United, We Solve: The Effects of Social Bonding on Task Performance

Emily Dinnerman, Madison Holdredge, Erika Legernes, Sarah Slain

University of California, Santa Cruz

Abstract

This study explored whether level of bondedness in a partnership would influence the partnership’s efficiency in completing a team-based task. We also examined whether the gender-pairing types (i.e. male-to-male, female-to-female, female-to-male) effected a difference in their task performance. Randomly selected psychology undergraduate students at a public California university (N=32) were randomly paired and given a set of questions to discuss. Control group pairs received generic, straight-forward questions, where the experimental group pairs received intimate, introspective questions. The pairs were then instructed to complete as many Tangram puzzles together as they could in five minutes. We found no significant main effect for the level of bondedness when we compared the two groups using an independent sample t-test (p = .303.). However, we found that same-gender pairs, on average, completed more puzzles than the coed pairs. The results suggest that teamwork efficiency may not depend on bondedness levels between paired groups, but by the gender of the partnership.

Keywords: social bonding, task performance, teamwork, partnership, gender

The Effects of Partner Bonding on Task Completion

LinkedIn: The world’s largest online professional network grows at the rate of two users per second, and boasts a 4.5 out of 5 stars in employee satisfaction from the Glassdoor--statistics that exceed those of Google and Facebook (LinkedIn, 2015; Glassdoor, 2015). LinkedIn has clearly won its title alongside social networking moguls, Twitter and Facebook, but to what does the company owe its success? The answer is found in their mission statement: “Connect the world’s professionals to make them more productive and successful” (LinkedIn, 2015). The key word is connection. When sizing up from four to 7,000+ employees, connecting professionals within the company is just as essential as providing their service to the public. To nurture this connection, CEO Jeff Weiner created the ideal environment for the team-oriented company through open-office floor plans, monthly community service events, and designated hours for his employees (Weiner 2012). Weiner believed in free communication, trust and the healthy exchange of ideas--the foundation to social connection. Taking from Weiner’s example, the purpose of this research experiment becomes clear: we aim to provide members of team-oriented careers with greater incentive to form social connection. Thus, we experimented with varying degrees of social bonding to study their effects on teamwork productivity.

Past research supports the impact of environmental factors in the workplace on team productivity. A study by Curtis and Upchurch (2008) focused on factors that would create a fun and comfortable environment in hospitality and customer service jobs. Four thousand randomly-selected human resource professionals completed a comprehensive, six-part survey assessing the policies, practices, and employee needs that would create a fun work environment. The study showed that two factors, (1) attachment--the recognition of personal milestones and fun social events--and (2) involvement--like games, humor, opportunities for personal development, and friendly competition--were vital to achieving a strong service culture. Curtis and Upchurch illustrated that work-life quality translates forward in their service quality. Another study, by Sliter, Kale, and Yuan (2014), focused on how “coping humor”, or humor used as a coping mechanism to deal with trauma as a workplace stressor, between co-workers, buffered co-worker relationships. One-hundred seventy-nine firefighters were given two surveys in a pretest-posttest study. The results indicated that coping humor helped overcome symptoms of burnout and posttraumatic stress disorder (PTSD). Therefore, internal workplace factors, like coping mechanisms, may promote co-worker relationships and a healthier workplace environment, which is crucial to trauma-inducing professions, like firefighting, emergency medical responders, and the police department.

Social bonding is based on a sense of trust, team identity, and positive attitudes between co-workers in a workplace (Lee, Gillespie, Mann, & Wearing, 2010; Henttonen, Johanson, & Janhonen, 2013). In a leadership and trustworthiness study by Lee et al. (2010), 34 engineering teams were given survey questionnaires to express their feelings about their team members. The study revealed that increased willingness to disclose sensitive information, known as trust, lead to an increase in communication, team performance, and the team’s collective knowledge. Lee et al. offers insight on how trust plays a role in team communication and, thusly, task performance. Another study on social bonding, lead by Henttonen et al. (2013), examined the correlation between social network bonding and performance efficiency. The researchers discovered that increased levels of network bonding and shared “team identity” lead to increased teamwork performance. This study illustrates that more than one factor (e.g. “team identity” and social network bonding) can play a role in determining team-based task performance in the workplace.

