Trust within Teams: the relative importance of Ability, Benevolence and Integrity Douglas A. Beatton BBus (Finance and Management) School of Management Queensland University of Technology A Dissertation in fulfilment of the requirements for the Degree of Master of Business (Research) 2007
187
Embed
Trust within Teams: the relative importance of Ability ...eprints.qut.edu.au/16651/1/Douglas_A._Beatton_Thesis.pdf · Trust within Teams: the relative importance of Ability, Benevolence
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Trust within Teams:
the
relative importance
of
Ability, Benevolence and Integrity
Douglas A. Beatton
BBus (Finance and Management)
School of Management
Queensland University of Technology
A Dissertation
in fulfilment of the requirements
for the
Degree of Master of Business (Research)
2007
iii
Abstract
Trust between team members is important: Research has shown that teams with
higher levels of trust have a propensity to be higher performers. This study built on
contemporary trust theory by examining initial interpersonal trust development
between a new team member and a newly formed work-team using experimental
rather than correlation-based survey methods. Undergraduate students from a
metropolitan Australian university participated in a vignette experiment examining
the effect of teams with varying levels of Ability, Benevolence and Integrity on trust
development. It was hypothesised that these antecedents of trust do not have similar
effect on our Intention to Trust as is currently depicted in Mayer, Davis and
Schoorman’s (1995) integrative model of organisational trust. Their model is
developed by hypothesising that the type and magnitude of the information we
receive about a trustee moderates the relationship between our Intention to Trust and
its antecedents. Initial examination of the traditional scales identified overlaps that
needed clarification. This was completed by informing existing scales and the
vignette manipulations with the context specific information that emerged from the
thematic analysis of structured interviews. Subsequent analyses of the questionnaire
data used ANOVA and Structural Equation Modelling techniques. In testing the
hypotheses, Ability was found to be most salient in the development of Intention to
Trust. This research contributes methodologically by developing a vignette-based
experimental method that improves the reliability of existing trust scales. The study
contributes theoretically by further explaining the salience of the trust antecedents
and practically by identifying that the judgment and decision-making of new work-
team members can be distorted by halo bias wherein they ignore the Benevolence
traits of team members of a group that exhibits high levels of Ability.
Table 1: Research results from a recent study by Serva et al. (2005) .................................... 43
Table 2: Effect and sample size estimates calculations from similar trust-related studies .... 57
Table 3: Examples of the low Cronbach Alpha results for the trust scales from recent studies............................................................................................................................................... 61
Table 4: Theoretical basis for the thematic analysis of the structured interview transcripts . 70
Table 5: Summary of results for the thematic analysis of the structured interview transcripts............................................................................................................................................... 71
Table 6: The five-point Likert-scaled Propensity to Trust items used in the Pilot Stage of the study....................................................................................................................................... 72
Table 7: The five-point Likert-scaled Intention to Trust items used in the Pilot and Final stages of this study................................................................................................................. 75
Table 8: Examples of the good Cronbach Alpha results for the Ability, Benevolence and Integrity scales used in recent studies.................................................................................... 76
Table 9: The five-point Likert-scaled Ability items used in the Pilot and Final stage of this study....................................................................................................................................... 76
Table 10: The five-point Likert-scaled Benevolence items used in the Pilot and Final stages of this study............................................................................................................................ 77
Table 11: The five-point Likert-scaled Integrity items used in the Pilot and Final stages of this study ................................................................................................................................ 77
Table 12: 2x2 Factorial design for the manipulation of the dimensions of Ability and Benevolence........................................................................................................................... 82
Table 13: 2x2 Factorial design for the manipulation of dimensions of Ability and Integrity 82
Table 14: Descriptive statistics and correlations from Study Stage 2, N = 187 .................... 87
Table 15: Comparison of ITT study results from this research and recent studies employing similar measurement scales based on the Mayer et al. (1995) theoretical foundation........... 87
Table 16: Ability information provided in the Pilot Study Low Ability vignette.................. 88
Table 17: Ability information provided in the Pilot Study High Ability vignette ................. 88
Table 18: Grading scale for the university population........................................................... 89
Table 20: Univariate statistics from the missing variable analysis of the Stage 3 data ......... 93
Table 21: Descriptive statistics and correlations from Study Stage 3, N = 224..................... 95
Table 22: Comparison of ITT study results from this research and recent studies employing similar measurement scales based on the Mayer et al. (1995) theoretical foundation; N = 224............................................................................................................................................... 96
Table 23: A comparison with recent trust studied reveals the improved scale reliability for Intention to Trust.................................................................................................................... 99
Table 24: Measures and suggested thresholds applied to assess SEM model fit................. 102
Table 26: Squared Multiple Correlations for the Y–Variables in the initial model with the RCMF; N = 224 ................................................................................................................... 107
viii
Table 27: Y–Variable Squared Multiple Correlations - the respecified ITT model; N = 224..............................................................................................................................................110
Table 28: Model-based measures of scale reliability - the respecified ITT model, N = 224110
Table 29: Summary of model fit improvement between the initial ITT model and the respecified ITT model; N = 224 ...........................................................................................110
Table 30: Squared Multiple Correlations for the X–Variables; N = 224..............................111
Table 31: Table A3.1: Squared Multiple Correlations for the X – Variables for the respecified Ability Congeneric model; N = 224...................................................................113
Table 32: Table A3.2: Model-based measures of scale reliability for the respecified Ability Congeneric model; N = 224..................................................................................................113
Table 33: Summary of model fit improvement between the initial and respecified Ability models; N = 224 ...................................................................................................................113
Table 34: Squared Multiple Correlations for X – Variables; N = 224 .................................114
Table 35: Squared Multiple Correlations for the X–Variables in the respecified Benevolence model....................................................................................................................................116
Table 36: Model-based measures of scale reliability for the respecified Benevolence model..............................................................................................................................................116
Table 37: Summary of Model Fit Improvement between the initial and respecified Benevolence models; N = 224..............................................................................................116
Table 38: Squared Multiple Correlations for the X – Variables in the initial Integrity model..............................................................................................................................................117
Table 39: Squared Multiple Correlations for the X – Variables in the initial Integrity model..............................................................................................................................................119
Table 40: Model-based measures of scale reliability for the respecified Integrity model....119
Table 41: Summary of Model Fit Improvement between the initial and respecified Integrity models; N = 224 ...................................................................................................................119
Table 42: Squared Multiple Correlations for the X–Variables for the respecified three-factor CFA model; N = 224 ............................................................................................................121
Table 43: Model-based measures of scale reliability for the respecified three-factor CFA model; N = 224.....................................................................................................................121
Table 44: Summary of model fit difference between the respecified three-factor CFA model and the χ2 threshold for 24 degrees of freedom at p = 0.05..................................................121
Table 45: Variable loading combinations for the comparison of rival First Order CFA Models..................................................................................................................................122
Table 46: Fit indices for the competing First Order CFA models of Ability, Benevolence and Integrity identifying the superiority of the three-factor oblique CFA model; N = 224 .......123
Table 47: Squared Multiple Correlations for the X and Y–Variables in the respecified path analysis model; N = 224.......................................................................................................125
Table 48: Summary of Model Fit Improvement between the initial and respecified path analysis; N = 224. .................................................................................................................125
Table 49: Summary of Model Fit Improvement between the respecified path analysis model and the re-specified path analysis model excluding item Int4; N = 224...............................126
Table 50: Fit indices identifying the superiority of the hypothesised three-factor path analysis model; N = 224.....................................................................................................................127
ix
Table 51: 2x2 Factorial design for the manipulation of the High and Low Information Levels for Ability and Benevolence ................................................................................................ 129
Table 52: 2x2 Factorial design for the manipulation of High and Low Information Levels for Ability and Integrity............................................................................................................. 129
Table 53: Descriptive Statistics of the Dependent Variable: Intention to Trust for Groups with High and Low Information Levels of Ability (ability) and Integrity (integ) ............... 131
Table 54: Descriptive Statistics of the Dependent Variable: Intention to Trust for Groups with High and Low Information Levels of Ability (ability) and Benevolence (benev)....... 134
x
xi
Table of Figures
Figure 1: Stages of small group development; adapted from Robbins (2006, p. 489)........... 22
Figure 2: Antecedent of Trust – Ability................................................................................. 25
Figure 3: Antecedents of Trust – Ability and Integrity.......................................................... 27
Figure 4: A proposed model of the antecedents of trust – Ability, Benevolence and Integrity............................................................................................................................................... 29
Figure 5: McKnight et al's. (1998) model of initial trust formation....................................... 31
Figure 6: Integrative model of organizational trust (Mayer et al., 1995, p. 715).................. 33
Figure 7: Example dimensions of Situational Strength; adapted from (Gill et al., 2005)...... 36
Figure 8: The proposed Information Levels of the trust antecedents..................................... 37
Figure 9: The proposed theoretical model for the study of how the antecedents of trust affect our Intention to Trust in the context of newly forming work-teams...................................... 38
Figure 10: The Three study Stages ........................................................................................ 47
Figure 11: Initial 3x2 factorial design depicting Information Levels of Ability, Benevolence and Integrity........................................................................................................................... 59
Figure 12: The Study Stages .................................................................................................. 65
Figure 13: Study Stage 1........................................................................................................ 67
Figure 14: Study Stage 2........................................................................................................ 85
Figure 15: Study Stage 3........................................................................................................ 91
Figure 16: Box plot of Intention to Trust for Stage 2 ............................................................ 98
Figure 17: Box plot of Intention to Trust for Stage 3 ............................................................ 98
Figure 18: A priori conceptualisation of the study model to be tested using LISREL 8.3 .. 101
Figure 19: Standardised Solution for the respecified Congeneric CFA Model of the endogenous factor Intention to Trust; items Itt2, Itt3, Itt7 and Itt9 excluded ...................... 109
Figure 20: Standardised solution for the respecified congeneric CFA model of the trust antecedent Ability; items Ab1, Ab5 and Ab6 excluded....................................................... 112
Figure 21: Standardised Solution for the Re-specified Congeneric CFA Model of the trust Antecedent Benevolence; items Ben2 and Ben3 excluded..................................................115
Figure 22: Standardised Solution for the respecified Congeneric CFA model of the trust Antecedent Integrity; Items Int4, Int5 and Int6 excluded .................................................... 118
Figure 23: Standardised Solution for the respecified first order CFA model of the trust antecedents; Ability, Benevolence and Integrity; N = 224 .................................................. 120
Figure 24: Standardised solution for the respecified path analysis model of Intention to Trust and its antecedents Ability, Benevolence and Integrity; N = 224........................................ 124
xii
xiii
Glossary of Terms and Abbreviations AB Ability
ANOVA Analysis of Variance
BEN Benevolence
CFA Confirmatory Factor Analysis
DV Dependent Variable
HAHB High Ability combined with High Benevolence Information Level
HALB High Ability combined with Low Benevolence Information Level
HAHI High Ability combined with High Integrity Information Level
HALI High Ability combined with Low Integrity Information Level
INT Integrity
ITT Intention to Trust
IV Independent Variable
LAHB Low Ability combined with High Benevolence Information Level
LAHI Low Ability combined with High Integrity Information Level
LALB Low Ability combined with Low Benevolence Information Level
LAHI Low Ability combined with Low Integrity Information Level
LISREL Linear Structural Relations
ML Maximum Likelihood
PTT Propensity to Trust
QUT Queensland University of Technology
SEM Structural Equation Modelling
SPSS Statistical Package for the Social Sciences
xiv
xv
Statement of Original Authorship
The work contained in this thesis has not been previously submitted for a Degree or
Diploma at any other higher education institution. To the best of my knowledge and
belief the thesis contains no material previously published or written by another
person except where due reference is made in the thesis itself.
Signed : ___________________________ Date: __________
Douglas A. Beatton
xvi
xvii
Acknowledgements
Many people contributed to the completion of this thesis. To all those involved I
offer my sincere gratitude and thanks. There are some people whose contribution was
exceptional; I would like to acknowledge their considerable assistance. First, thank
you to my wife, Debbie, and our two sons, for their understanding and patience. It
was your love and support that has been the greatest driving force. To my
supervisors, Dr. Kerrie Unsworth, Dr. Neal Knight-Turvey, thank you for your
guidance and advice, for believing in me, for helping me develop my research skills,
and especially to Kerrie for your continuous support, even as you readied for your
firstborn.
To Dr. Stephen Cox, thank you for your responsiveness and willingness to answer
my interminable questions regarding method and quantitative analysis. Stephen, your
proactive assistance helped me improve my analytical approach and enabled me to
realise additional benefits from my study effort. To Associate Professor Lisa Bradley
whose knowledge and experience in the development of vignettes was invaluable. To
Dr Chrys Gunasekara for his ongoing mentoring and support, thank you. To my
fellow research students for the helpful advice and debate that contributed to my
understanding. In particular, thank you to Francis Chan, Sukie Sawang, Tony Niven,
Mark Keogh, Laxman Samtani and the recently completed Dr Jack Keegan and Dr
Glen Murphy who were always available to share the load when things became
difficult.
My thanks go to the hundreds of anonymous Queensland University of Technology
(QUT) Business Faculty students who willingly, and without reward, offered
themselves as subjects for this study. Finally, the assistance of the School of
Management, the Faculty of Business, the Research Students Centre and the Library
staff at QUT is gratefully acknowledged.
xviii
1
Chapter One
1.1 Introduction and Overview
Trust touches nearly every facet of our lives. As a baby, we are initiated to trust
through a deep and abiding faith that emerges from the relationship with those who
care for us (Raths, 1972). In our daily lives, we extend interpersonal trust to a person
or group in expectation that they can be relied upon (Rotter, 1967). ‘Trust is essential
for stable social relations’ (Blau, 1964, p. 99). In business, we trust in the benefits
that are expected to result from our cooperative day-to-day interpersonal relations
et al., 1998; Rousseau et al., 1998; Schoorman et al., 1996b; Serva et al., 2005) that
trust studies should be not undertaken without adequately considering the context in
which the trust takes place, an extra dimension has been added to the trust
antecedents in the Mayer et al. (1995) model.
The extra dimension added to the Mayer et al. (1995) model addresses the dynamics
associated with initial trust development in the context of this study, newly forming
work-teams. This addition addresses the concerns of Williams (2001, p.379) who
states that while ‘current models clarify many of the processes that are fundamental
to interpersonal trust development,(sic) they do not jointly address both the cognitive
and the affective influences of social group membership on trust’. In the context of
this study, a potential team member would cognitively assess what they know about
team members before making a decision to join the team. This is as far as this cross-
sectional study goes; it does not proceed into the dynamic life of the team wherein
ongoing feedback may bring about affective changes in the trust of fellow team
members. This is not to say that the affective factor of feedback should not be
included in the model for future studies; particularly those longitudinal studies that
expose trustors and trustees to multiple trusting episodes over time. Given the cross-
sectional nature of this study, and the single trusting context of a new member
making the decision to join a newly formed work-team, the affective factor of
feedback will not be included in the model for this particular study.
What has been included in this study is a context moderation factor that incorporates
the different types of information we could expect to receive about the members of a
new team we were considering to join. For example, if asked to join a team of
students working on an assignment, we would seek information about their
knowledge (Ability), whether they complied with university rules about plagiarism
(Integrity), and whether they looked after the best interests of fellow team members
(Benevolence). Some students do work to a high standard, others to a lower standard.
Thus, there are levels of information that could be communicated concerning a work-
team’s Ability, Benevolence, or Integrity. The question is how to operationalise
antecedent information levels.
36
In seeking to clarify the relationship between trust and its antecedents, Kramer &
Cook (2004) propose numerous contextual moderators of the trust antecedents,
Situational Strength among them. Gill et al. (2005) defined Situational Strength as
the quantity and type of information about the trustee available to the trustor prior to
the trustor extending trust to the trustee in a given situation. For example, a trustor
may be informed that a team is highly skilled and experienced. Perceiving a team
with high levels of Ability, the trustor may have Intention to Trust the team in the
expectation of an acceptable outcome. Alternatively, if the trustor is advised that the
team have low Ability, the trustor may choose not to trust the team. Gill et al. (2005)
incorporated a Situational Strength factor into their study by providing levels of
information about a trustee’s Ability, Benevolence, and Integrity; they termed the
levels Strong, Weak and Ambiguous (Figure 7).
Figure 7: Example dimensions of Situational Strength; adapted from (Gill et al., 2005)
In the (Gill et al., 2005) study, when all the antecedent information was bad (Low),
the situation was viewed as Weak and when there was a mix of good (High) and bad
(Low) information, the situation was seen as Ambiguous. When the trustor
considered all the information about the trustee was good (High), the situation was
termed Strong. The Gill et al. (2005) notion of Situational Strength is flawed because
a situation where all the trust antecedent information is bad could also be considered
as strong, strongly negative. It is argued that we would be ambivalent about
IntegrityBenevolenceAbility
Situational Strength(Moderator)
Weak
Ambiguous
Strong
LowHighLowHighLowHigh
IntegrityBenevolenceAbility
Situational Strength(Moderator)
Weak
Ambiguous
Strong
LowHighLowHighLowHigh
37
ambiguous information. We humans naturally focus on marginal differences when
we make decisions {Easterlin, 2005 #960}). In the context of this study, we would
focus on the margin of the trust information, the good or bad information we assess
when making a decision to join a work-team.
Therefore, unlike the Gill et al. (2005) study, it is asserted that under normal
circumstances a trustor would consider the marginal Information Level about a
trustee on two dimensions; good (High) and bad (Low) (Figure 8). For this study, the
terms of good and bad will be avoided because goodness or badness is in the eye of
the beholder, what some consider good, may be considered bad by others with
different values and beliefs. Instead, the notion of High and Low Information Level
is applied.
Figure 8: The proposed Information Levels of the trust antecedents
Let us view some examples of High and Low Information Levels. A person who
always looks after the best interests of their fellow team members could be seen to
exhibit High Benevolence. The same person could possess fewer skills and therefore
be considered to have Low Ability. While having fewer skills, if we viewed this
person as an honest person, they would attribute them with High Integrity. In this
way, the antecedents of trust have varying information levels that affect our personal
decision to trust. Therefore, it is offered that under a context where we can choose to
join a newly forming work-team, we would rationally consider {van Praag, 1999
#961} the marginal difference in the trust-related information communicated about
the teams. Based our cognitive assessment of the marginal difference, the level of
information, we would make our decision to extend trust and join the work-team, or
IntegrityBenevolenceAbilityInformation Level
Low
High
IntegrityBenevolenceAbilityInformation Level
Low
High
38
vice versa. Therefore it is proposed to incorporate High and Low Information Levels
into the Mayer et al. (1995) Integrative Model of Organisational Trust (Figure 9).
