The effect of employee affective and cognitive trust in leadership on organisational citizenship behaviour and organisational commitment: Meta-analytic findings and implications for trust research

Abstract

Trust is a contemporary topic, as society is losing trust in prominent institutions. Understanding trust in the workplace is critical, yet, a consensus around trust as unidimensional or multidimensional has not emerged in the literature. Some measure trust globally, while others measure its dimensions. This article builds on organisational trust research by exploring the relationships between a model of trust multidimensionality and organisational citizenship behaviour, organisational commitment and its facets. Findings from this meta-analysis of 11 studies indicated that trust dimensions have different strengths of relationship with organisational citizenship behaviour and commitment. Although the number of studies included is small, similar meta-analyses are considered valuable and worth exploring for the purpose of theory development. In the context of inconsistent trust definition and measurement, these findings support confirmation that trust is a multidimensional construct.

JEL Classification: L2

Keywords

Leadership organisational citizenship behaviour organisational commitment trust dimensions trust measurement

1. Introduction

Recent worldwide events have seen a progressive decline of trust in societal institutions (Biddle and Reddy, 2019; Stevenson and Wolfers, 2011). Commentators argue that this is a consequence of the Great Recession and fears about immigration and economic dislocation caused by globalisation and automation (Edelman, 2019). People doubt major public institutional leadership, as evidenced by the recent findings from the Australian Banking Royal Commission (Henrique-Gomes, 2019; Stevenson and Wolfers, 2011). Regarding national elections, the influence of social media platforms has shifted public trust from a traditional media preference to one in favour of learning from peers or experts whom they have self-selected to ‘follow’ (Cacciatore et al., 2016; Edelman, 2019). The 2016 US presidential election is a recent, significant example of this emerging phenomenon (Hollibaugh, 2016; Torabi Asr, 2019). The public is reorienting their trust to more local sources who they perceive as more reliable, with the employer emerging as the most trusted entity (Biddle and Reddy, 2019; Edelman, 2019).

The academic literature supports the position that trust is fundamental in the workplace. Trust is a determinant for many employee outcomes essential for organisational performance, such as job performance and satisfaction (Yang and Mossholder, 2010; Zhu et al., 2013), organisational commitment (OC; Hon and Lu, 2010; Miao et al., 2014b), organisational citizenship behaviour (OCB; Lu, 2014; Mushonga et al., 2014) and psychological safety (Frazier et al., 2017). Organisations spend time and resources on leadership development to improve outcomes, behaviours and culture. Hence, a solid understanding of trust may translate to an increase in leader ability to ensure these important employee and organisational performance outcomes occur.

1.1. Rationale and aim

Historically, organisational trust has been difficult to define. The literature defines trust as a trait, a state and a cognitive process (Rousseau et al., 1998; Schoorman et al., 2007). Research has used derivations of Rousseau et al.’s (1998) definition that describes trust as a state of vulnerability to another person (Agote et al., 2016; Chen et al., 2012; Colquitt, 2001; Kim et al., 2018). Furthermore, despite the well-acknowledged importance of organisational trust, the literature is not consistent in how employee trust in leaders is measured. Some examples in the literature measure trust as an overall global construct (Conchie, 2013), while others measure specific dimensions of trust separately (Hon and Lu, 2010). Moreover, these dimensions have inconsistent descriptions in the literature. The most common nomenclature for dual dimensions of trust are affective and cognitive trust (Carter and Mossholder, 2015; Lu, 2014; Miao et al., 2014a, 2014b; Newman et al., 2014; Zhu et al., 2013). Affective trust is dependent on the quality of relationships, while cognitive trust comprises the leader’s displayed behavioural integrity (Dirks and Ferrin, 2002; McAllister, 1995; Schaubroeck et al., 2011).

This article reports a meta-analysis of the current literature examining whether distinct trust dimensions between employees and leaders differently impact organisational outcomes. This article follows a future research recommendation from Dirks and Ferrin (2002) who examined differences in the way employee trust in leaders affects organisational outcomes. Their paper was unable to examine trust dimensions’ relationships with organisational outcomes separately, as there was insufficient literature at the time. Since then relevant literature has emerged to justify and conduct this level of analysis using a meta-analytic approach. This article intended to advance trust theory in order to demystify its current obscure conceptualisation. As with Dirks and Ferrin (2002), the focus was employee–leader relationships, rather than peer-to-peer relationships.

Dirks and Ferrin’s (2002) meta-analysis showed trust has been measured both as a global construct and as a multidimensional construct, with varying terminology to describe those dimensions, and this has been supported by more recent research (Dietz and Den Hartog, 2006; McEvily and Tortoriello, 2011; Schoorman et al., 2007). The most common names for trust dimensions are affective and cognitive trust (Carter and Mossholder, 2015; Ertürk, 2010; Lu, 2014), and the most common measure of those dimensions is McAllister’s (1995) affective and cognitive trust scale (McEvily and Tortoriello, 2011). Dirks and Ferrin (2002) identified a need for future research to clarify the distinctions between trust dimensions and to explore relationships between trust dimension and outcomes, thus confirming or challenging trust’s multidimensionality.

This aim of this article is to bring together the contemporary body of trust literature and determine whether trust is a multidimensional construct. A meta-analytic statistical approach is used to aggregate published trust relationship findings and evaluate whether multidimensional trust measurement approaches produce similar or different relationships with key organisational outcomes. If the meta-analysis findings show significant differences between the two dimensions of trust and organisational outcome variables, then this would support the theory that trust is multidimensional.

1.2. A review of organisational trust

The antecedents and outcomes of organisational trust have been of interest to practitioners and academics for almost 70 years (Dirks and Ferrin, 2002). Dirks and Ferrin (2002) noted in their meta-analysis that Rousseau et al’.s (1998) definition of trust was the most commonly cited version in the literature at that time. More contemporary organisational trust reviews emerged that expanded trust definition beyond Rousseau et al’.s (1998) to include trust as interpersonal risk, expectations of another’s benevolence and an attribution of positive intentions of the words or actions of others (Dietz and Den Hartog, 2006; McEvily and Tortoriello, 2011). Furthermore, McEvily and Tortoriello (2011) found 129 unique measures of organisational trust in their review of organisational trust measurement and noted little replication between these measures. These measures were based on any one of the definitions discussed here and either measured trust as a unidimensional or multidimensional construct.

McAllister (1995) was the first to propose a multidimensional trust model that consists of affective and cognitive bases. Affective trust results from relationships and the emotional bonds that exist between individuals, while cognitive trust is based on reasoning that occurs to determine if someone is trustworthy and develops through the observation of behaviour (McAllister, 1995). McAllister’s (1995) affective and cognitive trust scale assesses the dual dimensional approach to trust with 11 items. Example items include ‘We can both freely share our ideas, feelings and hopes’ and ‘This person approaches his or her job with professionalism and dedication’. The first item describes an aspect of the relationship between the two individuals (affective trust) and the second is an observation of positive workplace behaviour (cognitive trust).