While interpersonal skills help, external cues, such as race, gender, and age, also play a role in team-based task efficiency. Dumas, Phillips, and Rothbard (2013) conducted a survey study that assessed whether the level of professional bonding was influenced by racial differences between co-workers in an office. The findings concluded a positive association between relationship closeness within racial groups, but no correlation in relationship closeness between racial groups. These findings highlight the importance of designing the right kind of interactions for bonding between participants. So, we took into account the effects of gender-to-gender pairing against task performance in our experiment.

One study by Pain & Harwood (2009) indicated that the type of team-building exercises mattered for improving relationships between members. This study facilitated a number of open discussions within a soccer team after administering highly structured performance reviews to each player after each match. The results suggested that the intervention was effective in improving perception of team functioning, training quality, self-understanding, player ownership, and team performance. Players also developed positive opinions toward the intervention procedure, associating it with themes of honesty, trust, openness, and improved communication. In other words, team-building exercises that encourage open discussion significantly improve teamwork performance and attitudes between members. From this study, we learned to design our experiment around discussion-based activities in hopes of evoking a sense of social bonding between group pairs.

One significant application for our study on team bonding and productivity lies with surgical mortality rates. A health services study by Neily et al. (2010) conducted a surgical quality improvement program, which required surgeons to work and train as a team for two years, in 78 veteran hospitals across the United States. The results demonstrated a decline 50% greater in surgical mortality-rate in the trained group over the untrained group. This training program provides a legitimate example that bonded teams output higher task efficiency. Fortunately, bonding benefits can be extended to any field that involves team-related work: students in universities, Olympic athletes, cancer research labs, software engineering firms, etc.!

Research supports that communication between team members cultivates trust, positive attitudes, and a greater sense of team identity. From this fact, we predicted that the more teammates feel bonded to one another, the greater their group task efficiency will become. In our study, we manipulated “bondedness” levels for two groups of psychology research students and measured the effects on the groups’ task performance in a puzzle game of Tangrams. We wanted to explore whether stronger team “bondedness” affects team efficiency in a group task. Thusly, we hypothesized that, because members of a team may be more trusting and share more positive attitudes toward each other with higher levels of bonding, higher “bondedness” levels lead to greater efficiency in teamwork-based tasks.

Method

Design

We conducted a posttest only, independent design experiment. Our independent variable was the level of partner bonding--nonbonded or bonded. Our dependent variable was the number of tangram puzzles completed by each pair (from zero to six completed puzzles).

Though our main hypothesis did not focus on the gender of the pairs, we also took note of how many pairs were female-female, male-male, and coed to see if gender had an interaction effect on our data. With this other independent variable added, we looked at a 3 x 2 x 2 independent groups factorial experiment, with our two independent variables being the gender of the pairs and the amount the pair bonded, and our dependent variable being the number of tangram puzzles completed.

Participants

A total of 32 participants took part in our experiment on Thursday, May 28th, 2015 from 12:00pm-1:45pm. All participants were undergraduates enrolled in a Research Methods for Psychology class at the University of California, Santa Cruz. Each student in the class was both a researcher as well as a participant; half of the class conducted their experiments on Tuesday, May 26th, 2015 and used the other half of the class as their participants, while students switched roles for the following Thursday. Participants were selected randomly from the professor for which experiments they would take part in by sitting in the front rows of the lecture hall while the professor picked among them randomly based on how many participants a specific researcher needed.  Each participant was required to take part in at least seven experiments, and were awarded extra credit it they completed more. Among our own participants, 44% identified as male and 56% identified as female. Once selected, participants were randomly paired off by asking each person to pick a card that was face-down from a selection that was fanned out by one of our researchers. They were then instructed to find the other participant with their matching card (e.g. the participant with an “Ace” paired with the other participant with an “Ace”).

Materials

We arranged our 32 participants into pairs using cards numbered Ace-8 from two suits in a pack of regular playing cards. Every word spoken by the researchers was read off of our protocol sheet (see Appendix A) to ensure that the only difference among the experimental and control groups was the questions they discussed with one another before completing the puzzles. Each pair of partners was handed a piece of paper with a list of questions to discuss, depending on if they were part of the control group or experimental group (see Appendix B). Identical sets of aminated seven-piece tangrams were delivered to each pair along with a stapled packet of six different tangram puzzles, one on each sheet, which could be completed using all seven puzzle pieces (see Appendix D). Researchers were equipped with solution sets for the tangram puzzles to assess the completion and accuracy of the pairs’ puzzles (see Appendix E).

