Inglês (pdf)
Artigo em XML
Referências do artigo
Tradução automática
Enviar este artigo por email
Citado por SciELO 
Citado por Google 
Similares em
SciELO 
Similares em Google 
Permalink
1. Introduction
Literature addresses conflicts from different perspectives. In the organizational context, conflicts can be classified by relationship or socio-affective aspect; that is, (i) concerning the task, called cognitive or substantive; and (ii) derived from emotional or affective aspects of interpersonal relationships, called affective (Jehn, 1995). Cognitive conflicts are generally treated as disagreements between group members that may occur during the execution of a task due to difference of opinions and attitudes (Puck & Pregernig, 2014). Affective conflicts comprise incompatibility or tensions between group members and can trigger animosity and annoyance, since debates minimize the best solutions and praise individual interests (Ehie, 2010).
In the organizational context, conflicts can raise disagreements on whether they are constructive (beneficial) or destructive (harmful) for groups and organizations (Jehn, 1995; Um & Oh, 2021). A stream of research (e.g., Granados Ruiz & Llanos, 2024; Hjerto & Kuvaas, 2017) highlights that cognitive conflicts negatively affect groups and organizations. Another stream of research (e.g., Jehn, 1995; Um & Oh, 2021) points to their positive effects because, in conflict, the individuals who make up the groups are able to learn from each other and contribute to achieving common benefits, such as stimulating creativity and innovation to obtain competitive advantage (Clauss & Ritala, 2023; Tjosvold, 1997). It is argued that when conflicts are absent, groups neglect inefficiencies and continue with existing forms of work (Kakar, 2018).
Different types of conflict can influence aspects related to work and performance. In this study, the one that refers to intragroup information sharing is relevant (McCarter et al., 2020; Tsai & Bendersky, 2016). Information sharing can help reduce uncertainty and increase motivation to use information (Beuren et al., 2020; Hwang et al., 2013). Research has observed that task-related conflicts lead to greater information sharing and improved decision-making, which does not occur in affective conflicts (Van Greunen et al., 2021).
However, research gaps are observed regarding predictors and cognitive conflicts, their influence on information sharing, and the interference of affective conflicts in this relationship. Literature indicates that these are strictly related to cognitive conflicts (Granados Ruiz & Llanos, 2024; O'Neill & Mclarnon, 2018). Thus, this study examines the role of conflict predictors and cognitive conflicts on information sharing, moderated by intra-workgroup affective conflicts at cooperatives; the motivation to investigate those conflicts stems from the ambiguity of manager roles in these organizations, as observed by Beuren et al. (2018).
Role ambiguity refers to uncertainties about an individual's role in the organization and is established when there is a lack of information or insufficient authority for the role (Glazer, 2021). At cooperative organizations, drivers based on cooperative principles can accentuate this conflict (Anzilago et al., 2018; Serdyukov & Grima, 2024), mainly due to the ambiguous roles of the different stakeholders in the organization (Beuren et al., 2018; Teixeira et al., 2020). For instance, the associate can be simultaneously a customer, supplier, and manager; therefore, the presumption is that intragroup cognitive conflicts are common within cooperatives, which makes them a favorable scenario for carrying out this research.
This context aligns with the assumptions of Collective Action Theory. It assumes that individuals, as rational social actors, make cooperative decisions based on the actions of others (DeMarrais & Earle, 2017). Thus, it is hypothesized that increased collective action can lead to successful achievement of results from shared resources (Murunga et al., 2021). These considerations are aligned with cooperatives' context: everyone cooperates and shares the same values to achieve common goals (Ferreira da Silva et al., 2022; Pansera & Rizzi, 2020).
The results of this research contribute to the literature by addressing the relationships between conflict predictors and cognitive conflicts (e.g., Chhajer & Dutta, 2021; Hsu, 2018), cognitive conflicts and information sharing (e.g., Du & Xu, 2018), and moderation of affective conflicts between cognitive conflicts and information sharing (e.g., De Wit et al., 2013); also, by providing empirical evidence, and revealing different sets of relationships that can bring insights for further investigation; by incorporating efforts to reduce divergences in the findings of studies on the impact of cognitive conflicts, which are not always presented as beneficial. The contribution to a subject little addressed in the literature -the field of cooperatives (Anzilago et al., 2018; Mannes et al., 2022), despite its idiosyncrasies, cooperative principles, and representativeness in the world economic context (Beuren et al., 2020)- is also noteworthy.