Figure 9: The proposed theoretical model for the study of how the antecedents of trust affect our Intention to Trust in the context of newly forming work-teams
Figure 9 identifies the theoretical model for this study. In selecting formal definitions
for the factors in the study model, consideration of the competing definitions of trust
has identified that the Mayer et al.'s (1995) trust theory incorporates the six themes of
trust distilled from the review of the literature (Chapter 2). Therefore, this research
will build on the previous work of Mayer et al. (1995) and others who have
subsequently sought to advance the integrative model of organisational trust.
Consistent with their research, the definition of Intention to Trust applied to this
study is:
Ability
Benevolence Trustor’s Intention To Trust
Integrity
Antecedent Factors of the
Trustee
Contextual
Information Level
+
+
+
Ability
Benevolence Trustor’s Intention To Trust
Integrity
Antecedent Factors of the
Trustee
Contextual
Information Level
+
+
+
39
‘the willingness of a party to be vulnerable to the actions of another
party based on the expectation that the other will perform a particular
action important to the trustor, irrespective of the Ability to monitor or
control that other party’ (Mayer et al., 1995, p. 712).
The trust antecedents are defined as:
Ability is ‘the skills competencies and characteristics that enable a party to have
influence over some specific domain’ (Mayer et al., 1995, p. 717)
Benevolence is ‘the extent to which a trustee is believed to want to do good to the
trustor’ (Mayer et al., 1995, p. 718).
Integrity is ‘the trustor’s perception that the trustee adheres to a set of principles
that the trustor finds acceptable’ (Mayer et al., 1995, p. 719).
In consideration of the proposition that the level of information about a trustee
affects our Intention to Trust, this study contributes Information Level which is
defined as the dimensional combinations (High to Low) of the good or bad
information communicated to a trustor about the Ability, Benevolence and
Integrity of a trustee prior to the trustor initially extending trust to that trustee.
3.2 Chapter Three Summary
Chapter 3 compared and contrasted two competing models to develop the study
model of initial trust formation between a potential team member and a newly
forming work-team. In doing so, it was shown that the kernel of the McKnight et al.
(1998) model was similar to the Mayer et al. (1995) integrative model of
organisational trust but that the McKnight et al. (1998) model failed to adequately
consider the role that the themes of Context and Dynamics play in the formation and
development of our Intention to Trust. It was identified that the Mayer et al. (1995)
model better addressed Dynamics by incorporating a Feedback loop which informs
the trustor about the probability of success or risk in future trusting episodes.
However, the Mayer et al. (1995) model inadequately considered the context in
40
which trusting episodes occur. In consideration of the varying levels of the good or
bad information that feeds back to a trustor after extending trust to a trustee, a High
and Low Information Level was incorporated into the study model and model
variables were subsequently defined in preparation for hypotheses development.
41
Chapter Four
4.1 Further Developing the Hypotheses for this Stud y
One thing becomes apparent from a review of the literature and the selection of a
study model; scholars operationalise trust differently depending upon the focus and
phase of the trust study. On the one hand, trust could be operationalised as a single
phase; a one-off transaction that is typically derived from a weighed calculus of how
much we think we will gain and lose (Williamson, 1975). On the other hand, we
have an ‘intra- or interpersonal trust continuum’ (Rousseau et al., 1998, p. 395) that
operationalises trust as an ongoing relationship where we question not "How much
do I trust?" but "In what areas and in what ways do I trust?" (Lewicki, McAllister,
& Bies, 1998, p. 443). While it is acknowledged that some ongoing relations are
transactional, with a focus on gains and losses, Rousseau et al. (1998, p. 398) remind
us that ‘both history and the nature of the interaction between the parties can shape
the form that trust takes’. This study operationalises the theoretical model identified
in Figure 9 of Chapter Three and focuses on the contextual situation where the new
team member (trustor) and the newly formed work-team (trustee) are unknown to
one another. In this context, the trustor needs to cognitively assess how the High and
Low Information Levels of information about the trustee affect their decision to
initially extend trust to the trustee.
Prior to the trustor having any information about the trustee, it has been argued that a
trustor’s Propensity to Trust should be most salient. Our Propensity to Trust is our
‘general willingness to trust others’ (Mayer et al., 1995, p. 715). Propensity to Trust
is that level of trust we extend before we have any information about the trustee, akin
to our innate or natural trust (Raths, 1972). The information or feedback that we
receive from our relationships and interactions with others readjusts our trust
perception. Considerable research effort has identified that, once we receive
information about a trustee, our Propensity to Trust wanes in favour of or Intention to
Trust (Costa, 2003; Costa et al., 2001; Gill et al., 2005; Huff & Kelley, 2003; Mayer
& Davis, 1999; Serva et al., 2005; Tan & Tan, 2000). Because considerable research
effort has previously been applied to understanding Propensity to Trust, this study
42
focuses on how we cognitively determine our Intention to Trust in the context of a
trustor receiving initial information about a trustee.
So what happens when we initially receive information about a trustee, which trust-
antecedents most affect our Intention to Trust? Table 1 identifies results from a
recent study by Serva et al. (2005). This longitudinal study was conducted in the
setting of two interacting teams developing software for a website. What is
interesting about this study is that the dynamic nature of trust that was revealed via
the changing correlation of Intention to Trust (ITT) with its antecedents. At Time T1,
the forming stage of team development when the management team communicated
functional requirements to the development team, the trust antecedents of Ability (r
= .70, p = .001), Benevolence (r = .51, p = .05) and Integrity (r = .62, p = .001)
correlated with Intention to Trust. At Time T2, the antecedent data from the previous
time, T1, no longer significantly correlated with ITT. Similarly, at Time T3, the
antecedent data from the previous time, T2, no longer significantly correlated with
ITT. The Intention to Trust of the teams had changed over time. Depending on the
stage of team development, the antecedents appeared to have varying effects on trust.
43
Table 1: Research results from a recent study by Serva et al. (2005)
halla
This table is not available online. Please consult the hardcopy thesis available from the QUT Library
44
This is not surprising because there has been scholarly disagreement over the relative
strength of the relationship between trust and its antecedents. In 1999, Davis found
that Integrity (r = .818, p = .01) had the strongest relationship with Intention to Trust.
More recently, Gill et al. (2005, p. 99) proposed that ‘it is possible that Ability,
Benevolence, and Integrity contribute differently to the establishment of trust’. The
results of the Serva et al. (2005) study supports this proposition but, unlike this study,
their research did not measure trust and its antecedents in the pre-stage. This study
seeks to address the gap in the Serva et al. (2005) study by clarifying our
understanding of how the trust antecedents affect our Intention to Trust during the
pre-stage of the life of a work-team. Just as the effect of the trust antecedents differed
in the later stages of team development, it is hypothesised that:
H2: At the pre-stage of team development, the trust antecedents of
Ability Benevolence and Integrity will have differential effects on our
Intention to Trust members of a newly forming work-team.
Pursuing this hypothesis avails us of an opportunity to better understand trust by both
clarifying and the contributing to our current knowledge of trust development in the
context of newly forming work-teams. As a person with an opportunity to join a
newly forming work-team, we would seek information about the other group
members. How competent (Ability) are they? Will they lie (Integrity) and get me into
trouble with authority? In addition, will they do what is in my best interests
(Benevolence)? From the perspective of Ability, it would be easy for us to gauge the
knowledge of team-members; we could simply assess physical examples from their
previous work assignments; reports, products or perhaps customer feedback. A team
member devoid of the necessary skills and experience would not normally be able to
produce examples of previously completed quality work.
While manifestations of Ability would be physically obvious, it would be more
difficult to gauge the Integrity or Benevolence of other team members. Benevolent
behaviour would be less obvious than Integrity because issues of Integrity should
emerge during the employee hiring process. If we exhibit low Integrity behaviour by
lying and deceiving our peers we may get away with it on the first or perhaps the
second occasion, but it would be reasonable to assume that our peers would be less
45
likely to trust us after each additional episode of low Integrity behaviour.
Deceitfulness and dishonesty are memorable events (Arvey & Wanek, 2006); our
honesty image reflects our honest and dishonest behaviours and these are highly
visible to others (Van Iddekinge, Taylor, & Eidson Jr., 2005). We can all recall
episodes where someone has disappointed us when they do not do what they say they
will do. Because humans tend to have a behavioural trait towards trustworthiness
(Rotter, 1980), if we were contacted by an employer checking on the credentials of a
person previously in our employ, most of us would truthfully inform the new
employer if our past employee had exhibited low Integrity behaviour (Bulmash,
2006). Therefore, we would be more likely to find out about episodes of Low
Integrity.
Identifying Benevolent behaviour would be more difficult. For example, we first
need to extend trust to a person, group or organisation and wait for the feedback that
informs us that they have, or have not, behaved in our best interest (Benevolence).
This takes time. For example, when we first meet a new member of a work-team, we
have not previously trusted them; we have not received feedback from a previous
trusting episode. We could ask other people who had worked with them, but, with so
many of today’s teams comprising temporary, part-time or contract staff, we may not
have the opportunity to gather the information we need to adjudge other team
member’s Benevolence. They are strangers to us. Therefore, it is hypothesised that:
H3: Ability will have a more salient effect on our Intention to Trust
members of a newly forming work-team than Integrity and
Benevolence.
46
4.2 Chapter Four Summary
Chapter Four applied the study model to the context of a member making a decision
to extend trust by joining a newly forming work-team. At this stage of trust
development, the feedback that informs us that the trustee performed as we expected
has not come into play because we have not yet shared a trusting episode. Before we
make a decision to join a team we would request information about the Ability,
Benevolence and Integrity of the other team members. In seeking to understand
which of these factors is most salient in the development or our Intention to Trust,
previous studies have provided different results. There is scholarly disagreement over
which antecedent of trust is most important in trust development between members
of a team. In the context of this study, it was hypothesised that the salience of the
antecedents of our Intention to Trust is Ability, Integrity then Benevolence. The
research design is now justified.
47
Chapter Five
5.1 The Research Design
The research design is the plan that structures the investigation to obtain answers to
the research questions (Punch, 2000). Holistically, this research is to be conducted in
three stages (Figure 10).
Figure 10: The Three study Stages
To gain an understanding of the study setting and to collect data to assist with the
development of the scales and the vignette manipulation, the Vignette Development,
Stage 1, involved structured interviews. Additionally, descriptive statistics from
similar trust-related studies were applied to identify typical effect sizes that were
used to estimate sample size requirements for the subsequent study stages. The Pilot
Stage 2 pre-tested the vignettes, experimental manipulations, and measurement
scales on a sample frame from the population. The Final Data Collection stage
applied the refined scales and vignettes to collect data for final analysis. The
following pages seek to justify and define the research design choices for the study
setting, the type of measures, the unit of analysis, the time horizon, and, the research
paradigm (Cavana, Sekaran, & Delahaye, 2001, pp. 285-382).
Final Data Collection• Quantitative• Questionnaire• Manipulation questionnaire• n = 224• Data analysis
68
To maximise the external validity of the vignettes (Gliner et al., 1999), the thematic
analysis sought to identify and formulate vignette content that satisfied the
informational and rationale requirements (Shapiro, Buttner, & Barry, 1994) that a
potential team member would normally use to fairly decide (Bies, Shapiro, &
Cummings, 1988) if they would join a newly forming work-team. Therefore, to make
the vignettes more believable, the subject’s statements were incorporated into the
vignettes used to manipulate the dimensions of the antecedents of Intention to Trust.
To begin vignette development, Undergraduate Student subjects from the Business
Faculty population of a metropolitan Australian university were convenience
sampled as they entered the main entrance of the Business Faculty building. The
nature of the study was not revealed to the subjects, the structured interview
technique sought to avoid projecting the purpose of the study and the attitude of the
researcher onto the subject. Consistent with the goal for Stage 1 of the study,
capturing the subject's attitudes towards group work was the stated goal of the
interview. Without reward, the anonymous subjects voluntarily agreed to participate,
or not participate, in the individual structured interview that initially confirmed they
were undergraduate students from the Business Faculty. Those who satisfied the
undergraduate student criterion and confirmed their previous experience with
working in teams on group assignments were considered suitable for Stage 1 of this
study. Those persons who did not meet the criterion were thanked and advised that
they did not satisfy the parameters of the study population of Undergraduate Student
from the Business Faculty of a metropolitan Australian university.
Having confirmed they were part of the target population, the subjects were cued by
asking them to ‘caste your mind back to your group assignment experiences where
all members of the group received the same grade’. This statement sought to
cognitively position the subjects ready to respond to the following three questions:
1. Do you consider there are risks associated with working on group
assignments?
This closed question (Cavana et al., 2001, p. 142) sought to confirm that the subject
perceived there was, or was not, risk associated with group work. Risk is a
69
theoretical foundation for trust, without risk in an interrelationship there is no need to
trust (Golembiewski & McConkie, 1975). Subjects who did not confirm there were
risks associated with group work would not need to extend trust to another and
therefore were considered unsuitable for this trust-measuring study. Interestingly, all
those who agreed to participate in a structured interview perceived risks with group
work; all subjects stated there was more to lose than gain. It would appear that
students might not like group assignments; perhaps this is a future research
opportunity.
Those subjects who perceived risks with group work were then asked the open-ended
question:
2. What were they?
The subjects were neither prompted nor interrupted; a remote position was
maintained with respect to the subjects. In seeking to avoid invoking researcher bias,
the interviewee’s exact comments were recorded. The final question sought to
confirm a second theoretical foundation for Intention to Trust, expectation (Deutsch,
1960):
3. What do you expect from other team members when working on a
group assignment?
Responses were recorded in exactly the same manner as question two. The subjects
were asked if they had any other comments, these were recorded. Finally, the
subjects were thanked for their participation. On average, each structured interview
was completed in ten minutes.
With just three factors to consider, the interviewee statements were analysed and
sorted using a Microsoft Excel database (Microsoft, 2007). Excel’s database, sorting
and other data manipulation routines were more than adequate for this qualitative
analysis task. Thematic analysis of the interview transcripts synonymously identified
the words and phrases associated with each of the trust antecedents from the
70
theoretical foundation for this study; Mayer et al.'s (1995) integrative model of
organisational trust (Table 4).
Table 4: Theoretical basis for the thematic analysis of the structured interview transcripts
Trust Antecedent:
High High HighHigh High LowHigh Low HighHigh Low Low
Low High HighLow High LowLow Low HighLow Low Low
competent incompetent do good not do good high principled low principled
skilled unskilled helpful not helpfulacceptable standards
non-acceptable standards
knowledgeable un-knowledgeable altruistic selfish high integrity low integrityexpert less expert good behaviour bad behaviour similar values dissimilar values
good judgement bad judgement good intentions bad intentions similar beliefs dissimilar beliefsdependent independent openness closedgroup goals self goals fair unfair
caring uncaring congruent non-congruentreliable unreliable moral immoral
doesn't lie liesconsistent inconsistent
Benevolence IntegrityAbility
Synonyms: Based on the
Integrative Model of Organisational Trust (Mayer et al., 1995)
High LowSynonym Dimensionality:
Low
High Low High Low
Information Level:
High Low HighHigh Low
The interviewee statements were categorised as pertaining to Ability, Benevolence or
Integrity based on the largest number of words in the statement that concurred with
the synonyms for each factor. For example, a statement such as ‘sometimes the other
students don’t tell the truth, they don’t do what they say they will do’ was themed as
an Integrity-related statement. Statements like ‘they don’t consider my opinion’ and
‘they don’t care about what is important to me’ were themed as a negative comment
relating to Benevolence. Other statements such as ’they may not have the necessary
knowledge’ and ‘the group may have lower grade expectations than me’ were
themed as Ability-related. The largest number of comments, 75, related to the
Integrity sub-themes (Table 5); many subjects appeared to be concerned that other
students did not share their values and standards. Benevolence sub-themes came
second, 51; many students had emerged dissatisfied from previous group assessment
encounters. Surprisingly, the trust antecedent that attracted the fewest number of
comments, 9, was Ability. If the volume of the comments is any guideline to the
salience of the trust antecedents, Integrity may be more important than Benevolence
and Ability could be viewed as least relevant.
71
Table 5: Summary of results for the thematic analysis of the structured interview transcripts
Antecedent of Trust
Ability Benevolence Integrity
Number of Comments (+ positive; - negative)
Ability Sub-themes
Benevolence Sub-themes
Integrity Sub-themes
+ - + - + -
Competency 5 4 Cooperative 3 4 Reliability 7 4
Group participation
3 1
Delivering work to the standard promised
3 5
Rapport between team members 3 4
Delivering work when it was promised
2 9
Sharing in decisions
3 5 Commitment
8 3
Communication 9 6
Share the workload
13 13
Recognition 1
Accepting responsibility for outcomes
2 3
Contribution involvement
3 Honesty
1 2
Treatment of fellow team members
6
Subtotals: 5 4 31 20 36 39 Totals: 9 51 75
At this stage of the study, this researcher’s interpretation of the interviewee
transcripts varies considerably from the study hypotheses and the results that
emerged from the Final Stage data analysis. However, this is understandable because
the interviewees had not been exposed to the experimental manipulation; one would
expect their responses to the structured interview questions to be reflective of their
past group assignment experiences. Interim research observations aside, selected
comments from the structured interviews were included in the wording of the
vignettes and the questionnaire items for the Intention to Trust, Ability, Benevolence,
Integrity and Propensity to Trust scales.
72
6.1.1.1 The Propensity to Trust Scale
The Propensity to Trust (PTP) scale, used in Stage 2 of this study, was founded upon
modified items from Gillespie (2003), Schoorman et al. (1996a), Mayer & Davis
(1999) and Rotter (1967). Based on the thematic analysis of the structured
interviews, it was determined that certain items from the original scale were not
applicable to the study setting. For example, ‘most salespeople are honest in
describing their products’ (Mayer & Davis, 1999, p. 136) is clearly out of context for
a university study setting. In replacement, more applicable items were incorporated.