Procedure

Our experiment took place in a quiet hallway, with a long desk lining the wall and stools for participants to sit on. In front of each set of two stools were placed the question set for the correct group (bonding or nonbonding), a set of seven tangram puzzle pieces in a ziplock bag, and an upside-down puzzle packet. The experiment took place during one of our hour and a half psychology class lectures, where half of the class was conducting experiments, and the other half was participating in experiments. One of our researchers acted as a runner, who went from the location of the experiment to the lecture hall, where she would ask the professor to select eight participants randomly, and then bring back to the location of the experiment. This process occurred four times, with two groups of eight participating under the control conditions and another two groups of eight participating under the experimental conditions. As soon as a group of eight participants arrived, they were randomly divided into pairs using playing cards and told to sit down in ascending order on the wall (Ace, 2, 3, 4) with their partners. They were then told to spend three minutes discussing with their partner the questions provided to them (see Appendix B). The questions were the only part of the experiment that varied between control and experimental conditions. After the three minutes, each pair was given five minutes to complete as many tangram puzzles as possible. Two researchers were each given two pairs to watch over, so they would check if a puzzle was correctly solved once a pair raised their hands after working it out. If the puzzle was completed correctly, the researcher would say “correct” and record the puzzle as completed next to the pair’s identity on the Record sheet (see Appendix C). If the puzzle was not solved correctly, the researcher would say “That is not the correct solution”, and the pair would continue to solve the puzzles. After the five minutes, the participants were dismissed and brought back to class by the runner, with the other researchers setting up the experiment for the next set of eight. This process would repeat until all four sets were completed.

Results

                Our hypothesis was that more socially bonded individuals would be able to complete more puzzles in a given time frame over less socially bonded pairs. Among the 16 pairs involved in this experiment, eight contained both a male and a female, three were both males, and five were both females. Our experimental group and our control group contained eight pairs each. The average number of puzzles solved among all 16 pairs was 3.19 out of a possible 6 puzzles (SD = 1.8). As illustrated in Figure 1, the average number of puzzles solved in the control group was 3.25 (SD = 2.19), while the average number of puzzles solved in the experimental group was 3.13 (SD = 1.46). There was not a significant main effect for the level of bondedness when we compared the two groups using an independent sample t-test, t(14) = .134, p = .303. Thus, because this result does not support our hypothesis, teamwork efficiency may not depend on bondedness levels between paired groups. However, some gender differences were noted: among the mixed-gender pairings, the average number of puzzles solved was 2.75 (SD = 1.75), whereas the average number for male-to-male pairs was 3.67 (SD = 2.52) and 3.6 (SD = 1.67) for female-to-female pairs, as illustrated in Figure 2. Although these results among the gender pairings were not statistically significant, the outcome that same-gender pairs, on average, completed more puzzles than mixed-gender pairs is still worthy of noting.

Discussion

Our study examines the relationship between social bondedness and task performance efficiency.  We predicted that participants who were more socially bonded would complete more puzzles within a given timeframe.  We hypothesized that social bondedness would positively correlate with task completion efficiency.

Our results did not support this prediction. There was no significant relationship between bondedness and task performance. The total puzzles completed by more socially bonded pairs (experimental group) was just one more than the less bonded teams (control group). We also found that same-gender pairs completed slightly more puzzles than coed pairs, which positively correlates with the hypothesis that social bondedness improves task efficiency among teams.

Every individual has varied propensities to trust or bond with others. Depending on their life experiences, cultural background, socio-economic factors and personality types, they may exhibit a different response to the exchange of information part of the experiment. Levels of closeness could have varied depending on answers to questions.

Trust may not be a deciding factor in a team’s success, but it may directly influence a pair’s motivation and attention toward shared goals.  It is worth mentioning that trust was probably not formed during the stint of more intimate questions. We cannot ensure that “team identity” was formed within this short timeframe either.

Some of the discussion questions for the more socially bonded pairs included personal achievements and individual values.  Based on previous research, these questions should have helped build a stronger social environment, though, not necessarily, team task efficiency.  The more socially bonded pairs also engaged in more laughter, supported by above the research, that bondedness buffers stress and facilitates working relationships.