Understanding cognitive conflicts and their causes is essential for managerial practice since, expressed in the form of debates, they can promote positive behavioral interactions and more information sharing (Bedford et al., 2022; Du & Xu, 2018). The latter is a source of advantage for group work and figures as a critical antecedent of decision-making optimization and organizational results improvement, especially in complex and dynamic environments (Uitdewilligen & Waller, 2018). Thus, the relevance of the research results in the practical context of the organizations studied and those with similar structures, especially for the critical role of agricultural cooperatives as agents of local and social development (Anzilago et al., 2018; Ferreira da Silva et al. 2022). Our aim is to highlight that cognitive conflicts can be beneficial by establishing conditions to expand information sharing with a view to the effectiveness of intragroup communication.
2. Theoretical basis and hypotheses
2.1 Predictors and cognitive conflicts
Cognitive conflicts or task conflicts seek to improve the quality of group work, making the group avoid conformity because divergent opinions and the involvement of individuals with different capacities and knowledge can lead to alternative ideas, which enhances the creation of more innovative solutions (Chai et al., 2020; Mooney et al., 2007). Such conflicts can have positive effects by enabling greater identification and discussion of different perspectives, which increases understanding of the task, optimizes the decision-making process, and yields favorable results (Maia & Lima, 2020). At cooperatives, conflicts may trigger problems that can negatively impact performance and minimize the benefits of the shared business model (Shantz et al., 2020).
Literature has addressed the different causes of conflicts considering elements of an individual and organizational nature. For this research, we included questions concerning task types or structures (Jehn, 1995), task interdependence (Van Der Vegt et al., 2003), and mutuality among group members (Amason & Sapienza, 1997) as conflict predictors. These are related to different task types and involve work group issues like cooperative decisions (DeMarrais & Earle, 2017), resource sharing (Murunga et al., 2021), and role ambiguity (Bhardwaj & Sharma, 2024).
Task types represent the categorization of tasks as simple and routine or complex and non-routine (Jehn, 1995). They can lead to cognitive conflicts or task conflicts and are related to group performance because they affect individual attitudes towards them (Pelled et al., 1999; Beuren et al, 2018). Task conflict helps improve performance on more complex decision-making tasks, which does not occur on less complex tasks (Liu et al., 2022). The ambiguity of roles can strengthen conflicts in tasks completion and negatively impact performance, as observed by Beuren et al. (2018) at cooperatives. Despite these divergent situations, it is assumed that:
H1a: Task types directly and positively influence cognitive conflicts.
Task interdependence is associated with exchanging resources, information, or materials and how much individuals should be involved in solving problems at their workplace (Mooney et al., 2007; Pitafi et al. 2020). It can influence the course and consequences of organizational conflicts (Hsu, 2018; Van Der Vegt et al., 2003). Groups that have more collaborative communication, for example, are less likely to experience the negative effects of conflicts, in addition to avoiding barriers that hinder teamwork (Brett et al., 2020; Pitafi et al., 2020).
Individuals act in isolation within the group in situations of low interdependence, with few interactions or communication; therefore, the possibility of conflict is low (Neck et al., 1996). Higher levels of task interdependence, in which individuals need to cooperate and share, bring them closer together and make them more supportive and influential, encouraging them towards healthy conflict profiles (O'Neill & Mclarnon, 2018; Um & Oh, 2021). The need for rapprochement to cooperate can be explained by the triggering of more complex and ambiguous tasks (Rahman et al., 2019). A more significant occurrence of cognitive conflicts is expected, given the greater need for communication between individuals at cooperatives. Therefore, it is assumed that:
H1b: Task interdependence directly and positively influences cognitive conflicts.
Mutuality among group members occurs when different individuals feel jointly responsible, share goals, and can manifest as a desire to accommodate other individuals to benefit the group through positive social connections aimed at the collective (Amason & Sapienza, 1997; Hartmann et al., 2021). When there is mutuality, there is a greater feeling of communication among agents, who come to believe that they are similar to their partners, which can improve task performance (Chhajer & Dutta, 2021). Consistent with cooperative principles, this is a guiding position of the social responsibility of cooperatives (Beuren et al., 2020).
With high levels of mutuality, communicative agents benefit from favorable judgments, while lower levels can lead to communication errors and even trigger conflicts among group members (Amason & Sapienza, 1997). These authors adduce that higher levels of mutuality may be associated with lower levels of cognitive conflicts. Thus, it is assumed that:
H1c: Mutuality among group members directly and positively influences cognitive conflicts.
2.2 Cognitive conflicts and information sharing
Conflicts can increase information sharing in groups of decision-makers since individuals are asked to describe and justify their positions to defend the opposing points of view (Tjosvold, 1997). When discussing their positions, they may perceive possible inadequacies in their ideas and values, which favors the combination of possibilities that had not been previously considered in isolation, in addition to encouraging discussion and understanding of the aspects covered (Liao et al., 2021).
