International Journal of Appreciative Inquiry
The purpose of the field experiment was to study the effectiveness of a generative Appreciative Inquiry versus ‘traditional’ AI and a problem-solving approach in fostering generativity among organizational members. The results empirically supported the thesis that synergenesis was the most generative of all the approaches.
Download the Research Notes here or read below
May 2013
Generativity, as defined by Kenneth Gergen in 1978, is the capacity to challenge what is taken for granted and propose fresh alternatives for future action. Appreciative Inquiry (AI) is generative in several ways: it values the strengths of the organisation rather than looking at its weaknesses; uses a collaborative approach to facilitate dialogue; creates new possibilities for the future; and energises people to take action. As there have been very few, if any, studies systematically exploring the process of AI and what makes it generative, I was motivated to undertake this research. The research was conducted in a government agency with an attempt to help facilitate change in the organisation at the same time. Conducting AI-based research was both exciting and challenging, because the organisation was unfamiliar with the AI process and had traditionally depended on a deficit-based approach. It is also the organisation where I work.
Begin with the positive
Since it was a field experiment that I conducted in my organisation, I had to approach the study as a consulting project. For any change initiative to be successful, it is important to have the support and approval of the senior leadership. When I first started describing the research methodology, there was an immediate resistance to the idea of research within because of the apprehension around uncovering of internal secrets. I was able to convince the senior management about my research only when I told them that AI implies working on strengths of the people and the organisation. This focus on the positive and appreciation made my research sound harmless, enabling its acceptance.
I realised that, although I was researching how AI is more than positive, it is better to begin with the focus on positive, especially in an organisation that is new to the AI process or has always relied on deficit-based problem-solving approaches to change. The positive and strength-focused approach provides a new lens for organisational members who are accustomed to problem-finding ways of change.
Stakeholders
The next step in the research process was identifying stakeholders. Getting the whole system in the room is essential for change (Weisbord, 2004). For everyone in the organisation to support the change program, it is important for individuals to understand their unique contribution and how it fits in the overall success of the organisation (Ludema et al, 2003). AI literature provides guidelines on how to select stakeholders or participants for research. The rule of thumb is “anyone affected by the study”.
I thought I had included everyone, until I realised that a manager who had not been included was very upset and trying to impede the progress of the study. I learned that it was hard and sometimes impractical to bring the whole organisation together in a room, but good practice was asking the people in the room to identify who was missing and who else should be included in the study.
This helps in getting a good representation of the organisation. Also, consistent communication to the organisation about the change and getting feedback from them at various stages of the change process can be an alternative to including the whole system.
Logical positivist vs. social constructionist
I had used surveys to measure some of the constructs of my research. As a researcher trained in this positivistic mindset, I was very excited about administering surveys. But when participants have undergone an interesting AI session and are energised by the power of dialogue and ideas generated from the intervention, surveys can kill the spirit of AI. When I administered the surveys after conducting AI sessions, the energy level suddenly dropped. This may or may not affect the results, but the difference between story mode and a logical comprehensive mode is considerable.
The logical positivist approach considers knowledge to be valid when it is quantitatively measured and validated. However as AI is based on the social constructionist paradigm, I feel it is essential to consider methods other than quantitative surveys to describe the effectiveness of AI. The narrative method that is ingrained in the AI approach needs to be kept alive and be utilised for this purpose. Asking participants to keep diaries, writing stories about theirexperiences or drawing pictures could be alternatives means of measuring the effectiveness of AI.
It is hard to think of these approaches without worrying about the potential threats of subjective evaluations and biases. However there are ways to manage them, for example using multiple rater evaluations, rater training and so on. It puts the onus on us as AI researchers and practitioners to be more creative and rigorous in the ways we measure AI without affecting the enthusiasm of the participants.
Storytelling and working with stories
According to Yaeger and Sorensen (2001), one of the factors that matters most in the effectiveness of AI is storytelling. In my research, I found that storytelling is essential to make AI generative though I did not measure storytelling as a variable in the study. I was anxious about how a group of serious, analytically minded engineers would receive AI and synergenesis sessions.