The fact that same-gender pairs completed more puzzles on average than the mixed-gender pairs suggests that perceived shared backgrounds may have contributed to faster task completion.  As mentioned, those with similar backgrounds developed closer bonds; so it is safe to assume similar application with same-gender teams.

Even though we did not find a significant effect, research widely supports the power of bonding on team relationships’ ability to perform better with a sufficient bond.

There are multiple possible reasons we achieved our results.  The environment may have played a role—if participants were isolated from other teams, confounds, like social comparison and anxiety could have been reduced.  Also, repeated exchanges would have been more effective in establishing social bondedness and producing improved task performance than the content of the information itself.  For instance, teammates could meet two or three times and exchange information.  Variables, like social anxiety and levels of attraction, are one of the many contributors that may have altered team performance.

The team’s performance could have been affected by feelings toward one another after sharing certain pieces of information.  If a participant’s newly formed concept of his or her teammate is negative, this could have a slowing effect on the team’s ability to perform a task efficiently together.  The fact that every individual has different background and propensities is a confounding variable that is difficult to control for.

We could have improved construct validity by administering anonymous surveys after the experiment to better understand how individuals viewed their teammate and further evaluate our results.  Some teams may have had previous exchanges prior to the experiment, affecting the study’s internal validity.

In a future study, pairs could meet for interims of different length and multiple times to strengthen their relationships.  If these teams performed separately from others and are collected more randomly, significant results may be yielded.  This study could investigate reciprocal effects between trust and sharing of knowledge, and explore the specificity of the answers and how they relate to the performance.

Since bonding and building relationships changes the dynamics of a team, it is important to shed light on the processes that mediate performance effectiveness and understand the relationship between team dynamics and performance levels.  In order to improve performance a manager may want to focus on bridging social connectedness between teams.  As discussed earlier, these effects are utilized by noteworthy businesses, though they can be applied to medicine, academics, sports teams and any organization striving for success.  

References

Aron, A., Melinat, E., Aron, E. N., Vallone, R. D., & Bator, R. J. (1997). The experimental generation of interpersonal closeness: A procedure and some preliminary findings. Personality and Social Psychology Bulletin, 23(4), 363-377.

Curtis, C. R., & Upchurch, R. S. (2008). A case study in establishing a positive service culture: Attachment and involvement in the workplace. Journal of Retail and Leisure Property, 7(2), 131-138. doi: http://dx.doi.org/10.1057/rlp.2008.5

Dumas, T. L., Phillips, K. W., & Rothbard, N. P. (2013). Getting closer at the company party: Integration experiences, racial dissimilarity, and workplace relationships. Organization Science, 24(5), 1377-1401. doi: http://dx.doi.org/10.1287/orsc.1120.0808

Glassdoor. (2015). LinkedIn. Glassdoor, 1. Retrieved from http://www.glassdoor.com/Reviews/LinkedIn-Reviews-E34865.htm

Henttonen, K., Johanson, J., & Janhonen, M. (2013). Work-team bonding and bridging social networks, team identity and performance effectiveness. Personnel Review, 43(3), 330-349. doi: http://dx.doi.org/10.1108/PR-12-2011-0187  

Lee, P., Gillespie, N., Mann, L., & Wearing, A. (2010). Leadership and trust: Their effect on knowledge sharing and team performance. Management Learning, 41(4), 473-491. doi:10.1177/1350507610362036

LinkedIn. (2015). About us. LinkedIn, 1. Retrieved from https://www.linkedin.com/about-us

Neily, J., Mills, P., Young-Xu, Y., Carney, B., West, P., Berger, D., … Bagian, J. (2010). Association between implementation of a medical team training program and surgical mortality. The Journal of the American Medical Association, 304(15), 1693-1700.

Pain, M., & Harwood, C. (2009). Team building through mutual sharing and open discussion of team functioning. The Sport Psychologist, 23(4), 523-542.

Sliter, M., Kale, A., & Yuan, Z. (2014). Is humor the best medicine? The buffering effect of coping humor on traumatic stressors in firefighters. Journal of Organizational Behavior, 35(2), 257-272. doi: http://dx.doi.org/10.1002/job.1868

 Weiner, J. (2012). From vision to values: The importance of defining your core. LinkedIn, 1. Retrieved from https://www.linkedin.com/pulse/20121029044359-22330283-to-manage-hyper-growth-get-your-launch-trajectory-right

Appendix A
Social Bondedness vs. Task Efficiency Protocol

Subjects and Intro:

1. Sarah Runner: Retrieve 8 randomly selected students from the class and bring them to Jack Baskin Engineering 2 hallway.
2. Madison Reader:

1. “For this experiment, you will be randomly paired with a partner and be given a few minutes to get to know your partner with some given questions. Afterwards, you will work on a puzzle task with your partner in a game called Tangrams.”

