunnette (Ed.), Handbook of industrial and
organizational psychology (pp. 63-130). Chicago: Rand McNally.

Campbell, J. J ., Dunnette, M. D., Lawler, E. E., & Weick, K. E. (1970). Managerial
behavior, performance, and effectiveness: New York, NY, US; McGraw-Hill.

Carr, J. Z., Schmidt, A. M., Ford, J. K., & DeShon, R. P. (2003). Climate perceptions
matter: A meta-analytic path analysis relating molar climate, cognitive and

affective states, and individual level work outcomes. Journal of Applied
Psychology, 88(4), 605-619.

Cheung, G. W., & Rensvold, R. B. (1998). Cross-cultural comparisons using non-
invariant measurement items. Applied Behavioral Science Review, 6(1), 93-110.

105

Church, A. H., & Waclawski, J. (1998). Designing and using organizational surveys: A
seven-step process. San Francisco: Jossey-Bass.

Day, D. V., & Bedeian, A. G. (1991). Work climate and type a status as predictors of job

satisfaction: A test of the interactional perspective. Journal of Vocational
Behavior, 38(1), 39-52.

Denison, D. R. (1990). Corporate culture and organizational ejfectiveness. New York:
John Wiley & Sons.

Denison, D. R., Janovics, J ., Young, J ., & Cho, H. J. (2006). Diagnosing organizational
cultures: Validating a model and a method: Unpublished manuscript.

Denison, D. R., & Mishra, A. K. (1995). Toward a theory of organizational culture and
effectiveness Organization Science, 6(2), 204-223.

Dooney, J ., & Smith, N. (2006). Society for human resource management human capital
benchmarking study. Alexandria, VA: Society for Human Resource Management.

Elizur, D. (1984). Work values: A structural analysis of work outcomes. Journal of
Applied Psychology, 69(3), 379-389.

Festinger, L. (1957). A theory of cognitive dissonance. Stanford University Press.

French, W. L., & Bell, C. H. (1995). Organizational development. Englewood Cliffs, NJ:
Prentice-Hall.

Friedlander, F., & Margulies, N. (1969). Multiple impacts of organizational climate and
individual value systems upon job satisfaction. Personnel Psychology 22(2), 171-
1 83. .

Gelade, G. A., & Ivery, M. (2003). The impact of human resource management and work
climate on organizational performance. Personnel Psychology, 56(2), 383-404.

Grewal, D., & Shanna, A. (1991 ). The effect of salesforce behavior on customer

satisfaction: An interactive framework. The Journal of Personal Selling & Sales
Management, 11(3), 13-23.

Griffeth, R., Hom, P., & Gaertner, S. (2000). A meta-analysis of antecedents and
correlates of employee turnover: Update, moderator tests, and research
implications for the next millennium. Journal of Management, 26(3), 463-488.

Guest, D. (1997). Human resource management and performance: a review and research
agenda. International Journal of Human Resource Management, 8(3), 263 -276.

Guest, D., Michie, J ., Conway, N., & Sheehan, M. (2003). Human resource management
and corporate performance in the UK. British Journal of Industrial Relations,
41(2), 291-314.

106

Hatter, J ., Schmidt, F ., & Hayes, T. (2002). Business-unit-level relationship between
employee satisfaction, employee engagement, and business outcomes: A meta-
analysis. Journal of Applied Psychology, 87(2), 268-279.

Hackman, J. R., & Oldham, G. R. (1976). Motivation through the design of work: Test of
a theory. Organizational Behavior and Human Decision Processes, 16, 250-279.

Hays, R., Marshall, G., Wang, E., & Sherbourne, C. (1994). Four-year cross-lagged
associations between physical and mental health in the medical outcomes study.
Journal of Consulting and Clinical Psychology, 62(3), 441-449.

Hermstein, R. J. (1970). On the law of effect. Journal of the Experimental Analysis of
Behavior, 13, 243—266.

Heskett, J. L., Jones, T. O., Loveman, G. W., Sasser, W. E., & Schlesinger, L. A. (1994).
Putting the service-profit chain to work. Harvard Business Review, 72(2), 164-
174.

