News
Human Resources: Creating high performing teams in a pulp mill

As organizations are becoming more team-based and leaders are continuously striving for high performance, how then do you create and sustain a high performing team? This article describes an action-research project designed to answer that question…

As organizations are becoming more team-based and leaders are continuously striving for high performance, how then do you create and sustain a high performing team? This article describes an action-research project designed to answer that question. The author took a model developed by J. Richard Hackman (2002) and tested it in the Alberta Pacific Forest Products (Al-Pac) mill in central Alberta.¹

Chances for the effective performance of a team will increase if certain conditions are in place, according to Hackman. He states that there are five main conditions, namely; Team Task, Compelling Direction, Supportive Context, Enabling Structure and Expert Coaching that influence the effectiveness of a team.

Applying the theory

Al-Pac’s original strategy was to build a plant that was designed on a team-based structure. Ten years ago, the start-up managers created self-directed work teams with the intent that along with new technology, the Al-Pac project would become world-class. Teams would save money, increase productivity and improve communications. A team design would produce better resource utilization and better decision-making. A team structure would result in increased job satisfaction. The combination of technology, organization design, and good leadership in such areas as community/aboriginal relations, and environmental practices, has resulted in the Al-Pac operation exceeding expectations. Production has increased each year and Al-Pac holds world level daily output records for a single line pulp machine. It holds records for total compensation dollars per ton of pulp produced. They have a secure fibre supply, strong relationships with their aboriginal neighbors, and a full market for all their production. They have one of the lowest costs per units in the North American industry. The technical, environmental, and community accomplishments of Al-Pac are well-documented. Strategically it was important for Al-Pac to know how to create and sustain effective performance in the teams in their core area in order to continue to enhance the organizational performance.

In Al-Pac, training is a major part of the culture. More hours and dollars are spent per team member developing people than would normally be found in the forest industry. For example, resources are expended to train all team members in team functions and team skills. All team members attend training in team skills. There are six subjects taught, namely: team interactions, learning from conflict, problem-solving, direct dealing, coaching, and conducting team meetings. Each training session is half a day. After the classroom training, all team members are required to complete a process that demonstrates the application of the various skills / knowledge learned in the classroom. The belief is that if all the team members have completed their team skills, classroom training, and application activity, that the team will be more effective.

Much engineering analysis has been completed (and published in Pulp and Paper Canada), testing for the relationship of such physical science variables as water temperature, and pressure on the quality and output of a tonne of paper or pulp. Behavioral science research, on the other hand, is relatively less common in the industry. This study is intended to address this research gap by addressing the question, ‘ what are the conditions in a Pulp operation that relate to achieving a high performing team?’ In particular, how do Hackman’s conditions and team skills affect the team outcomes?

As depicted in the figure above, Hackman argues that a leader’s role is to address certain conditions. The first is to create a real team rather than a team in name only. Then, there must be a compelling purpose, and finally an enabling structure. Next the leader should provide a supportive context, and expert coaching. Managing these five conditions will create the likelihood of an effective team, which performs well in client satisfaction, team growth and individual member learning.

Hackman describes sub-conditions for each of his main conditions.

Hypothesis

1. That the values for Hackman’s main and sub-conditions are not the same across the teams.

2. That Hackman’s conditions have a positive effect on the effectiveness of a team.

3. That other variables like; team skills, attendance, or team development activities, have a positive effect on the effectiveness of a team.

The method

The study examined the relationship between the various team condition variables and the outcome variable, EFFECT, which measures team effectiveness along the three Hackman dimensions — client satisfaction, team growth and individual learning. The project involved 162 team members from 19 core production teams. Initial 81 interviews were systematically conducted to obtain terminology from team members that described team characteristics of high performance. An 80-question survey was designed from the interview data and the Hackman model. The data were entered into a statistical software program (SPSS). Reliability analyses were done on the multiple indicators for each of the Hackman sub-conditions (alpha ranging from 0.55 to 0.79). Analysis of variance (ANOVA) was performed to test the variations among the team conditions while correlations and regression analyses were conducted to identify the relationships between the various independent variables and the outcome variable.

The findings

Generally, the findings from this project supported Hackman’s model. The data show that the stronger the scores of some of the conditions, the better the teams’ effectiveness score.

The ANOVA results showed that other than for Supporting Context, there is significant variation across the 19 teams on all the main conditions and the team skills composite variable at the conventional level (p-value < 0.05). The p-value for the outcome variable EFFECT is marginally significant at 0.052. The significant statistics indicate that the null hypothesis of equal means across teams can be rejected. Hence, the evidence generally support hypothesis 1 that there is a significant variation across teams on these variables, except for the variable Supporting Context. This means that the teams were not all performing at the same level nor are they subject to the same conditions, as perceived by the members.

Correlation analyses were completed between each of the Hackman’s main conditions and the outcome variable, EFFECT. A similar correlation analysis was done for the team skills composite variable and the outcome variable. The Pearson correlation coefficients computed at both the individual level and at the team level (using average scores for the team members) are all significant at the conventional level indicating a strong correlation between each of Hackman’s main conditions and the outcome, as well as between the team skills and the outcome.