Besides exchanging information between team members, cognitive conflicts related to tasks can assist in problem-solving, idea generation, and decision-making (Pelled et al., 1999; Zhou et al., 2022). The relevance of cognitive conflicts stands out, especially in decision-making, given the benefits of enabling information that corresponds to the complexity and variety associated with this strategic task (Maia & Lima, 2020).
Cognitive or task conflicts are positively related to innovative behaviors and knowledge sharing (Kakar, 2018), challenging established thinking to enhance organizational unlearning (Zhang et al., 2024). It is postulated that cognitive conflicts arising from constructive discussions about different perspectives impact the effect of information sharing on the manager's degree of influence (Du & Xu, 2018). It follows that cognitive conflicts tend to increase the information sharing necessary to achieve different organizational objectives such as innovation (Bedford et al., 2022). In the cooperative environment, information sharing is relevant because it enables organizational learning, which is essential for improving performance (Singh et al., 2021). Thus, it is expected that:
H2: Cognitive conflicts directly and positively influence information sharing.
2.3 Moderation of affective conflicts between cognitive conflicts and information sharing
Cognitive conflicts influence information sharing; however, there are other aspects interfering in this process. Affective conflicts may also interfere; cognitive conflicts have disadvantages when associated with affective conflicts because these are dysfunctional and can compromise the development of tasks (Claro et al., 2018; Granados Ruiz & Llanos, 2024). Problems such as friction, negative emotions, and interpersonal animosity, representing affective conflicts, can spread and determine how group members react in the debate concerning a specific task, which causes greater rigidity during cognitive conflicts (De Wit et al., 2013).
The strong beliefs or fragile egos of the group members, neuralgic points of affective conflicts, can cause strong attachment to ideas so that individuals end up not recognizing contributions from others. This can trigger negative interpersonal behaviors and lead to distancing, which prevents the effective incorporation of information (Devine, 1999). Affective conflict can prevent the effective incorporation of groups, as it discourages involvement among its members, which can impact the information sharing and knowledge (Lee et al., 2019).
Affective conflicts can lead team members to disregard information from other individuals and focus only on their point of view, especially when there is some inconsistency or lack of motivation to process the different perspectives (De Wit et al., 2013). Therefore, information processing can become more biased, with less commitment to the group and more competitive social motivation, which can affect the quality of decisions (Flores et al., 2018).
Affective conflicts can disrupt communication and cooperation among group members, reducing receptivity to ideas promoted by peers, and negatively impacting group cohesion (Chen et al., 2017; Todorova, 2021) and compromise strategic decisions (Prasad & Junni, 2017). Thus, it is expected that the relationship between cognitive conflicts and information sharing is negative in the presence of affective conflicts, which leads to the following hypothesis:
H3: Affective conflicts negatively moderate the relationship between cognitive conflicts and information sharing.
2.4 Mediation of cognitive conflicts between conflict predictors and information sharing
Cognitive conflicts tend to improve strategic decision making because they facilitate the exchange of information among higher-level managers (Amason & Sapienza, 1997; Loughry & Amazon, 2014). However, there is evidence that the benefits of cognitive conflicts depend on the integration of divergent views of the group, as they come from debate and the exchange of ideas (Clauss & Ritala, 2023; Mooney et al., 2007).
Workgroups that perform routine tasks may have disagreements that impair their functioning, which is not perceived in groups with non-routine tasks, where disagreements can be beneficial, as they provide opportunities for the development of new ideas (Jehn, 1995; Liu et al., 2021). It is possible to obtain significant information gains in cognitive conflicts or moderate task conflicts, which leads individuals to feel more active, interested, and energized; those emotions influence job satisfaction (Todorova, 2021).
Tsai and Bendersky (2016) found that task conflicts or cognitive conflicts are expressed through debates and associated with greater information sharing. There is a perception of receptivity concerning divergent opinions among individuals. Information sharing between group members allows improvements in management, thus enhancing innovation capacity and competitiveness (Bedford et al., 2022).
Regardless of the level of cognitive conflicts, both high and low, facilitate (hinder) sharing (Kakar, 2018). In this perspective, a mediating effect of cognitive conflicts between conflict predictors and information sharing is postulated as follows:
H4a: Cognitive conflicts mediate the relationship between task types and information sharing.
H4b: Cognitive conflicts mediate the relationship between task interdependence and information sharing.
H4c: Cognitive conflicts mediate the relationship between mutuality among group members and information sharing.
Figure 1 presents the theoretical model with the constructs and the hypotheses.
