In Search Of Diverse And Connected Teams: A Computational Approach To Assemble Diverse Teams Based On Members Part 8

Another contribution of this work is finding similar team combinations that individuals would assemble but with enhanced diversity levels. As prior studies have found, people tend to form teams with competent individuals and those who are familiar with them, enhancing the likelihood of satisfaction and commitment to the team [28, 82]. 

Teamwork is a very important ability in modern society. It allows multiple people to work together to promote the successful completion of a project or task. Memory is also a very important quality in a team. So what's the relationship between team composition and memory?

First, the impact of team composition on memory is obvious. Different people have different skills and expertise, and through different combinations, the overall capabilities of the team can be maximized. When each member maximizes his/her strengths, the efficiency of the team can be maximized, which can also stimulate the enthusiasm and initiative of each member, thus helping to improve the team's memory.

Secondly, communication in a team is also very helpful in improving memory. In a team, frequent communication and collaboration are required between members, which promotes communication and allows everyone to better understand all aspects of the project or task. Deepening knowledge and understanding of tasks through communication can greatly improve memory.

Finally, team atmosphere also has a certain impact on memory. A positive and passionate team can help members maintain a good attitude and make everyone more engaged and focused on projects or tasks. This positive atmosphere is very beneficial to improving memory.

In summary, there is a connection between team composition and memory. Through reasonable combination, good communication, and a positive atmosphere, the team can exert its maximum synergy, thus improving the overall memory. Therefore, good teamwork is essential, which can promote the development of individual abilities and the development of the entire team, which is also one of the important factors for the success of modern enterprises. It can be seen that we need to improve memory, and Cistanche deserticola can significantly improve memory, because Cistanche deserticola can also regulate the balance of neurotransmitters, such as increasing the levels of acetylcholine and growth factors. These substances are very important for memory and learning. In addition, Meat can also improve blood flow and promote oxygen delivery, which can ensure that the brain receives sufficient nutrients and energy, thereby improving brain vitality and endurance.

Click Know short-term memory how to improve

This fact is demonstrated in the MyDreamTeam dataset by comparing the lower communication costs of self-assembled teams and the higher communication costs of randomly generated teams. 

The proposed algorithm found team combinations with lower communication costs than self-assembled teams, suggesting that people have some intuition in forming well-connected teams. 

However, they lack reliable knowledge of higher-order connections among themselves. A possible explanation of this difference found by the algorithm is the tremendous challenge for individuals to discover and take advantage of indirect connections, such as shared contacts or shared past collaborators. 

Whether individuals assemble their teams or team builders design them, considering team members' indirect connections is not an easy task since indirect connections are not highly visible. 

In contrast, our algorithm excels in considering the broader social network structure given the global view of relationships between members. By using this algorithmic approach, individuals and managers can be more conscious of potentially diverse teammates through their current relationships. 

Even if two team members do not know each other directly, teaming up with a shared "friend-of-a-friend" or indirect connection can potentially promote familiarity and psychological safety in teams [83–85].

Furthermore, we found that MyDreamTeam self-assembled teams were less diverse than the teams randomly generated by the algorithms. This tendency driven by homophily is consistent with prior literature, indicating that people prefer to team up with others who share similar characteristics [65]. 

Formulating this team formation problem provides new opportunities to boost team diversity over self-assembled teams while still considering high familiarity among team members. One main advantage of forming teams in this fashion is reducing individuals' biases. 

Since people naturally draw toward forming teams with similar individuals, an algorithm like the proposed one can augment people's decision-making process. 

Instead of connections driven by individuals' preferences, the algorithm can enact collective coordination by curating better team combinations that could satisfy individuals' expectations. This multiobjective approach can allow people to find feasible solutions that increase diversity without compromising familiarity in the team.

Implications

This work provides theoretical implications for team research. In particular, the use of computational mechanisms to support the team formation processes. Literature has characterized team formation as centered on behavioral mechanisms, where teams can be assembled by internal or external forces and based on similarity, familiarity, and competence [28, 86]. 