To my surprise, they loved telling stories. I realised that, if the intervention is structured well, the topic of inquiry is of interest to participants and they want to do something about it, everyone enjoys storytelling. This taught me the value and art of creating a good balance.
As for the approach to working with stories, I compared (1) synergenesis (Bushe, 2010), (2) the classical AI Discovery phase and (3) a problem-solving approach. On the surface I did not find classical AI and synergenesis to be very different. Both engaged participants equally. But the results communicated a different story. I found that synergenesis was more generative: the listener writes the story, adding his/her own reflections, thus empathising more with the stories. The stories are used as a catalyst to generate ideas in the group, unlike classical AI where stories are used to understand mainly the root causes of success. The storytelling in synergenesis helps not only to generate new and innovative ideas but also changes the narrative held by the group.
Conclusion
As a researcher and practitioner of AI, I realised that not everything goes as planned. As an AI researcher, I faced emerging challenges, such as: gaining the support of the stakeholders at every stage; finishing the research in the given amount of time and resources; organisational structures, processes and other change initiatives that hampered the progress of the study.
My first session was so successful that I was confident that rest of the process would go without a glitch but, sooner rather than later, my illusion was shattered! In the next session, the participants refused to sign the confidentiality agreement, so I had to wrap up the session without any data. However, even though I was not getting data for my research, I used the opportunity to use AI questions to focus the attention on the positive aspects of the organisation and also introduce them to the change process.
Similarly, there were times when I wanted the inquiry to be longer than the allocated time, just because of the participants’ enthusiasm, but as a researcher I had to be consistent about the time given to all the AI sessions. I made up by having follow-ups after I finished my research. This taught me the value and art of creating a good balance between the required research rigour and the desired flexibility of AI sessions.
May 2017: The research revisited
In 2013, I conducted Appreciative Inquiry and problem-solving sessions in a large public transportation organization to understand if AI was a more generative process than problem solving. Seventy-six members participated in the study. The groups were divided among three interventions (synergenesis, classical AI and problem solving). In order to provide heterogeneity, each intervention was composed of employees from the four divisions of the company: operations, non-operations, exempt employees and non-exempt employees.
Identification of the need and design of the interventions
The senior management identified a need for an employee recognition program at the organization in order to engage employees, reduce absenteeism and improve performance. After the need was identified, I formed an advisory committee of twenty employees across the organization who had experience in designing such programs. The purpose of forming the committee was threefold.
First, the committee was important for increasing awareness among employees about the need for a recognition program and for reducing resistance about conducting an organization-wide change. Second, the advisory committee helped define the protocols for the three interventions. Third, the committee was used to encourage employees to participate in the intervention. The questionnaire consisted of:
- Best experiences of employee recognition at the organization
- Experience of communication that connects all levels of the organization
- Most inspiring team experience, and
- Dreaming an organization.
The first three themes were also used for synergenesis. The main themes for the problem-solving approach were:
- Reasons for failure of employee recognition program
- Possible ways of recognition
- Removing communication barriers, and
- Decreasing absenteeism to engage employees.
In other words, the affirmative approach of inquiry in the classical AI and synergenesis was converted to a deficit-based approach in the problem-solving intervention. The results indicated that compared to problem solving, AI was more generative, i.e., participants were able to come up with new ideas, challenge old ways of thinking and foster possibilities of a collective future.
The positivity of AI makes it attractive as an organization design (OD) intervention, particularly in an organization that is riddled with negativity and disgruntled employees. The culture of the organization at the time was that it was undergoing major changes, there were layoffs, the senior leadership was a revolving door and there was a fair degree of skepticism and distrust regarding leaders.
On the positive side, many employees had been with the company for more than 20 years and had historical knowledge of the company. The employees were proud of the old days, and “how it was before”. This unique mix of negativity regarding the present and pride in the past, I believe, made it ideal for an AI intervention. The questions in AI are focused around describing the “best” in the past. In the AI sessions, the participants had an opportunity to relive some of those best moments together with people with similar historical knowledge.