Heskett, J. L., Sasser, W. E., & Schlesinger, L. A. (1997). The service profit chain. New
York: The Free Press.

Hofmann, D. A., & Stetzer, A. (1996). A cross-level investigation of factors influencing
unsafe behaviors and accidents. Personnel Psychology, 49(2), 307-339.

Hu, L., & Bentler, P. M. (1998). Fit indices in covariance structure modeling: Sensitivity

to underparameterized model misspecification. Psychological Methods, 3(4), 424-
453.

Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives. Structural Equation
Modeling, 6(1), 1-55.

Huselid, M. A. (1995). The impact of human resource management practices on turnover,
productivity, and corporate financial performance. Academy of Management
Journal, 38(3), 635-672.

Huselid, M. A., Jackson, S. E., & Schuler, R. S. (1997). Technical and strategic human
resource management effectiveness as determinants of firm performance.
Academy of Management Journal, 40(1), 171-188.

James, L. A., & James, L. R. (1989). Integrating work environment perceptions:
Explorations into the measurement of meaning. Journal of Applied Psychology,
74(5), 739-751.

James, L. R. (1982). Aggregation bias in estimates of perceptual agreement. Journal of
Applied Psychology, 67(2), 219-229.

107

James, L. R. , Choi, C., Ko, C., McNeil, P., Minton, M., Wright, M., et al. (2008).
Organizational and psychological climate: A review of theory and research.
European Journal of Work and Organizational Psychology, 17(1), 5-32.

James, L. R., Demaree, R. G., & Wolf, G. (1984). Estimating within-group interrater
reliability with and without response bias. Journal of Applied Psychology, 69(1),
85-98.

James, L. R., Hartman, A., Stebbins, M. W., & Jones, A. P. (1977). Relationships
between psychological climate and a vie model for work motivation. Personnel
Psychology, 30(2), 229-254.

James, L. R., & Jones, A. P. (1974). Organizational climate: A review of theory and
research. Psychological Bulletin, 81 (12), 1096-1112.

James, L. R., & Jones, A. P. (1976). Organizational structure: A review of structural
dimensions and their conceptual relationships with individual attitudes and
behavior. Organizational Behavior & Human Performance, 16(1), 74-113.

Katz, D. (1964). The motivational basis of organizational behavior. Behavioral Science
9(2), 131-146.

Katz, R., & Van Maanen, J. (1977). The loci of work satisfaction: Job, interaction, and
policy. Human Relations, 30(3), 469-486.

Kenny, D. A., & Harackiewicz, J. M. (1979). Cross-lagged panel correlation: Practice
and promise. Journal of Applied Psychology, 64(4), 372-379.

Kessler, R. C., & Greenberg, D. F. (1981). Linear panel analysis: Models of quantitative
change. New York: Academic Press.

Kline, R. B. (1998). Principles and practice of structural equation modeling. New York:
Guilford Press.

Kohli, A. K., & J aworski, B. J. (1990). Market orientation: The construct, research
propositions, and managerial implications. Journal of Marketing, 54(2), 1-18.

Kopelman, R. E., Brief, A. P., & Guzzo, R. A. (1990). The role of climate and culture in
productivity. In B. Schneider (Ed.), Organizational climate and culture (pp. 282-
318). San Francisco, CA: Jossey-Bass.

Koys, D. J. (2001). The effects of employee satisfaction, organizational citizenship
behavior, and turnover on organizational effectiveness: A unit-level, longitudinal
study. Personnel Psychology, 54(1), 101-114.

Kozlowski, S. W. J ., & Doherty, M. L. (1989). Integration of climate and leadership:
Examination of a neglected issue. Journal of Applied Psychology, 74(4), 546-553.

108

Kozlowski, S. W. J ., & Farr, J. L. (1988). An integrative model of updating and
performance. Human Performance, 1, 5-29.

Kozlowski, S. W. J ., & Hattrup, K. (1992). A disagreement about within-group

agreement: Disentangling issues of consistency versus consensus. Journal of
Applied Psychology, 77(2), 161-167.