Providing the impetus for the current paper, Dirks and Ferrin’s (2002) meta-analysis presented evidence to demonstrate trust’s multidimensionality and trust dimensions’ different relationships with organisational outcomes. Their meta-analysis (k = 106; N = 27,103) examined research on trust between employees and leaders (Dirks and Ferrin, 2002); 94% of the studies in the meta-analysis measured either the cognitive aspect of trust or global trust. Few studies measured affective trust and the authors noted that cases investigating affective trust ‘were too dispersed across relationships’, resulting in an insufficient number of studies to conduct analyses. Thus, Dirks and Ferrin (2002) could not examine the operational differences between affective and cognitive trust; they identified this limitation as an important gap to fill when the necessary research emerged.

1.3. The impact of multidimensional trust on organisational outcomes

The Dirks and Ferrin (2002) meta-analysis demonstrated that cognitive trust, when measured alone, had differing relationships with organisational outcomes compared with global trust. For example, global trust positively influenced OCB and job performance, but cognitive trust did not. Dirks and Ferrin (2002) argued that there must be some other aspect within global trust that was unaccounted for when measuring cognitive trust alone. This missing element could be a dimension of trust based on emotional connection developed through relationships. Other research affirmed the importance of the emotional connection in trust, albeit without empirically testing trust multidimensionality (Dietz and Den Hartog, 2006; McEvily and Tortoriello, 2011; Schoorman et al., 2007). There are reasons throughout the academic literature, albeit in different contexts, to support a hypothesis that affective trust will have a different strength of relationship with organisational outcomes than cognitive trust, regardless of Dirks and Ferrin’s (2002) inability to demonstrate that in their paper.

The literature has shown the significance of the social, reciprocal exchange processes between leaders and their followers (Podsakoff et al., 2000; Settoon et al., 1996). This suggests that affective trust, the more relational dimension, may play a large role in employees and organisational outcomes. Settoon et al. (1996) demonstrated the significance of the employee–leader relationship on the organisational outcomes examined in this article. They created a theoretical ground point to form a hypothesis that affective trust, the more relational trust dimension, will have a different strength of relationship with organisational outcomes than cognitive trust, thereby extending Dirks and Ferrin’s (2002) findings. These outcomes are OCB and OC; two key organisational outcomes found to be strong determinants of organisational effectiveness (Organ et al., 2006; Podsakoff et al., 2009).

OCBs are the discretionary behaviours employees demonstrate at work that are beyond the expectations of the role or the job description (Organ et al., 2006). Research has shown trust to have a positive relationship with OCB and that trust mediates the relationship between leadership behaviours and OCB (Miao et al., 2014a; Newman et al., 2014; Wong et al., 2012). Miao et al. (2014a) and Newman et al. (2014) both explored affective and cognitive trust and its mediational relationship between leadership styles and OCB. Both studies found positive relationships with affective and cognitive trust, which is consistent with Settoon et al’.s (1996) findings on the social exchange processes between employees and leaders.

OC is the employee’s psychological connection to the organisation (Allen and Meyer, 1990; DeConinck and Bachmann, 2007; Tett and Meyer, 1993) and trust mediates the relationship between organisational variables and OC (Indartono and Chen, 2011; Liu and Wang, 2013). This mediational role of trust means that it plays a key role in several mechanisms leading to OC, suggesting that employees who trust their leaders are more committed to their workplaces. OC is multidimensional and consists of three distinct types: affective, continuance and normative commitment. According to Meyer and Allen (1991), affective commitment is the employee’s emotional attachment to, identification with and involvement in the organisation; continuance commitment is the costs employees associate with leaving the organisation and normative commitment is the employee’s feelings of obligation to remain with the organisation. As OC is both a unidimensional and multidimensional psychological state, this article explores all OC relationships with trust dimensions.

Hon and Lu (2010) and Chiang and Lin (2016) found that both affective and cognitive trust had a positive relationship with OC. However, Yang (2005) found differing relationships in relation to the two dimensions of trust and the facets of OC. Both dimensions of trust had a positive relationship with affective commitment, but not continuance commitment. Yang (2005) did not explore the relationship with normative commitment. Miao et al. (2014b), however, found that all three facets of OC had a positive relationship with cognitive trust, but only affective and normative commitment had a positive relationship with affective trust.

In sum, the literature indicates potential for an overview of multidimensional trust in organisations, therefore this article uses a meta-analytic approach to investigate whether affective and cognitive trust between employees and leaders differs in strength of impact on specific organisational outcomes. The organisational outcomes investigated are of OCB, OC and/or its facet. It is hypothesised that there will be significant differences in the strength of relationship between the two dimensions of trust and OCB, OC and/or its facets.

2. Method

Authors conducted a systematic review of academic literature to identify relevant studies that investigated the relationship between the two dimensions of trust (affective and cognitive), OCB, OC and its facets. The review explicitly focused on employee–leader relationships in the workplace. It was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher et al., 2010), and inclusion and exclusion criteria were decided prior to commencement of the literature search.

2.1. Literature search

Authors conducted quarterly computer searches on the following databases from October 2015 to May 2019: Business Source Complete, CINAHL Complete, ERIC, Health Business Elite, Medline Complete, PsycINFO and SocINDEX with Full Text. Authors searched each database individually, with duplicates removed once all searches were completed and combined. Authors examined reference lists of extracted articles by hand for additional relevant articles. The same inclusion and exclusion criteria applied to the original database search were applied to reference list examinations and subsequent searches.

The search terms used were affective trust AND cognitive trust, OR emotional trust AND rational trust OR relationship-based trust AND character-based trust. The rationale to search using the term combinations listed followed Dirks and Ferrin’s (2002) meta-analysis on employee trust in leadership. Dirks and Ferrin (2002) noted that previous research had used these combinations of terms to refer to a dual dimension of trust. For example, Schoorman et al. (2007) used the terms emotional and rational trust to describe the dual dimension. This further supported the search term choice. In order to cover a wider scope of research, authors purposefully excluded the terms ‘OCB’ and ‘OC’ from the database search. After removing duplicates and excluding articles that did not measure both trust dimensions, authors examined articles for their measurement of OC and/or OCB.

2.2. Inclusion and exclusion criteria

Authors defined inclusion criteria prior to the literature search. Articles’ titles and abstracts were screened and retained if (a) they were written in the English language; (b) they were peer-reviewed; (c) the full text was available; (d) they measured affective trust, cognitive trust,¹ OC and/or OCB; (e) they were published in or after 2002² and (f) their participants were considered employed adults (e.g. aged 18 years or above). Student populations used as samples were only included if the study mentioned that they were also employed while studying, to ensure participants were able to rate trust in a leader from real-life work experience. Authors placed no limits on the sample size or recruitment method used as long as they fit the participant criteria of being employed adults. Moreover, authors did not exclude sample types, such as gender type, country of origin or a specific industry.

