Institutional Knowledge at Singapore Management University A within-person perspective on feedback seeking about task performance

In organisations, feedback about multiple performance dimensions is often available. Consequently, employees have to decide on which performance dimensions they will be seeking feedback. In a lab experiment 126 students indicated on which performance dimensions they wanted to receive feedback after completing a computerized in-basket task. Results showed that partici- pants especially sought feedback about their best and most important performance dimensions. Individuals with a high learning goal orientation sought more feedback about their least important performance dimensions as compared to individuals with a low learning goal orientation. In general, results indicated that previous findings obtained in between-person studies of feedback seeking hold relatively well at a within-person level of analysis. The results of the current study illustrate how adopting a within-person perspective can broad-en our understanding of the feedback-seeking process in organisations.

In the last two decades, several studies have demonstrated that employees in organisations do not passively wait for feedback, but often proactively seek performance feedback themselves (for reviews, see Ashford, Blatt, & VandeWalle, 2003;Morrison, 2002). Active feedback seeking is a valuable self-regulation strategy for obtaining useful information about task performance as supervisors and co-workers in organisations are often reluctant to provide feedback to employees. Research has shown that by seeking feedback, employees can adjust their goal-directed behaviour (Morrison & Weldon, 1990), better assess their capabilities (Ashford & Tsui, 1991), enhance their future effectiveness (Renn & Fedor, 2001), and "learn the ropes" of a new job (Morrison, 1993).
Although substantial advancements in the feedback-seeking domain have been made over the last years, it is noteworthy that empirical studies have looked almost exclusively at variables predicting the frequency of feedbackseeking behaviour or, in other words, focused on employees' decision whether or not to seek feedback. In addition, almost all empirical studies have relied on a between-person approach for studying the frequency of feedback-seeking behaviour (for noteworthy exceptions, see Morrison & Vancouver, 2000;Trope, Gervey, & Bolger, 2003;Vancouver & Morrison, 1995).
Yet, employees in organisations are confronted with many feedback-seeking options. There are several different types of feedback they can seek and several different ways in which they can do so. For instance, in an organisational context, performance feedback is typically multi-dimensional, including feedback information on multiple tasks, assignments, skills, and behaviours (Ashford & Northcraft, 2003). Individuals not only have to decide whether they will seek performance feedback or not, but they also have to decide on which performance dimensions they will be seeking feedback. Till now, very little research has paid attention to feedback-seeking choices across performance dimensions. Congruent with more general theories of behavioural choice, it is appropriate that researchers adopt a within-person approach when studying this kind of intra-individual decision-making processes (Kanfer, 1990;Pelham, 1993;Stevenson, Busemeyer, & Naylor, 1990).
This study aims to examine individuals' feedback-seeking decisions across different performance feedback dimensions using a within-person perspective. Specific hypotheses are grounded in a theoretical framework delineating the motivational basis of feedback-seeking behaviour . Previous studies have highlighted the important role of two motives underlying feedback seeking (e.g., the ego-based motive and the instrumental motive). A next step is to understand how these motives not only affect the decision to seek feedback or not, but also affect decisions about which performance dimensions to seek feedback.