Kozlowski, S. W. J ., & Hults, B. M. (1987). An exploration of climates for technical
updating and performance. Personnel Psychology, 40(3), 539-563.

Kozlowski, S. W. J ., & Klein, K. J. (2000). A multilevel approach to theory and research
in organizations: Contextual, temporal, and emergent processes. In K. J. Klein &
S. W. J. Kozlowski (Eds), Multilevel theory, research, and methods in
organizations: Foundations, extensions, and new directions (pp. 3-90). San
Francisco, CA: Jossey-Bass.

LaBarbera, P. A., & Mazursky, D. (1983). A longitudinal assessment of consumer

satisfaction/dissatisfaction: The dynamic aspect of the cognitive process. Journal
of Marketing Research, 20(4), 393-404. ‘

Lance, C., Butts, M., & Michels, L. (2006). The sources of four commonly reported

cutoff criteria: What did they really say?. Organizational Research Methods, 9(2),
202-220.

Lawler, E. E., Hall, D. T., & Oldham, G. R. (1974). Organizational climate: Relationship
to organizational structure, process and performance. Organizational Behavior &
Human Performance Vol. 11(1), 139-155.

LePine, J. A., Erez, A., & Johnson, D. E. (2002). The nature and dimensionality of
organizational citizenship behavior: A critical review and meta-analysis. Journal
of Applied Psychology, 87(1), 52-65.

Lev, B. (2001). Intangibles management, measurement, and reporting. Washington, DC:
Brookings Institution Press.

Lewin, K. (1951). Field theory in social science. Oxford, England: Harpers.

Likert, R. (1961). New patterns of management: New York, NY, US; McGraw-Hill.

Lipsey, M. (1990). Design sensitivity: Statistical power for experimental research.
Thousand Oaks, CA US: Sage Publications, Inc. Retrieved from PsycINFO
database.

Litwin, G. H., & Stringer, R. A., Jr. (1968). Motivation and organizational climate:
Oxford, England; Harvard U., Graduate School of Busi.

109

Locke, E. A. (1976). The nature and causes of job satisfaction. In M. Dunnette (Ed.),

Handbook of industrial and organizational psychology (pp. 1297-1349). Chicago:
Rand McNally.

Locke, E. A., & Latham, G. P. (1990). A theory of goal setting & task performance:
Upper Saddle River, NJ, US; Prentice-Hall, Inc.

Madon, S., Willard, J ., Guyll, M., Trudeau, L., & Spoth, R. (2006). Self-fulfilling
prophecy effects of mothers' beliefs on children's alcohol use: Accumulation,

dissipation, and stability over time. Journal of Personality and Social Psychology,
90(6), 911-926.

Mathieu, J. E., & Kohler, S. S. (1990). A cross-level examination of group absence
influences on individual absence. Journal of Applied Psychology, 75(2), 217-220.

Mathieu, J. E., & Zajac, D. M. (1990). A review and meta-analysis of the antecedents,

correlates, and consequences of organizational commitment. Psychological
Bulletin, 108(2), 171-194.

Maxham, J. G. (2001). Service recovery's influence on consumer satisfaction, positive
word-of-mouth, and purchase intentions. Journal of Business Research, 54(1), 11-
24.

Maxham, J. G., & Netemeyer, R. G. (2003). Firms reap what they sow: The effects of
shared values and perceived organizational justice on customers' evaluations of
complaint handling. Journal of Marketing, 67(1), 46-62.

McGregor, D. (1960). The human side of enterprise: New York, NY, US; McGraw-Hill.

Meyer, J., & Herscovitch, L. (2001). Commitment in the workplace: Toward a general
model. Human Resource Management Review, 11(3), 299-326.

Meyer, J. P., Stanley, D. J ., Herscovitch, L., & Topolnytsky, L. (2002). Affective,
continuance, and normative commitment to the organization: A meta-analysis of

antecedents, correlates, and consequences. Journal of Vocational Behavior, 61(1),
20—52.