Authors retained studies if they measured affective trust and cognitive trust as constructs³ within the context of employee–leader relationships in the workplace. Studies were required to specifically address the target audience being rated (e.g. employees rating their direct managers or senior leaders), and how affective and cognitive trust were measured in order to be included. Type of study was restricted to those that empirically measured both trust dimensions.

Figure 1 outlines how the PRISMA model was used to identify relevant studies. Two hundred and fifteen studies were initially identified and once duplicates were removed this reduced to 89. After further analysis, based on inclusion criteria, 11 eligible articles remained. Although the number of studies exploring the relationships in this article is small, meta-analyses of this nature are considered efficacious and worth exploring for the purpose of new theory development (Turner et al., 2013; Valentine et al., 2010). In the context of inconsistent trust definition and measurement, this meta-analysis will contribute to new theory development.

Figure 1.

Search strategy based on the PRISMA model.

2.3. Coding

Authors coded the remaining studies for quality. To establish coding reliability, two authors independently coded the remaining studies as recommended by Rosenthal (1995). Authors created a quality assurance coding audit tool based on Rosenthal’s (1995) recommendations and assessed each study against that tool to ensure quality in study inclusion. The tool outlined the following quality assurance criteria: (1) trust was measured using McAllister’s (1995) affective and cognitive trust scale or version of that scale, (2) the desired organisational outcome variables were also measured (OCB, OC and its facets), (3) effect sizes were reported as r⁴ and (4) each scale received Cronbach’s alpha scores above .75. The remaining studies reasonably met all criteria (see Table 1). Coders reached complete agreement and achieved an internal consistency rating of κ = 1. There was no quality weighting assigned to any study, so no quality weightings were assigned in the meta-analysis calculations (Rosenthal, 1995).

Table 1.

Description of the studies included.

Author and year	N	Trust measure	Direct outcome measure	Reliability scores (Cronbach’s alpha)
1. Berkovich (2018)	654 public primary school teachers in Israel on the topic of trust in principal	• McAllister (1995): affective and cognitive measure of trust	• Teachers’ extra-role behaviour at school was measured using 23 items of the Somech and Drach-Zahavy (2000) OCB scale	• Trust in direct manager ○ Affective .88 ○ Cognitive .92 • OCB .89
2. Chiang and Lin (2016)	402 front-line employees in four- and five-star hotels in Taiwan	• McAllister (1995): affective and cognitive measure of trust	• Nine items were used from an organisational commitment scale (Allen and Meyer, 1990) • Adjusted questionnaire based on hospitality scholar and hotel employee opinions	• Trust in direct manager ○ Affective .94 ○ Cognitive .94 • Organisational commitment .77
3. Hon and Lu (2010)	286 employees from hotel companies in the Guangzhou region in China who reported directly to an expatriate supervisor	• Adaptation of McAllister (1995): affective and cognitive trust measure • Affective trust – five items • Cognitive trust – five items	• Used a shortened scale of six items adapted from Mowday et al. (1979) Organisational Commitment Questionnaire • This six-item scale has been validated by previous scholars in a Chinese context	• Trust in direct manager ○ Affective .83 ○ Cognitive .81 • Organisational commitment .79
4. Lu (2014)	104 supervisor-subordinate dyads employed in public-sector organisations in a city in eastern China	• Used the scale that (Yang et al., 2009) developed, which was based on McAllister’s (1995) affective and cognitive trust scale	• Williams and Anderson’s (1991) measure of OCB	• Trust in direct manager ○ Affective .93 ○ Cognitive .82 • OCB .85 and .79
5. Miao et al. (2014a)	247 supervisor-subordinate dyads in a manufacturing organisation located in mainland China	• McAllister (1995): affective and cognitive measure of trust	• OCB was assessed using a scale developed by Farh et al. (1997)	• Trust in direct manager ○ Affective .79 ○ Cognitive .81 • OCB .87
6. Miao et al. (2014b)	239 civil servants in a Chinese public-sector organisation	• McAllister (1995): affective and cognitive measure of trust	• Used the 18-item organisational commitment scale developed by Meyer et al. (1993) with three six-item subscales measuring affective, normative and continuance commitment	• Trust in direct manager ○ Affective .94 ○ Cognitive .95 • Organisational commitment ○ Affective .93 ○ Normative .87 ○ Continuance .79
7. Mushonga et al. (2014)	317 employees from various organisations located in urban areas in the southeastern region of the United States	• Used a scale that was based on McAllister’s (1995) affective and cognitive trust scale	• OCB was measured using a seven-item organisational citizenship behaviour-supervisor directed (OCBs) scale from Williams and Anderson’s (1991) original seven-item OCB-individual directed scale	• Trust in direct manager ○ Affective .87 ○ Cognitive .88 • OCB .89
8. Newman et al. (2014)	184 supervisor-subordinate dyads from organisations in the Zhejiang Province in southeastern China	• McAllister’s (1995) affective and cognitive trust scale	• OCB was measured by a 16-item scale developed and validated by Lee and Allen (2002)	• Trust in direct manager ○ Affective .88 ○ Cognitive .91 • OCB .91
9. Yang (2005)	353 employees from two regional medical centres in the Southeastern United States	• Yang created his own adaptation of McAllister’s (1995) affective and cognitive trust scale, with more emphasis on three foci: direct manager, senior manager and co-worker	• Used Clugston’s (2000) shortened version of Meyer and Allen’s (1991) questionnaire (16 items to 10 items) with each facet estimated by five items	• Trust in direct manager and senior leader ○ Affective .92 ○ Cognitive .91 • Organisational commitment ○ Affective .88 ○ Continuance .84
10. Yang and Mossholder (2010)	210 supervisor–subordinate dyads from a state-operated medical healthcare system in the southeastern United States	• The authors developed their own 20-item cognitive and affective trust measure based on McAllister’s (1995) affective and cognitive trust scale that pointed to two referent groups: direct supervisors and senior management	• Used a task-focused citizenship behaviour instrument developed by Settoon and Mossholder (2002) • Affective commitment was measured by a five-item scale (Meyer and Allen, 1984)	• Trust in direct manager ○ Affective .94 ○ Cognitive .95 • Trust in senior leader ○ Affective .95 ○ Cognitive .94 • OCB .96 • Affective commitment .89
11. Zhu et al. (2013)	318 supervisor–subordinate dyads from a manufacturing organisation located in mainland China	• McAllister (1995): affective and cognitive trust	• OCB was measured using (Farh et al., 1997) scale. • Affective commitment was measured using Meyer et al. (1993) six-item scale	• Trust in direct manager ○ Affective .83 ○ Cognitive .79 • OCB .88 • Affective commitment .75

OCB: organisational citizenship behaviour.