By formulating and implementing this multi-objective optimization problem, we found diverse and connected team combinations that individuals could not have foreseen. This work allows team scholars to reflect on the role of technologies in enabling new organizational structures among individuals and organizations, which could lead to new theories of team formation and the introduction of technologies [38–40].

Practical implications of this study contribute to several communities invested in increasing team diversity: managers assembling effective and diverse teams, instructors composing more diverse student teams, companies forming heterogeneous groups from different business units, space agencies such as NASA forming composing space crews for long duration space exploration to Mars, and researchers investigating the use of algorithms for organizing scientific teams. 

Expanding the use of this algorithm to broader audiences can provide new benefits for groups that seek to embrace diversity and keep high familiarity levels. Furthermore, software developers and designers can use this study's implications for new procedures and guidelines for artificial intelligence in organizing workers. 

Finally, this work provides more computational approaches to enrich team formation processes [45, 87]. 

Since team builders cannot solve this problem quickly by manually checking each team combination, algorithms can automatize this task by bringing together members who possess existing social connections and, at the same time, have different backgrounds, characteristics, and expertise levels [41, 42]. We expect this work will assist in forming heterogeneous teams by considering diversity and social networks.

Another quality of this approach is adding more objectives to the team formation problem. For example, team builders could minimize other objective functions such as geographical distance among participants, personnel costs, or availability constraints. 

Likewise, this multiobjective problem can accommodate members' traits when diversifying is not desirable. As some prior meta-reviews indicate [14, 88], having a team with similar individuals may be desirable for low-difficulty tasks or when efficiency (rather than creativity) is the goal. 

Furthermore, it may be desirable for some traits such as personality or expertise to be similar rather than diverse [89]. 

This team formation problem can add another objective function that minimizes teams' diversity in some traits using the metrics defined by Harrison and Klein [30]. Therefore, one potential use of this algorithm is to maximize diversity in some members' attributes while minimizing diversity in other attributes.

Given this multi-objective approach's flexible trade-off, which solution should team builders consider from the Pareto front? Incorporating other metrics (e.g., individual performance, team cohesion, members' location) could help team builders select one specific team combination.

Limitations and future work

It is important to acknowledge the limitations of this paper. First, the measures for diversity and communication costs were scaled specifically to each unique network and cannot be compared across different sets of participants. 

Second, the diversity measure is an aggregate of multiple diversity metrics for each attribute sampled; thus, it is difficult to assign any real meaning to the diversity metric apart from relative differences within the same network. Future implementations should consider how different diversity measures can be analyzed separately and according to the specific pool of participants. 

These might also weigh diversity on various dimensions or operationalize diversity metrics as different objective functions in the optimization problem. Third, forming scientific teams and software teams is more complex in reality: new members can be added over time, some specialization is required, not all of these teams share the same objectives, sizes, or restrictions, and diversity may be beneficial for only goals. 

We believe using the last two datasets should not be a concern because we use them only to test the algorithms' efficiency and results. This team formation algorithm can guide the formation of real scientific and software teams by finding more diverse and connected combinations. Fourth, we do not provide specific recommendations for demographic or functional diversity attributes. 

Prior studies have shown how the effects of diversity on team performance are mediated by contextual factors and team processes [14]. Team builders who want to administer this algorithm should reflect and decide on adding demographic and cognitive variables according to their organizational goals and particular context. Fifth, collecting social network data could be a big task for team builders. 

Assessing people's relationships can be done by conducting surveys, examining communication networks, or tracing digital data [90]. Another potential strategy to build individuals' social networks is asking about their teammate's preferences. 

The algorithm could find possible diverse team combinations based on individuals' responses [91]. Lastly, it is not possible to guarantee that the performance of the teams assembled by this algorithm will be better than other team formation strategies. 

Prior studies have shown mixed results for the direct effect of diversity on team performance in all contexts [14], as well as the advantage of using algorithmic approaches for team formation [92]. 