When the organization was undergoing change, the morale was low, but AI gave an opportunity to rethink, to view things in a positive light. Specifically, AI sessions helped them realize what they most appreciated about the company, why they were still with the company, and which positive aspects from the past could be brought to the present.
Three main lessons learned
You have to be mindful of the questions asked during any intervention. My research showed that a question focused on problems results in deficit thinking, whereas questions focused on finding the positive help people redefine the present and imagine a new social reality. I had asked participants to respond in writing to “What are your thoughts on having an employee recognition program in the organization? What should it include?” and repeated the same question after the sessions. Not knowing if I would find anything of significance, to my astonishment, when I analyzed the results, I discovered that majority of the participants had much more favorable opinions after AI than after problem solving for having an employee recognition program.
Also, after problem-solving session people did not write as much, and what they wrote had a pronounced “negative” sentiment. As David Cooperrider has said, change begins with the first question you ask. Problem solving does have a time and place, but if the goal of the initiative is for participants to reimagine the present and be more hopeful about the future, it is important to include generative questions.
It is possible to produce many generative ideas in short of amount of time with AI. Critics of Appreciative Inquiry complain that it takes several days to conduct an AI session and, for the same reason, sometimes companies are hesitant to commit to long sessions. My research used only a two-hour AI Discovery session and produced a significant number of generative ideas. In a classical AI, stories are generated in the Dream stage and then brainstormed in Design stage. I used the AI approach “synergenesis” (a term coined by Gervase Bushe), in which stories were a catalyst to provoke ideas for the topic of inquiry. People read each story aloud and brainstormed ideas until they produced an exhaustive list of ideas, and then moved on to the next story and repeated the same process. This approach helped not only generate new ideas in a shorter time but also shifted the narrative held by group members in the direction of the new ideas and new future imagined during the conversations.
There is more to AI than positivity. People are drawn to the positive aspects of AI, but more than being limited to being positive, my research was able to prove that AI can be generative, i.e. it changes people’s mindset and compels them to act in new ways. There are instances where an AI intervention is conducted and employees feel good during the intervention, but the excitement wanes in a few months because of lack of outcome as a result of the intervention. Therefore, it is important that practitioners define clearly the purpose of the inquiry, the nature of the questions and the design of the intervention in order to make it more generative, and to have a sustainable effect.
Although I was not able to follow through with the research, the company successfully launched an employee recognition initiative from the ideas generated during the AI sessions. Appreciative Inquiry’s focus on storytelling is a compelling aspect and, during challenging times, practitioners should find ways to engage people in stories that bring the best from the past and rethink the future.
References
- Bushe, G.R. (1995) Advances in Appreciative Inquiry as an Organization Development Intervention. Organization Development Journal, 13, 14–22.
- Bushe, G. R. (2010) Generativity and the Transformational Potential of Appreciative Inquiry, Organizational Generativity: Advances in Appreciative Inquiry, Volume 3. Bingley, UK: Emerald.
- Gergen, K. J. (1978) Toward Generative Theory, Journal of Personality and Social Psychology, 36(11), 1344–1360. doi: 10.1037/0022-3514.36.11.1344
- Ludema, J. D., Whitney, D., Mohr, B. J., and Griffin, T. J. (2003) The Appreciative Inquiry Summit: A Practitioner’s Guide for Leading Large-Group Change. San Francisco, CA: Berrett-Koehler Publishers.
- Weisbord, M. (2004). Productive Workplaces Revisited. San Francisco, CA: Jossey-Bass, Inc.
- Yaeger, T. and Sorensen Jr, P. (2001) What Matters Most in Appreciative Inquiry: Review and ThematicAssessment. Ed. D. Cooperrider, P. Sorensen, Jr T. Yaeger, D. Whitney. Appreciative Inquiry: An Emerging Direction for Organization Development, 129–142.
Neelima Paranjpey, PhD
Neelima is an I/O psychologist with a doctorate degree in OD. She works as a consultant in a talent management consulting firm in Chicago. She has a passion for helping organizations identify their strengths and leverage those to engage employees and increase productivity.
Contact: neelimaparanjpey@gmail.com