For example, the results for the correlations with Effect at the team level are:

Real Team r = .544, at Sig. .016

(2-tailed)

Compelling Directionr = .804, at Sig. .000

Enabling Structurer = .860, at Sig. .000

Supportive Contextr = .467, at Sig. .001

Expert Coachingr = .854, at Sig. .000

Team Skillsr = .710, at Sig. .001

Correlation analysis was also done on other team level variables computed from archival data. Neither the attendance, nor the Team Development Stage Assessment2 demonstrated any significant relationship with EFFECT or team safety (a measure of the level of safe behaviors of the team members). A business optimization program introduced by Competitive Edge Management3 was the only variable in this category found to be significant with EFFECT at the 0.05 level using a one-tailed test.

A set of linear regression analysis was systematically completed with the Hackman condition variables as well as team skills and demographic variables as independent variables and the team effectiveness variable EFFECT as the dependent variable. The sub conditions — task, challenging, design, reward system, problem-solving a
nd coaching skills — were found to be significant predictors.4 With an R-square of over 0.60, it means that over 60% of the variance in the dependent variable could be explained by these six predictor variables. Hence, for predictive purposes, one should be able to arrive at a fairly good estimate of the outcome measure by knowing the scores for these six significant sub conditions, without the need to have the measures of all the sub conditions covered in the Hackman model. The regression formula for estimation is:

Predicted EFFECT value = 0.196 + 0.121 x Task value + 0.176 x Challenging value + 0.206 x Design value + 0.188 x Reward System value + 0.173 x Problem Solving value + 0.148 x Coaching value.

As an example, a one unit increase in Task value (say, from 3 to 4 out of a 5-point scale), will be associated with a 0.121 unit increase in the predicted Effect value, while a one unit increase in Design value will be associated with a 0.206 unit increase in the predicted Effect value. Such information may be helpful for the organization to determine areas of improvements on which to focus.

The regression analyses provide empirical evidence that the Hackman conditions are positively associated with team effectiveness. At least one sub condition under each main condition area was found to be significant except for Expert Coaching (which was also a significant predictor in a more detailed analysis where the sub condition outcome variable, Individual Learning, was used instead of the overall EFFECT). This finding lends support to Hypothesis 2 that all of Hackman’s conditions are positively related to team effectiveness.

As for team skills, Problem Solving and Coaching were found to be good predictors of team effectiveness. Hence, Hypothesis 3 is supported at least for these two skill areas.

Conclusion

The results of this project proved the hypotheses and support Hackman’s model that there are certain conditions which, when present, will increase the likelihood that one can create and sustain high performing teams. The project applied a theoretical framework from J. Richard Hackman in an action-research experiment in Al-Pac and found that there is a relationship between certain team conditions and developmental activities and the team outcome.

First, the project results showed that team effectiveness is a factor of certain conditions. By applying Hackman’s model to 19 teams in the pulp operations function, the results showed a positive relationship between effectiveness and the following predictive conditions: team task, a challenging direction, the design of the team, the reward system, expert coaching, and how the team members have applied two team skills competencies, namely; problem solving, and coaching.

Secondly, the results above should not discourage Al-Pac from continuing to peruse resources in some of the other Hackman conditions that were included in the study but not found specifically significant. The study found that there were strong correlations among the various Hackman conditions, however, for predictive purposes, the regression analysis could be effectively utilized to identify certain key factors, like the design of the work. This project will add value to the action planning of the nineteen (19) teams who participated in the study. As well, because Al-Pac strongly believes in continuous improvement, this research can help guide decisions on team design and performance, in other parts of the organization.

Leading teams

As Hackman said in his book5, the main message for leaders and members of work teams is to get in place the basic conditions that foster team effectiveness. Once these conditions are in place then leaders can make small adjustments to facilitate the group’s progress towards meeting their goal. Although proven in this experiment to be an effective tool in the Al-Pac pulp mill, there is every reason to believe that Hackman’s model can be of global significance and applicable to any business.P&PC

1 The author collected data for this study as an internal Al-Pac initiative. The analyses and results were also presented in an applied project that is part of the MBA program requirement at Athabasca University.

2 The Blanchard Corporation (1998) – The TDSA is a self-scoring tool to help teams identify their stage of development.

3 Competitive Edge Management – www.cem.ca

4 It should be noted that the analysis does not necessary indicate that other variables are not important as many variables are correlated to varying extent. What the analyses showed is that the significant sub condition variables found in that model can be used as good predictors.

5 Hackman, J.R., 2002, Leading Teams: Setting the Stage for Great Performance, Boston Mass. Harvard Business School Press.

At the time that this project was completed, Doug Moynihan was the Director of Human Resources for Al-Pac. He is currently a consultant with North West Training and Development and can be contacted at nwtd@tbaytel.net

The Model

The Conditions for Team Effectiveness

Expert Coaching

Team

Effectiveness

– Client satisfaction

– Team growth

– Members learn

Enabling

Structure

Compelling

Direction