For example, issues with reliance on the work of other students were consistently
raised during the structured interviews. Therefore, the following item was included
from the Gillespie (2003, p.32) scale; ‘ I would rely on task-related judgements made by
others’. Table 6 lists the twelve Propensity to Trust (PTP) items.
Table 6: The five-point Likert-scaled Propensity to Trust items used in the Pilot Stage of the study
The following questions ask about your general trus t in others:
Disagree Strongly
Disagree Neither Agree
nor Disagree
Agree Agree Strongly
1. In dealing with strangers, it is better to be cautious until they have provided evidence that they are trustworthy ..............................................................
2. Most people keep their promises ................................
3. Most people answer questions honestly................................
4. Most people say what they believe themselves and not what they think you want to hear ...............................
5. Most people tell the truth about the limits of their knowledge ................................................................
6. Most people cannot be counted on to do what they say they will do...............................................................
7. These days, you must be alert or someone is likely to take advantage of you................................
8. I normally rely on the task-related skills and abilities of others................................................................
9. I would not follow the advice of others on important issues................................................................
10. I confidently allow other people to make decisions for me during an absence................................
11. I would rely on task-related judgements made by others ....................................................................................
12. I usually monitor others when they have to do something for me ................................................................
73
Reviewing Table 6, items 1 and 2 were derived directly from the Rotter (1967) scale.
Items 3 to 7 came from the original PTP scale of Schoorman et al. (1996a) and were
modified for study context. Items 8 to 11 are from the Gillespie (2003) study and
addressed trust-affecting issues raised by subjects during the structured interviews;
namely the risk of having to rely on the advice and judgement of others to make
decisions. Item 12 was added because the need to monitor the work of fellow group
members emerged as a strong theme during the structured interviews. Item 12 was
also included because the Mayer et al. (1995) definition of trust identifies monitoring
behaviours as a reflection of a lower Propensity to Trust. The notion of monitoring is
also included in Mayer et al. (1995) Intention to Trust scale.
74
6.1.1.2 The Intention to Trust Scale
The Intention to Trust (ITT) scale was constructed upon the foundation of items from
the initial Mayer et al. (1995) theory-based studies (Schoorman et al., 1996a; Mayer
& Davis, 1999). The ten ITT items were modified to reflect the study setting (Table
7). Currall & Inkpen (2002) recommend that context and the level of theorizing
needs to be considered for proper trust measurement. In line with that
recommendation, the questionnaire items used in this study were adjusted to reflect
the contextual issues that emerged from the thematic analysis of the structured
interviews with the student subjects during the Vignette Development Stage of this
study. For example, to mimic a theme that arose during the structured interviews,
item 38 was modified by removing the words ‘top management’ and replacing them
with ‘this group’.
‘If I had my way I wouldn’t let top management have any
influence over issues that were important to me’ (Mayer & Davis, 1999)
was modified to:
38. If I had my way I wouldn’t let this group have any
influence over issues that were important to me’
Interviewees in Stage 1 were not concerned about university management; they were
concerned that the behaviour of fellow group members would detrimentally affect
the grade they expected to receive for their group work. Other trust-related issues
raised by the subjects during the structured interviews were incorporated into the ITT
questionnaire items. These issues included: students proving they could do the work;
whether or not the other group members could be counted on to complete the work;
whether the students would be honest about their knowledge; whether social loafers
would leave the other group members to do all the work; whether the skills, abilities
and efforts of the other group members would result in work of an acceptable
standard, and; whether members would take notice of the opinions of fellow group
members or force decisions on them.
75
To make the questions understandable and more relevant to the student subjects, the
face validity of the questionnaire items 42 through 47 were improved (Cavana et al.,
2001, p214) by using the words and phrases spoken by the subjects when describing
issues raised during the structured interviews. Table 7 lists the ten Intention to Trust
questionnaire items employed in the Pilot and Final stages of the study.
Table 7: The five-point Likert-scaled Intention to Trust items used in the Pilot and Final stages of this study
Think about the information presented in the email and answer the following questions about how you perceive this group:
Disagree Strongly
Disagree
Neither Agree
nor Disagree
Agree Agree Strongly
38. If I had my way I wouldn’t let this group have any influence over issues that were important to me ..............................................................................................
39. I would be willing to let this group have complete control over an assignment that was critical to me .................
40. If I joined this group, I feel I need to keep an eye on what other group members do ..............................................................................................
41. I would be comfortable giving this group a task or problem which was critical to me, even if I could not monitor the group’s actions ................................
42. In dealing with this group, it would better to be cautious until they have provided proof that they are trustworthy ................................................................
43. I feel this group could be counted on to do what they say they will do ..............................................................................................
44. This group should be honest about their knowledge ..............................................................................................
45. I will be cautious or this group could take advantage of me .............................................................................................
46. I feel I could rely on the skills and abilities of this group ..............................................................................................
47. I would allow this group to make decisions for me ..............................................................................................
76
6.1.1.3 Ability, Benevolence and Integrity Scales
With the composition of the DV scales for Propensity to Trust and Intention to Trust
complete, a decision had to be made concerning the IV scales for the trust
antecedents of Ability, Benevolence and Integrity. Previous studies based on the
Mayer et al. (1995) trust theory have used the IV scales developed for the initial
Schoorman et al. (1996a) study; they have consistently exhibited good scale
reliability (Table 8).
Table 8: Examples of the good Cronbach Alpha results for the Ability, Benevolence and Integrity scales used in recent studies
Because of their good reliability, the same scales were used in this study. Tables 9,
10, and 11 itemise the Ability, Benevolence, and Integrity questions employed in
both Stages 2 and Stage 3 of this study. Like the PTP and ITT scales, they were
modified to reflect the study setting. For example, the subject in items 48 to 53 was
changed from ‘manager’ to ‘group’.
Table 9: The five-point Likert-scaled Ability items used in the Pilot and Final stage of this study
Continue to think about the information presented in the email and answer the following questions about how you perceive this group:
Disagree Strongly
Disagree
Neither Agree
nor Disagree
Agree Agree Strongly
48. The group is very capable of performing its job .............................................................................................
49. The group appears to be successful at the things it tries to do ................................................................
50. The group has much knowledge about the work that needs to be done ............................................................
51. I feel very confident about the group’s skills ..............................................................................................
52. The group has specialized capabilities that could increase my performance.......................................................
53. The group is well qualified ..............................................................................................
77
Table 10: The five-point Likert-scaled Benevolence items used in the Pilot and Final stages of this study
Disagree Strongly
Disagree
Neither Agree
nor Disagree
Agree Agree Strongly
54. The group will be very concerned about my welfare...................................................................................
55. My needs and desires will be very important to the group ................................................................
56. The group would not knowingly do anything to hurt me ..................................................................................
57. The group will really look out for what is important to me................................................................
58. The group will go out of its way to help me ..............................................................................................
Table 11: The five-point Likert-scaled Integrity items used in the Pilot and Final stages of this study
Disagree Strongly
Disagree
Neither Agree
nor Disagree
Agree Agree Strongly
59. The group has a strong sense of justice ..............................................................................................
60. I will never have to wonder whether the group will stick to its word ................................................................
61. The group tries to be fair in dealings with others ..............................................................................................
62. The group's actions and behaviours are not very consistent ................................................................
63. I like the group’s values ..............................................................................................
64. Sound principles seem to guide the group’s behaviour................................................................
78
6.1.1.4 Vignette Development
Like the scales, the vignettes were similarly developed from the thematic analysis of
the structured interviews. The research design called for all three antecedents of trust
to be concurrently manipulated within each vignette. With the goal of optimising a
subject’s comprehension, the vignettes were theoretically founded and composed to
reflect the specific words and phrases used by the subjects during the structured
interviews in Stage 1. For example, the theoretical foundation for this study (Mayer
et al., 1995) proposes that confidence in another’s Ability and their intent to act on
your behalf lies at the core of trusting another individual, group or organisation.
These issues emerged in the structured interviews conducted in Stage 1. There were
concerns that ‘group members would deliver inferior work’ or that they ‘lacked the
knowledge’ or would not ‘consider the welfare’ or ‘the needs and desires’ nor act in
the ‘best interests of other group members’. The customisation of vignette content
was expected to enhance subject understanding of vignette content; this tends to
maximise relevance to the subjects and engages their interest (Hughes & Huby,
2004, p. 40). The following is an example of how the subject’s words and phrases
were incorporated into a vignette depicting High Ability and High Benevolence:
• The group has good subject knowledge. • In the past, the members of this group have shown concern for the welfare of
other group members.
• Your needs and desire should be important to the group.
• The group members really look out for what is important to other group members.
The following vignette is an example of the concurrent manipulation of three trust
antecedents in a vignette depicting a work-team with High Ability, High
Benevolence and Low Integrity.
79
6.1.1.5 Example Preliminary Vignette
Please read the following case and imagine that you are in the situation described. Then answer the questions in the following questionnaire. You are an undergraduate student enrolled in a compulsory unit for your Business Degree. The unit assessment includes a group assignment worth forty percent. You receive the following email from the unit coordinator: From: [email protected] Sent: Monday, 10 May 2006 9:02 AM To: BSB311_student@ student.uni.edu.au Subject: BSB311 Assessment item #2: The Group Assignment As unit coordinator, I reviewed the records for the unit and noticed that you are not currently assigned to a group for the written assignment. Receiving a pass grade for the assignment is a mandatory unit requirement. The Unit Outline states that ‘Students who do not receive a pass mark for their group assignment will fail the unit’. To receive your degree, you must pass this unit. Therefore, to pass the unit, you must immediately join an existing group. The assignment is due in three weeks; all members of a group receive the same mark for the assignment. A couple of the groups have room for another student. Group #6 members volunteered the following information about their group.
• The group has good subject knowledge. • The group exhibits good time management and decision-making skills. • The group’s written work is good. • Based on official University records, authenticated by me, the group
members volunteered the following information concerning their grades:
Student Grades from their last three Group Assignments
No. 1 No. 2 No. 3
Student Average for their Group Assignments
Student Grade Point
Average Student A 6 5 6 5.67 5.70 Student B 7 7 5 6.33 6.30 Student C 6 7 5 6.00 5.90
Overall Group Assignment Average: 6.00 Overall Student Grade Point Average for the Group: 5.97
• The group communicates well. For example, all group members promptly answer emails.
• Group members show concern for the welfare of others. For example, group
members always arrive on time for meetings and never make hurtful comments to other group members.
80
• The members of this group are team-orientated and cooperative. For example,
they listen to the ideas of the other group members, are flexible, and openly discuss and resolve differences of opinion. They learn from one another and go out of their way to help fellow group members.
• With this group, things should get done the way you want them to be done.
• Some group members can have a questionable sense of justice. For example,
some group members may bend university rules and they may not value the extra effort put in by other group members.
• When it comes to doing the work, some group members can be unfair. For
example, when they finally agree on what has to be done, they won’t share the work equally.
• Sometimes, a group member may be untruthful; they sometimes don’t do
what they say.
• The weaker principles of this group will let you do anything you want. Now complete the following questionnaire. It will indicate to me if you should join the group or whether I need to arrange for you to join another group. Please begin the questionnaire now; it will be collected in about ten minutes. Thank You BSB311 Unit Coordinator Faculty of Business
81
6.1.1.6 Reason for changing from a 3x2 to a dual 2x2 factorial design
Piloting the initial vignettes proved problematic, concurrent manipulation of the three
trust antecedents was a flawed design strategy; the vignettes suffered from face
validity issues. While the vignettes were written in familiar speech, that was easier to
understand and used oft-repeated familiar words and phrases from the structured
interviews, pre-test subjects had difficulty comprehending their meaning; they voiced
confusion. It became obvious that concurrent manipulation of all three antecedents
was illogical. The subjects doubted the validity of vignette content that depicted a
group with dimensionally opposite attributes of Integrity and Benevolence. The
following is an example of the vignette content for:
Low Integrity
• Some group members can have a questionable sense of justice. For example, some group members may bend university rules and they may not value the extra effort put in by other group members.
• When it comes to doing the work, some group members can be unfair . For
example, when they finally agree on what has to be done, they won’t share the work equally.
• Weaker principles guide the behaviour of this group. Sometimes, a group
member may be untruthful ; they sometimes don’t do what they say.
High Benevolence
• In the past, the members of this group have shown concern for the welfare of other group members.
• Your needs and desire should be important to the group.
• The group has not knowingly done anything to hurt anyone in the past.
• The group members really look out for what is important to other group
members.
Subjects voiced concern that the initial vignettes were not believable. This is not
unusual, because adults consistently detect lies and anomalies in vignettes (Bussey,
1992). A review of the above example revealed that; on the one hand it is easy to
imagine that group members of High or Low Ability could exhibit attributes of Low
Benevolence by, in the words of the vignettes, being ’untruthful’ and ‘showing little
concern for fellow group members’. On the other hand, it is hard to believe that a
Low Integrity group that are ‘untruthful’ and ‘unfair’ could exhibit High
82
Benevolence at the same time by ‘looking out for what is important to other group
members’. To satisfy the basic prerequisite for a trustworthy vignette, the vignettes
must depict a believable situation familiar to the subjects (Bussey, 1992). In response
to pre-test subject feedback, the factorial design of the study was altered to facilitate
dimensional manipulation of the independent variables without compromising the
believability of the vignettes. The a priori 3x2 factorial design, where the three trust
antecedents were concurrently manipulated, was changed to a dual 2x2 factorial
design. In the (2x2) factorial design, four different vignettes manipulated High and
Low Information Levels of Ability and Benevolence (Table 12) and the other four
vignettes manipulated High and Low Information Levels of Ability and Integrity
(Table 13).
Table 12: 2x2 Factorial design for the manipulation of the dimensions of Ability and Benevolence
Trust Antecedent
Ability Benevolence High Low High Low
X X X X X X X X
Table 13: 2x2 Factorial design for the manipulation of dimensions of Ability and Integrity
Trust Antecedent
Ability Integrity High Low High Low
X X X X X X X X
83
The vignettes were rewritten to concurrently depict High or Low Ability with either a
dimension of Benevolence or a dimension of Integrity. The updated vignettes were
piloted on different pre-test subjects; none provided feedback that the new vignettes
were unbelievable or confusing. The dual 2x2 factorial design was applied in study
stages two and three.
With the design updated plus the vignettes and scales completed, the last page of the
questionnaire requested demographic information. The demographic questions
initially sought confirmation that the subject was a student from the target population
of Business Faculty undergraduate students. To ensure that the same subject did not
submit multiple questionnaires, personal information was requested; gender, age,
birth month and year, and mother’s maiden name. The students were advised they
would remain anonymous, no names or student numbers were collected. The subjects
were provided with sufficient questionnaire space to voluntarily provide feedback
and comment. The resulting Stage 2 questionnaire incorporated the Propensity to
Trust items prior to the vignette manipulation and the Intention to Trust, Ability,
Benevolence and Integrity items post the vignette. The demographic items were
included at the end of the questionnaire. To provide data for future research, items
addressing the following variables were included prior to the vignette manipulation:
• Preference for Teamwork (Kirkman & Shapiro, 2001)
• Individualism (Maznevski & DiStefano, 1995 in Kiffin-Petersen &
Cordery 2003)
• Self-assessed Ability (developed for this study and based upon
Schoorman et al. (1996a)
• Self-assessed Benevolence (developed for this study and based upon
Schoorman et al. (1996a)
• Self-assessed Integrity (developed for this study and based upon
(Schoorman et al., 1996a)
• Locus of Control from Carlopio, Andrewartha, Armstrong, & Whetten
(2001, p. 99) based on Rotter’s (1967) internal-external Locus of Control
Scale
84
6.1.1.7 Summary of Stage 1 Results
The thematic analysis of the transcripts from the structured interviews identified
words and phrases synonymous with the theoretical foundation for this study.
Once categorised, interviewee comments were applied to develop the vignettes
that would manipulate the levels of the independent variables, Ability,
Benevolence and Integrity. Similarly, the questionnaire items for the IVs and the
DVs were refined to reflect the context of the study. In pre-testing the
questionnaires, the concurrent manipulation of all three IVs appeared
unbelievable to the respondents. In response, the planned 3 x 2 factorial study
design was re-engineered to a two 2x2 factorial design. Ability and Benevolence
were manipulated together and Ability and Integrity were concurrently
manipulated with separate vignettes. Finally, demographic items were added to
the questionnaire to be used for data collection in Stage 2.
85
6.1.2 Analysis and Results of Study Stage 2: Pilot of the Test Instruments The goal of Stage 2 was to strengthen internal validity and make the vignettes
realistic, by piloting and refining them (Hughes & Huby, 2004) prior to their final
application in stage three of this study. One hundred and eighty seven Undergraduate
Student subjects from the Business Faculty population of a metropolitan Australian
university were sampled to test the vignette manipulations and questionnaire items
developed in Stage 1 (Figure 14).
Figure 14: Study Stage 2
The 100 female and 87 male respondents, not involved in Study Stage 1, satisfied the
experimental requirement for previous experience with group-based assessment
where all team members received the same grade from their group work. The
respondents individually completed the questionnaire prior to their scheduled lecture
or tutorial. The subjects self-selected by choosing to accept or reject a questionnaire
as they entered their classroom. The subjects were offered neither recognition nor
reward for completing a questionnaire. Randomisation occurred via the random
nature of the subjects’ arrival at their class and their equal chance of receiving one of
the randomly sorted questionnaires that contained one of the eight different vignette
Level of significance: * p < .05 ** p < .01 *** p < .001
In support of hypotheses H1, Ability ( r = .538, p = .01), Benevolence (r =.465, p =
.01) and Integrity (r = .593, p =.01) were found to have a positive relationship with
Intention to Trust (Table 14). Overall, these results were consistent with previous
studies (Table 15); there was a positive relationship between the trust antecedents,
Ability, Benevolence and Integrity, and Intention to Trust.
Table 15: Comparison of ITT study results from this research and recent studies employing similar measurement scales based on the Mayer et al. (1995) theoretical foundation
Correlation of
Intention to Trust with
Ability
Benevolence Integrity
Stage 2 Results from this Study .538 * .465 ** .593 **
Level of significance: * p < .05 ** p < .01 *** p < .001
88
However, the significance of the Ability relationship with Intention to Trust was
lower than previous studies. A lower than expected level of significance prompted a
check of the effect of the vignette manipulation and a review of subject comments. A
review of questionnaire feedback comments identified that some subjects perceived
little or no significant difference between the High Ability and Low Ability
information in the vignettes (Tables 16 and 17).