2.3.1. Variables

Authors classified OC according to whether it was studied as a global variable or as a facet variable (i.e. affective, normative and continuance commitment). All included studies measured OCB as a global variable.

2.3.2. Statistics reviewed

Authors assessed studies if they reported effect size data in terms of a Pearson correlation coefficient, r, or equivalent. All screened studies met this criterion. To preserve independence, authors only used one effect size from each study pertaining to each relationship between trust dimensions and organisational outcomes (Rosenthal, 1995). Reliability scores were acceptable to good (e.g. .75 and above) for all measures of OCB, OC and its facet variables and can be found in Table 1. Effect sizes for the studies included in the meta-analysis are in Table 2.

Table 2.

Correlation coefficients of affective and cognitive trust impacts on workplace outcomes.

Direct outcome	Affective trust	Cognitive trust
Organisational citizenship behaviour	• Direct manager: r = .35, p < .01 (Berkovich, 2018) • Direct manager: r = .525, p < .01 (Lu, 2014) • Direct manager: r = .21, p < .01 (Miao et al., 2014a) • Direct manager: r = .35, p < .01 (Mushonga et al., 2014) • Direct manager: r = .34, p < .01 (Newman et al., 2014) • Direct manager: r = .33, p < .01 (Yang and Mossholder, 2010) • Direct manager: r = .21, p < .01 (Zhu et al., 2013)	• Direct manager: r = .17, p < .01 (Berkovich, 2018) • Direct manager: r = .43, p < .01 (Lu, 2014) • Direct manager: r = .15, p < .01 (Miao et al., 2014a) • Direct manager: r = .33, p < .01 (Mushonga et al., 2014) • Direct manager: OCB: r = .24, p < .01 (Newman et al., 2014) • Direct manager: r = .28, p < .01 (Yang and Mossholder, 2010) • Direct manager: r = .15, p < .01 (Zhu et al., 2013)
Organisational commitment	• Direct manager: r = .72, p < .01 (Chiang and Lin, 2016) • Direct manager: r = .37^a, p < .01 (Hon and Lu, 2010)	• Direct manager: r = .61, p < .01 (Chiang and Lin, 2016) • Direct manager: r = .31^a, p < .01 (Hon and Lu, 2010)
Affective commitment	• Direct manager: r = .42, p < .01 (Miao et al., 2014b) • Senior leader: r = .43^a, p < .01 (Yang, 2005) • Direct manager: r = .49, p < .01 (Yang, 2005) • Senior leader: r = .40^a, p < .01 (Yang and Mossholder, 2010) • Direct manager: r = .46^a, p < .01 (Yang and Mossholder, 2010) • Direct manager: r = .33, p < .01 (Zhu et al., 2013)	• Direct manager: r = .29, p < .01 (Miao et al., 2014b) • Senior leader: r = .39^a, p < .01 (Yang, 2005) • Direct manager: r = .40, p < .01 (Yang, 2005) • Senior leader: r = .31^a, p < .01 (Yang and Mossholder, 2010) • Direct manager: r = .37^a, p < .01 (Yang and Mossholder, 2010) • Direct manager: r = .26, p < .01 (Zhu et al., 2013)
Normative commitment	• Direct manager: r = .40, p < .01 (Miao et al., 2014b)	• Direct manager: r = .35, p < .01 (Miao et al., 2014b)
Continuance commitment	• Direct manager: r = .10, p = ns (Miao et al., 2014b) • Senior leader: r = −.02, p = ns (Yang, 2005) • Direct manager: r = −.04, p = ns (Yang, 2005)	• Direct manager: r = .24, p < .01 (Miao et al., 2014b) • Senior leader: r = .05, p = ns (Yang, 2005) • Direct manager: r = −.02, p = ns (Yang, 2005)

Corrected r – removal of common method variance.

2.4. Analysis procedures

For those relationships between trust dimensions and organisational outcomes that were explored in only one study (e.g. trust dimensions and normative commitment), a significance test of the difference between two dependent sample correlations with one variable in common was conducted using Lee and Preacher’s (2013) calculation. First, authors converted each correlation coefficient into a z-score using Fisher’s r-to-z transformation. Then, authors calculated the asymptotic covariance of these estimates using Steiger’s (1980) equations.

For relationships examined in multiple studies, for example, trust dimensions and OCB, OC, continuance commitment and affective commitment, authors used Hunter and Schmidt’s (2004) random-effects approach. Authors selected this method as most appropriate for the number of studies that explored the relationships of interest, as the random-effects model assumes that the studies sampled from populations differ from each other in ways that could affect the treatment effect. First, the average mean (r) and average weighted mean (ρ), or the population effect size, were calculated (Field and Gillet, 2010). Next, the variance of sample effect sizes, the sampling error variance and the variance in population (τ²) were calculated (Field and Gillet, 2010). The method estimated variance in population effect sizes by correcting the variance in sample effect sizes by the sampling error. To quantify the true heterogeneity in a random-effects model, the between-studies variance (τ²) reflects how much the true population effect sizes estimated in the single studies differ (Deeks et al., 2008). Finally, authors computed a 95% credibility interval (CV) using each population effect size to provide an estimate of variability of the individual effect sizes across studies (see Table 3; Field and Gillet, 2010). Hunter and Schmidt (2004) recommend correcting for CVs, which are the interval in which an artefact has a given probability.

Table 3.

Meta-analytic estimates for trust dimensions and organisational outcome relationships.

Relationship	k	N	r	SDr	ρ	Τ	τ²	95% CV
Relationship	k	N	r	SDr	ρ	Τ	τ²	LL	UL
Organisational citizenship behaviour–affective trust	8	2034	.33	.11	.32	.06	.003	.21	.43
Organisational citizenship behaviour–cognitive trust	8	2034	.25	.11	.22	.006	.004	.10	.34
Organisational commitment–affective trust	2	688	.55	.25	.57	.17	.03	.24	.91
Organisational commitment–cognitive trust	2	688	.46	.21	.49	.14	.02	.20	.77
Affective commitment–affective trust	4	1120	.41	.06	.41	.01	.0002	.38	.44
Affective commitment–cognitive trust	4	1120	.32	.06	.32	.01	.0002	.30	.35
Continuance commitment–affective trust	2	592	−.10	.28	−.14	.19	.04	−.51	.23
Continuance commitment–cognitive trust	2	592	.10	.19	.08	.12	.01	−.16	.31

k = number of correlations; N = combined sample size; r = average correlation of studies sampled; SDr = average sample variance; ρ = population effect size based on estimated true weighted correlation; τ = standard deviation of the distribution of population effect sizes; τ² = population effect size variance; CV = credibility interval at 95%.

