Feedback är ett tveeggat svärd

Följ bloggkonversation mellan Mikael Nygren och Avraham N. Kluger. Avraham (Avi)

Mikael:

Dear Avi, it is an honor for me to have this conversation with you. Your work in the field of feedback and feedforward has been influential for me and my colleagues. We at the Centre for Leadership meet organizations and managers as well as other consulting firms who all strongly believe it is important and vital to create a feedback culture, and they create different systems of performance appraisals for both managers and employees. Feedback as a concept is considered to be an incredibly powerful way to influence others behavior and performance. The term is widely used in different contexts within organizations. My first question to one of the leading researchers in the field will therefore be: What really is feedback?

Avi:

As the name feedback indicates it is something with two elements "feed" and "back". The "feed" means some information that is sent "back" to some inner standard or goal for comparison of the information against the standard. For example, if I give you feedback in the form of a course grade indicating 95%, you are likely to implicitly compare the information (the "feed") of 95% against a standard of 100%. You may also compare it to a different standard of 80%, which might have been your grade in the previous exam in the same course. Yet, if I will add information to the "feed" and tell you "your grade is 95% and it was the lowest grade in class", I will further change your "back" or background to evaluate the information and the standard may be "being normal or being OK or being average". Thus, feedback is a dual phenomenon and the properties of feedback must always be understood as a combination of information pertaining to some standard. Only from understanding the pair we can start to understand possible outcomes of feedback.

Mikael Nygren:

Ok, thank you for this informative first answer of yours regarding what feedback really is. Basically then, as I see it, feedback is some form of information given to an individual or a group regarding some aspect of this person or group, their behavior, their performance on a task or the result of this performance. This information (i. e., feed), I guess, can come directly from other human beings (for instance ones manager, partner, teacher, coach etc.) as well as indirectly by comparing one’s own behavior with that of others. The information can also come from other, non-human, sources such as the score a person gains in a videogame, the time it takes for an athlete running 100 meters, the amount of money a seller working solely on provision receives during a month etc.

Moreover, this information is probably quite meaningless in itself if not compared to some form of standard (the back in feedback, which can probably be internal as well as external goals and norms) regarding expected results, performance and way of being. I guess then, as you also said in the final sentence of your answer, that one needs to include both the feed and the back and their relationship in order to understand possible outcomes of feedback. By that means I guess that whether the quality, source and amount of the feed (information given) will have a positive, negative or neutral influence on the individual or group depends on the locus (internal or external) and level (higher or lower than actual performance) of the standard (the “back”). If this is correct, can you tell me a little about factors that might moderate the effect of feedback and how these interact? Or, to use other words, which kind of feedback by which kind of source in which kind of situation for which kind of people is positive, negative or neutral?

Avi:

Indeed the effect of feedback depends both on the "feed" and the "back". One can think of various possible "back" or standard. To my best judgment the most parsimonious approach to understand standards is to consider general motivation theories. I will take a short detour to explain this issue of motivation and come back to what I know from my own research about "feed" and "back".

My work with Dr. Liat Levontin (in progress) noted the following:

A curious regularity is frequently observed when comprehensive sets of measures of concepts such as motivations (Ronen, 1994), values (Schwartz, 1992), organizational effectiveness (Quinn & Rohrbaugh, 1983), leadership behaviors (Lawrence, Lenk, & Quinn, 2009), and goals (Grouzet et al., 2005) are subjected to multidimensional scaling. Specifically, a two-dimension solution frequently emerges in which two consistent oppositions are noticed: Preservation (security, maintenance) opposing change (self-actualization, development), and competition (power, self-interest) opposing cooperation (self-transcendence, affiliation). More than one theory proposed independently a similar two-dimensional conflict. In some cases we have only a theory available (Nohria & Lawrence, 2002) while in other cases the theory is supported with extensive data (e.g., Schwartz, 1992).

Thus, we can ask what how various feed affect people under each of these four motivational states: when competing vs. trying to bond, and when trying to protect and vs. when trying to pursue self actualization goals or follow one's passion. It seems to me that in many organizational contexts the salient goals are competition and protection (psychological safety). Communications in organizations that emphasizes normative performance (either the highest possible rating on a scale or the mean of all other rates) directs attention to competitive goals and to psychological safety goals. These would be considered, in goal-orientation theory language (Dweck & Leggett, 1988; Elliot & Church, 1997), either as performance proving goals (competition) or performance avoidance goals (security). Yet, we could ask what happens when people pursue other goals (cooperation or self actualization).