Table 16: Ability information provided in the Pilot Study Low Ability vignette
Student Grades from their last three Group
Assignments
No. 1 No. 2 No. 3
Student Average for their Group
Assignments
Student Grade Point
Average Student A 4 4 4 4.00 4.30 Student B 4 4 5 4.33 4.42 Student C 5 4 5 4.67 4.57
Overall Group Assignment Average: 4.33 Overall Student Grade Point Average for the Group: 4.43
Table 17: Ability information provided in the Pilot Study High Ability vignette
Student Grades from their last three Group Assignments
No. 1 No. 2 No. 3
Student Average for their Group
Assignments
Student Grade Point
Average Student A 6 5 6 5.67 5.70 Student B 7 7 5 6.33 6.30 Student C 6 7 5 6.00 5.90
Overall Group Assignment Average: 6.00 Overall Student Grade Point Average for the Group: 5.97
In response to this qualitative information that appeared to identify that subjects
similarly perceived the Low and High Ability vignette information, the vignettes
were referred for expert review. The expert, who had extensive experience in
lecturing and teaching students from the target population, suggested that the
students would view the low grades depicted in the Low Ability vignettes as
acceptable grades. This may occur because the university allows students to graduate
with a small number of failing grades; a conceded pass of ‘3’ (Table 18); the
phenomenon that students may accept substandard marks as a passing grade is
worthy of further pedagogical research.
89
Table 18: Grading scale for the university population
Grade Example Grade
Percentages 1 Fail 0 to 14% 2 Fail 15 to 29% 3 Conceded Pass 30 to 49% 4 Pass 50 to 64% 5 Credit 65 to 74% 6 Distinction 75 to 84% 7 High Distinction 85 to 100%
In response to the advice of the vignette expert, the Information Level in the Low
Ability vignettes was changed to include more failing grades (Table 19).
a Number of cases outside range (Q1 - 1.5*IQR, Q3 + 1.5*IQR).
b indicates the inter-quartile range (IQR) is zero.
94
Subjects did not consistently fail to answer a particular question or sequence of
questions. From 224 responses, no item had more than two missing values; the
missing data appeared to be missing at random. Therefore, particular items were not
considered to be problematic (Cohen, 1988, 432-433). This was expected because
published and previously well-tested scales have been used (Robinson, Shaver, &
Wrightsman, 1991). The first two criteria for ML selection had been satisfied.
In addressing the last criteria for the appropriate use of ML, it was noted that values
were missing on not just independent variables, but also on the dependent variables.
For example, in the collected data there were seven values missing across six of the
ten ITT variables. When data are missing on more than one variable there is a
concern that study conclusions may be limited, particularly when the data are
missing on dependent variables. This raises the concern that a sample with missing
values may not be representative of the population (Cohen, 1988). However, in
support of the use of the ML procedure, ML ‘doesn’t care’ whether data are missing
on the independent or dependent variables; ‘one is not limited in generalization of
one’s findings’; the ML model estimating procedure provides the best possible,
unbiased, estimates for each missing value by considering all the data values in a
model (Cohen et al., 2003, p. 442). The final criteria for ML was satisfied, data
values were missing across both the independent and dependent variables. Therefore,
with the sample meeting the ‘large size requirement of n >200’ (Cohen et al., 2003,
p. 440), the missing at random nature of the absent values, and because those absent
values occurred in both the independent and dependent variables, the decision was
made to retain the estimated missing values from the application of the maximum
likelihood imputation (ML) estimation technique using the EM algorithm (Cohen et
al., 2003).
However, while the missing data had been replaced, its use for SEM analysis is
founded upon a key assumption; the variables in an SEM model must be distributed
multivariate normal. This is particularly important because Maximum Likelihood
estimation technique, ML, was the chosen analytic strategy for assessment of the
structural equation models. Prior to using the PRELIS 2.30 software (Joresborg &
Sorbom, 1996) to satisfy the SEM requirement for multivariate normality, the
outliers in the variables were assessed using the SPSS ‘Frequencies’ command. A
95
preliminary screen of the data (Tabachnick & Fidell, 2007 pp. 72-74) identified the z
scores were all less than 3.29, p = .01, two tailed; there appeared to be no potential
outliers. Histograms of the data for each variable appeared normal and no univariate
outliers stood out in the box plots. Initial SPSS examination of the data for
univariate and multivariate normality applied conservative alpha levels of .01 or .001
(Tabachnick & Fidell, 2007, pp. 80-810); the kurtosis and skewness statistics were
close to zero. In final preparation for SEM analysis, subsequent data screening used
the PRELIS 2.30 software (Joresborg & Sorbom, 1996). While the skewness and
kurtosis values are optimally zero, the demonstrated skewness of the IV and DV
variables fell within a PRELIS-accepted range of 2.00 to 3.00, and the kurtosis
values fell within the acceptable range of 7.00 to 21.00 (Curran, West, & Finch,
1996, p. 28). Repeating this analyses with and without missing data was considered
unnecessary because of the small number of missing values (Tabachnick & Fidell,
2007, p. 72). Therefore, the data are generally approximating a normal distribution
(Byrne, 1998). To get a feel for the results to come, the descriptive statistics of the
normalised Stage 3 data were next analysed.
In gaining a feel for the normalised Stage 3 data, initial analysis using SPSS 13.0 for
Windows (SPSS, 2004) identified descriptive statistics similar to Stage 2 (Table 21)
Table 21: Descriptive statistics and correlations from Study Stage 3, N = 224
Mean Standard Deviation
Intention to
Trust Ability Benevolence Integrity Intention to
Trust 28.49 5.93
Ability 19.15 4.93 .653 ** 1
Benevolence 14.39 3.80 .357 ** .203 ** 1
Integrity 18.63 4.04 .639 ** .488 ** .678 ** 1
Level of significance: * p < .05 ** p < .01 *** p < .001
Similar to Stage 2, hypotheses H1 was supported in Stage 3. In comparing Stage 2
and Stage 3 results (Table 22):
• the positive relationship between Intention to Trust and Integrity
strengthened slightly from r = .593, p = .01, in Stage 2 to r = .639, p = .01 in
Stage 3;
96
• Ability had similarly strengthened and the significance level improved from r
= .538, p = .05, to r = .653, p = .01, but;
• the ITT relationship with Benevolence had weakened slightly from r = .465,
p = .01, to r = .357, p = .01.
These differences could have arisen from sample differences. Alternatively, the
differences could have arisen from the Stage 2 refinement of the vignettes and
questionnaire items which sought to contribute to the reliability of the trust scales.
Table 22: Comparison of ITT study results from this research and recent studies employing similar measurement scales based on the Mayer et al. (1995) theoretical foundation; N = 224
Correlation of Intention to Trust
with
Ability Benevolence Integrity
Stage 3 results from this Study .653 ** .357 ** .639 ** Stage 2 results from this Study .538 * .465 ** .593 **
Model fit to the data could only be achieved after the removal of six Y-Variables
and, subsequently, less than half of the variance extracted was explained by Intention
to Trust. Examination of the Y-Variables excluded from the models identified that
four of the six items were the negative-worded questions.
106
Negative-worded questionnaire items complicate the interpretation of CFA results
(Williams, Ford, & Nhung, 2002, p. 376). Studies have identified that wording
effects impact on the factor structure (Harvey, Billings, & Nilan, 1985) and this can
detrimentally affect construct variance. Negative-worded items have also been found
to complicate the interpretation of alternative analysis methods like Exploratory
Factor Analysis (EFA) (Williams et al., 2002, pp. 377-379). Negatively worded
questions introduce an affective factor that has the potential to reduce trait variance
by up to 20% (Rogelberg, 2002). The cognitive response from subjects differs when
questions are positive or negatively worded. For example, a subject tends to look for
how much is in the glass when we use a positive-worded question like; ‘Is the glass
half full?’. Alternatively, a negative-worded question like ‘Is the glass half empty?’
affects a subject differently and results in an altered cognitive response (Williams et
al., 2002).
To bypass the issue one could singularly ‘use positively worded questionnaire items
to avoid the irrelevant construct factor’ (Rogelberg, 2002, pp. 377-379). However,
negatively worded questions contribute to the internal validity of a study by
minimising common method variance and consistency bias (Cavana et al., 2001, p.
229). Negatively worded questions make the subject think and provide direct
feedback to the researcher on whether the subjects understood the questions. An
alternative method provides for the benefits of negatively worded questions as well
as the inclusion of an item-wording factor that has been found to improve CFA
model fit and subsequent interpretation. The inclusion of an orthogonal Reverse-
Coded Method Factor (RCMF) has been has been shown to account for the affective
factor that can confound models incorporating negative-worded questionnaire items
(Williams et al., 2002, p.375). In consideration of a subject’s altered cognitive
response to negative-worded questionnaire items and a resultant detrimental effect on
variance, an RCMF factor was included in the initial and respecified Intention to
Trust CFA models.
In support of retaining negatively-worded questionnaire items, the initial Intention to
Trust CFA model included all the questionnaire items and the RCMF factor that
sought to explain the negatively worded questionnaire items; Itt1, Itt3, Itt5 and Itt8.
Based on their standardised parameter estimates, these items loaded more adequately
107
onto the latent factor, Intention to Trust (Table 26). Other than item Itt7 which
remained non-significant, the individual standardised parameter estimates all
exceeded .41 and were statistically significant (Bollen, 1989a; Joresborg & Sorbom,
1996). The Squared Multiple Correlation (R2) for item Itt7 continued to be
problematic; the R2 values for the other items ranged between .23 and .51. These
preliminary results indicated that the variables were more adequately serving as
measures to the underlying construct (Bollen, 1989a).
Table 26: Squared Multiple Correlations for the Y–Variables in the initial model with the RCMF; N = 224
Item Itt1 Itt2 Itt3 Itt4 Itt5 Itt6 Itt7 Itt8 Itt9 I tt10 Standardised Parameter Estimates with ITT
.43 .51 .47 .68 .45 .73 .02 .80 .67 .67
Standardised Parameter Estimates with ITT and RCMF
.41 .44 .60 .41 .
R2 with ITT
.226 .240 .323 .466 .380 .512 .000 .400 .517 .425
R2 with RCMF and ITT
.358 .259 .409 .460 .596 .54 .000 .460 .64 .452
However, the overall measures of fit suggested problems with the model (χ2 (31)
117.70 at p = .00; RMSEA = .112; AGFI = .831; CFI = .877; IFI = .877; NNFI =
.821). The covariance matrix of the hypothesised model was significantly different to
the covariance matrix of the sample data (Bollen, 1989a) and suggests that re-
specification is needed to improve the model fit to the sample data. Examination of
the modification indices identified that four items were problematic; Itt2, Itt3, Itt7
and Itt9. Item Itt2, I would be willing to let this group have complete control over an
assignment that was critical to me; item Itt3, If I joined this group, I feel I need to
keep an eye on what other group members do; item 7, This group should be honest
about their knowledge, and; item Itt9, I feel I could rely on the skills and abilities of
this group. Even though these items were theory-based, they were found to have
correlated errors and high fitted residuals. Therefore I conclude that, in the context
108
projected by the vignettes, this sample from the population of Business Faculty
Undergraduate Students are associating or confusing the meaning of these items.
The items could be classically dealt with by including within-variable correlated
errors to improve model fit (Joresborg & Sorbom, 1996). However, the correlation of
measurement errors and the additional parameters arising from their use generates
additional error terms, which can result in a corresponding loss of the meaning and
dilute the substantive conclusions that can be drawn from a model (Gerbing &
Anderson, 1984). A more parsimonious first order CFA structure with fewer
parameters to satisfactorily account for the observed covariance matrix is preferred
(Gerbing & Anderson, 1984, p. 579). Additionally, the use of fewer parameters
avoids degrees-of-freedom penalties that can detrimentally affect the GFI and AGFI
absolute indices (Byrne, 1998, p. 116; Kelloway, 1998, pp. 32-33). Therefore, it was
decided not to correlate the measurement errors in the models but instead to test a
series of plausible theory-based models (Bollen, 1989a, p. 71) that progressively
excluded Intention to Trust items Itt2, Itt3, Itt7 and Itt9 to identify a respecified
congeneric CFA model that optimally fitted to the observed covariance matrix.
109
6.1.3.3.3 The Re-specified Intention to Trust Congeneric Mode that excludes Items Itt2, Itt3, Itt7 and Itt9
The respecified congeneric model excluded items Itt2, Itt3, Itt7 and Itt9 from the
published scale (Mayer & Davis, 1999) for the Intention to Trust endogenous factor
(Figure 19). The individual standardised parameter estimates for ITT (Figure 19) all
exceeded .46 and were statistically significant (Bollen, 1989a).
Figure 19: Standardised Solution for the respecified Congeneric CFA Model of the endogenous factor Intention to Trust; items Itt2, Itt3, Itt7 a nd Itt9 excluded
The items appeared to load adequately onto the latent factor (Table 27). The Squared
Multiple Correlations (R2) for each variable ranged between .33 and .63 indicating
that the variables were reasonably serving as adequate measures to the underlying
construct (Bollen, 1989a). Additionally, the model-based measures of scale
reliability, construct reliability and variance extracted (Table 28) were within
suggested thresholds (Coote, 2006).
110
Table 27: Y–Variable Squared Multiple Correlations - the respecified ITT model; N = 224
For the respecified model, all measures of absolute, incremental fit improvement,
and parsimony were within acceptable thresholds (Kelloway, 1998) (RMSEA = .053;
GFI = .963; AGFI = .930; NNFI = .976). However, the respecified model reflecting
the congeneric CFA models of Ability, Benevolence, and Integrity failed to achieve
non-significance (χ2 = (24) = 38.82, p = 0.029) at the p = 0.05 or higher level (Table
44).
Table 44: Summary of model fit difference between the respecified three-factor CFA model and the χχχχ2 threshold for 24 degrees of freedom at p = 0.05
Model χχχχ2 df p = 0.05
Respecified three-factor CFA model 38.82 24 0.029 χ2
THRESHHOLD (24) p = 0.05 36.42 24 0.05 Chi-square Difference at p = .05 2.40 0 0.05
There was a difference between the covariance matrix of the hypothesised model and
that of the observed data (Bollen, 1989b). However, the Chi-square difference
122
(χ2 (1)
= 2.4, p = 0 .05) was considered insufficient to warrant model refitting. Model
differences in the order of 2 are considered insufficient justification for the re-
specification of a model (Kelloway, 1998). Given the minimal Chi-squared
difference, the model was assessed as supporting the hypothesis that the respecified
three-factor CFA model fits the covariance matrix of the observed data. The
following section seeks to operationalise the model by comparing rival models.
6.1.3.3.15 Operationalising of the CFA Model
To clarify relationships between the observed variables and the exogenous factors,
rival models comparison tested the multidimensionality of the hypothesised three-
factor (oblique) CFA model. The comparison of the rival CFA models with the
hypothesised three-factor (oblique) CFA model sought to support the foundation
theory for this study, that the antecedents of Ability, Benevolence and Integrity
explain our Intention to Trust. However, the high positive correlation (.86) between
Benevolence and Integrity (Figure 23) indicate that, when a member joins a newly
formed work-team, Benevolence and Integrity could collapse into a single exogenous
factor. This was tested using a Two-factor oblique model (Table 45).
Table 45: Variable loading combinations for the comparison of rival First Order CFA Models
Model
Ability variables Ab2, Ab3 and Ab4
loaded on
Benevolence variables Ben1, Ben4 and Ben 5
loaded on
Integrity variables Int1, Int2 and Int3
loaded on Hypothesised Model:
Three-factor (oblique)
Ability BEN INT
Two-factor (oblique)
Ability BENINT
123
The Fit indices from the rival models converged, suggesting that the hypothesised
(three oblique factor) model reflecting the theoretical foundation for this study was
superior (Table 46).
Table 46: Fit indices for the competing First Order CFA models of Ability, Benevolence and Integrity identifying the superiority of the three-factor oblique CFA model; N = 224
Model: χ2 df RMSEA RMR GFI AGFI CFI IFI NNFI
Hypothesised
Model:
Three-factor
38.82 24 .053 .036 .963 .930 .984 .984 .976
Two-factor
Ability-
BENINT
73.41 26 .09 .051 .932 .882 .960 .961 .943
In particular, comparison with the rival nested model identifies that the hypothesised
three-factor oblique model provides a superior fit to the data than the competing two-
factor model; the RMSEA measure of absolute measure of fit error has increased to
.09 in the two-factor model (Table 46). The three-factor oblique model better fits the
population covariance matrix from the sample data; all its fit indices are within
acceptable limits.
In conclusion, subsequent to fitting the measurement models of the exogenous
factors and consideration of rival first order CFA models, consistent with the
underlying theory for this study (Mayer et al., 1995), the hypothesised three-factor
(oblique) model better fitted the covariance matrix of the observed data from the
population of Business Faculty Undergraduate Students. The next section pursues a
path analysis of the complete model hypothesising that Ability Benevolence and
Integrity are antecedents of Intention to Trust.
124
6.1.3.3.16 Path Analysis of the Intention to Trust Model
The goal of the path analysis model was a test of Hypothesis H1 that Ability,
Benevolence and Integrity are antecedents of Intention to Trust. The initial path
analysis model that included all the X and Y-Variables did not adequately fit the
covariance matrix of the data: (χ2 (318) = 898.1, p = .000; RMSEA = .09; GFI = .77;
AGFI = .73; CFI = .85; IFI = .85; NNFI = .84). Subsequently, it was hypothesised
that a respecified model concatenating the results of the congeneric models and the
first order CFA models would better fit the covariance matrix of the observed data
(Figure 24).
Figure 24: Standardised solution for the respecified path analysis model of Intention to Trust
and its antecedents Ability, Benevolence and Integrity; N = 224
The items in the respecified model loaded adequately onto the latent factors (Table
47). The standardised parameter estimates of the X and Y-Variables in the
measurement model were statistically significant (Bollen, 1989a); (Joresborg &
Sorbom, 1996); and the Squared Multiple Correlations (R2) for each variable ranged
between .32 and .83 indicating that the variables were reasonably serving as adequate
measures to the underlying construct (Bollen, 1989a).
125
Table 47: Squared Multiple Correlations for the X and Y–Variables in the respecified path analysis model; N = 224
χ2diff = (237) = 755.93, p = 0.05, between the initial and respecified models (Table 48)
was significantly larger than the χ2 CRIT (237) p = 0 .05: 236.33, the respecified path
analysis model has improved fit to the covariance matrix of the observed data.