Next, authors conducted significance testing for each of these relationships using Preacher’s (2002) calculation for the test of the difference between two correlation coefficients. First, the method converted each correlation coefficient into a z-score using Fisher’s r-to-z transformation. Then, the method compared these z-scores, making use of the sample size employed to obtain each coefficient (Cohen and Cohen, 1983). Table 4 shows the effect sizes, which were calculated using Cohen’s q (1988).

Table 4.

Significance and effect size of the relationship between affective and cognitive trust on outcome variables.

Outcome variable	Sig-test between affective and cognitive trust	Effect size (Cohen’s q)
Organisational citizenship behaviour	• Z = 3.83, p < .001	Small (q = .108)
Organisational commitment	• Z = 2.06, p < .05	Small (q = .111)
Affective commitment	• Z = 2.46, p < .01	Small (q = .115)
Normative commitment*	• Z = –0.61, p = .54	None
Continuance commitment	• Z = –3.79, p < .001	Small (q = .221)

Note. Trust dimensions on this outcome variable* were only measured in one study; therefore, a significance test of the difference between two dependent sample correlations with one variable in common was used. For the meta-analytic findings, an approach for independent samples was used.

3. Results

Of the 11 included studies, 6 included samples from China, 3 from the United States, 1 from Israel and 1 from Taiwan. Sample sizes ranged from 104 (Lu, 2014) to 654 (Berkovich, 2018), with a median of 302 employees. Participants were from a variety of organisations, including public, private, agriculture, manufacturing, healthcare and hospitality industries. All studies were published after 2005, with the majority (91%) in the last decade. All studies included in the analysis used McAllister’s (1995) affective and cognitive trust scale or some adaptation of it (e.g. removal of items that might not be relevant to the sample population). All studies used reliable measures of affective and cognitive trust, OCB, OC and its facets reporting Cronbach’s alpha scores between .75 (Zhu et al., 2013) and .96 (Yang and Mossholder, 2010).

3.1. Associations between trust dimensions on organisational outcomes

Studies showed mixed relationships between trust dimensions (affective and cognitive trust) and organisational outcomes (OCB, OC and its facets). The studies examining trust dimensions and OCB demonstrated moderate to strong positive relationships (Berkovich, 2018; Lu, 2014; Miao et al., 2014a; Mushonga et al., 2014; Newman et al., 2014; Yang and Mossholder, 2010; Zhu et al., 2013). Chiang and Lin (2016) and Hon and Lu (2010) reported that both affective and cognitive trust showed moderate to strong and positive associations with OC. Trust dimensions were also shown to have moderate to strong and positive relationships with affective commitment and normative commitment (Miao et al., 2014b; Yang, 2005; Yang and Mossholder, 2010; Zhu et al., 2013). Studies exploring continuance commitment showed weak to moderate relationships with trust dimensions (Miao et al., 2014b; Yang, 2005).

The meta-analytic findings for trust dimensions and OCB, OC, affective commitment and continuance commitment are in Table 3. Both affective and cognitive trust displayed moderate to strong and positive relationships with OCB, OC and affective commitment. However, the association between trust dimensions and continuance commitment was weaker.

In random-effects meta-analysis, the between-study variance is referred to as tau-squared, τ² (Borenstein et al., 2007; Deeks et al., 2008). τ² is the variance of the effect sizes across the population of studies in this meta-analysis, which reflects the variance of the true effect sizes (heterogeneity), while tau (τ) is the estimated standard deviation of underlying true effects across studies (Deeks et al., 2008). The summary meta-analysis effect and τ, as standard deviation, describes the distribution of true effects in Table 3 (Borenstein et al., 2007).

Authors conducted significance testing on the relationships between trust dimensions and organisational outcomes (see Table 4). Results mostly support this study’s hypothesis, as four out of the five relationships demonstrated significant results. Affective trust was found to have a significantly stronger relationship with OCB (Z = 3.83, p < .001) and OC (Z = 2.06, p < .05) than cognitive trust. Regarding the facets of OC, affective trust showed a significantly stronger relationship with affective commitment than cognitive trust (Z = 2.46, p < .01), while continuance commitment had a stronger relationship with cognitive trust than affective trust (Z = −3.79, p < .001). There was no difference found between the trust dimensions relationships with normative commitment, albeit this meta-analysis had only one study to draw from that explored this relationship. Table 4 reports effect sizes of these relationships in Cohen’s q.

4. Discussion

This meta-analysis explored trust dimensions (affective and cognitive) relationships with critical organisational outcomes (OCB, OC and/or its facets) related to organisational effectiveness. Affective trust is emotionally based and forms through relationships, while cognitive trust is based on the employee’s analysis of observation of trustworthy leader behaviour (McAllister, 1995). Results of the meta-analysis showed that affective and cognitive trust have significantly different strengths of relationships on four out of five organisational outcomes explored in this study.

The findings showed that OCB has a stronger positive relationship with affective trust than cognitive trust. This suggests employees may go beyond what is required of their role when they have stronger emotional, relational-based trust in their leaders than if their leaders displayed trustworthy behaviours alone. This finding supports that of Dirks and Ferrin (2002), who demonstrated that global trust positively influenced OCB, but cognitive trust did not. When the cognitive trust dimension is isolated, it does not appear to have as strong an influence on OCB as the affective trust dimension or global trust.

The literature differed in how OC was conceptualised in each study included in this meta-analysis. Some studies examined OC as a global construct, while others measured the three facets (affective, continuance and normative) separately. When OC was a global measure, affective trust appeared to have a stronger positive relationship than cognitive trust. An emotional trust connection to the leader appears to encourage stronger commitment to the organisation.

As with trust dimensions and OC, the same relationship was evident between trust dimensions and the facet of affective commitment, but the significance of difference was much higher between affective and cognitive trust. This could be due to affective commitment conveying an emotional connection to the organisation. It makes intuitive sense that a more emotionally driven trust dimension, such as affective trust, has a stronger relationship than a less emotionally derived trust dimension (i.e. cognitive trust) with this OC facet.

Interestingly, continuance commitment appeared to have a stronger, positive relationship with cognitive trust than affective trust. Affective trust appeared to have a negative relationship with the outcome. Continuance commitment is the cost–benefit analysis an employee makes about staying or leaving the organisation. It appeared that only cognitive trust positively influenced the cost–benefit analysis for employees. Perhaps this is because employees weigh up the likelihood of staying with a manager where they know what behaviours to expect, rather than moving to a new team where they may not know what to expect.