The theoretical model predicts a direct relationship among task types (H1a), task interdependence (H1b), and mutuality among group members (H1c) in cognitive conflicts. H2 assumes a direct relationship between cognitive conflicts and information sharing, while H3 conjectures moderation of affective conflicts in this relationship. The model also predicts indirect relationships through cognitive conflicts among task types (H4a), task interdependence (H4b), and mutuality among group members (H4c) in information sharing. In addition, it includes control variables.
3 Methodological procedures
3.1 Data collection and sampling
A survey was conducted with professionals working in intermediate management positions at cooperatives, found in LinkedIn through searches for the terms 'manager', 'coordinator', 'supervisor'. The assumption is that employees working in these roles are familiar with and manage the processes related to cognitive conflicts and information sharing. Cooperatives were chosen due to their form of management, guided by cooperative principles, which is presumed to create conflicts between managers due to the lack of clarity of roles. Agricultural cooperatives were selected because agriculture is among the most prominent segments listed in the Organization of Brazilian Cooperatives (OCB), which in 2020 had R$ 160.1 billion in assets, with investments of R$ 239 million (OCB, 2022).
Thus, we focused on managers of Brazilian agricultural cooperatives registered on LinkedIn, 855 invitations were sent and 404 were accepted. The access link to the questionnaire on the QuestionPro platform was sent to these managers. The survey link was also sent to 693 professionals from Brazilian agricultural cooperatives. There was support from the OCB and affiliates in disseminating the research to managers registered in their associate networks. Professionals were contacted between August and October, 2021. The final sample resulted in 94 valid responses, a satisfactory sample size according to the parameters of Hair et al. (2021).
The profile of the respondents indicates that 75% are male, with ages ranging between 24 and 72 years old: 44% between 40 and 49 years old and 38% between 30 and 39 years old. As for the positions occupied, 36% exercise the function of coordinator, 27% of managers, and 26% of supervisors. Respondents work at cooperatives located mainly in the South (48%), Southeast (30%), and Midwest (15%) regions. Concerning the segment of activity of agricultural cooperatives, supply goods and inputs stood out (39%), followed by industrialized products of animal origin (24%) and industrialized products of plant origin (24%). Each cooperative may act in more than one segment.
3.2 Constructs and research instrument
The theoretical model of the research consists of six constructs: conflict predictors (task types, task interdependence, and mutuality among group members), cognitive conflicts, affective conflicts, and information sharing. The research instrument consists of assertions on a seven-point Likert scale, which indicates the degree of agreement for each statement ranging from 1 = strongly disagree to 7 = strongly agree. Questions about the profile of the respondents and the organizations they work for were included at the end of the questionnaire.
Regarding conflict predictors, in the first block, we sought to identify the task types that can trigger cognitive conflicts, with 20 assertions by Jehn (1995) that were rephrased from interrogative to affirmative. In the second block, four assertions on task interdependence by Van Der Vegt et al. (2003) were presented, in which the scale was changed from five points to seven for standardization purposes. The third block was composed of four assertions by Amason and Sapienza (1997) to identify mutuality among group members, and the scale, which had five points, was changed.
In the fourth block, six assertions about cognitive conflicts were presented, four adapted from Mooney et al. (2007) and two from Shah and Jehn (1993). The assertions were rephrased from interrogative to affirmative. In the fifth block, questions related to the identification of affective conflicts were presented, six were assertive, four were adapted from Kakar (2018), and two from Mooney et al. (2007). The adaptations implied rephrasing from interrogative to affirmative.
Finally, the sixth block comprised six assertions about information sharing, three adapted from Bunderson and Sutcliffe (2002), and three adapted from Kakar (2018). The adaptations consisted of changes in the scale used; the one by Kakar (2018) had nine points, focusing on information sharing, and the other focused on knowledge sharing.
The research instruments used were in English, so their translation into Portuguese was performed. Assertions did not consider the context of cooperatives; hence, the research instrument was evaluated by two researchers from an accounting graduate program, and a pre-test was conducted by two professionals in the field before applying it to professionals from agricultural cooperatives.
All data were collected considering a single method, and the same professionals responded to the assertions about all constructs, which can trigger common method bias (Podsakoff & Organ, 1986). A cover letter was attached to the research instrument with clarifications regarding the purpose of the research and information about the items of each construct and their respective scales to minimize possible problems concerning the common method bias. The Harman single-factor test was performed after data collection, which resulted in a total explained variance of 43.10%, considered acceptable according to the 50% limit (Podsakoff & Organ, 1986).
3.3 Control variables
Group dynamics influence the thinking and behavior of individuals within the group, and this dynamic is influenced by the individual characteristics of each person who composes it (Díaz-García et al., 2013). Therefore, research includes control variables of an individual nature, such as age and gender (Qi & Armstrong, 2019; Tremblay, 2017).