Table 48: Summary of Model Fit Improvement between the initial and respecified path analysis; N = 224.
Model χ2 df p = .05 RMSEA RMSEA Improved?
Initial Model 898.01 318 .00 .106
Respecified Model 142.08 81 .00 .058 yes
Chi-squared Difference 755.93 237 .05
However, the AGFI (.88) measure of absolute fit was marginally outside the
acceptable limit of .9 (Kelloway, 1998, p. 26). In addition, above limit modification
indices (18.65) were associated with item Int2. A model re-specification that
excluded Int2 from the model did not improve the model fit (Table 49).
126
Table 49: Summary of Model Fit Improvement between the respecified path analysis model and the re-specified path analysis model excluding item Int4; N = 224
Model χ2 df p = .05 RMSEA RMSEA Improved?
Respecified Model 142.08 81 .00 .058
Respecified Model excluding item Int2
116.76 68 .000 .057 NO
Chi-squared Difference 25.32 13 .05
The Chi-square difference, χ2diff = (13) 25.32, between the models (Table 53) is
smaller than the χ2CRIT (13) p = 0 .05: 38.8852. Therefore, the respecified model that
tentatively excluded item Int2 did not significantly improve model fit. Additionally,
in consideration of model re-specification, (Kelloway, 1998, p. 37) argues that
‘model differences of one measure probably do not provide sufficient information to
warrant such action’. Therefore, the hypothesised path analysis model in Figure 25
was considered to better fit the covariance matrix of the observed data.
While the measures of fit could be considered acceptable, the test statistic ratio of the
parameter estimate and its standard error, Est/SE, using maximum likelihood
estimation (Bollen, 1989a; Joresborg & Sorbom, 1996), indicated that all parameter
estimates in the structural model were not statistically significant. Benevolence
appeared not to adequately explain Intention to Trust. The parameter (γ12 = - .40, t
value (81) = - 1.88) estimate marginally failed to achieve statistical significance at t (1,
.05) = +/- 1.96 (Kelloway, 1998, p.29). A non-significant parameter could indicate that
Benevolence was unimportant to the model (Byrne, 1998). A competing ‘Two Factor
Model 1’, that excluded the Benevolence exogenous factor, was tested (Table 50).
The absolute measures of model fit indicated the models similarly fitted the
covariance matrix of the observed data; the RMSEA and RMR values of the
competing models were within .002. From a theoretical perspective there is
considerable support for Benevolence as an antecedent of Intention to Trust (Davis,
1999; Serva et al., 2005; Schoorman et al., 1996a); exclusion of Benevolence from
the model would contradict the theoretical foundation for this study (Mayer et al.,
1995) and previous correlation-based support that Benevolence is an antecedent of
Intention to Trust.
127
Table 50: Fit indices identifying the superiority of the hypothesised three-factor path analysis model; N = 224
Trust Antecedents
Model: Ability BEN INT χ2 df RMSEA RMR GFI AGFI CFI IFI NNFI
Reviewing the hypothesised Three-factor model (Figure 25), the non-significant
negative parameter (γ12 = -.40) between Benevolence and Intention to Trust together
with the high Integrity to Intention to Trust parameter (γ13 = .92) appeared to be
evidence of suppression effect. Suppression effect is present when a parameter is much
larger than the sum of its contributing factors or when positive contributory factors result
in a negative parameter (Cohen, 1988, pp. 76-79; MacKinnon, Krull, & Lockwood,
2000). In the case of the hypothesised Three-factor model (Figure 24), the variables
explained by Benevolence had positive parameters but resulted in a negative parameter
between Benevolence and Intention to Trust. Additionally, in relation to the correlation
of Intention to Trust with its antecedents of Ability (r = .653), Benevolence (r = .357)
and Integrity (r = .639), the Integrity to Intention to Trust parameter was
disproportionately large (γ13 = .92). This evidence, together with the high correlation (r
= .86) between the exogenous factors of Benevolence and Integrity in the first order
CFA model of the trust antecedents, support the presence of suppression effect.
One way to deal with suppression effect is to follow the method of Yuan et al. (2005)
and combine the two highly correlated measures; in this case Benevolence and Integrity.
This method was tested earlier and was unsuccessful; a Two-factor CFA model was an
inferior fit to the covariance matrix of the observed data (Table 45). The hypothesised
Three-factor model better fitted the covariance matrix of the observed data. Consistent
with the theory, Ability, Benevolence and Integrity are considered antecedents of
Intention to Trust. While, the preceding tests supported the hypothesis that Benevolence
does have an effect on Intention to Trust and discounted the (Yuan et al. (2005) method
of dealing with suppression effect, these analyses do not provide evidence concerning
Hypothesis H3, which of the Intention to Trust antecedents is most salient. The following
Analysis of Variance (ANOVA) seeks evidence by examining the simple main and
indirect effects of the trust antecedents of Ability, Benevolence and Integrity on
Intention to Trust.
129
6.1.4 ANOVA of Intention to Trust and its Antecedents of Ability, Benevolence and Integrity
To review the reasons for the dual 2x2 ANOVA; Stage 1 of the study identified that
vignettes depicting dimensionally opposite Information Levels of Benevolence and
Integrity were perceived by the subjects as being unrealistic. This resulted in a dual 2x2
factorial design where a subject was exposed to a single vignette depicting High or Low
Information Levels of Ability and Benevolence or a vignette depicting High or Low
Information Levels of Ability and Integrity (Tables 51 and 52).
Table 51: 2x2 Factorial design for the manipulation of the High and Low Information Levels for Ability and Benevolence
Trust Antecedent
Ability Benevolence High Low High Low
X X X X X X X X
Table 52: 2x2 Factorial design for the manipulation of High and Low Information Levels for Ability and Integrity
Trust Antecedent
Ability Integrity High Low High Low
X X X X X X X X
130
6.1.4.1 Manipulation Checks To add weight to the SEM analysis that identified that the experimental manipulations
embedded in the vignettes had been effective, a series of manipulation checks were
conducted using SPSS for Windows version 13 (SPSS, 2004). Each experimental
condition was assessed using a series of t-tests that supported a claim that the
participants rated the groups differently and congruent with the direction of
manipulation. For each test, the IV value was the summated scale for the variable as
completed after the participant read the vignette. The results indicate that:
• The High Ability Information Level group exhibited a significantly different
mean than the Low Ability Information Level group; (t(178.62) = -15.061, p =
.000).
• The High Benevolence Information Level group exhibited a significantly
different mean than the Low Benevolence Information Level group; (t(119) =
-9.237, p = .000).
• The High Integrity Information Level group exhibited a significantly different
mean than the Low Integrity Information Level group; (t(89.687) = -5.73, p = .000).
These results support the previous claim that all three experimental manipulations had
been successful. The effects of Ability and Integrity on Intention to Trust is next
analysed.
131
6.1.4.2 ANOVA of the effects of Ability and Integrity on Intention to Trust A 2x2 ANOVA evaluated the effects of combinations of High and Low Information
Levels of Ability and Integrity on Intention to Trust. The means and standard deviations
for Intention to Trust as a function of the two variables are presented in Table 53.
Table 53: Descriptive Statistics of the Dependent Variable: Intention to Trust for Groups with High and Low Information Levels of Ability (ability) and Integrity (integ)
ability integ Mean Std. Deviation N 1.00 low 26.0000 6.25833 25 2.00 high 27.7452 5.12036 26
1.00 low
Total 26.8897 5.71711 51 1.00 low 30.0400 6.37293 26 2.00 high 32.7597 3.88109 26
2.00 high
Total 31.3998 5.40168 52 1.00 low 28.0596 6.57781 51 2.00 high 30.2524 5.16191 52
Total
Total 29.1667 5.97880 103
A Levene’s Test of Equality of Error Variances indicated that the error variance of the
dependent variable was homogenous across the four groups and suggests this stated
assumption of ANOVA was not broken; (F(3,99) = 1.623, p = .189). The ANOVA
indicated a significant between-subjects main effect for Ability (F(3,99) = 17.485, p =
.000, η2 = .19); 19% of Intention to Trust was explained by Ability. Integrity also had a
significant main effect (F(3,99) = 4.252, p = .042) but the effect size (η2 = .04) was small
(Tabachnick & Fidell, 2007, p. 55). The interaction between Ability and Integrity was
trivial (η2 = .002) and not significant (F(3,99) = .203, p = .65). Plots (Figures 25 and 26)
of the interaction between the independent variables on the dependent variable identified
that High and Low Information Levels of Ability and Integrity made a significant
difference in the estimated marginal means of the DV, Intention to Trust.
132
Figure 25: Estimated Marginal Means of Intention to Trust with Low or High Integrity
highlow
Ability
33.00
32.00
31.00
30.00
29.00
28.00
27.00
26.00
Est
imat
ed M
argi
nal M
eans
high
lowIntegrity
highlow
Ability
33.00
32.00
31.00
30.00
29.00
28.00
27.00
26.00
Est
imat
ed M
argi
nal M
eans
high
lowIntegrity
133
Figure 26: Estimated Marginal Means of Intention to Trust for groups with Low or High Ability
highlow
Integrity
33.00
32.00
31.00
30.00
29.00
28.00
27.00
26.00
Est
imat
ed M
argi
nal M
eans
high
lowAbility
highlow
Integrity
33.00
32.00
31.00
30.00
29.00
28.00
27.00
26.00
Est
imat
ed M
argi
nal M
eans
high
lowAbility
134
6.1.4.3 ANOVA of the effects of Ability and Benevolence on Intention to Trust
The 2x2 ANOVA evaluated the effects of combinations of High and Low Information
Levels of Ability and Benevolence on Intention to Trust. The means and standard
deviations for Intention to Trust as a function of the two variables are presented in Table
54.
Table 54: Descriptive Statistics of the Dependent Variable: Intention to Trust for Groups with High and Low Information Levels of Ability (ability) and Benevolence (benev)
A Levene’s Test of Equality of Error Variances indicated that the error variance of the
dependent variable was homogenous across the four groups and suggests this stated
assumption of ANOVA was not broken; (F(3,117) = 1.202, p = .312). The ANOVA
indicated a significant between-subjects main effect for Ability (F(3,121) = 47.293, p =
.000, η2 = .29); 29% of Intention to Trust was explained by Ability. Benevolence also
had a significant main effect (F(3,121) = 4.533, p = .035) but the effect size (η2 = .04) was
small (Tabachnick & Fidell, 2007, p. 55). The interaction between Ability and
Benevolence was small (η2 = .04) but significant (F(3,121) = 4.722, p = .032). Based on
these results, the Intention to Trust of the sample appears to be affected to a greater
extent by Ability than Benevolence, which appears to have a weak effect on Intention to
Trust.
22.8433 5.46458 29
26.6552 5.12191 29
24.7493 5.59035 58
30.8622 5.08779 3130.8233 3.70887 32
30.8425 4.40494 63
26.9864 6.60786 60
28.8418 4.87389 61
27.9218 5.84919 121
benev1.00 low
2.00 high
Total
1.00 low
2.00 high
Total
1.00 low
2.00 high
Total
ability1.00 low
2.00 high
Total
Mean Std. Deviation N22.8433 5.46458 29
26.6552 5.12191 29
24.7493 5.59035 58
30.8622 5.08779 3130.8233 3.70887 32
30.8425 4.40494 63
26.9864 6.60786 60
28.8418 4.87389 61
27.9218 5.84919 121
benev1.00 low
2.00 high
Total
1.00 low
2.00 high
Total
1.00 low
2.00 high
Total
ability1.00 low
2.00 high
Total
Mean Std. Deviation N
135
A plot (Figure 27) of the interaction between the independent variables on the dependent
variable identified that High and Low Information Levels of Ability has a significant
difference on Intention to Trust for groups with High (F(1,117) = 40.641, p = .000) and
Low (F(1,117) = 11.149, p = .001) Information Levels of Benevolence. However, the
effect of a High Information Level of Ability was so high (Observed Power = 1.0) that
High or Low Information Levels of Benevolence had little or no effect on Intention to
Trust
Figure 27: Estimated Marginal Means of Intention to Trust with Low or High Benevolence
highlow
Ability
32.00
30.00
28.00
26.00
24.00
22.00
Est
imat
ed M
argi
nal M
eans
high
lowBenevolence
highlow
Ability
32.00
30.00
28.00
26.00
24.00
22.00
Est
imat
ed M
argi
nal M
eans
high
lowBenevolence
136
However, High and Low Information Levels of Benevolence resulted in a significant
difference in the estimated marginal means of Intention to Trust for groups with Low
(F(1,117) = 8.887, p = .003) but not High (F(1,117) = .001, p = .975) Information Levels of
Ability (Figure 28); the Benevolence of a work-team appeared not to affect Intention to
Trust. On the one hand, for participants in the High Ability condition, the level of
Benevolence had no effect on their Intention to Trust. On the other hand, for those in the
Low Ability condition, High Levels of Benevolence resulted in higher Intention to Trust.
The perception is that work-teams with High Ability appears to buffer against any effect
their Benevolence has on Intention to Trust.
highlow
Benevolence
32.00
30.00
28.00
26.00
24.00
22.00
Est
imat
ed M
argi
nal M
eans
high
lowAbility
highlow
Benevolence
32.00
30.00
28.00
26.00
24.00
22.00
Est
imat
ed M
argi
nal M
eans
high
lowAbility
Figure 28: Estimated Marginal Means of Intention to Trust with Low or High Ability
137
6.2 Summary of the Stage 3 Results Two hundred and twenty four subjects were matched to both phases of the data
collection. From the perspective of the secondary goal of this study, contribution to
method, the application of the SEM provided strong evidence that this three-stage study
had developed vignettes, questionnaires and scales that were firmly based on the theory
under test. Subsequent omnibus ANOVA results supported the inference that the refined
vignettes had contributed to subjects better differentiating the High and Low levels of
Ability communicated via the vignettes. In seeking to overcome limitations in similar
trust-related studies, the refinement of the Intention to Trust questionnaire items had
significantly contributed to improving the inter-item consistency and reliability of the
Intention to Trust scale.
Subsequent SEM analysis continued method contribution by exampling a process to
refine the trust scales to optimally reflect a study context. This process involved
Congeneric CFA models that identified the variables explained by the exogenous
factors, Ability, Benevolence, and Integrity as well as the endogenous factor, Intention
to Trust. The application of a Reverse Coded Method Factor contributed to overcoming
the affective factor that arose from the negatively worded Intention to Trust
questionnaire items. The subsequent test of rival First Order CFA models supported the
hypothesis that the earlier defined Congeneric CFA models of Ability, Benevolence and
Integrity reliably explained the trust antecedents. Subsequently, the First Order CFA
Model was applied to a full LISREL model to analyse the path analysis of the Intention
to Trust model. Rival single and two-factor models converged towards a fitted model
that provided support for the hypothesis that the three trust antecedents, Ability,
Benevolence and Integrity, explain Intention to Trust. However, a marginal non-
significant negative Benevolence-Intention to Trust parameter and an Integrity- Intention
to Trust parameter that appeared larger than the sum of its contributory factors heralded
the possibility of suppression effect in the model. A subsequent ANOVA sought to
explain hypothesis H3 by examining the simple main and indirect effects of the trust
antecedents of Ability, Benevolence and Integrity on Intention to Trust.
138
In pursuit of the primary study goal of answering the research questions, the between-
groups ANOVA supported the inference that our Intention to Trust increases when a
Low Integrity work-team has High Ability. Similarly, our Intention to Trust further
increases when a team has High Integrity together with High Ability. Ability had a
similar effect on groups with High and Low Levels of Benevolence. However,
Benevolence had little effect on Intention to Trust when a work-team had a High
Information Level of Ability. To the participants in this study, the competence,
knowledge and skills of a High Ability work-team appears to buffer against the work-
team’s Benevolent traits. Therefore, one could conclude that the study sample from the
population of undergraduate students from the Business Faculty of an Australian
university appear to consider Ability most important when cognitively assessing their
Intention to Trust a work-team that they could choose to join.
139
Chapter Seven
7.1 Overall Summary of Study Findings
In the dyadic relationship of a subject choosing to join a work-team, this research sought
to address a gap in our current understanding of the antecedents that affect co-worker
trust development. In doing so, this study pursued two goals. The first goal of the study
was to enhance our understanding of trust development in newly forming groups by
focussing on the research questions and the hypotheses. From the perspective of the
secondary goal, the three-stage methodology employed in this study improved the
reliability of the published trust scales by customising them to reflect the context of the
study. The Cronbach Alpha statistic for Intention to Trust improved from the low levels
of previous studies (ά = .56) to an acceptable level in Stage 2 (ά = .77) and further
improved to a good result (ά = .83) in Stage 3. The resultant trust scale for Intention to
Trust improved the internal validity of this study and strongly supported the claim that
instruments, particularly vignette manipulations, employed in a study need to
realistically reflect the study context. This supports the argument that context is
inexorably linked to our Intention to Trust.
While the research design of this vignette experiment offers method and scale
improvement alternatives that could be useful for future research, the primary study goal
was to address the research gap that identified a need to enhance our understanding and
seek explanation to the research questions:
R1: How do we develop interpersonal trust in the context of newly forming work-
teams?
Or more specifically
R2: Which trust antecedents are more important when initiating trust development
between the team members?
140
Three hypotheses provided the primary focus for the study effort:
H1: Benevolence and Integrity and Ability will be positively associated
with our Intention to Trust members of a newly forming work-team.
Hypotheses H1 was supported: in Stage 2, Ability (r = .538, p = .01) and Benevolence (r
=.465, p = .01) and Integrity (r = .593, p =.01), and; in Stage 3 Ability (r = .653, p =
.01), Benevolence (r =.357, p = .01) and Integrity (r = .639, p = .01). In the context of
newly forming work teams, Benevolence and Integrity and Ability are positively
associated with Intention to Trust.
H2: At the pre-stage of team development, the trust antecedents of
Ability, Benevolence and Integrity have differential effects on our
Intention to Trust members of a newly forming work-team.
Study Stages 2 and 3 supported Hypothesis H2. Stage 2 results identified differential
correlation relationships between the trust antecedents and Intention to Trust; Ability (r
= .538, p = .05), Benevolence (r =.465, p = .01) and Integrity (r = .593, p = .01).