Mowday, R., Porter, L., & Steers, R. (1982). Employee-organization linkages. New
York: Academic Press.

Muchinsky, P. (1977). Organizational communication: Relationships to organizational
climate and job satisfaction. Academy of Management Journal, 20(4), 592-607.

Muthén, L.K. and Muthén, ED. (2002). How to use a Monte Carlo study to decide on
sample size and determine power. Structural Equation Modeling, 4, 599-620.

NADA. (2007). Sales Trends. Retrieved from http://www.nada.org/NR/rdonlyres
/CDDE9FFF-73B0-4F15-AEDA-DF00723A25CF/O/June07SalesTrends.pdf.

110

Nadler, D. A. (1996). Setting expectations and reporting results: Conversations with top
management. In A. I. Kraut (Ed.), Organizational surveys (pp. 177-203). San
Francisco: Jossey-Bass.

Naumann, S. E., & Bennett, N. (2000). A case for procedural justice climate:
Development and test of a multilevel model. Academy of Management Journal,
43(5), 881-889.

Neal, A., & Griffin, M. A. (2006). A study of the lagged relationships among safety
climate, safety motivation, safety behavior, and accidents at the individual and
group levels. Journal of Applied Psychology, 91(4), 946-953.

Nicholas, J. M. (1982). The comparative impact of organization development

interventions on hard criteria measures. Academy of Management Review, 7, 53 l-
542.

Oliver, A., Cheyne, A., Tomas, J. M., & Cox, S. L. (2002). The effects of organizational
and individual factors on occupational accidents. Journal of Occupational and
Organizational Psychology, 75(4), 473-488.

Ostroff, C. (1992). The relationship between satisfaction, attitudes, and performance: An
organizational level analysis. Journal of Applied Psychology, 77(6), 963-974.

Ostroff, C. (1993). The effects of climate and personal influences on individual behavior
and attitudes in organizations. Organizational Behavior and Human Decision
Processes, 56(1), 56-90.

Ostroff, C., & Bowen, D. E. (2000). Moving HR to a higher level: HR practices and
organizational effectiveness. In K. J. Klein & S. W. J. Kozlowski (Eds),
Multilevel theory, research, and methods in organizations: Foundations,
extensions, and new directions (pp. 211-266). San Francisco, CA: Jossey-Bass.

Ostroff, C., Kinicki, A. J ., & Tamkins, M. M. (2003). Organizational culture and climate.
In W. C. Borman, D. R. Ilgen & R. J. Klimoski (Eds), Handbook of psychology:

Industrial and organizational psychology (Vol. 12, pp. 565-593). Hoboken, NJ:
John Wiley & Sons.

Paauwe, J. (2009). HRM and performance: Achievements, methodological issues and
prospects. Journal of Management Studies, 46(1), 129-142.

Paauwe, J ., & Boselie, P. (2005). HRM and performance: What next?. Human Resource
Management Journal, 15(4), 68-83.

Parker, C. (1999). A test of alternative hierarchical models of psychological climate:

PC[sub]g[/sub], satisfaction, of common method variance?. Organizational
Research Methods, 2(3), 257-274.

111

Parker, C. P., Baltes, B. B., Young, S. A., Huff, J. W., Altmann, R. A., Lacost, H. A., et
al. (2003). Relationships between psychological climate perceptions and work

outcomes: A meta-analytic review. Journal of Organizational Behavior, 24(4),
389-416.

Payne, R. L., & Pugh, D. D. (1976). Organizational structure and climate. In M. Dunnette
(Ed.), Handbook of industrial and organizational psychology (pp. 1125-1172).
Chicago: Rand McNally.

Perrow, C. (1970). Organizational analysis: A sociological veiw. Belmont, CA:
Wadsworth.

Price, L. L., & Feick, L. F. (1984). The role of interpersonal sources in external search:
An informational perspective. In T. Kinnear (Ed), (Vol. 2, pp. 250-255). Provo,
UT: Association for Consumer Research.