For one of these two conflicts we already have some answer (Van-Dijk & Kluger, 2004; Van Dijk & Kluger, 2009, August). The conflict between change and preservation or between self-actualization and security was described by Higgins (Higgins, 1997, 2000) as resulting from the operation of the promotion focus versus the prevention focus in self regulation. To put in everyday words, one can distinguish between things we do because we "want to" (promotion) or because we "have to" (prevention).

Armed with the concepts of promotion vs. prevention (or self actualization vs. security) we can predict how different "feed"s affect motivation and performance. We can consider the most basic dimension of the feed, that is, whether or not it exceeds the standard. If the feed exceed the standard we have positive feedback and if it falls short of the standard we have negative feedback. Negative feedback or the outcome of loss and failure is congruent with the strategy of avoiding loss under prevention focus, whereas the outcome of gain and success is congruent with the strategy of approaching rewards under promotion focus. The idea of the congruency is that the prevention system is activated to take action most when it recognizes negative outcome, because negative outcome is congruent with its purpose (avoiding punishment).

Similarly, the promotion system is activated to take action most when it recognizes positive outcome because this outcome is congruent with its purpose (approaching rewards). In contrast to activation by congruency, incongruence is less likely to activate the systems because the salient outcomes do not fit their purposes (for example, when one is focused on prevention, positive outcome does not indicate any imminent punishment, and hence no increase in effort is needed). For example, if your accountant told you that all your papers are very well organized for the yearly tax report, you would be calm and likely do nothing.

In summary, we expect that positive feedback will increase motivation and performance under promotion focus, but debilitate motivation and performance under prevention focus. For negative feedback we predict the opposite: it will support motivation and performance under prevention focus but debilitate motivation and performance under promotion focus. This prediction was repeatedly supported in our various studies (Van-Dijk & Kluger, 2004; Van Dijk & Kluger, 2009, August) where promotion and prevention focus were conceptualized in several different ways.

Specifically, we operationalized promotion focus as a job "I always wanted to do", as holding values of self-direction, as being in either artistic or investigative occupation, and as performing a task that calls for creativity; we operationalized prevention focus as a job "I was obliged to do", as holding values of conformity and security, as being in conventional occupation, and as performing a task that calls for error detection. With each of these operationalizations we found the expected effect of feedback sign (either positive or negative) as expected.

What does it mean for practice? If you want your personnel to show more courage, creativity, or to take initiative consider how to bombard it with authentic positive feedback. In contrast, if you want them to come on time, to be punctual or orderly, spare your positive feedback and consider negative feedback and punishments. Yet, of course this is not that simple, but that could be deferred to another conversation.

To summarize my answer on different "feed"s and different "back"s, I have shown that knowing both is crucial for predicting feedback effects on performance and exposed our ignorance regarding what are the various "feed" (we only consider positive or negative aspects of it), and the various standards (we only considered two out of four prominent motivations).

Mikael Nygren:

Ok, so let me try to summarize your previous answer very briefly. If I understood you correctly, whether feedback will have a positive or negative impact of performance depends on the level of congruency between the feed (information given) and the back (the standard). For instance, as you demonstrate, the motive or goal (back) a person has for engaging in an activity determines whether a specific feed (positive or negative) will improve subsequent performance. If the motive is focused on prevention (avoidance of mistakes, stability, security etc.) the person will respond positive to negative feeds and neutral or negative to positive feeds. If, on the other hand, the person has a motive that is more focused on promotion (development, achievement, self-actualization etc.) the feed ́s given should be positive in order to have a positive impact on subsequent performance.

Mikael Nygren:

If this is the case then it seems important for anyone, a manager for instance, who plan to give feed to his or her subordinates that he or she knows about their underlying motive for their work if he or she wants the feed given to improve their work performance. It should probably also be of importance to consider the kind of work task that are at focus. If the aim of the task is to prevent from any mistakes to happen (like for instance the crew in an airplane), negative feeds should be used and if the aim of the task is to create or develop something (like for instance a brain storming session for coming up with ideas) positive feeds should be used.For a manager then, it seems important to do a detailed task and person analysis before starting to give information (feed’s) about performance. My personal experience is that most managers never take time to do that, their feeds will have a positive impact only one third of the times, and neutral or negative impact two thirds of the times. Statistically then, it would probably be better if they were not giving any feed ́s at all. That is, if they don’t analyze the task and persons beforehand.