Similarly, different IV correlations were witnessed in Study Stage 3; Ability (r = .653, p
= .01), Benevolence (r =.357, p = .01) and Integrity (r = .639, p = .01). More
importantly, the results from ANOVA and pairwise comparisons of the main and
interactive effects of the trust antecedents identified that the sample from a population of
undergraduate students from the Business Faculty of an Australian university appear to
consider Ability more important than Benevolence or Integrity when cognitively
assessing their Intention to Trust a work-team that they could choose to join. These
results support the hypothesis that the trust antecedents of Ability, Benevolence and
Integrity have differential effects on our Intention to Trust.
141
H3: Ability will have a more salient effect on our Intention to Trust
members of a newly forming work-team than Integrity or Benevolence.
Based on correlation results, support for Hypothesis H3 was initially inconclusive. The
pilot of the test instruments in Study Stage 2 resulted in a marginally stronger Intention
to Trust relationship with Integrity (r = .593, p = .01) than Ability (r = .538, p = .05) or
Benevolence (r =.465, p = .01). This correlation relationship was reversed in Stage 3; the
Intention to Trust relationship with Ability (r = .653, p = .01) was marginally stronger
than Integrity (r = .639, p = .01) and Benevolence (r =.357, p = .01). However, focusing
on Ability and Integrity, the results from ANOVA and pairwise comparisons of their
main and interactive effects provided stronger support for a claim that Ability is more
salient than Integrity in the development of Intention to Trust. The ANOVA indicated a
significant between-subjects main effect for Ability (F(3,99) = 17.485, p = .000, η2 = .19);
19% of Intention to Trust was explained by Ability. Integrity also had a significant main
effect (F(3,99) = 4.252, p = .042, η2 = .04) but it explained less (4%) of the variance
associated with Intention to Trust. The interaction between Ability and Integrity was
trivial (η2 = .002) and not significant (F(3,99) = .203, p = .65). Therefore, in the context
of members of newly forming work-teams, Ability is considered to have a more salient
effect on Intention to Trust than Integrity.
Similarly, Ability has a more salient effect on the development of Intention to Trust than
Benevolence. The ANOVA indicated a significant between-subjects main effect for
Ability ( F(3,121) = 47.293, p = .000, η2 = .29); 29% of Intention to Trust was explained
by Ability. Benevolence also had a significant main effect (F(3,121) = 4.533, p = .035) but
the small effect size (η2 = .04) indicated that Benevolence only explained 4% of the
variance associated with Intention to Trust. The interaction between the independent
variables and the dependent variable identified that High and Low Information Levels
of Ability has a significant difference on Intention to Trust for groups with High (F(1,117)
= 40.641, p = .000) and Low (F(1,117) = 11.149, p = .001) Information Levels of
Benevolence. However, the effect of a High Information Level of Ability was so high
(Observed Power = 1.0) that High or Low Information Levels of Benevolence had little
142
or no effect on Intention to Trust. Based on these results, the Intention to Trust of the
sample appears to be affected largely by Ability than Benevolence, which appears to
have a weak effect on Intention to Trust.
In summary, for study participants in a sample from the population of undergraduate
students from the Business Faculty of an Australian university, the Benevolence of a
work-team has no effect on their Intention to Trust a work-team with High levels of
Ability but does have a minimal effect on their Intention to Trust a Low Ability work-
team. Additionally, Integrity has a lesser effect on Intention to Trust than Ability. The
ANOVA results support Hypothesis H3; in the context of a member joining a newly
forming work-team, Ability has a more salient effect on Intention to Trust than
Benevolence or Integrity.
143
Chapter Eight
8.1 Discussion, Limitations and Future Research
It has been argued that understanding how to better develop trust has the potential to
assist management to realise the positive outcomes that can benefit an organisation, its
employees and the customers they serve. This research sought to contribute to the
success of our organisations by deepening our understanding of trust development, the
foundation for high performing individuals, teams, and their organisations.. The
contributive-effort focused on two gaols. Firstly, an experimental method has been
tested and is offered for consideration by researchers who seek to satisfy the requirement
for vignette-based manipulations that must appear realistic to the study participants and
optimally reflect a study setting (Murphy et al., 1986). This method contribution also
sought and improved published trust-related scales because it is important that our scales
consistently and accurately measure the variable they purport to measure (Christensen,
2004, pp. 185-198). By improving the quality of the measurement scales, this study
contributed by providing a three-stage methodology that could prove helpful to others
who seek to refine questionnaire item wording and vignette development. In particular,
the method led to improvements in the internal consistency of the previously
problematic Intention to Trust measure by increasing its Cronbach Alpha reliability
coefficient to a magnitude not evidenced in recent trust-related studies. The method is
offered for consideration by other researchers.
While the method contribution may be useful to other researchers, the primary goal of
this study was to contribute to our knowledge of how to develop the trust that leads to
higher performing employees, the teams they work in, and the organisations that engage
them (Ana Cristina Costa, 2003; Kiffin-Petersen & Cordery, 2003). This is particularly
important in today’s work environment because of the predominance of part-time and
contract employees (ABS, 2005) who do not benefit from the trusting work environment
that emerges from the culture of organisations with permanent employees (Misztal,
2002). Today, the number of part-time and temporary employees in Australian
workplaces is increasing (ABS, 2006). In the shadow of contemporary work
144
environments that exhibit disparate views of trust (Elangovan & Shapiro, 1998; Jones &
George, 1998), this study effort addressed a gap in the literature. Previous studies have
not addressed the pre-stage of team development. This is understandable because as
permanent employees we would transition this pre-stage once, when we first join a new
organisation.
As a permanent employee, we would go through a pre-stage induction or training
process designed to socialise us into our new work-team environment. From the training,
we would gain an understanding of what the organisation expected of us, the principles
and standards that underpin the integrity of the organisation. As we interacted with our
fellow team members, we would gain an understanding of their strengths and
weaknesses, those areas where they had particular abilities or skills. The feedback we
received from our multiple interactions with fellow employees would inform us of how
they treated fellow team members, how benevolent they were to others. Having
progressed through this initial pre-stage of team development, we would join our work-
team. Even if we moved throughout the organisation, we would be joining teams who
shared similar principles and values to ourselves. In environments where permanent
employees are pre-socialised, the pre-stage of team development would be almost non-
existent. This is not the case with part-time or temporary employees. They would need
to go through the pre-stage of team development every time they accepted a temporary
position with people whom they had not previously worked. These temporary employees
would have no information upon which to found expectations about what team members
had what skills, who was reliable, and, how they would be treated by other team
members. If we accept that increased levels of trust leads higher performing work-teams,
then it is important to understand the pre-stage of team development and how to
optimally develop trust between work-teams of temporary or part-time employees.
This and a plethora of previous studies have provided considerable evidence that our
consideration of the abilities, benevolence and integrity of others contributes to our
decision to extend trust to others. Like this study, much previous research was based
upon the integrative model of organisational trust (Mayer et al., 1995) which equitably
145
viewed the trust antecedents of ability, benevolence and integrity (Costa et al., 2001;
Davis et al., 2000; Gill et al., 2005; Mayer & Gavin, 2005; Schoorman et al., 1996a;
Serva et al., 2005). However, these studies did not address the pre-stage of team
development. This raises the question addressed by this research; which trust antecedent
is most salient in the pre-stage of team development? In pursuing this question, this
study built upon the existing integrative model of organisational trust by incorporating
the contextual moderation factor of Information Level that accounts for the type and
volume of information a trustor applies when cognitively making a decision to extend
trust to a trustee. Experimental manipulation of high and low Information Levels about
potential team members provided insight into what trust related attributes we expect in
our fellow team members. In the context of a member joining a newly forming work
team, this study found that Ability had a greater effect on our Intention to Trust than
Benevolence or Integrity.
More specifically, while our Intention to Trust a Low Ability team increased when the
team exhibited either a High level of Benevolence or a High level of Integrity, this was
not the case when the team exhibited High Ability. When the team exhibited a High
level of Ability, Integrity had a lesser effect on our Intention to Trust but more
importantly High or Low Information Levels of Benevolence had little or no effect on
our Intention to Trust a High Ability team. This result was unexpected and is counter to
the current literature. We would reasonably expect that the way we are treated, our level
of Benevolence to fellow team members, would in some way affect how we extend trust.
These results would have us believe that fellow team members do not care about how
they are treated by fellow team members. Clearly, this is not the case. As we work with
people we seek the acceptance and recognition through our everyday interactions (Blau,
1964). Working with people who we like and who reciprocate by liking and treating us
fairly is a much more desirable work environment. So how did this unexpected result
emerge?
146
This discussion has identified that it is not normal for us to ignore how we are treated by
others. Previous studies identified that Ability Benevolence and Integrity are all
important in the development of trust (Costa, 2003; Costa et al., 2001; Earley, 1986;
Davis, 1996b; Serva, Benmati, & Fuller, 2005; Tan & Tan, 2000; Whitener et al., 1998).
Perhaps the lack of effect from Benevolence is a reflection of the research design for this
study. After all, in Study Stage 1, the thematic analysis of structured interviews
identified eight times more participant statements relating to the Integrity, and five times
more statements relating to the Benevolence, than the Ability of a work-team. Based on
this qualitative evidence, one could conclude that Ability is the least important of the
trust antecedents. However, in Study Stages 2 and 3, quantitative results identified that
Ability was perceived to be more salient than Benevolence or Integrity. In seeking
explanation, Murphy et al. (1986) remind us that vignettes, like those applied in Stages 2
and 3, project information that is less subject to the confounds that can contaminate
experimental studies. The design of this study specifically sought to exclude confounds
from Study Stages 2 and 3. Having excluded these confounds, it would be reasonable to
expect that the salience of the trust antecedents could be different from the qualitative
Stage 1.
While vignettes help to minimise confounds, they can amplify the signal strength of a
manipulation above that of real-life situations (Murphy et al. 1986). This can affect the
power of a statistical test and could have caused one or more of the trust antecedents to
have a greater effect on Intention to Trust than the other antecedents. This may have
occurred with Ability. The observed power for Ability was nearly twice the result of
Benevolence and Integrity. This was unexpected because a priori calculation of effect
size and power predicted similar levels for Ability, Benevolence and Integrity. However,
the studies used as a basis for the priori power calculations did not address the pre-stage
context of team development; this study focussed on the pre-stage. Rousseau et al.
(1998, p. 398) assert the importance of considering context in trust-related research, the
importance of certain factors change depending upon the situation in which we find
ourselves. Therefore, the unexpected result could be reflective of the pre-stage context
147
of this study. Seeking another explanation, perhaps the subjects perceived stronger
Ability information in the vignettes than Benevolence and Integrity information.
Great care was taken to ensure a balanced projection of Ability, Benevolence and
Integrity information in the vignette content. However, the Ability information did use
more physical page-space in the vignettes. Perhaps the tabular depiction of the grade
history of the other work-team members may have caused the subjects to retain the
Ability information more prominently in their short-term memory and prompted them
make an Ability-related decision to join the work-team (Lurie & Mason, 2007). Thus,
the subjects could have more readily recalled Ability information when responding to
the questionnaire. However, statistical tests supported the inference that Benevolence
and Integrity manipulations were successful. Benevolence and Integrity information did
affect a subject’s Intention to Trust when the work-team had Low Ability. Of course,
this was not the case when the work-team had High Ability. Perhaps the information in
the High Ability vignettes was considerably stronger than the Low Ability information
in the vignettes. While this is a plausible explanation, it is considered unlikely because
the manipulation checks supported the effectiveness of the High and Low Ability
Information Levels depicted in the vignettes. Therefore, the strength of the High and
Low Ability information is considered balanced and equitable. While the strength of the
vignette information could have affected the salience of the trust antecedents, this is not
supported by the statistical tests. Another explanation is that subjects from a population
of undergraduate students from the Business Faculty of an Australian university do
consider Ability more important when considering their Intention to Trust a newly
forming work-team.
It somehow appears logical that students would consider Ability more important. After
all, they are at university to gain a professional qualification and the Australian
university constantly reminds them that, post graduation, their grade point average is
positively correlated with the starting salary they could expect from their initial
employer (Sandvig, Tyran, & Ross, 2005). Given that High Ability students have
superior skills and greater knowledge and is rewarded with higher grades, it would be
148
reasonable to assume that if we joined a work-team with High Ability we could expect
to increase the probability that we too would be rewarded with a higher grade; thus
increasing our overall grade point average and the chances of getting a higher paid job.
In responding to whether or not they would agree to join the group depicted in the
vignette, all subjects offered a group with high Ability agreed to join the group. This was
not the case with those offered a low Ability group. Just 15% of those subjects agreed to
join the group and they were subjects offered a group with Low Ability and a High level
of Integrity. Clearly students appear to focus on the Ability of fellow team members and
generally ignore the team’s Benevolence and to a lesser extent their Integrity.
In general, we could conclude that students are more positively affected by a High
Ability work-team, will trust them more because of the higher probability of success and
reward, and, be more likely to join a High Ability work-team. In explanation, the
students could view the scenario depicted in the vignettes as an example of transactional
trust wherein they need to extend trust to the other members of their work-team for the
twelve-week duration of a semester after which they may never come into contact with
those students again. They do not care how they are treated, they just want a high grade.
This could be a manifestation of halo error (McShane & Travaglione, 2003) wherein a
student’s need, or what Enos (2002) terms their positive emotional attachment, to
achieve a high grade point average causes them to singularly focus on the Ability trait of
the work-team while ignoring the work-team’s Benevolence and Integrity traits.
Of course, this beggars the question: does a similar situation arise in our workplaces that
are employing more and more part-time or contract employees. Given that context is
important in how we develop trust (Rousseau et al. 1998) perhaps a comparison between
a student environment and that of a part-time worker will help us understand the
applicability of this research to the general workplace. Firstly, the dynamics of the
trusting relationship is similar in both contexts; both the student and the part-time
workers come together for a finite period of time after which they may not meet again.
From the perspective of motivation and reward, both the part-time worker and the
student would view the trusting episode as a transaction. As a student, we would focus
149
on optimising our grade that on average should improve if we work with students of
higher ability. Given that our workplace tends to pay higher wages for higher
performance, a part-time worker would expect personal rewards to be maximised when
they joined a high ability work-team that would normally result in higher level of
performance. Therefore, the students and part-time workers could both view Ability as
most important; ability leads to a higher grade and higher monetary rewards. The repeat
of this study in a workplace context offers opportunity for ongoing research.
If Ability is more salient in the development of trust between members of a newly
forming work-teams, and higher levels of trust leads to higher performance teams, then
how should managers develop trust between team members? Should a manager just
ignore Benevolence and Integrity and focus on having everyone in the team exhort the
excellence of their Abilities? Clearly not, because the study of Gill et al. (2005) and
many other researchers has identified that the salience of the trust antecedents varies
over the life of a work-team. In the performing stage of a team Gill et al. (2005) found
that the importance of Ability waned in favour of Benevolence. Once the team members
had a clear understanding of the skills, expertise and Abilities of fellow team members,
they were more concerned about the other team members delivering the work to the
expected standard at a time when it was due. This is understandable because, in the short
life of a team project there would be little time to retrain team members. We would need
to identify who had the ability to do what, allocate tasks based on skills and expertise,
then trust in the benevolent nature of all team members to deliver as expected. The
willingness of students to join the High Ability groups would tend to indicate that this
was probably the thinking of the subjects in this study.
Of course, this is a very short-term view by the subjects in this study; they would very
quickly find out that not everyone does what they say they will do. After all, this is why
we have project managers within teams, to keep the work on time and to budget.
Therefore, as managers we should not ignore Benevolence and Integrity. It is our
Integrity, our values and principles, which underpin our expectations of, and compliance
with, group norms of behaviour (McShane & Travaglione, 2003, p. 270). Ability
150
Benevolence and Integrity are all important in trust development. What this study has
identified is that in the pre-stage of team development we have a propensity to be
blinded by the skills and expertise of others. As managers, we need to help our work-
teams over the halo effects of Ability. Managers could encourage unfamiliar team
members to not only share their knowledge and skills but more importantly, their past
group experiences. In sharing past group work experiences, team members could better
adjudge how a fellow team member is likely to perform later on in the project. Perhaps
the provision of references to past team members could help confirm the claims of our
fellow team members. Such manager-initiated actions could go some way to mitigating
the risk and help lower the risk of our work-teams underperforming. Some managers
could consider such get-to-know-you sessions a waste of time, particularly with teams
that are only together for a short time. However, this study has provide considerable
evidence to support a claim that a manager cannot just rely on the judgement of part-
time employees to develop the trust that leads to higher performing work teams, as
managers, we need to help them overcome their limitations.
Like our part-time workers, this study also has limitation. The most obvious limitation
emerges from the generalisability of the results; Business Faculty undergraduate
students from a metropolitan Australian university may not accurately reflect the
parameters of our part-time workforce. It has already been identified that future research
should consider trust development over the complete life of a work-team; the pre-stage,
the forming stage, the storming stage, the norming stage, the performing stage, and the
adjourning stage. Such research could help managers better understand how to fine-tune
their skills to continually develop greater levels of trust over the life of their work-teams.
Another limitation of this study is the use of vignette manipulation. Perhaps this vignette
experiment reveals a deeper method-related question; do we take more notice of written
information about how we will be treated by others or do we prefer personal contact?
Perhaps such a research question should focus on the psychometric limitation in this
study by questioning whether written vignette manipulations, paper people, are a valid
proxy for the interpersonal communication of information about fellow team members.
151
The use of vignette manipulation may not be single limitation in this experimental study.
A second limitation could be the thematic analysis of the structured interview transcripts
in Study Stage 1. They are the qualitative interpretations; analysis by another researcher
could reveal alternative meanings. A third limitation; because the study setting related
directly to a group assessment context the questionnaire responses of some student
subjects could have been positive or negatively affected by their previous assessment
experiences. In particular, a negative affective bias may have emerged because there
were more negative comments about group assessment in the feedback section of the
questionnaires than positive assessment-related comments. A fourth limitation arises
from a possible pre-testing bias in the Stage 2 questionnaire that concurrently collected
Intention to Trust and Propensity to Trust responses. However, future studies could
address this limitation by following the example of Study Stage 3 and collecting the data
at different times. The final limitation arises from Structural Equation Model fitting
interpretations and the decisions of this researcher to include or exclude certain
variables. Although the model fitting adjustments were theoretical and empirically
founded, it would be reasonable to expect that another researcher could make different
model trimming decisions. Such methodological refinements could lead to future
research opportunities.