Pritchard, R. D., & Karasick, B. W. (1973). The effects of organizational climate on
managerial job performance and job satisfaction. Organizational Behavior &
Human Performance Vol. 9(1), 126-146.

Quinn, R. E., & Rohrbaugh, J. (1983). A spatial model of effectiveness criteria: Towards

a competing values approach to organizational analysis. Management Science, 49,
917-947.

Raju, N., Laffitte, L., & Byme, B. (2002). Measurement equivalence: A comparison of
methods based on confirmatory-factor analysis and item response theory. Journal
of Applied Psychology, 87(3), 517-529.

Rentsch, J. R. (1990). Climate and culture: Interaction and qualitative differences in
organizational meanings. Journal of Applied Psychology, 75(6), 668-681.

Richins, M. L. (1983). Negative word-of-mouth by dissatisfied consumers: A pilot study.
Journal of Marketing, 47(1), 68-78.

Richins, M. L. (1987). A multivariate analysis of responses to dissatisfaction. Journal of
the Academy of Marketing Science, 15, 24-31.

Rosnow, R., Rosenthal, R., & Rubin, D. (2000). Contrasts and correlations in effect-size
estimation. Psychological Science, 11(6), 446-453.

Russ-Efi, D., & Preskill, H. (2005). In search of the holy grail: Return on investment

evaluation in human resource development. Advances in Developing Human
Resources, 7(1), 71-85.

Ryan, A. M., Horvath, M., & West, B. (2003). The survey feedback process in employee
opinion surveying: Unpublished manuscript.

112

Ryan, A., & Ployhart, R. (2003). Customer service behavior. Handbook of psychology:
Industrial and organizational psychology, Vol. 12 (pp. 377-397). Hoboken, NJ
US: John Wiley & Sons

Ryan, A. M., Schmit, M. J ., & Johnson, R. (1996). Attitudes and effectiveness:
Examining relations at an organizational level. Personnel Psychology, 49(4), 853-
882.

Schein, E. H. (1985). Organizational culture and leadership (1 ed.). San Francisco:
Jossey-Bass.

Schein, E. H. (1990). Organizational culture. American Psychologist, 45, 109-119.

Schein, E. H. (1992). Organizational culture and leadership (2 ed.). San Francisco:
Jossey-Bass.

Schein, E. H. (2000). Sense and nonsense about culture and climate. In N. M. Ashkanasy,
C. P. M. Wilderom & M. F. Peterson (Eds), Handbook of organizational culture
& climate (pp. xxiii-xxx). Thousand Oaks, CA: Sage.

Schiffman, L. G. (1971). Sources of information for the elderly. Journal of Advertising
Research, 11, 33-37.

Schmit, M. J ., & Allscheid, S. P. (1995). Employee attitudes and customer satisfaction:

Making theoretical and empirical connections. Personnel Psychology, 48(3), 521-
536.

Schneider, B. (1975). Organizational climates: An essay. Personnel Psychology, 28(4),
447-479.

Schneider, B. (1987). The people make the place. Personnel Psychology, 40(3), 43 7-453.

Schneider, B. (1990). The climate for service: An application of the climate construct. In
B. Schneider (Ed.), Organizational climate and culture (pp. 383-412). San
Francisco, CA: Jossey-Bass.

Schneider, B., & Bowen, D. E. (1985). Employee and customer perceptions of service in
banks: Replication and extension. Journal of Applied Psychology, 70(3), 423-433.

Schneider, B., Ehrhart, M. G., Mayer, D. M., Saltz, J. L., & Niles-Jolly, K. (2005).
Understanding organization-customer links in service settings. Academy of
Management Journal, 48(6), 1017-1032.

Schneider, B., Goldstein, H. W., & Smith, D. B. (1995). The asa framework: An update.
Personnel Psychology Theory and literature, 48(4), 747-773.

113

Schneider, B., Hanges, P. J ., Smith, D. B., & Salvaggio, A. N. (2003). Which comes first:
Employee attitudes or organizational financial and market performance? Journal
of Applied Psychology, 88(5), 836-851.