Avi:

I agree. Yet, even this recommendation has qualiriers. Consider negative feedback. Negative feedback should improve motivation and performance for prevention tasks, under prevention situations, and for people characterized by chronic prevention focus. At the same time, negative feedback, per se, may have detrimental effects that may offset any benefit found in my previous work.

Consider for example taking a course in statistics, which many people take because they have to (prevention focus), not because they desired to study it. Now, imagine that you failed the midterm exam. Your reaction might be complex. On the one hand, as I showed in my research with Dina Van Dijk, you will have a heightened concern with statistics which will push you to try harder. On the other hand, ruminating thoughts ("perhaps I am an idiot who cannot master Statistics") may undermine your belief in your abilities (self-efficacy) and will reduce your motivation. Suppose further that despite of the rumination you increased your effort and worked hard but then failed also in the final exam. This time, you may plunge into learned helplessness. Thus, our theory and findings may be limited to initial reaction to negative feedback or for tasks for which failure is not likely to undermine beliefs in ability. Thus, the job of the manager is doubly complex, he or she has not only to gauge whether the person is operating under promotion or under prevention focus, she or he also have to gauge whether the person believe in his or her ability to perform.

Mikael Nygren:

Besides matching the feed to the task and the person, I would suggest that a manager has an important part to play in setting and influencing the back for the person receiving the feed. Would you agree, and if so, in what way can a manager influence the back for a person?

Avi:

I could not agree more. The task of the manager is to direct attention to goals that are compatible with the strategy of his or her unit. In many modern organizations there is a cry for creativity, innovativeness and leadership. Yet using negative feedback in performance appraisals is futile because it (a) reduces their motivation (negative feedback on promotion tasks) and (b) it might generate prevention focus by trying to avoid negative evaluations because they are related to security needs of the subordinate.

The way to direct attention to promotion focus goal could be with feedforward. I am attaching a new paper (in press) that demonstrates how this could be done.

http://www.ledarskapscentrum.se/pub3/attachments/149_Bouskila- Yam__Kluger_2010_HRMR_in_press.pdf

Mikael Nygren:

Highly interesting paper indeed, I really look forward to see the results of the project when it is completed. I think that your ideas with feedforward is really interesting, we work very much in line with those ideas when consulting with our clients. If you are not familiar with it yet I recommend the literature around solutions focused methods which are very similar to the feedforward and appreciative inquiry way of working. To the readers of this conversation can you just briefly tell us a little about what feedforward is and the ideas behind it?

Avi:

Thank you for the reference to "Solution-Focused Brief Therapy" -- I have scanned a review including the "miracle question" -- I guess that is what you meant. Indeed it has some commonality with FFI and with AI.

The Feedforward Interview (FFI) is a multi-purpose interview protocol designed to overcome some of the limitation of feedback. FFI is a theory-based modification of the Appreciative Interview component of Appreciative Inquiry theory and method (Cooperrider & Srivastva, 1987). Before explaining the theory behind it, consider the five steps of the FFI protocol and imagine that you are the interviewee (for a detailed protocol see Kluger & Nir, 2009).

FFI protocol

Introduction. “I am sure that during your work (or any other domain of inquiry) you have had both negative experiences and positive experiences. Today, I would like to focus only on your positive experiences.”

Story. “Could you please tell me a story that happened at your work (life), during which you felt full of life (happy, energized), even before the results of your actions became known?”. Note the emphasis "before the results of your actions became known" to direct you to consider cases where the mere activity was good for you regardless of organization rewards or societal approval.

Peak. “What was the peak moment of this story? What did you think at the peak moment? How did you feel at that moment (including your physiological reaction)? “

The conditions (learning; inquiry). “What were the conditions, in yourself, others, and the organization (physical, temporal) that allowed this story to happen?”

The feedforward question. “Recall the conditions that allowed you to feel alive (at work). Consider these conditions as road signs or a beacon that shows you how to flourish (at work). To what degree is your current behavior (at work) or your plans for the immediate future takes you closer to, or further away from, the conditions that allowed you to be happy (at work)?”