We chose to constraint the respondents' age and gender to control individual factors that may influence information sharing. Age was measured in years and identified through an open-ended question to the respondents. Gender comprised two categories segregated by coding: "0" when respondents identified themselves as male and "1" when they identified themselves as female.
3.4 Analysis techniques and procedures
Data analysis began with a factor analysis of the research instrument to verify relationships and common factors of the constructs (Hair et al., 2017). In this analysis, one item of the task interdependence construct, two of the cognitive conflicts construct, and one of the information sharing construct were removed from the model because they do not meet the criterion proposed by Hair et al. (2017) of obtaining factor loads greater than 0.70 for each item. Subsequently, descriptive analyses were performed to assess the distribution of quantitative variable responses.
Consistent with the study by Jehn (1995), task type construct was measured as a score from the sum of the 20 items. Thus, the score could vary between 20 (20*1) and 140 (20*7) [20 items * scale range from 1 to 7, where 1 is the minimum possible and 7 is the maximum possible per respondent]. Cronbach's Alpha of the 20 items, calculated using the IBM SPSS software, was 0.75, which is satisfactory. This score represented one single item in the SmartPLS software.
Finally, Structural Equations Modeling (SEM) was applied to the hypothesis test, estimated from Partial Least Squares (PLS) (Hair et al., 2017). The modeling was applied using SmartPLS software 3.0. PLS-SEM is a multivariate analysis technique that combines factor analysis and multiple regression methods to examine the structure of interrelationships between constructs, mainly in complex models (Hair et al., 2021). This technique was used due to its robustness when the data do not present normality and is compatible with smaller samples (Hair et al., 2021) and also considering the exploratory stage of the research.
4. Result Description and Analysis
4.1 Measurement model
The structural equation modeling analysis began by verifying the measurement model, which assesses the convergent and discriminant validities and the composite reliability of the constructs (Hair et al., 2017). Table 1 shows the values corresponding to the measurement model's validity and reliability. (See Table 1).
Table 1 Validity and reliability of the measurement model.
| Cronbach’s Alpha | rho_A | CR | AVE | |
|---|---|---|---|---|
| 1. Task types | 1.000 | 1.000 | 1.000 | 1.000 |
| 2. Task interdependence | 0.710 | 0.832 | 0.832 | 0.632 |
| 3. Mutuality among group members | 0.842 | 0.871 | 0.891 | 0.672 |
| 4. Cognitive conflicts | 0.839 | 0.882 | 0.895 | 0.685 |
| 5. Affective conflicts | 0.893 | 0.929 | 0.916 | 0.647 |
| 6. Information sharing | 0.751 | 0.757 | 0.835 | 0.504 |
Note: Cronbach's Alpha (>0.70); rho_A (>0.70); CR = Composite Reliability (>0.70); AVE = Average Variance Extracted (>0.50)
Source: own elaboration
The model presents validity and reliability, with loads of Cronbach's Alpha, rho_A, and Composite Reliability (CR) greater than 0.70 for all constructs (Hair et al., 2017). The task type construct was analyzed as a single item, so all loads are equal to 1.000. The convergent validity, by the values presented for Average Variance Extracted (AVE), demonstrates reliability with loads of the constructs greater than 0.50. Table 2 shows the correlations of discriminant validity based on the Fornell-Larcker criteria and Heterotrait-Monotrait Ratio of Correlations (HTMT). (See Table 2).
Table 2 Discriminant validity correlations and results.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | |
|---|---|---|---|---|---|---|---|---|
| 1. Task types | - | 0.039 | 0.065 | 0.136 | 0.207 | 0.107 | 0.062 | 0.017 |
| 2. Task interdependence | -0.031 | 0.795 | 0.395 | 0.400 | 0.268 | 0.326 | 0.095 | 0.135 |
| 3. Mutuality among group members | -0.008 | 0.334 | 0.820 | 0.481 | 0.412 | 0.643 | 0.101 | 0.071 |
| 4. Cognitive conflicts | -0.126 | 0.336 | 0.437 | 0.828 | 0.298 | 0.644 | 0.191 | 0.162 |
| 5. Affective conflicts | 0.221 | -0.209 | -0.367 | -0.284 | 0.804 | 0.386 | 0.097 | 0.102 |
| 6. Information sharing | -0.071 | 0.249 | 0.512 | 0.524 | -0.348 | 0.710 | 0.166 | 0.189 |
| 7. Gender | 0.062 | 0.003 | 0.021 | -0.171 | 0.070 | -0.051 | - | 0.114 |
| 8. Age | 0.017 | -0.066 | 0.059 | 0.106 | -0.041 | 0.077 | -0.114 | - |
Note: Values in bold represent the square roots of the AVE; the lower diagonal indicates the correlations by the Fornell-Larcker criterion; and the upper diagonal indicates the values by the HTMT criterion
Source: own elaboration.