Normative commitment had positive relationships with both trust dimensions; however, the differences between affective and cognitive trust did not differ significantly. Normative commitment is the sense of obligation an employee feels to stay with the organisation. In the context of this article, trust dimensions appear to be important for positively influencing employees’ sense of obligation to stay with an organisation. There was only one study identified that examined this relationship in the literature, which is discussed in the limitations. As more literature emerges, future exploration may inform the true relationship between normative commitment and trust dimensions.

Cohen (1988) suggests an effect size measure with the denomination q to interpret the difference between two correlations. Exploring effect sizes of trust dimensions on OCB (q = .108), OC (q = .111), affective commitment (q = .115) and continuance commitment (q = .221) revealed that not only do trust dimensions’ relationships significantly differ but that the difference has a small effect. Although this paradigm may not be obvious without careful analysis, affective trust’s relational and emotional qualities do have a greater impact on OCB, OC and affective commitment than trust that develops from only demonstrating trustworthy behaviour (e.g. cognitive trust). Trustworthy behaviour seems to have a greater influence on the cost–benefit analysis an employee makes about staying or leaving the organisation.

4.1. Implications

These findings provided evidence that trust dimensions have stronger and differing relationships with the organisational outcomes examined in this study. Using a meta-analytic statistical approach, this study’s findings reinforce trust as a multidimensional, not unidimensional, construct.

From the scientist–practitioner perspective, it is worth exploring the possibility that leaders’ efforts are better placed developing relationships with employees to increase the emotional connections, through which affective trust cultivates. This may be more efficacious than focussing on displaying trustworthy behaviour alone, if they wish to have stronger effects on organisational performance outcomes. Organisational development practitioners who build programmes to change leader behaviour may also consider whether these behaviours are resulting in strong leader–employee relationships and, by extension, emotional bonds. These findings suggest that employee commitment to their organisation more strongly links to their emotional connection to their leaders. Organisational development practitioners and organisational leaders should tap into this in light of the employer and other entities emerging as the most trusted entity for the public (Biddle and Reddy, 2019; Edelman, 2019). Outcomes from this article, in support of trust theory advancement, will help leaders to better build and sustain trust in teams within this shifting societal paradigm. It may never have been more important than right now to do so.

4.2. Limitations

Although meta-analyses based on small numbers of studies are valid (Valentine et al., 2010), more studies measuring affective and cognitive trust between employees and leaders would support theory advancement, to demonstrate the gravity of the findings. The relationship between trust dimensions and continuance commitment, as well as global OC, showed larger credibility interval spans than other relationships explored in this meta-analysis. The systematic search, using PRISMA guidelines, only identified two studies examining trust dimensions and global OC, and their findings were significant and thus included in the analysis. In addition, the systematic search only identified two studies examining trust dimensions and continuance commitment, but those studies showed non-significant results. The search also only yielded one study on trust dimensions and normative commitment, and therefore, no meta-analysis was performed on that relationship. Regardless, authors deemed the studies reasonable to include, as it was important to explore all facets of OC and trust dimensions. With more studies to analyse, there would be stronger meta-analytic findings for these specific relationships.

McAllister’s (1995) affective and cognitive trust scale is the foremost measure of trust dimensionality (McEvily and Tortoriello, 2011). The literature has not reviewed this measure for validity since McAllister (1995) originally published it and this is a potential limitation for this study. McAllister (1995) did not consider employees’ perceptions of how they come to trust their leaders or what their perceptions of trust were based on. Therefore, no evidence is apparent in McAllister’s (1995) paper regarding whether he conducted qualitative analysis of employee perception of trust in leaders to determine which items to include. This is a critical practice in valid, reliable scale development (DeVellis, 2016). The extent to which this measure is valid in today’s working context is questionable given the vast variability of organisational environments and types. The nature of leadership is changing in terms of whether the leader leads a team in a traditional sense or virtually. There has been a rise in flexible working arrangements, virtual teams, virtual workplaces, virtual leadership and activity-based working due to globalisation, developments in technology and rising prices of space for organisations to occupy (Atkinson and Hall, 2011; Galea et al., 2014). The benefit of this new way of working is that teams can be assembled according to the knowledge, skills and abilities required, from anywhere in the world, enabling an organisation to become more agile, adaptable and competitive in a global market. The challenge is that it is not known if or how virtual leadership and teams affect building and sustaining organisational trust.

A final limitation is that examining longitudinal studies to make concrete conclusions about causation would have been ideal. However, no longitudinal studies were identified that met the inclusion criteria, thus no causal conclusions can be made. A longitudinal study of trust dimensions and organisational outcomes would improve understanding of trust’s multidimensionality.

4.3. Future research

Understanding organisational trust is critical, as research to date has not aligned around a single definition or means to measure, despite its well-established influence on organisational outcomes (Dirks and Ferrin, 2002; McEvily and Tortoriello, 2011). As more research emerges that measures multidimensional trust against organisational outcomes, future meta-analyses should expand the findings in this article. Moreover, the only measure of multidimensional trust appears to be McAllister’s (1995) affective and cognitive trust scale, thus future research should explore alternative multidimensional trust measures as they surface in the literature.

Future research should confirm or contextualise the validity of multidimensional trust within the current working climate. The working climate has shifted since the original scale was developed (McAllister, 1995). The rise of virtual teams, agile workforces and globalisation may affect how employees conceptualise trust (Ferazzi, 2012). Although the field is relatively new, there has been some research on the importance of trust building in teams led by virtual leaders (Caulat, 2012). Antonakis and Atwater (2002) described the emerging organisational condition that is becoming a norm and the challenges it presents for trust development. There is some evidence that the trust-building process between employees and virtual leaders is similar to the process in a traditional, face-to-face leader–employee relationship, but there are also potential differences that must be explored (Hirschy, 2011; Malhotra et al., 2007; Merriman et al., 2007; Oertig and Buergi, 2006; Pauleen, 2003). These differences could relate to the way trust develops, what constitutes trust in virtual employee–leader relationships and to what extent the current model of multidimensional trust is sufficient.

5. Conclusion

The aim of this meta-analysis was to investigate whether affective and cognitive trust between employees and leaders differs in strength of impact on the organisational outcomes of OCB, OC and/or its facets. This study found that the strength of the relationship between affective trust and cognitive trust was different for four out of the five organisational outcomes explored. This suggests that trust is multidimensional and that the dimensions have different strengths in terms of influencing positive organisational outcomes in some circumstances; however, this requires further exploration.

This meta-analysis also showed that the two trust dimensions are measured predominantly using McAllister’s 11-item scale of affective and cognitive trust. Not since Dirks and Ferrin’s (2002) paper has any form of dual dimension model of employee trust in leaders been explored in a known meta-analysis. Therefore, this study maintains that trust is indeed multidimensional and that the two dimensions (affective and cognitive) impact workplace outcomes differently in some cases, but not others.