Studying Feedback Seeking at a Within-Person Level of Analysis
The question whether to use a between-subjects design or a within-person design is a much debated issue in the behavioural sciences (e.g., Allport, 1937;Bem, 1983;Rosenzweig, 1986;Runyan, 1983). In recent years, consensus has been reached that these different levels of analysis are both legitimate avenues of study and are in fact both necessary for developing a thorough understanding of human behaviour (e.g., Fleeson, 2004;Rosenzweig & Fisher, 1997;Van Kampen, 2000). Several scholars have argued that the choice for between-versus within-person designs should depend on the research question and the theoretical framework under study (e.g., Cervone, 2005;Keren & Raaijmakers, 1988). For instance, after reviewing a number of studies in personality psychology taking a within-person approach, Pelham (1993) concluded that important questions about the relations among multiple traits within a person can only be addressed using within-person designs. In line with this conclusion, we argue that the research question in this study ("About which performance dimensions do people seek feedback?") is more consistent with a within-person approach.
Studies examining feedback seeking in organisations have traditionally adopted a between-person approach wherein respondents are asked to provide self-reports of the frequency of their feedback-seeking behaviour in the past (e.g., Ashford & Cummings, 1985;Renn & Fedor, 2001;Roberson, Deitch, Brief, & Block, 2003;VandeWalle, Challagalla, Ganesan, & Brown, 2000). In this research tradition, differences in feedback seeking across persons are examined. For instance, a typical research question in previous studies (e.g., Morrison & Cummings, 1992) has been: "Do people with high performance expectations seek more feedback than people with low performance expectations?". This question is important from an organisational perspective because it points to strategies for organisations to encourage feedback seeking.
Despite the usefulness of a between-person approach, there are also limitations. For example, consider John who has an opportunity to seek feedback after completing a managerial assessment instrument. When considering feedback seeking, John will probably not ask himself how his preference for feedback about his Coordinating abilities compares with the preference for feedback about Coordinating of his colleagues. Instead, he is very likely to ask himself whether he prefers to receive feedback about his Coordinating abilities or about his Information Management abilities. This example helps to illustrate that people's everyday evaluations and decisions are typically made from their own frame of reference rather than from the perspective of researchers (who typically compare evaluations and decisions with those made by other people). It should be clear that research showing, for instance, a positive relationship between performance expectations and feedback seeking, does not automatically enable to conclude that similar effects will be found at the within-person level when multiple performance dimensions are available. Research on self-motives in social psychology has shown that within-person designs are more appropriate to investigate this kind of idiographic decision making (Pelham, 1993).
The current study is a first step to better comprehend how individuals direct their attention to specific performance dimensions. Previous studies adopting a within-person perspective in the feedback-seeking domain (e.g., Morrison & Vancouver, 2000;Vancouver & Morrison, 1995) have shed new light on how individuals selectively seek different types of information (e.g., task, role, social, organisational, performance) and utilise different sources. The current study further extends this line of work by taking a closer look at feedback seeking across performance dimensions. Morrison and Vancouver (2000) conceptualised feedback seeking about performance as a unidimensional activity, whereas research suggests that performance is a multidimensional construct, consisting of various performance dimensions (e.g., Borman & Brush, 1993;Campbell, McCloy, Oppler, & Sager, 1993;Tett, Guterman, Bleier, & Murphy, 2000). It remains unclear to what extent people differentiate between these different performance dimensions when seeking feedback and whether the same factors drive feedback-seeking behaviour across performance dimensions as feedback seeking across different types of information.
The focus of the present study is also important from a practical perspective. Not all performance dimensions are equally valued by an organisation. Organisations are looking for development of these individual competencies that closely align with the competencies required by their strategic intent (e.g., Huselid, 1995;Wright, Dunford, & Snell, 2001). Therefore, a better understanding of feedback-seeking decisions across performance dimensions in organisations is crucial. Insight in this decision process might assist organisations in designing strategies that direct employee feedback seeking towards the specific performance dimensions that are valued in the organisation and that contribute to the overall effectiveness of the organisation. For instance, when participating in a developmental assessment centre, participants can seek feedback about a number of different dimensions (e.g., interpersonal, leadership, sales, analytical skills) that are targeted by the various assessment exercises (e.g., in-basket simulations, role-play, presentations) (Abraham, Morrison, & Burnett, 2006). A particular organisation might be especially interested in developing employees' leadership skills and therefore like to stimulate feedback seeking after assessment towards those specific performance dimensions. When we know which factors drive these feedback-seeking decisions, organisations might develop interventions to influence these decisions (e.g., by changing employees' implicit beliefs, see Heslin, Latham, & VandeWalle, 2005).
In sum, this study addresses a question that has not been previously studied. It looks at individuals' feedback-seeking decisions when feedback about multiple performance dimensions is available after completing an assessment centre exercise. This study also employs a methodological approach that has not typically been used in the feedback-seeking literature. Rather than assessing between-person relationships, it assesses relationships at a withinperson level.

The Ego-Based Motive
In a recent review of research on feedback-seeking behaviour in organisations,  presented the main motives that underlie feedback seeking. One of these motives is the ego-based motive to protect and enhance one's ego. The ego-based motive, also known as a self-enhancement motive, has a long tradition in social psychology (Sedikides & Strube, 1997). Across diverse measures and situations, this literature offers sound support for the self-enhancement motive. People seek, process, remember, and judge self-related information in a manner that will place the self in the best possible light. These findings have also been confirmed in feedback-seeking research in organisations. Employees avoid seeking feedback if they expect the feedback to be negative or threatening (Morrison & Cummings, 1992;Northcraft & Ashford, 1990). Similarly, when people are self-confident and expect feedback to be positive, they seek feedback more frequently (Ashford, 1986). On the basis of these findings about frequency of feedback seeking, we expect to find a similar pattern when people can choose between feedback about several performance dimensions. We expect that people will be more likely to seek feedback about those performance dimensions they think they are best at, because feedback about these dimensions will most likely be favourable and ego-enhancing.
Hypothesis 1: People will seek more feedback about their best performance dimensions as opposed to their worst performance dimensions.