The discriminant validity by the Fornell-Larcker criterion was met, considering that the square root of the AVE is higher than the correlation between the constructs (Hair et al., 2017). The HTMT discriminant validity criterion was also met, which consists of evaluating the mean correlations of the items between the constructs, with values lower than 0.85 being expected when more conceptually distinct constructs are observed (Hair et al., 2021), as in the case of this research.
4.2 Structural model
Path coefficients were initially estimated using the structural model. Bootstrapping with 5,000 resamples, bias-corrected and accelerated confidence interval, and a two-tailed test at the significance level of 0.05 were considered to analyze the model and determine the significance of the relationships between the latent constructs (Hair et al., 2017). Blindfolding was used to verify predictive relevance (Q2). Table 3 shows the results of the tests performed. (See Table 3).
Table 3 Structural model results.
| Hypotheses | Beta (β) | Error | T-value | VIF | P-value | Decision | |
|---|---|---|---|---|---|---|---|
| H1a | Task types -> Cognitive conflicts | -0.117 | 0.098 | 1.191 | 1.001 | 0.234 | Not Accept |
| H1b | Task interdependence -> Cognitive conflicts | 0.210 | 0.111 | 1.886 | 1.126 | 0.059* | Accept |
| H1c | Mutuality among group members -> Cognitive conflicts | 0.366 | 0.118 | 3.096 | 1.125 | 0.002*** | Accept |
| H2 | Cognitive conflicts -> Information sharing | 0.425 | 0.090 | 4.698 | 1.207 | 0.000*** | Accept |
| H3 | Cognitive conflicts * Affective conflicts -> Information | 0.124 | 0.081 | 1.537 | 1.084 | 0.124 | Not Accept |
| sharing | |||||||
| - | Gender -> Information sharing | 0.034 | 0.105 | 0.327 | 1.047 | 0.743 | - |
| - | Age -> Information sharing | 0.033 | 0.108 | 0.306 | 1.025 | 0.759 | - |
Note: Significance at the level of *p<0.10; **p<0.05; *** p<0.01.
Evaluation of the structural model: R2: Cognitive conflicts = 0.220; Information sharing = 0.304.
Predictive relevance (Q2): Cognitive conflicts = 0.140; Information sharing = 0.137.
Source: own elaboration.
There was no statistical support to accept H1a (ß = -0.117), which predicted a relationship between task types and cognitive conflicts. The other hypotheses that assumed direct relationships, H1b (ß = 0.210; p < 0.10), H1c (ß = 0.366; p < 0.01), and H2 (ß = 0.425; p < 0.01) were accepted. To analyze the relationship between variables, a significance level of up to 10% was considered, which is acceptable for exploratory research (Hair et al., 2017). It is worth noting that the size of the research sample (n=94) allows accepting the hypotheses at a 10% significance level, considering direct relationships (Hair et al., 2021). For it to be significant, the minimum path coefficient expected is between 0.21 and 0.30 (Hair et al., 2021). H3 (ß = 0.124), which predicted moderation of affective conflicts in the relationship between cognitive conflicts and information sharing, did not present statistical significance; therefore, it was rejected. When considering the control variables (gender and age) on information sharing no significance was observed either.
The Variance Inflation Factor (VIF) values to identify multicollinearity issues were observed in addition to the path coefficients. The VIF values presented by the research constructs varied between 1.001 and 1.207, which suggests the absence of multicollinearity (Hair et al., 2017). The variance explained (R2) presented significant explanatory power of cognitive conflicts and information sharing constructs. Values greater than zero were found when determining the constructs' predictive relevance (Q2) (Hair et al., 2017).
We further determined the potential specific indirect effects in the structural model to complement the analysis, as shown in Table 4.(See Table 4).
Table 4 Specific indirect effects.
| Beta (β) | Error | T-value | P-value | Decision | ||
|---|---|---|---|---|---|---|
| H4a | Task types -> Cognitive conflicts -> Information sharing | -0.050 | 0.043 | 1.157 | 0.247 | Not Accept |
| H4b | Task interdependence -> Cognitive conflicts -> | 0.089 | 0.049 | 1.813 | 0.070* | Not Accept |
| Information sharing | ||||||
| H4c | Mutuality among group members -> Cognitive conflicts | 0.156 | 0.067 | 2.317 | 0.021** | Accept |
| -> Information sharing | ||||||
Note: Significance at the level of *p<0.10; **p<0.05; *** p<0.01.