There are implications for future research, organisational leaders and organisational development practitioners, based on these findings. Trust is complex and may be misunderstood. Current research warrants confirming the model and measure of trust dimensions for relevance in the present working climate to better guide organisational leaders and organisational development practitioners in their work.

Footnotes

Final transcript accepted 16 December 2019 by Peter Jordan (AE Organisational Behaviour).

Funding

The author(s) received no financial support for the research, authorship and/or publication of this article.

ORCID iD

Sarah Fischer

Notes

References

Agote

Aramburu

Lines

(2016) Authentic leadership perception, trust in the leader, and followers’ emotions in organizational change processes. The Journal of Applied Behavioral Science 52: 35–63.

Allen

Meyer

(1990) The measurement and antecedents of affective, continuance and normative commitment to the organization. Journal of Occupational Psychology 63: 1–18.

Antonakis

Atwater

(2002) Leader distance: A review and a proposed theory. The Leadership Quarterly 13: 673–704.

Atkinson

Hall

(2011) Flexible working and happiness in the NHS. Employee Relations 33: 88–105.

*Berkovich

(2018) Typology of trust relationships: Profiles of teachers’ trust in principal and their implications. Teachers and Teaching 24: 749–767.

Biddle

Reddy

(2019) ANU Poll 29 Universities in Australia: Attitudes and Challenges (Computer File). Canberra, ACT, Australia: Australian Data Archive, The Australian National University.

Borenstein

Hedges

Rothstein

(2007) Meta-analysis: Fixed effect vs. random effects. Meta-analysis.com: 1–162.

Cacciatore

Meng

Boyd

, et al. (2016) Political ideology, media-source preferences, and messaging strategies: A global perspective on trust building. Public Relations Review 42: 616–626.

Carter

Mossholder

(2015) Are we on the same page? The performance effects of congruence between supervisor and group trust. Journal of Applied Psychology 100: 1349–1363.

10.

Caulat

(2012) Virtual Leadership. Faringdon: Libri Publishing Ltd.

11.

Chen

Lam

Zhong

(2012) Effects of perceptions on LMX and work performance: Effects of supervisors’ perception of subordinates’ emotional intelligence and subordinates’ perception of trust in the supervisor on LMX and, consequently, performance. Asia Pacific Journal of Management 29: 597–616.

12.

*Chiang

Lin

(2016) Motivating organisational commitment in hotels: The relationship between leaders and employees. Journal of Human Resources in Hospitality & Tourism 15: 462–484.

13.

Clugston

(2000) The mediating effects of multidimensional commitment on job satisfaction and intent to leave. Journal of Organizational Behavior 21: 477–486.

14.

Cohen

(1988) Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates.

15.

Cohen

(1983) Applied Multiple Regression/correlation Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum Associates.

16.

Colquitt

(2001) On the dimensionality of organisational justice: A construct validation of a measure. Journal of Applied Psychology 86: 386–400.

17.

Conchie

(2013) Transformational leadership, intrinsic motivation, and trust: A moderated-mediated model of workplace safety. Journal of Occupational Health Psychology 18: 198–210.

18.

DeConinck

Bachmann

(2007) The impact of equity sensitivity and pay fairness on marketing managers’ job satisfaction, organisational commitment and turnover intentions. Marketing Management Journal 17: 134–141.

19.

Deeks

Higgins

Altman

(2008) Analysing data and undertaking meta-analyses. In: Higgins

JPT

Green

(eds) Chapter 9: Cochrane Handbook for Systematic Reviews of Interventions: Cochrane Book Series. Chichester: John Wiley & Sons, Inc., pp. 243–296.

20.

DeVellis

(2016) Scale Development: Theory and Applications (Vol. 26). Newbury Park, CA: SAGE.

21.

Dietz

Den Hartog

(2006) Measuring trust inside organisations. Personnel Review 35: 557–588.

22.

Dirks

Ferrin

(2002) Trust in leadership: Meta-analytic findings and implications for research and practice. Journal of Applied Psychology 87: 611–628.

23.

Edelman

(2019) Trust at work. Edelman.com, 21 January. Available at: https://www.edelman.com/insights/trust-at-work

24.

Ertürk

(2010) Exploring predictors of organizational identification: Moderating role of trust on the associations between empowerment, organizational support, and identification. European Journal of Work and Organizational Psychology 19: 409–441.

25.

Farh

Earley

Lin

(1997) Impetus for action: A cultural analysis of justice and organisational citizenship behavior in Chinese society. Administrative Science Quarterly 42: 421–444.

26.

Ferazzi

(2012) How to build trust in a virtual workplace. Harvard Business Review, 13 March. Available at: https://hbr.org/2012/10/how-to-build-trust-in-virtual

27.

Field

Gillett

(2010) How to do a meta-analysis. British Journal of Mathematical and Statistical Psychology 63: 665–694.

28.

Frazier

Fainshmidt

Klinger

, et al. (2017) Psychological safety: A meta-analytic review and extension. Personnel Psychology 70: 113–165.

29.

Galea

Houkes

De Rijk

(2014) An insider’s point of view: How a system of flexible working hours helps employees to strike a proper balance between work and personal life. The International Journal of Human Resource Management 25: 1090–1111.

30.

Henrique-Gomes

(2019) Kenneth Hayne says trust in politics and institutions ‘damaged or destroyed’. The Guardian, 8 August. Available at: https://www.theguardian.com/australia-news/2019/aug/08/kenneth-hayne-says-trust-in-politics-and-institutions-damaged-or-destroyed

31.

Hirschy

(2011) Virtual team leadership: A case study in Christian Higher Education. Christian Higher Education 10: 97–111.

32.

Hollibaugh

Jr (2016) Presidential appointments and public trust. Presidential Studies Quarterly 46: 618–639.

33.

*Hon

(2010) The mediating role of trust between expatriate procedural justice and employee outcomes in Chinese hotel industry. International Journal of Hospitality Management 29: 669–676.

34.

Hunter

Schmidt

(2004) Methods of Meta-analysis: Correcting Error and Bias in Research Findings. 2nd ed. Newbury Park, CA: SAGE.

35.

Indartono

Chen

CHV

(2011) Moderating effects of tenure and gender on the relationship between perception of organisational politics and commitment and trust. South Asian Journal of Management 18: 7–36.

36.

Kim

Wang

Chen

(2018) Mutual trust between leader and subordinate and employee outcomes. Journal of Business Ethics 149: 945–958.

37.

Lee

Preacher

(2013) Calculation for the test of the difference between two dependent correlations with one variable in common (Computer software). Available at: http://quantpsy.org

38.

Lee

Allen

(2002) Organizational citizenship behavior and workplace deviance: The role of affect and cognitions. Journal of Applied Psychology 87: 131–142.

39.