Feedback interventions compared to FFIs

Both feedback and the FFI appear to induce a tension between a standard (a goal) and some information pertaining to one's standing relative to the standard, yet feedback focuses on external standards and performance information while the FFI focuses on internal standards and performance information. Specifically, whereas typical feedback involves both an external standard of evaluation (e.g., what constitute an A in a course is set by the instructor) and external information regarding the distance from the standard (e.g., an actual grade received), the FFI seeks to establish both an internal standard of excellence based on past performance and to generate internal information regarding the distance from the standard that currently exist or expected to exist in the immediate future. There additional differences between feedback and FFI and three notable ones are discussed next. First, feedback often, albeit not always, is focused on outcomes, whereas FFI attempts to glean from a successful outcome a standard of superior processes. Second, the experience of being listened to is dramatically different from the experience of "being told" what to do. Given that listening is so crucial for human development (Pasupathi, 2001; Rogers & Roethlisberger, 1991/1952) it is not surprising that it leads to increased mutual liking between interviewer and interviewee, which is opposite of the documented destructive effects of feedback in organizations for human relationships (Coens & Jenkins, 2000). Third, FFI is based on stories. Stories are not only much easier form of communication for people, but they recruit experiential knowledge anchored in the constraints in which one operates. As such, it provides much richer guidelines from improvement that is possible from a typical feedback in organizations, which tend to be focused on abstractions (e.g., "you need to improve your teamwork"). In summary, FFI allows the interviewee to discover on her or his own their code for superior performance, emphasizing superior processes needed for superior performance, while building relationships and offering rich contextual guidelines.

Mikael Nygren:

I want to start with a brief summary of our conversation so far. Please correct me if I missed out on something important.

We started off with clarifying the concept of feedback as a dual phenomenon where the feed (information given) is compared to the back (standard of evaluation). We then went on to discuss what moderating factors influence the effects of feedback on subsequent performance, concluding that to understand the impact of feedback (positive/negative) it’s necessary to elucidate the motive state of the individual receiving the feedback as well as the characteristics of the task. Under prevention focus negative feedback appears to increase performance and positive feedback will have a positive effect on performance under promotion focus. Yet, we noted some limitations to this conclusion. Negative feedback per se might undermine performance and motivation. This led us to conclude that managers need not only to match feedback to task and person but also to influence the internal standard (back) of the person receiving the feedback. It seems optimal to direct attention to promotion focus goals and this could be accomplished using feedforward.

I’ll try to summarize your previous answer concerning feedforward very briefly. Using feedforward one seeks to establish an internal standard of excellence as well as internal performance evaluation. This distinguishes feedforward from feedback in that feedback focuses on external information and standards. Feedback also tends to focus on outcomes whereas feedforward emphasize the standard of processes through which a successful outcome is reached. Using the Feedforward Interview the interviewee experiences being listened to instead of being told to, which further extends the advantages of feedforward.

Mikael Nygren:

What positive outcomes can be expected from using feedforward?

Avi:

I just received an early draft from Marie-Helene Budworth and Gary P. Latham. They ran a field experiment on the effects of FFI relative to traditional performance appraisal feedback. They trained 13 managers with FFI and had 12 managers perform traditional performance appraisal who evaluated their 145 subordinates. Next the performance of each subordinate was assessed by a peer who could provide an anonymous and unbiased performance assessment. The performance of the subordinate of managers who were trained with FFI was .41 SD higher than the performance of the other group. These data is impressive because the performance measure is obtained independently from both the trained supervisor and the direct report. So, the first answer is the one outcome is improved performance. YET, this improved performance is likely to be accompanied with the following benefits: improved supervisor-subordinate relationships, more positive atmosphere at the workplace, new strategic insights for the superior, and increase job satisfaction via increased liking of one's supervisor.

Mikael Nygren:

How would you recommend managers to implement feedforward in their organizations?

Avi:

Feedforward could be implemented in variety of ways: as a component of strength-based employee performance (Bouskila-Yam & Kluger, in press), as a selection and placement tool in interviewing candidates (see Kluger & Nir, 2010), as a tool for learning what makes customer especially appreciate one's products and services and as a coaching technique for HR to work with senior managers and for managers to develop their subordinate.

Questions?

Mikael Nygren

+46 70 699 19 88