Source: own elaboration.
The task types predictor does not show statistical significance. When considering the mediation of cognitive conflicts in the relationship between task interdependence and information sharing (ß = 0.089, p-value = 0.070), there is no support for accepting the hypothesis, considering the quantitative insufficiency of the sample (Hair et al., 2021). In turn, it was observed that mutuality indirectly affects information sharing through cognitive conflicts (ß = 0.156, p-value = 0.021).
4.3 Result discussion
Hypothesis H1a predicted a direct and positive relationship between task types and cognitive conflicts, which did not have statistical support. This finding contrasts with Pelled et al. (1999), who propose that task types can lead to cognitive conflicts or task conflicts. Jehn (1995) states that routine tasks negatively influence cognitive conflicts, while non-routine tasks have a positive influence. This research analyzed task types by score (Jehn, 1995). However, no basis was found for the additional analyses considered by the author, who classified tasks as routine and non-routine, which may have led to the non-significant relationship between task types and cognitive conflicts. It is also possible that there is a lack of clarity about the existence of cognitive conflicts, given the low values attributed by cooperative managers to assertions such as "I disagree about the way things are done in my work group."
H1b presumes task interdependence's direct and positive influence on cognitive conflicts, which presented statistical support. The interdependence of tasks means that individuals need to reconcile their duties while seeking to optimize relationships in the workplace to generate positive results (Vidyarthi et al., 2014). This aligns with that proposed by Neck et al. (1996), who stated that task interdependence can influence conflicts. This implies continuous monitoring of managers regarding the interactions between individuals during their tasks, corroborated by the high values assigned to assertions in the questionnaire, such as "different ideas and opinions are expressed during the tasks", with an average of 5.57. Furthermore, the ambiguity of manager roles may be more prominent at cooperatives (Beuren et al., 2018), signaled in the survey by higher values for assertions, such as "members of my group debate or discuss different ideas about how things should be done."
H1c predicts a direct and positive relationship between mutuality among group members and cognitive conflicts, which was statistically confirmed. This finding is consistent with that of Amason and Sapienza (1997). Mutuality considers the degree of involvement with the contributions of each individual and the connected discourse; high mutuality is associated with the perception of positive aspects concerning collaborative learning experiences (Lai et al., 2016). Given the positive results, mutuality in the working group's tasks is inferred, mainly due to the benefits of creating conditions for generating conflicts expressed in the form of debates (Du & Xu, 2018). The need for rapprochement at cooperatives is triggered by complex and ambiguous tasks (Anzilago et al., 2018).
H2 conjectures that cognitive conflicts directly and positively influence information sharing and it is supported.
According to Du and Xu (2018) and Bedford et al. (2022), cognitive conflicts can boost information sharing. Good communication provides conditions for solving problems and generating new ideas, which can influence performance (Valiyeva & Thomas, 2022). Cognitive conflicts expressed in debates can be associated with greater information sharing, even generating more significant contribution interest by working group members (Tsai & Bendersky, 2016). The role of managers to promote an environment that offers conditions to ease communication is highlighted (Mannes et al., 2022). The higher values attributed by managers to assertions indicate a perception of fluency in the debate among group members in the cooperatives.
H3 assumes the existence of negative moderation in the relationship between cognitive conflicts and information sharing, which was not supported. This finding contradicts literature, which proposed that affective conflicts reduce the performance of cognitive conflicts in information sharing (Kakar, 2018; Lu et al., 2011). However, Esbati and Korunka (2021) highlight that affective and cognitive conflicts are rarely investigated in isolation, generating ambiguous results and indicating that affective conflicts are dysfunctional and cognitive ones are functional. It is possible that affective conflicts do not negatively affect cooperative processes due to the mutuality among group members and conflicts expressed in debates (Du & Xu, 2018). This argument is supported by Silva et al. (2022) for cooperatives, they state that managers debate on improvements and critical issues, create a common language, and unify the organizational vision around objectives.
H4 postulated a mediating effect of cognitive conflicts in the relationship of predictors with information sharing. As in direct relationships, there was no statistical support to task types (H4a). Therefore, it is conjectured that, despite the ambiguity of the roles of cooperative managers observed in the study by Beuren et al. (2018), the standardization of tasks may be an explanatory factor for the results of this research, that "there are specific standards that I must meet when doing my job." There was also no statistical support when considering the mediating effect of cognitive conflicts on task interdependence (H4b). However, the mediating effect of cognitive conflicts was noted in the relationship between mutuality among group members and information sharing (H4c).