The Instrumental Motive
A second motive identified by  is the instrumental motive to achieve a goal or perform well. People seek feedback because it has informational value that can help them to meet goals and regulate their behaviour. Support for the existence of the instrumental motive stems from research on antecedents that are known to make the instrumental value of feedback more salient. One of the antecedents that highlights the instrumental value of feedback is the importance of goal-attainment. The higher the importance of goal-attainment to the performer, the more frequently the performer seeks feedback (Ashford, 1986;Tuckey, Brewer, & Williamson, 2002). On the basis of these findings, we hypothesise that when confronted with a choice between several performance dimensions when seeking feedback, people will be more likely to seek feedback about dimensions that are personally important to them as opposed to dimensions that have low personal importance. This would occur because important traits and abilities are closely associated with people's goals and ambitions (Pelham, 1991). Important traits are instrumental for achieving long-term desired outcomes and thus, diagnostic information about these traits is highly valued (Trope, 1986).
Hypothesis 2: People will seek more feedback about their most important performance dimensions as opposed to their least important performance dimensions.
On the basis of the role of the instrumental motive, we expect that individuals' learning goal orientation will interact with the importance of the performance dimensions in predicting feedback-seeking decisions across performance dimensions. Recently, goal orientation has received a lot of attention in the performance feedback area. Goal orientations are personal goal preferences in achievement-related situations. Two broad classes of underlying goals have been distinguished: (a) a learning goal orientation to develop competence by acquiring new skills and mastering new situations, and (b) a performance goal orientation to demonstrate and validate the adequacy of one's competence by seeking favourable judgments and avoiding negative judgments about one's competence (VandeWalle, 2003). Especially the former has been very successful in predicting feedback-seeking behaviour: Learning goal orientation has consistently been found to be related to the frequency of feedback seeking (Tuckey et al., 2002;VandeWalle & Cummings, 1997;VandeWalle et al., 2000). The relationship between learning goal orientation and feedback seeking can be explained by the instrumental motive: "A learning goal orientation (…) increases the salience of the instrumental motive, and accordingly, feedback-seeking behavior" (Ashford et al., 2003, p. 778). As importance already activates an instrumental motive, we expect this effect to be more pronounced for people with a high learning orientation.
Hypothesis 3: People with a high learning goal orientation will seek more feedback about the most important performance dimensions than people with a low learning goal orientation.

Participants
One hundred twenty-six I/O psychology master students participated in this study. They were given extra course credit for their voluntary participation. Participants had an average age of 22.9 years (SD = 1.8); 70% were female, 30% male.

Procedure
Participants were given the task to complete a computerized in-basket exercise that simulated daily work activities. We chose this task because it provides participants with a realistic environment that might instigate high involvement and motivation. In addition, in-basket exercises are often used for selection and development purposes in organisations. The in-basket exercise used was developed by Tett, Steele, and Beauregard (2003) to measure eight basic managerial performance dimensions that are included in a recently developed taxonomy of managerial performance (Tett et al., 2000): Coordinating, Decisiveness, Task Focus, Composure, Information Management, Problem Awareness, Quantity Concern, and Trustworthiness.
Prior to completing the computerized in-basket exercise, participants rated their standing on the eight performance dimensions and the perceived importance of these dimensions in a self-assessment questionnaire. Completing the computerized in-basket exercise took participants on average one hour. Upon completion of the exercise, participants were told that they would have the opportunity to read through a feedback report generated by the computer. However, as there would not be enough time for participants to review the report in its entirety, they were asked to specify those portions they most wanted to examine. Next, the feedback-seeking measure was administered. Finally, two weeks later, participants received a feedback report with quantitative and narrative feedback about their actual performance on the in-basket.