This research has already revealed a plethora of future research opportunities. The first
is to strengthen the generalisability of the results by repeating this study across a
population that is more reflective of today’s work environment. Secondly, the
phenomenon glimpsed in Study Stage 2 may worthy of future research; do universities
that allow students to graduate with a small number of failing grades result in students
and universities that are accepting of lower academic standards? The problems that
emerged in Stage 3, with respect of matching multistage data collections from
anonymous subjects, offers an opportunity for future research. Additionally, the
collection of additional data on personal attribute factors offers the opportunity to
identify the personality types that lead to certain trusting behaviours; such knowledge
could help management select the most suitable employees for work-team assignments.
The most obvious research opportunity emerges from the unexpected results that
152
identified the minimal effect of Benevolence and Integrity on Intention to Trust when a
work-team exhibits High Ability. This question in particular motivates a desire to pursue
evidence of causation, perhaps a longitudinal study in a more generalisable setting that
offers the opportunity to strengthen the notion and explain why the dynamic nature of
trust leads to transient and varying antecedent effects on our Intention to Trust.
153
Reference List
Abrams, L. C., Cross, R., Lesser, E., & Levin, D. Z. (2003). Nurturing interpersonal trust in knowledge-sharing networks. Academy of Management Executive, 17(4), 64-77.
ABS. (2003). Work - Paid work: Changes in labour force participation across
generations. Retrieved October 20, 2005, from http://www.abs.gov.au.ezp02.library.qut.edu.au/Ausstats/[email protected]/0/C39B360652A7A8BFCA256D39001BC355?Open
ABS. (2005). Labour Force, Australia, 6202.0.55.001. Retrieved May 4, 2006, from
ABS. (2006). Labour Force, Australia, 6202.0. Retrieved January 4, 2007, from
http://www.abs.gov.au/AUSSTATS/[email protected]/mf/6202.0?OpenDocument Akaike, H. (1987). Factor analysis and AIC. Psychometrika, 52(3), 317-332. Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: a
review and recommended two-step approach. Psychological Bulletin, 103(3), 411-423.
Arvey, R. D., & Wanek, J. E. (2006). Workplace honesty: an honest review.
PsycCRITIQUES. Aryee, S., Budhwar, P. S., & Chen, Z. X. (2002). Trust as a mediator of the relationship
between organizational justice and work outcomes: Test of a social exchange model. Journal of Organizational Behavior, 23(3), 267-286.
Atwater, L. E. (1988). The relative importance of situational and individual variables in
predicting leader behavior: The surprising impact of subordinate trust. Group & Organization Studies, 13(3), 290-310.
Audi, R., & Murphy, P. E. (2006). The many faces of integrity. Business Ethics
Quarterly, 16(1), 3-21. Auh, S. (2005). The effects of soft and hard service attributes on loyalty: the mediating
role of trust. Journal of Services Marketing, 19(2), 81-92. Becker, T. E. (1998). Integrity in organizations: Beyond honesty and conscientiousness.
Academy of Management Review, 23(1), 154-161.
154
Bell, G. G., Oppenheimer, R. J., & Bastien, A. (2002). Trust deterioration in an international buyer-supplier relationship. Journal of Business Ethics, 36(1-2), 65-78.
Bentler, P. M., & Chou, C.-P. (1987). Practical issues in structural modeling.
Sociological Methods & Research, 16(1), 78. Bews, N. F., & Rossouw, G. J. (2002). A role for business ethics in facilitating
trustworthiness. Journal of Business Ethics, 39(4), 377-390. Bhattacharya, R., & Devinney, T. M. (1998). A formal model of trust based on
outcomes. Academy of Management Review, 23(3), 459-472. Bies, R. J., Shapiro, D. L., & Cummings, L. L. (1988). Causal accounts and managing
organizational conflict: Is it enough to say it's not my fault? Communication Research, 15(4), 381-399.
Bigley, G. A., & Pearce, J. L. (1998). Straining for shared meaning in organization
science: problems of trust and distrust. Academy of Management Review, 23(3), 405-421.
Bijlsma-Frankema, K., & Costa, A. C. (2005). Understanding the trust-control nexus.
International Sociology, 20(3), 259-282. Blau, P. M. (1964). Exchange and power in social life. New York: John Wiley. Blum, S. D. (2005). Five approaches to explaining "truth" and "deception" in human
communication. Journal of Anthropological Research, 61(3), 289-315. Bollen, K. A. (1989a). A new incremental fit Index for general structural equation
models. Sociological Methods Research, 17(3), 303-316. Bollen, K. A. (1989b). Structural equations with latent variables. New York: Wiley. Buchan, N. R., Croson, R. T. A., & Dawes, R. M. (2002). Swift neighbors and persistent
strangers: a cross-cultural investigation of trust and reciprocity in social exchange. American Journal of Sociology, 108(1), 168-206.
Buchner, A., Erdfelder, E., & Faul, F. (1997). How to use G*Power. Retrieved October
12, 2005, from http://www.psycho.uni-duesseldorf.de/aap/projects/gpower/how_to_use_gpower.html
Bulmash, H. (2006). Corporate theft: crooks might be closer than you think. Canadian
Consulting Engineer, 47(7), 42-46. Bussey, K. (1992). Lying and truthfulness: Children's definitions, standards, and
Butler Jr., J. K. (1991). Toward understanding and measuring conditions of trust:
evolution of a conditions of trust inventory. Journal of Management, 17(3), 643. Byrne, B. M. (1998). Structural equation modeling with LISREL, PRELIS, and SIMPLIS
: basic concepts, applications, and programming. Mahwah, N.J.: L. Erlbaum Associates.
Byrne, B. M., & Bazana, P. G. (1996). Investigating the Measurement of Social and
Academic Competencies for Early/Late Preadolescents and Adolescents: A Multitrait--Multimethod Analysis. Applied Measurement in Education, 9(2), 113.
Cardona, P., & Elola, A. (2003). Trust in management: the effect of managerial
trustworthy behaviour and reciprocity. Working Paper. Unpublished manuscript. Carlopio, J. R., Andrewartha, G., Armstrong, H., & Whetten, D. A. (2001). Developing
management skills : a comprehensive guide for leaders (2nd ed.). Frenchs Forest, N.S.W.: Pearson Education.
Carpenter, P. A., Miyake, A., & Just, M. A. (1995). Language comprehension: sentence
and discourse processing. Annual Review of Psychology, 46, 91-100. Casielles, R. V., Alvarez, L. S., & Martin, A. M. D. (2005). Trust as a key factor in
successful relationships between consumers and retail service providers. Service Industries Journal, 25(1), 83-101.
Cavana, R. Y., Sekaran, U., & Delahaye, B. L. (2001). Applied business research :
qualitative and quantitative methods. Milton, Qld.: John Wiley & Sons Australia. Champy, J. A. (1997). Preparing for organisational change. In F. Heselbein, M.
Goldsmith & R. Beckhard (Eds.), The Organization of the Future (pp. 9-18). San Francisco: Jossey-Bas Publishers.
Chansler, P. A., Swamidass, P. M., & Cammann, C. (2003). Self-managing work teams -
an empirical study of group cohesiveness in "natural work groups" at a Harley-Davidson Motor Company plant. Small Group Research, 34(1), 101-120.
Child, J. (2001). Trust--The fundamental bond in global collaboration. Organizational
Dynamics, 29(4), 274-288. Child, J., & Mollering, G. (2003). Contextual confidence and active trust development in
the Chinese business environment. Organization Science, 14(1), 69-80. Christensen, L. B. (1994). Experimental methodology (6th ed.). Boston: Allyn and
Bacon.
156
Christensen, L. B. (2004). Experimental methodology (9th ed.). Boston, MA: Allyn and Bacon.
Cliff, N. (1983). Some cautions concerning the application of causal modeling methods.
Multivariate Behavioral Research, 18(1), 115-126. Cohen, J. (1977). Statistical power analysis for the behavioral sciences (Rev. ed.). New
York: Academic Press. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.).
Hillsdale, N.J.: L. Erlbaum Associates. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple
regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, N.J.: L. Erlbaum Associates.
Coletti, A. L., Sedatole, K. L., & Towry, K. L. (2005). The effect of control systems on
trust and cooperation in collaborative environments. Accounting Review, 80(2), 477-500.
Connell, J., Ferres, N., & Travaglione, T. (2003). Engendering trust in manager-
subordinate relationships: predictors and outcomes. Personnel Review, 32(5), 569-587.
Cook, J., & Wall, T. (1980). New work attitude measures of trust, organizational
commitment and personal need non-fulfillment. Journal of Occupational Psychology, 53(1), 39-52.
Coote, L. (2006). An introduction to structural equation modelling; course notes. St
Lucia: University of Queensland. Coppola, N. W., Hiltz, S. R., & Rotter, N. G. (2004). Building trust in virtual teams.
IEEE Transactions on Professional Communication, 47(2), 95-104. Cortina, J. M., Chen, G., & Dunlap, W. P. (2001). Testing interaction effects in LISREL:
examination and illustration of available procedures. Organizational Research Methods, 4(4), 324.
Costa, A. C. (2003). Measuring trust in teams: Development and validation of the trust
team inventory. Australian Journal of Psychology, 55, 120. Costa, A. C. (2003). Work team trust and effectiveness. Personnel Review, 32(5), 605-
622. Costa, A. C., Roe, R. A., & Taillieu, T. (2001). Trust within teams: The relation with
performance effectiveness. European Journal of Work & Organizational Psychology, 10(3), 225.
157
Costigan, R. D., Ilter, S. S., & Berman, J. J. (1998). A multi-dimensional study of trust
in organizations. Journal of Managerial Issues, 10(3), 303-317. Craig, S. C., Martinez, M. D., Gainous, J., & Kane, J. G. (2006). Winners, losers, and
election context: voter responses to the 2000 presidential election. Political Research Quarterly, 59(4), 579-592.
Currall, S. C., & Inkpen, A. C. (2002). A multilevel approach to trust in joint ventures.
Journal of International Business Studies, 33(3), 479-495. Curran, P. J., West, S. G., & Finch, J. F. (1996). The robustness of test statistics to
nonnormality and specification error in confirmatory factor analysis. Psychological Methods, 1(1), 16-29.
Daft, R. L. (2004). Organization theory and design (8th ed.). Cincinnati, Ohio: South-
Western College Publishing. Das, T. K., & Teng, B. S. (2001a). Relational risk and its personal correlates in strategic
alliances. Journal of Business and Psychology, 15(3), 449-465. Das, T. K., & Teng, B. S. (2001b). Trust, control, and risk in strategic alliances: an
integrated framework. Organization Studies, 22(2), 251-283. Das, T. K., & Teng, B. S. (2004). The risk-based view of trust: a conceptual framework.
Journal of Business and Psychology, 19(1), 85-116. Davis, A., & Rothstein, H. (2006). The effects of the perceived behavioral integrity of
managers on employee attitudes: a meta-analysis. (Report No. 01674544): Springer Science & Business Media B.V.
Davis, G. M.-T. W. (1999). A test of an interpersonal trust model. (integrity,
benevolence). Univ Microfilms International. Davis, J. H., Schoorman, F. D., Mayer, R. C., & Tan, H. H. (2000). The trusted general
manager and business unit performance: empirical evidence of a competitive. Strategic Management Journal, 21(5), 563-576.
De Dreu, C. K. W., Giebels, E., & Van de Vliert, E. (1998). Social motives and thrust in
integrative negotiation: the disruptive effects of punitive capability. Journal of Applied Psychology, 83(3), 407-422.
Deutsch, M. (1960). The effect of motivational orientation upon trust and suspicion.
Human Relations, 13, 123-139. Dirks, K. T. (1999). The effects of interpersonal trust on work group performance.
Journal of Applied Psychology, 84(3), 445-455.
158
Drucker, P. F. (1997). Introduction: toward the new organisation. In F. Hesselbein, M.
Goldsmith, R. Beckhard & Peter F. Drucker Foundation for Nonprofit Management. (Eds.), The organization of the future (1st ed., pp. xvi, 397). San Francisco, California: Jossey-Bass Publishers.
Drucker, P. F. (1999). Management challenges for the 21st century (1st ed.). New York:
HarperBusiness. Earley, P. C. (1986). Trust, perceived importance of praise and criticism, and work
performance: an examination of feedback in the United States and England. Journal of Management, 12(4), 457.
Elangovan, A. R., & Shapiro, D. I. (1998). Betrayal of trust in organizations. Academy of
Management Review, 23(3), 547-566. Elsbach, K. D., & Elofson, G. (2000). How the packaging of decision explanations
affects perceptions of trustworthiness. Academy of Management Journal, 43(1), 80-89.
Enders, C. K. (2006). Analyzing structural equation models with missing data. In G. R.
Hancock & R. O. Mueller (Eds.), Structural equation modeling : a second course (pp. 313-344). Greenwich, Conn.: IAP.
Enos, M. (2002). Defining emotion with measures of intensity and range of experience.,
ProQuest Information & Learning, US. Ferres, N., Connell, J., & Travaglione, A. (2004). Co-worker trust as a social catalyst for
constructive employee attitudes. Journal of Managerial Psychology, 19(6), 608-622.
Ferrin, D. L., & Dirks, K. T. (2003). The use of rewards to increase and decrease trust:
mediating processes and differential effects. Organization Science, 14(1), 18-31. Fitzgerald, B. J., Pasewark, R. A., & Noah, S. J. (1970). Validity of Rotter's
interpersonal trust scale: a study of delinquent adolescents. Psychological Reports, Vol. 26(1), 163-166.
Foster, G. (2006). Do students want to succeed? peer group choice, social influence, and
undergraduate performance, Working paper. Adelaide: University of South Australia.
Franta, P. J. (2000). A validation study of Shaw's assessment of organizational
trustworthiness (Robert Shaw). Univ Microfilms International. Frost, T., Stimpson, D. V., & Maughan, M. R. (1978). Some correlates of trust. Journal
of Psychology: Interdisciplinary and Applied, 99(1), 103-108.
159
Fukuyama, F. (1995). Trust : the social virtues and the creation of prosperity. New
York: Free Press. Gabarro, J., & Athos, J. (1976). Interpersonal relations and communications.
Englewoods Cliffs, NJ: Prentice-Hall. Galford, R., & Drapeau, A. S. (2003). The enemies of trust. Harvard Business Review,
81(2), 88. Gambetta, D. (2003). Can we trust trust? In E. Ostrom & T. K. Ahn (Eds.), Foundations
of social capital (pp. 274-290). Cheltenham, UK Northampton, MA: Edward Elgar Pub.
Gerbing, D. W., & Anderson, J. C. (1984). On the meaning of within-factor correlated
measurement errors. Journal of Consumer Research, 11(1), 572. Giffin, K. (1967). The contribution of studies of source credibility to a theory of
interpersonal trust in the communication process. Psychological Bulletin, 68(2), 104-120.
Gill, H., Boies, K., Finegan, J., & McNally, J. (2005). Antecedents Of trust: establishing
a boundary condition for the relation between propensity to trust and intention to trust. Journal of Business & Psychology, 19(3), 287-302.
Gillespie, N. (2003). Measuring interpersonal trust in work relationships. Australian
Journal of Psychology, 55, 124. Gliner, J. A., Haber, E., & Weise, J. (1999). Use of controlled vignettes in evaluation:
does type of response method make a difference? Evaluation and Program Planning, 22, 313-322.
Golembiewski, R. T., & McConkie, M. (1975). The centrality of interpersonal trust in
group processes. In C. L. Cooper (Ed.), Theories of group processes (pp. 131-185). New York: Wiley.
Handy, C. B. (1996). Beyond certainty : the changing worlds of organizations. Boston:
Harvard Business School Press. Harvey, R. J., Billings, R. S., & Nilan, K. J. (1985). Confirmatory Factor Analysis of the
Job Diagnostic Survey : Good News and Bad News. Journal of Applied Psychology, 70(3), 461-468.
Hill, C. W. L., & Jones, G. R. (2004). Strategic management : an integrated approach
(6th ed.). Boston: Houghton Mifflin.
160
Hoelter, J. W. (1983). The analysis of covariance structures: goodness-of-fit Indices. Sociological Methods Research, 11(3), 325-344.
Hosmer, L. T. (1995). Trust: the connecting link between organizational theory and
philosophical ethics. Academy of Management Review, 20(2), 379-403. Hu, L. T., & Bentler, P. M. (1995). Evaluating model fit. In R. H. Hoyle (Ed.),
Huff, L., & Kelley, L. (2003). Levels of organizational trust in individualist versus
collectivist societies: a seven-nation study. Organization Science: A Journal of the Institute of Management Sciences, 14(1), 81-90.
Hughes, R., & Huby, M. (2004). The construction and interpretation of vignettes in
social research. Social Work & Social Sciences Review, 11(1), 36-51. Jarvenpaa, S. L., Knoll, K., & Leidner, D. E. (1998). Is anybody out there? antecedents
of trust in global virtual teams. Journal of Management Information Systems, 14(4), 29.
Jones, G. R., & George, J. M. (1998). The experience and evolution of trust:
Implications for cooperation and teamwork. Academy of Management Review, 23(3), 531-546.
Joresborg, K., & Sorbom, D. (1996). Lisrel 8: user's reference guide. Chicago. IL.:
Scientific Software International. Joresborg, K. G. (1997). Statistical significance of parameter estimates. In B. M. Byrne
(Ed.) (pp. 104, Structural equation modeling with LISREL, PRELIS, and SIMPLIS : basic concepts, applications, and programming). Mahwah, N.J.: L. Erlbaum Associates.
Kantowitz, B. H., Roediger, H. L., & Elmes, D. G. (1994). Experimental psychology :
understanding psychological research (5th ed.). Minneapolis/St. Paul: West Pub. Co .
Keen, P. G. W. (1990). Telecommunications and organizational choice. In J. Fulk & C.
W. Steinfield (Eds.), Organizations and communication technology (pp. 295-312). Newbury Park, Calif: Sage Publications.
Kelloway, E. K. (1998). Using LISREL for structural equation modeling : a researcher's
guide. Thousand Oaks: Sage Publications. Kenny, D. A., & Kashy, D. A. (1992). Analysis of the multitraite multimethod matrix by
Kerlinger, F. N. (1986). Foundations of behavioral research (3rd ed.). Fort Worth: Holt Rinehart and Winston.