Liu

Wang

(2013) Perceived risk and organisational commitment: The moderating role of organisational trust. Social Behavior and Personality: An International Journal 41: 229–240.

40.

*Lu

(2014) Ethical leadership and organisational citizenship behaviour: The mediating roles of cognitive and affective trust. Social Behavior and Personality: An International Journal 42: 379–389.

41.

McAllister

(1995) Affect-and cognition-based trust as foundations for interpersonal cooperation in organizations. Academy of Management Journal 38: 24–59.

42.

McEvily

Tortoriello

(2011) Measuring trust in organisational research: Review and recommendations. Journal of Trust Research 1: 23–63.

43.

Malhotra

Majchrzak

Rosen

(2007) Leading virtual teams. Academy of Management Perspectives 21: 60–70.

44.

Merriman

Schmidt

Dunlap-Hinkler

(2007) Profiling virtual employees: The impact of managing virtually. Journal of Leadership & Organizational Studies 14: 6–15.

45.

Meyer

Allen

Smith

(1993) Commitment to organizations and occupations: Extension and test of three-component conceptualization. Journal of Applied Psychology 78: 538–555.

46.

Meyer

Allen

(1984) Testing the ‘side-bet theory’ of organisational commitment: Some methodological considerations. Journal of Applied Psychology 69: 372–378.

47.

Meyer

Allen

(1991) A three-component conceptualization of organisational commitment. Human Resource Management Review 1: 61–89.

48.

*Miao

Newman

Huang

(2014a) The impact of participative leadership on job performance and organisational citizenship behavior: Distinguishing between the mediating effects of affective and cognitive trust. The International Journal of Human Resource Management 25: 2796–2810.

49.

*Miao

Newman

Schwarz

, et al. (2014b) Servant leadership, trust, and the organisational commitment of public sector employees in China. Public Administration 92: 727–743.

50.

Moher

Altman

Liberati

, et al. (2010) Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. International Journal of Surgery 8: 336–341.

51.

Mowday

Steers

Porter

(1979) The measurement of organisational commitment. Journal of Vocational Behavior 14: 224–247.

52.

*Mushonga

Thiagarajan

Torrance

(2014) Fairness in the workplace: The mediating role of trust in the relationship between supervisory justice and work outcomes. SAM Advanced Management Journal 79: 17–25.

53.

*Newman

Kiazad

Miao

, et al. (2014) Examining the cognitive and affective trust-based mechanisms underlying the relationship between ethical leadership and organisational citizenship: A case of the head leading the heart? Journal of Business Ethics 123: 113–123.

54.

Oertig

Buergi

(2006) The challenges of managing cross-cultural virtual project teams. Team Performance Management: An International Journal 12: 23–30.

55.

Organ

Podsakoff

MacKenzie

(2006) Organisational Citizenship Behavior: Its Nature, Antecedents, and Consequences. London: SAGE.

56.

Pauleen

(2003) An inductively derived model of leader-initiated relationship building with virtual team members. Journal of Management Information Systems 20: 227–256.

57.

Podsakoff

Blume

Whiting

, et al. (2009) Individual- and organizational-level consequences of organisational citizenship behaviors: A meta-analysis. Journal of Applied Psychology 94: 122–141.

58.

Podsakoff

MacKenzie

Paine

, et al. (2000) Organisational citizenship behaviors: A critical review of the theoretical and empirical literature and suggestions for future research. Journal of Management 26: 513–563.

59.

Preacher

(2002) Calculation for the test of the difference between two independent correlation coefficients (Computer software). Available at: http://quantpsy.org

60.

Rosenthal

(1995) Writing meta-analytic reviews. Psychological Bulletin 118: 183–192.

61.

Rousseau

Sitkin

Burt

, et al. (1998) Not so different after all: A cross-discipline view of trust. Academy of Management Review 23: 393–404.

62.

Schaubroeck

Lam

Peng

(2011) Cognition-based and affect-based trust as mediators of leader behavior influences on team performance. Journal of Applied Psychology 96: 863–871.

63.

Schoorman

Mayer

Davis

(2007) An integrative model of organisational trust: Past, present, and future. Academy of Management Review 32: 344–354.

64.

Settoon

Bennett

Liden

(1996) Social exchange in organizations: Perceived organisational support, leader-member exchange, and employee reciprocity. Journal of Applied Psychology 81: 219–227.

65.

Settoon

Mossholder

(2002) Relationship quality and relationship context as antecedents of person- and task-focused interpersonal citizenship behavior. Journal of Applied Psychology 87: 255–267.

66.

Somech

Drach-Zahavy

(2000) Understanding extra-role behavior in schools: The relationships between job satisfaction, sense of efficacy, and teachers’ extra-role behavior. Teaching and Teacher Education 16: 649–659.

67.

Steiger

(1980) Tests for comparing elements of a correlation matrix. Psychological Bulletin 87: 245–251.

68.

Stevenson

Wolfers

(2011) Trust in public institutions over the business cycle. The American Economic Review 101: 281–287.

69.

Tett

Meyer

(1993) Job satisfaction, organisational commitment, turnover intention, and turnover: Path analyses based on meta-analytic findings. Personnel Psychology 46: 259–293.

70.

Torabi Asr

(2019) The language gives it away: How an algorithm can help us detect fake news. The New York Times, 25 August. Available at: https://www.pbs.org/newshour/science/the-language-gives-it-away-how-an-algorithm-can-help-us-detect-fake-news

71.

Turner

Bird

Higgins

(2013) The impact of study size on meta-analyses: Examination of underpowered studies in Cochrane reviews. PLoS ONE 8: e59202.

72.

Valentine

Pigott

Rothstein

(2010) How many studies do you need? A primer on statistical power for meta-analysis. Journal of Educational and Behavioral Statistics 35: 215–247.

73.

Williams

Anderson

(1991) Job satisfaction and organisational commitment as predictors of organisational citizenship and in-role behaviors. Journal of Management 17: 601–617.

74.

Wong

Ngo

(2012) The effects of trust in organisation and perceived organisational support on organisational citizenship behaviour: A test of three competing models. The International Journal of Human Resource Management 23: 278–293.

75.

*Yang

(2005) The role of trust in organizations: Do foci and bases matter? PhD Thesis, Nanjing University, Nanjing, China.

76.

Yang

Mossholder

Peng

(2009) Supervisory procedural justice effects: The mediating roles of cognitive and affective trust. The Leadership Quarterly 20: 143–154.

77.

*Yang

Mossholder

(2010) Examining the effects of trust in leaders: A bases-and-foci approach. The Leadership Quarterly 21: 50–63.

78.

*Zhu

Newman

Miao

, et al. (2013) Revisiting the mediating role of trust in transformational leadership effects: Do different types of trust make a difference? The Leadership Quarterly 24: 94–105.