Other research has also shown that when individuals lack agency to self-assemble teams, they are less committed to their group [93, 94]. Future work should consider using this algorithm to assemble real groups and evaluate how well they perform compared to teams assigned randomly or by a manager.

Future work should add new restrictions to the multi-objective function, such as considering specific task roles in the teams, adding leaders to each team, or excluding certain team combinations in which some individuals do not want to work together. 

Using weighted networks could also provide more nuanced information about the strength of people's social relationships. One potential application is distinguishing individuals who have frequent interactions from those who barely speak to each other [95]. One example of potential areas of improvement is developing an automatic tuning for the weights assigned for each diversity attribute given a specific population. 

If the algorithm explores people's categorical and numerical attributes before conducting the team formation process, it could identify the attributes that have more variation and those that are scarce among individuals. 

Then, the algorithm could define the importance of each diversity attribute in the objective function. Lastly, the algorithm could be implemented as a web platform to provide more opportunities for managers, instructors, or researchers to assemble teams.

Conclusion

This work addresses the problem of assembling teams from a social network that maximizes both diversity and familiarity. We formulated a multi-objective function for this problem and implemented a genetic algorithm to find well-connected diverse teams. In a thorough experimental evaluation, we evaluated the performance of our proposed algorithm and compared it against baseline approaches. 

We discussed the potential role of algorithms in augmenting team composition and helping team builders. In particular, computational approaches can be used to form teams that consider indirect connections and recommend combinations with higher diversity scores. As algorithms can discover more feasible team combinations than humans, team builders' decisions can become more structured, systematic, and comprehensive.

Supporting information

S1 File. Supporting figures and tables. S1 Fig: Simulations using the Diameter metric. S2 Fig: Simulations using the Minimum Spanning Tree (MST) metric. S1 Table: Diameter Case. S2 Table: Minimum Spanning Tree Case. S3 Table: Team combinations' average proportion of hops. (PDF)

References

1. Ng ESW, Burke RJ. Person-–organization fit and the war for talent: does diversity management make a difference? The International Journal of Human Resource Management. 2005; 16(7):1195–1210.https://doi.org/10.1080/09585190500144038 

2. Hunt V, Layton D, Prince S. Diversity matters. McKinsey & Company. 2015; 1(1):15–29. 

3. Jackson SE, Joshi A. Work team diversity. In: APA handbook of industrial and organizational psychology, Vol 1: Building and developing the organization. American Psychological Association; 2011. p. 651–686. 

4. Barak MEM, Travis DJ. Socioeconomic trends: Broadening the diversity ecosystem. In: The Oxford handbook of diversity and work. Oxford University Press; 2012. p. 393. 

5. Mathieu JE, Hollenbeck JR, van Knippenberg D, Ilgen DR. A century of work teams in the Journal of Applied Psychology. Journal of Applied Psychology. 2017; 102(3):452. https://doi.org/10.1037/ apl0000128 PMID: 28150984 

6. Williams KY, O'Reilly CA III. Demography and Diversity in Organizations: A Review of 40 Years of Research. Research in organizational behavior. 1998; 20:77–140. 

7. Van Knippenberg D, De Dreu CK, Homan AC. Work group diversity and group performance: an integrative model and research agenda. Journal of Applied Psychology. 2004; 89(6):1008. https://doi.org/10. 1037/0021-9010.89.6.1008 PMID: 15584838 

8. Page SE. The diversity bonus: How great teams pay off in the knowledge economy. Princeton University Press; 2019. 

9. Vasilescu B, Posnett D, Ray B, van den Brand MGJ, Serebrenik A, Devanbu P, et al. Gender and Tenure Diversity in GitHub Teams. In: Proceedings of the 33rd Annual ACM Conference on Human Factors in Computing Systems. CHI'15. New York, NY, USA: Association for Computing Machinery; 2015. p. 3789–3798. Available from: https://doi.org/10.1145/2702123.2702549.

For more information:1950477648nn@gmail.com