1 of 19

Why Do Team Leaders Perform Better than Team Members? Evidence from an Experiment

Nguyễn Việt Cường

Mekong Development Research Institute; and National Economics University

2 of 19

Content

Background
Objective
Experiment design
Results
Conclusions

3 of 19

Students from National Economics University two years after graduation

4 of 19

Background

There is a vast literature on team performance and the role of team leaders (e.g., Durham, 1997; Heckman and Wageman, 2005).
Most studies focus on the characteristics of team leaders on team performance (e.g., Sarin and McDermott, 2003).
Team leaders often performance better than team members. However is there another way around?

5 of 19

Objective

Using the experiments with students in universities, this study aims to examine several reasons why team leaders can perform better the team members:

The effect of simply being a team leader on performance.
The effect of being expected.
The effect of having more responsibility.

6 of 19

Experiment design

The experiments were conducted in courses for seven classes in two universities: Trade Union University and Foreign Trade University in Hanoi, Vietnam.
The total number of students is 544.
The courses with the experiments are micro-economics (four classes studied this course), macroeconomics (two classes studies this course), and project management (one class studies this course).
All the courses were given by the same one teacher.

7 of 19

Experiment design

For each class, students are divided randomly into teams. Each team consists two, or three or four students. The number of teams of different sizes is equal in each class.
Each team on average contains three students.
The total number of teams is 179, of which 59 teams have two members, 60 have three members and 60 have four members.
179 members are randomly assigned as the team leaders.

8 of 19

Experiment design

In each class, three teams with the highest average score of the midterm exam received commend and a small prize (100 thousand VND) from the teacher.
Similar, three teams with the highest average score of the final exam also received commend and a small prize (100 thousand VND) from the teacher. However, one week before the final exam:

50% of team leaders received an email from the teacher saying that they were trusted as the team leaders.
50% of team leaders received an email saying that their score is double in calculation of the average score of the team.

9 of 19

Regression of interventions

Explanatory variables	Dependent variables
Explanatory variables	Team leaders	Team leaders receiving trust	Team leaders having double weight	Teams with three members	Teams with four members
University entrance exam score	0.0049	0.0049	0.0008	-0.0205	0.0050
University average score	-0.0013	-0.0051	0.0127	0.0049	0.0808
Constant	0.2239***	0.0704	0.1194	0.7651***	0.1366
Observations	541	541	541	541	541
R-squared	0.001	0.001	0.001	0.017	0.015

*** p<0.01, ** p<0.05, * p<0.1.

10 of 19

Outcomes

Groups	Midterm exam score	Final exam score	Interest degree in the course
By team leadership
Team members	7.81	6.27	69.67
Team leaders	8.17	6.86	72.43
By team size
Two members	8.13	6.33	70.87
Three members	8.06	6.55	70.77
Four members	7.75	6.47	70.35
Total	7.94	6.47	70.60

11 of 19

Outcome: midterm exam score

12 of 19

Outcomes: final exam score

13 of 19

Outcomes: interest in the course

14 of 19

OLS regression of outcomes

Explanatory variables	Midterm exam score	Final exam score	Final exam score	Midterm exam score	Final exam score	Final exam score
Team leader	0.185**	0.270***	0.168*	0.182**	0.255***	0.154*
Constant	0.000	-0.000	0.000	-0.076	-1.443***	-1.830***
Control variables	No	No	No	Yes	Yes	Yes
Observations	511	527	526	511	526	526
R-squared	0.009	0.017	0.006	0.046	0.118	0.101

Control variables includes University entrance exam score and University average score

*** p<0.01, ** p<0.05, * p<0.1.

15 of 19

OLS regression of outcomes

Explanatory variables	Final exam score	Final exam score
Team leader
Team leaders receiving trust	0.2156**	0.3075**
Team leaders having double weight	0.2080*	0.2949**
Other team leaders	0.1004	0.1109
Team leaders receiving trust × Team leaders having double weight		-0.2516
Constant	0.0105	-0.0000
Observations	527	527
R-squared	0.018	0.020

*** p<0.01, ** p<0.05, * p<0.1.

16 of 19

OLS regression with interactions

Explanatory variables	Dependent variable is the final exam score
Explanatory variables	Model 1	Model 2	Model 3
University entrance exam score	0.0138	0.0141	0.0148
University average score	0.4988***	0.4804***	0.4579***
Team leader	0.6360*
Team leaders receiving trust		0.6407	0.2041*
Team leaders having double weight		0.1908*	0.3838
Other team leaders		0.1150	0.1145
Team leader × University score	-0.1490
Team leaders receiving trust × University score		-0.1698
Team leaders having double weight × University score			-0.0747
Constant	-1.5636***	-1.5135***	-1.4718***
Observations	526	526	526
R-squared	0.120	0.120	0.119

*** p<0.01, ** p<0.05, * p<0.1.

17 of 19

OLS regression with team size

Explanatory variables	Midterm exam score		Final exam score
Explanatory variables	Without interactions	With interactions	Without interactions	With interactions
Two members	Reference
Three members	-0.0362	0.0663	0.1030	0.3098*
Four members	-0.1916*	-0.0602	0.0670	0.3098**
Team leaders		0.3076**		0.5869***
Team leaders × Three members		-0.1663		-0.3675
Team leaders × Four members		-0.2254		-0.4276**
Constant	0.1613*	0.0089	0.0266	-0.2567**
Observations	511	511	527	527
R-squared	0.008	0.016	0.001	0.027

*** p<0.01, ** p<0.05, * p<0.1.

18 of 19

Conclusions

Simply being team leaders makes people feel more important and responsible. It can increase performance of team leaders compared to team members.
Higher trust or expectation from the teachers can increase the performance of team leaders.
More responsibility can also increase the performance of team leaders.

19 of 19

Thank you!