No statistical evidence of the influence of the control variables (age and gender) on information sharing was found. The age of most respondents denotes little diversity and may have led to a non-significant relationship between age and information sharing. The gender variable may have presented a non-significant relationship with information sharing due to the predominance of male respondents. Gender diversity tends to favor group work and mitigate the impact of certain types of conflicts, in addition to the general characteristics of individuals which influence group dynamics (Díaz-García et al., 2013; Lee et al., 2018).
Given the above, it is worth noting that cognitive conflicts can be beneficial, including the alignment of individual and collective interests (Um & Oh, 2021). However, managers decide whether to assess the intensity and causes of conflicts in work groups to ensure that their efforts are not compromised by dysfunctional dynamics (Esbati & Korunka, 2021). They must promote conditions to prevent cognitive conflicts from resulting in affective conflicts and harming work group cohesion. The ambiguity of managers' roles (Beuren et al., 2018) encourages the evaluation of conflicts in the context of cooperatives.
This research reinforces the importance of debates in a cooperative environment. In group work, the bond of cooperative goals enables engagement in open discussions so that conflicts become a constructive force, and information processing occurs more judiciously (Du & Xu, 2018). Tension between individual and collective interests can undermine the effectiveness of group work in a cooperative environment (Lee et al., 2018). Results demonstrate that cognitive conflicts influence intragroup information sharing, which can impact the cooperatives performance (Anzilago et al., 2018).
Research results are permeated by aspects advocated by the Theory of Collective Action. Collective actions involve conflicts between individual and group rationality, so that individual trust and reciprocity potentially affect levels of cooperation and joint benefits (Pansera & Rizzi, 2020). This reveals the need to manage activities in the context of cooperatives, so that cognitive conflicts can generate positive impacts, including encouraging information sharing so that everyone can cooperate. Cooperation can lead to improvements in the management process, innovation, and adoption of new technologies for cooperatives (DeMarrais & Earle, 2017).
5 Conclusions.
This study investigated the role of conflict predictors and cognitive conflicts on information sharing, moderated by intra-work group affective conflicts, at cooperatives. Results demonstrate that task interdependence and mutuality among group members' conflict predictors have a positive and significant relationship with cognitive conflicts. In contrast, there was no statistical significance for task types. The same was found when analyzing the mediating effect of cognitive conflicts in this relationship. Furthermore, cognitive conflicts exert a positive influence on information sharing. Unlike those observed in the literature, the moderation of affective conflicts in the relationship of cognitive conflicts with information sharing did not present statistical significance.
Given the findings, cognitive conflicts can be beneficial in the observed context. They can generate conditions to minimize communication problems and those resulting from divergent thoughts when expressed in the form of debates. Thus, conditions are created for group work to achieve common goals, which leads to greater information sharing and improvements in decision-making to boost individual, group, and organizational results. However, this requires creating an appropriate environment for internal debates and attention to the types of conflicts and how to manage them.
Results support the considerations of the literature regarding the association between conflict predictors and cognitive conflicts, as well as the influence of cognitive conflicts on intragroup information sharing. The study offers new perspectives regarding the proposed relationships and considers a little explored field in management literature. Given the idiosyncrasies and relevance of cooperatives, especially in a social and economic perspective, there is a need for more studies that analyze this segment. In fact, literature addressing this context is relatively scarce compared to other types of ownership structures.
Previous studies have analyzed both positive and negative aspects of cognitive conflicts, to which this research adds; however, the non-convergent results of this research encourage further investigations. The study highlights the need to understand conflicts in different organizational environments to better manage their impact on individuals and the organization as a whole. In the cooperative context, it is necessary for conflicts to be positive, in the sense of providing conditions for everyone to work towards common goals, thus complying with cooperative principles.
This study also addresses information sharing in work groups. Managers must create conditions to facilitate communication between work group members to obtain results favorable to the group and the organization. In this way, this research reinforces the importance of debates in a cooperative environment. It is thought that information sharing provides advantageous exchanges in order to offer new insights and improve group and organizational performance.
The results presented here have limitations. Initially, the focus was on cognitive conflicts. Future research may consider other forms of conflicts or different approaches between cognitive and affective conflicts. Other internal and external issues can impact the conduct of cognitive conflicts at the individual, group, and organizational levels. Future research may also consider other elements, such as the influence of size, management styles, and organizational structure on generating different degrees of conflict and information sharing. Limitations still stem from the method, such as the common method bias, since all data were collected by survey, in which a single respondent reported on the dependent and independent variables. We recommend considering other research strategies, such as interviews and document analysis. Qualitative approaches can help deepen the understanding of cognitive and affective conflicts and information sharing.