Divide into pairs:

1. Madison Reader:

1. “Now please choose a card from the shuffled deck from Erika **point to Erika** and find the participant who also shares your card number. This person will be your partner for the experiment. For example, if you get an Ace, you will work with the participant who also gets an Ace from the deck. Once you’ve found your partner, see Sarah **point to Sarah** who will give you a slip of paper with some questions on it. Hold onto that until everyone gets one.”

2. Erika Cards spreads out cards and lets everyone choose. Sarah Paper hands out questions on slips of paper for either (1) experimental or (2) control group.

Bonding Time:

1. Madison Reader:

1. “Go through each question and provide an honest, thorough answer to your partner. You have 3 minutes for this task. Begin on my mark.”

2. Madison Timer: “Go” and start the timer, and announce when time’s up.

Tangram Time:

1. Madison Reader introduces Tangrams:

1. “Now that you’ve gotten to know your partner, you will work together on solving pattern puzzles in a game of Tangrams.”

2. Erika and Sarah hand out bags and patterns while Madison instructs.
3. Madison Reader:

1. “In this game, you have 7 pieces to form the patterns given on the paper. You can rotate the pieces for the desired shape, but none may overlap each other. All pieces must be used to make the patterns. All patterns can be solved with the black side up. Any questions about the game?”
2. Allow participants to ask questions about the game.

4. Erika and Sarah get solution sheets ready.
5. Madison Reader:

1. “Please open your bags and prepare your pieces. You have 5 minutes to work through as many patterns as possible with your partner. When you think you’ve solved a pattern, raise your hand, and an experimenter will come to check your pattern. If they give you the ‘OK’, you may move onto the next pattern. Please leave your cards visible next to your workspace so the experimenters may record your group data. Also, if you can’t solve a pattern, you can start another. We will let you know the half-way and last minute mark. Any questions?
2. Allow participants to ask questions about the procedure.
3. “Begin on my mark.”

6. Madison Timer: “Go” and starts timer, announces the 2:50 and 4:00 minute marks. Calls time at 5 minutes.
7. Erika and Sarah check solved puzzles, record team # and tally for each solved puzzle onto the Records sheet. Madison stands by to answer questions from participants and keep time.
8. Madison Reader:

1. “Alright, time’s up! Please hand all things to us (experimenters) and make your way back to the classroom. Thank you for participating in our experiment!”
2. Erika and Sarah collect experiment material.

Appendix B

Control Group Questions:

Take each question at a time so that you and your partner answer them before moving on to the next. It’s ok if you don’t get through all of them.

• What’s your name?
• How old are you?
• What year are you?
• Where is your hometown?
• What is your major?
• Do you live on or off campus?
• How many siblings do you have?
• What was your high school mascot?
• Did you or do you have braces?
• How many roommates do you live with currently?
• Why did you choose this major?

Experimental Group Questions*:

Take each question at a time so that you and your partner answer them before moving on to the next. It’s ok if you don’t get through all of them.

• Would you like to be famous? In what way?
• Do you have a secret hunch about how you will die?
• For what in your life do you feel most grateful?
• Given the choice of anyone in the world, whom would you want as a dinner guest?
• If you wake up tomorrow having gained one quality or ability, what would it be?
• What is the greatest accomplishment in your life?
• What do you value most in a friendship?
• Share 2 things you consider a positive characteristic of your partner.
• What would constitute a “perfect” day for you?

*The experimental group questions were taken from the list of 36 questions developed to create a close connection between blind date couples (Aron et al., 1997).

Appendix C

Team Tangram Records

Experimental Group

Control Group

Appendix D

Tangram Patterns

Appendix E

Tangram Solutions

Appendix F

Figure 1. Correct number of puzzles solved for both the experimental (bonded group) as well as the control (non-bonded) group.

Figure 2. Correct number puzzles solved across all three gender-paired groups: mixed (one male and one female), two males, and two females.