Kiffin-Petersen, S. A., & Cordery, J. L. (2003). Trust, individualism and job
characteristics as predictors of employee preference for teamwork. International Journal of Human Resource Management, 14(1), 93-116.
Kim, P. H., Ferrin, D. L., Cooper, C. D., & Dirks, K. T. (2004). Removing the shadow
of suspicion: the effects of apology versus denial for repairing competence- versus integrity-based trust violations. Journal of Applied Psychology, 89(1), 104-118.
Kirkman, B. L., Rosen, B., Gibson, C. B., Tesluk, P. E., & McPherson, S. O. (2002).
Five challenges to virtual team success: Lessons from Sabre, Inc. Academy of Management Executive, 16(3), 67-79.
Kirkman, B. L., & Shapiro, D. L. (2001). The impact of team members' cultural values
on productivity, cooperation, and empowerment in self-managing work teams. Journal of Cross-Cultural Psychology, 32(5), 597-617.
Klein, K. J., Lim, B.-C., Saltz, J. L., & Mayer, D. M. (2004). How do they get there? an
examination of the antecedents of centrality in team networks. Academy of Management Journal, 47(6), 952-963.
Koehn, D. (2005). Integrity as a business asset. Journal of Business Ethics, 58(1-3), 125-
136. Korsgaard, M. A., Brodt, S. E., & Whitener, E. M. (2002). Trust in the face of conflict:
the role of managerial trustworthy behavior and organizational context. Journal of Applied Psychology, 87(2), 312-319.
Kramer, R. M. (1999). Trust and distrust in organizations: emerging perspectives,
enduring questions. Annual Review of Psychology, 50, 569-598. Kramer, R. M. (2001). Organizational paranoia: origins and dynamics. Research in
Organizational Behavior, 23, 1-42. Kramer, R. M., & Cook, K. S. (2004). Trust and distrust in organizations : dilemmas
and approaches. New York: Russell Sage Foundation. Lawrence, T. B., & Corwin, V. (2003). Being there: the acceptance and marginalization
of part-time professional employees. Journal of Organizational Behavior, 24(8), 923-943.
Letki, N. (2006). Investigating the roots of civic morality: trust, social capital, and
institutional performance. Political Behavior, 28(4), 305-325.
162
Levin, D. Z., Cross, R., & Abrams, L. C. (2002). The strength of weak ties you can trust: the mediating role of trust in effective knowledge transfer. Academy of Management Proceedings, D1.
Lewicki, R. J., & Bunker, B. B. (1996). Developing and maintaining trust in work
relationships. In R. M. Kramer & T. R. Tyler (Eds.), Trust in organizations: Frontiers of theory and research. (pp. 114-139): Sage Publications, Inc.
Lewicki, R. J., McAllister, D. J., & Bies, R. J. (1998). Trust and distrust: new
relationships and realities. Academy of Management Review, 23(3), 438-458. Lewis, J. D., & Weigert, A. (1985). Trust as a social reality. Social Forces, 63(4), 967-
985. Lurie, N. H., & Mason, C. H. (2007). Visual representation: implications for decision
making. Journal of Marketing, 71, 160–177. MacKinnon, D. P., Krull, J. L., & Lockwood, C. M. (2000). Equivalence of the
mediation, confounding and suppression effect. Prevention Science, 1(4), 173-181.
Makridakis, S. (1992). Management in the 21st Century. London: Sage in association
with The Open University. Mankins, M. C., & Steele, R. (2006). Stop making plans start making decisions.
Harvard Business Review, 84(1), 76-84. Marsh, H. W., & Grayson, D. (1995). Latent variable models of multitrait-multimethod
data. In R. H. Hoyle (Ed.), Structural equation modeling: Concepts, issues, and applications. (pp. 177-198): Sage Publications, Inc.
Marshall, E. M. (2000). Building trust at the speed of change: the power of the
relationship-based corporation. New York: American Management Association. Martin, B. (2006). Fight or Flight. Retrieved February 16, 2006, from
http://psychcentral.com/lib/2006/02/fight-or-flight/ Martin, D. F. (1999). The impact of trust on leader-member exchange relationships.,
Univ Microfilms International. Mayer, R. C. (1990). Understanding employee motivation through organizational
commitment., Univ Microfilms International. Mayer, R. C. (1998). Trust, an asset in any field. (Cover story). Baylor Business Review,
16(2), 8.
163
Mayer, R. C., & Davis, J. H. (1999). The effect of the performance appraisal system on trust for management: A field quasi-experiment. Journal of Applied Psychology, 84(1), 123-136.
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of
organizational trust. Academy of Management Review, 20(3), 709-734. Mayer, R. C., & Gavin, M. B. (2005). Trust in management and performance: who
minds the shop while the employees watch the boss? Academy of Management Journal, 48(5), 874-888.
Maznevski, M. L., & DiStefano, J. J. (1995). Measuring culture in international
management: the cultural perspectives questionnaire. unpublished working paper. The University of Virginia, Charlottesville, VA.
McAllister, D. J. (1995). Affect- and cognition-based trust as foundations for
interpersonal cooperation in organizations. Academy of Management Journal, 38(1), 24-59.
McClelland, V. (1987). Mixed signals breed mistrust. Personnel Journal, 66(3), 24-27. McKnight, D. H., Choudhury, V., & Kacmar, C. (2002). Developing and validating trust
measures for e-commerce: an integrative typology. Information Systems Research, 13(3), 334-359.
McKnight, D. H., Cummings, L. L., & Chervany, N. L. (1998). Initial trust formation in
new organizational relationships. Academy of Management Review, 23(3), 473-490.
McShane, S. L., & Travaglione, A. (2003). Organisational behaviour on the Pacific rim.
Microsoft Corporation. Microsoft. (2007). Microsoft Office Excel. Retrieved February 27, 2007 Miller, G. J. (1992). Managerial dilemmas : the political economy of hierarchy.
Cambridge England New York: Cambridge University Press. Mills, P. K., & Ungson, G. R. (2003). Reassessing the limits of structural empowerment:
organizational constitution and trust as controls. Academy of Management Review, 28(1), 143-153.
Mishra, A. K., & Spreitzer, G. M. (1998). Explaining how survivors respond to
downsizing: The roles of trust, empowerment, justice, and work redesign. Academy of Management Review, 23(3), 567-588.
164
Misztal, B. A. (1996). Trust in modern societies : the search for the bases of social order. Cambridge, Mass: Polity Press.
Misztal, B. A. (2002). Trusting the professional: A managerial discourse for uncertain
times. London and New York: Whitehead, Routledge. Mollering, G. (2005). The trust/control duality - An integrative perspective on positive
expectations of others. International Sociology, 20(3), 283-305. Murphy, K. R., Herr, B. M., Lockhart, M. C., & Maguire, E. (1986). Evaluating the
performance of paper people. Journal of Applied Psychology, 71(4), 654-661. North, D. C. (1990). Institutions, institutional change, and economic performance.
Cambridge New York: Cambridge University Press. Nunnally, J. C. (1978). Psychometric theory (2nd ed.). New York: McGraw-Hill. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York:
McGraw-Hill. O'Connor, T. G., & Hirsch, N. (1999). Intra-Individual Differences and Relationship-
Specificity of Mentalising in Early Adolescence. Social Development Social Development J1 - Social Development, 8(2), 256-274.
Oliver, A. L. (1997). On the nexus of organizations and professions: networking through
trust. Sociological Inquiry, 67(2), 227-245. Ouchi, W. G. (1981). Theory Z : how American business can meet the Japanese
challenge. Reading, Mass: Addison-Wesley. Owen, H. (1996). Building teams on a display of trust. People Management, 2(6), 34. Parkhe, A. (1998). Understanding trust in international alliances. Journal of World
Business, 33(3), 219-240. Pasewark, R. A., Fitzgerald, B. J., Sawyer, R., & Fossey, J. (1973). Validity of Rotter's
interpersonal trust scale: a study of paranoid schizophrenics. Psychological Reports, 32(3), 982.
Paul, D. L., & McDaniel, R. R. (2004). A field study of the effect of interpersonal trust
on virtual collaborative relationship performance. MIS Quarterly, 28(2), 183-227.
Paul, F., & Erdfelder, E. (1992). GPOWER: a priori, post-hoc, and compromise power
analyses for MS-DOS (computer program). Bonn, Germany: Bonn University.
165
Popa, C. L. (2005). Initial trust formation in temporary small task groups: testing a model of swift trust., Univ Microfilms International.
Porges, S. W. (1998). Love: an emergent property of the mammalian autonomic nervous
system. Psychoneuroendocrinology, 23(8), 837-861. Prahalad, C. K. (1997). The work of new age managers in the emerging competitive
landscape. In F. Heselbein, M. Goldsmith & R. Beckhard (Eds.), The Organization of the Future (pp. 159-168). San Francisco: Jossey-Bas Publishers.
Punch, K. (2000). Developing effective research proposals. London Thousand Oaks,
Calif.: SAGE. Raths, L. E. (1972). Meeting the needs of children : creating trust and security.
Columbus, Ohio: Merrill. Ridings, C. M., Gefen, D., & Arinze, B. (2002). Some antecedents and effects of trust in
virtual communities. Journal of Strategic Information Systems, 11(3-4), 271-295. Rigdon, E. E. (2006). 5-point variables, Structural Equation Modeling Discussion
Group. http://bama.ua.edu/archives/semnet.html: LISTSERV® 14.5. Robbins, S. P. (2006). Management (4th ed.). Frenchs Forest, NSW: Pearson Education
Australia. Robinson, J. P., Shaver, P. R., & Wrightsman, L. S. (1991). Measures of personality and
social psychological attitudes. San Diego, Calif.: Academic Press. Robinson, S. L., & Rousseau, D. M. (1994). Violating the psychological contract: not
the exception but the norm. Journal of Organizational Behavior, 15(3), 245-259. Rogelberg, S. G. (2002). Handbook of research methods in industrial and
organizational psychology. Malden, Mass. Oxford: Blackwell Publishers. Rotter, J. (1967). A new scale for the measurement of interpersonal trust. Journal of
Personality, 35(4), 651-665. Rotter, J. B. (1980). Interpersonal trust, trustworthiness, and gullibility. American
Psychologist, 35(1), 1-7. Rotter, J. B., Chance, J. E., & Phares, E. J. (1972). Applications of a social learning
theory of personality. New York London: Holt Rinehart and Winston. Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C. (1998). Not so different after
all: a cross-discipline view of trust. Academy of Management Review, 23(3), 393-404.
166
Rus, A. (2005). Trust and performance: institutional, interpersonal and network trust. In K. Bijlsma-Frankem & R. K. Woolthuis (Eds.), Trust under pressure : empirical investigations of trust and trust building in uncertain circumstances (pp. 80-104). Cheltenham, UK Northampton, MA: Edward Elgar.
Sackett, P. R., & Wanek, J. E. (1996). New developments in the use of measures of
honesty, integrity, conscientiousness, dependability, trustworthiness, and reliability for personnel selection. Personnel Psychology, 49(4), 787-829.
Sandvig, J. C., Tyran, C. K., & Ross, S. C. (2005). Determinants of graduating MIS
students starting salary in boom and bust job markets. Communications of AIS, 2005(16), 604-624.
Santoro, M. D., & Saparito, P. A. (2003). The firm's trust in its university partner as a
key mediator in advancing knowledge and new technologies. IEEE Transactions on Engineering Management, 50(3), 362-373.
Satorra, A., & Bentler, P. M. (1988). Scaling corrections for chi-square statistics in
covariance structure analysis. Paper presented at the American Statistical Association, Business and Economics Section, Alexandria, VA.
Schoorman, P. D., Mayer, R. C., & Davis, J. H. (1996a). Empowerment in veterinary
clinics: the role of trust in delegation. Paper presented at the Society for Industrial and Organizational Psychology, 11th Annual Meeting, San Diego, CA.
Schoorman, P. D., Mayer, R. C., & Davis, J. H. (1996b). Organizational trust:
philosophical perspectives and conceptual definitions. Academy of Management Review, 21(2), 337-340.
Sekaran, U. (2003). Research methods for business : a skill-building approach (4th ed.).
New York: John Wiley & Sons. Serva, M. A., Benmati, J. S., & Fuller, M. A. (2005). Trustworthiness in B2C e-
commerce: an examination of alternative models. 36(3), 89-108. Serva, M. A., Fuller, M. A., & Mayer, R. C. (2005). The reciprocal nature of trust: a
longitudinal study of interacting teams. Journal of Organizational Behavior, 26(6), 625-648.
Shapiro, D. L., Buttner, E. H., & Barry, B. (1994). Explanations: What Factors Enhance
Their perceived Adequacy? Organizational Behavior & Human Decision Processes, 58(3), 346-368.
Sheppard, B. H., & Sherman, D. M. (1998). The grammars of trust: A model and general
implications. Academy of Management Review, 23(3), 422-437.
167
Simon, L. S. (1995). Trust in Leadership: Its dimensions and mediating role., Univ Microfilms International.
Sitkin, S. B., & Roth, N. L. (1993). Explaining the limited effectiveness of legalistic
"remedies" for trust/distrust. Organization Science, 4(3), 367. Sniderman, P. M., & Grob, D. B. (1996). Innovations in experimental design in attitude
surveys. Annual Review of Sociology, 22(1), 377. Spreitzer, G. M., Noble, D. S., Mishra, A. K., & Cooke, W. N. (1999). Predicting
process improvement team performance in an automotive firm: explicating the roles of trust and empowerment. In R. Wageman (Ed.), Research on managing groups and teams: Groups in context, Vol. 2. (pp. 71-92): Elsevier Science/JAI Press.
SPSS. (2004). SPSS 13.0 for Windows (Version Release 1 Sep 2004). Chicago: SPSS
Inc. Steiger, J. H., & Lind, J. C. (1980). Statistically based tests for the number of common
factors. Paper presented at the Psychometric Society Annual meeting, Iowa City, IA.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5th ed.). Boston:
Pearson/Allyn & Bacon. Tan, H. H., & Tan, C. S. F. (2000). Toward the differentiation of trust in supervisor and
trust in organization. Genetic, Social & General Psychology Monographs, 126(2), 241-260.
Tehan, G., & Tolan, G. A. (2007). Word length effects in long-term memory. Journal of
Memory & Language, 56(1), 35-48. Tidd, S. T., McIntyre, H. H., & Friedman, R. A. (2004). The importance of role
ambiguity and trust in conflict perception: unpacking the task conflict to relationship conflict linkage. International Journal of Conflict Management, 15(4), 364-380.
Tilley, A. J. (2004). An Introduction to research methodology and report writing in
psychology (2nd edition. ed.). Brisbane, Qld.: Pineapple Press. Tuckman, B. W., & Jensen, M. A. C. (1977). Stages of small-group development
revisited. Group and Organization Studies, 2, 419-442. Van Iddekinge, C. H., Taylor, M. A., & Eidson Jr., C. E. (2005). Broad versus narrow
facets of integrity: predictive validity and subgroup differences. Human Performance, 18(2), 151-177.
168
Wanek, J. E. (1996). The construct of integrity: Item level factor analysis of the dimensions underlying honesty testing and big-five measures of personality., Univ Microfilms International.
Watson, G. W., & Papamarcos, S. D. (2002). Social capital and organizational
commitment. Journal of Business & Psychology, 16(4), 537-552. Wekselberg, V. (1996). Reduced ''social'' in a new model of organizational trust.
Academy of Management Review, 21(2), 333-335. Whitener, E. M. (2001). Do "high commitment" human resource practices affect
employee commitment? A cross-level analysis using hierarchical linear modeling. Journal of Management, 27(5), 515-535.
Whitener, E. M., Brodt, S. E., Korsgaard, M. A., & Werner, J. M. (1998). Managers as
initiators of trust: An exchange relationship framework for understanding managerial trustworthy behavior. Academy of Management Review, 23(3), 513-530.
Widaman, K. F. (1985). Hierarchically nested covariance structure models for multitrait-
multimethod data. Applied Psychological Measurement, 9(1), 1-26. Williams, L. J., Ford, L. R., & Nhung, N. (2002). Basic and advanced measurement
models for confirmatory factor analysis. In S. G. Rogelberg (Ed.), Blackwell handbooks of research methods in psychology ; 1. (pp. xi, 520). Malden, Mass. Oxford: Blackwell Publishers.
Williams, M. (2001). In whom we trust: group membership as an affective context for
trust development. Academy of Management Review, 26(3), 377-396. Williamson, O. E. (1975). Markets and hierarchies, analysis and antitrust implications :
a study in the economics of internal organization. New York: Free Press. Wong, Y.-T., Ngo, H.-Y., & Wong, C.-S. (2002). Affective organizational commitment
of workers in Chinese joint ventures. Journal of Managerial Psychology, 17(7), 580-598.
Wong, Y.-T., Ngo, H.-Y., & Wong, C.-S. (2003). Antecedents and outcomes of
employees' trust in Chinese joint ventures. Asia Pacific Journal of Management, 20(4), 481-499.
Wong, Y.-T., Wong, C.-S., & Ngo, H.-Y. (2002). Loyalty to supervisor and trust in
supervisor of workers in Chinese joint ventures: a test of two competing models. International Journal of Human Resource Management, 13(6), 883-900.
Wright, T. L., & Tedeschi, R. G. (1975). Factor analysis of the interpersonal trust scale.
Journal of Consulting and Clinical Psychology, 43(4), 470-477.
169
Yuan, Y., Fulk, J., Shumate, M., Monge, P. R., Alison, B. J., & Matsaganis, M. (2005). Individual participation in organisational information commons. The impact of team level social influence and technology-specification competence. Human Communication Research, 31(2), 212-240.
Zaheer, A., McEvily, B., & Perrone, V. (1998). Does Trust Matter? Exploring the
Effects of Interorganizational and Interpersonal Trust on Performance. Organization Science: A Journal of the Institute of Management Sciences, 9(2), 123.
Zand, D. E. (1972). Trust and Managerial Problem Solving. Administrative Science
Quarterly, 17(2), 229-239. Zellman, G. L. (1990). Report decision-making patterns among mandated child abuse
reporters. Child Abuse & Neglect, 14(3), 325-336. Zucker, L. G. (1986). Production of trust: institutional sources of economic structure.