Schneider, B., Parkington, J. J ., & Buxton, V. M. (1980). Employee and customer
perceptions of service in banks. Administrative Science Quarterly, 25, 252-267.

Schneider, B., & Reichers, A. E. (1983). On the etiology of climates. Personnel
Psychology, 36(1), 19-39.

Schneider, 3., White, S. S., & Paul, M. C. (1998). Linking service climate and customer
perceptions of service quality: Tests of a causal model. Journal of Applied
Psychology, 83(2), 1 50-163.

Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-
experimental designs for generalized causal inference: Boston, MA, US;
Houghton, Mifflin and Company.

Shrout, P., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies:
New procedures and recommendations. Psychological Methods, 7(4), 422-445.

Shostack, G. (1977). Breaking Free from Product Marketing. Journal of Marketing, 41,
73-80.

Smith, M. E. (2003). Another road to evaluating knowledge assets. Human Resource
Development Review, 2(1), 6-25.

deerlund, M. (2002). Customer familiarity and its effects on satisfaction and behavioral
intentions. Psychology & Marketing Scandinavian experiences, 19(10), 861-880.

Solomon, R. J. (1976). An examination of the relationship between a survey feedback
O.D. Technique and the work environment. Personnel Psychology, 29(4), 583-
594.

Sowinski, D. R., F ortmann, K. A., Lezotte, D. V. (2008). Climate for service and the
moderating effects of climate strength on customer satisfaction, voluntary

turnover, and profitability. European Journal of Work and Organizational
Psychology, 1 7(1), 73-88.

Swan, J. E., & Oliver, R. L. (1989). Postpurchase communications by consumers.
Journal of Retailing, 65(4), 516-533.

Thorndike, E. L. (191 1). Animal Intelligence: Experimental Studies. New York:
MacMillan.

Vandenberg, R. J ., & Lance, C. E. (2000). A review and synthesis of the measurement

invariance literature: Suggestions, practices, and recommendations for
organizational research. Organizational Research Methods, 3(1), 4-69.

114

Vroom, V. H. (1964). Work and motivation. Oxford, England: Wiley.
Weick, K. E. (1995). Sensemaking in organizations. Thousand Oaks, CA: Sage.

Weiss, H. M., & Cropanzano, R. (Eds). (1996). Affective events theory: A theoretical
discussion of the structure, causes and consequences of affective experiences at
work. US: Elsevier Science/JAI Press.

Wiley, J. W. (1996). Linking survey results to customer satisfaction and business
performance. In A. I. Kraut (Ed.), Organizational surveys: Tools for assessment
and change (pp. 330-359). San Francisco: Jossey-Bass.

Wright, P. M., & Gardner, T. M. (2003). Theoretical and empirical challenges in studying
the HR practice-finn performance relationship. In D. Holman, T. D. Wall, C.
Clegg, P. Sparrow & A. Howard (Eds), The new workplace: A guide to the
human impact of modern working practices (pp. 311-328). Chichester: John
Wiley and Sons

Wright, P. M., Gardner, T. M., Moynihan, L. M., & Allen, M. R. (2005). The relationship
between hr practices and firm performance: Examining causal order. Personnel
Psychology, 58(2), 409-446.

Wright, P. M. and Haggerty, J. J. (2005), Missing variables in theories of strategic human

resource management: Time, cause, and individuals. The International Review of
Management Studies, 1 6(2), 164-1 73.

Yi, Y. (1990). A critical review of consumer satisfaction. In V. A. Zeithaml (Ed.), Review
of marketing (pp. 68-123). Chicago: American Marketing Association.

Zohar, D. (1980). Safety climate in industrial organizations: Theoretical and applied
implications. Journal of Applied Psychology, 65(1), 96-102.

Zohar, D. (2000). A group-level model of safety climate: Testing the effect of group
climate on microaccidents in manufacturing jobs. Journal of Applied Psychology,
85(4), 587-596.

115

   

MICHIGAN sure UNIVERSITY LIBRARIES
lllllllllllllllIlllllllllilllll|||lllllljlllllllllll|
3 1293 03063 5 78