Measures
To measure participants' self-views, we used a modified version of the Self-Attributes Questionnaire (SAQ) of Pelham and Swann (1989). This temporally stable measure (four month test -retest reliability r = .77;  taps people's self-views on a number of different dimensions (e.g., intellectual ability, social skills, leadership ability, discipline, etc.). For this study, these dimensions were replaced by the eight performance dimensions that are measured by the in-basket exercise. For each of the eight per-formance dimensions, participants received a brief definition (e.g., Decisiveness: does not hesitate in making though decision) and the following instruction "This questionnaire has to do with your perceptions of your own managerial competencies. For the 8 competencies below, you should rate yourself relative to other college students of your own age by using the following scale". Next, they rated themselves relative to other participants on scales ranging from A (bottom 5%) to J (top 5%). We identified for each participant a "worst", and a "best" competency on the basis of their self-ratings and only these competencies were used in subsequent analysis (see a more detailed description of this procedure in the Appendix). In the remainder, we will refer to this variable as self-assessed standing.
Participants also rated the extent to which each of the eight performance dimensions was personally important to them, using 9-point scales (1 = not at all important, 9 = extremely important). Again, we identified for each participant a most important, and a least important competency on the basis of their ratings and only these competencies were used in subsequent analyses (see Appendix). In the remainder, we will refer to this variable as selfassessed importance. Mean correlations between self-ratings and importance ratings (r = .51) were comparable to findings in previous research Visser, Krosnick, & Simmons, 2003). Although these constructs are moderately correlated, evidence shows that self-assessed standing and importance reflect different constructs and have different cognitive and behavioural consequences (Pelham, 1991;Visser et al., 2003).
Prior to working on the in-basket exercise, participants completed the academic learning goal orientation questionnaire (VandeWalle, Cron, & Slocum, 2001). This questionnaire included four 7-point scale items (α = .74). An example item of this scale is "I truly enjoy learning for the sake of learning".
Upon completion of the in-basket exercise, participants indicated which feedback they preferred. The feedback-seeking measure ("How interested are you in your performance on each of the 8 competencies?") was taken from Trope andNeter (1994, see also Trope, et al., 2003). For each of the 8 performance dimensions, participants indicated how much they wanted feedback about that particular dimension on a 7-point scale ranging from not at all (1) to very much (7).

Results
Given that the self-assessed standing and importance variables were nonindependent, we conducted two separate analyses for each variable. We first examined the effect of self-assessed standing on feedback seeking. A GLM analysis with self-assessed standing as a within-persons factor (high vs. low) showed that people sought more feedback about their best performance dimensions (M = 5.76, SD = 1.31) as compared to their worst performance dimensions (M = 5.09, SD = 1.74), F(1, 125) = 16.14, p < .001, η 2 = .11. Thus, hypothesis 1 was supported.
To examine the effect of self-assessed importance and the hypothesised moderating effect of learning goal orientation, we tested a GLM model with self-assessed importance (high vs. low) as a within-persons factor, learning goal orientation as a continuous covariate, the interaction effect between these two independent variables, and feedback seeking as dependent variable. We found a main effect of self-assessed importance, F(1, 124) = 8.69, p < .01, η 2 = .06. People sought more feedback about their most important performance dimensions (M = 5.78, SD = 1.28) as compared to their least important performance dimensions (M = 4.79, SD = 1.73). Thus, hypothesis 2 was supported.
This main effect was qualified by a significant self-assessed importance x learning goal orientation effect, F(1, 124) = 4.57, p < .05, η 2 = .04. For the purpose of interpreting this significant interaction effect, median splits were ANSEEL & LIEVENS Figure 1. The interaction effect between learning goal orientation and self-assessed importance on feedback seeking. performed on the learning goal orientation measure. The means of this interaction effect are displayed in Figure 1. Both people with a low and a high learning goal orientation sought more feedback about their most important performance dimensions as opposed to the least important performance dimensions. However, as shown in Figure 1, this effect was more pronounced for people with a low learning goal orientation. Interestingly, planned comparisons showed that people with a high learning goal orientation sought more feedback about the least important performance dimensions than people with a low learning goal orientation, p < .05, while there was no difference in feedback seeking about the most important performance dimensions, p > .05. Thus, hypothesis 3 was not supported.

Discussion
Previous feedback-seeking research typically used between-person approaches to examine variables influencing the frequency of feedback seeking. However, employees' decision making in the feedback process is not solely limited to the question whether to seek feedback or not. Often, feedback about several task performance dimensions is available and employees have to choose between the various performance dimensions when seeking feedback. Because little is known about this decision-making process, we built on previous models delineating how different motives affect the frequency of feedback-seeking behaviour for studying feedback seeking across performance dimensions (e.g., Tuckey et al., 2002).
In general, the results of the current study indicate that previous findings about motives of feedback-seeking behaviour hold relatively well when different performance dimensions are available for seeking feedback. First, we found support for the ego-based motive as one of the main drivers of intraindividual feedback-seeking decisions. Individuals especially sought feedback about the performance dimensions they thought they were best at. This finding supports previous research indicating that people seek feedback to enhance their self-image by seeking feedback when performance expectations are high (Morrison & Cummings, 1992;Northcraft & Ashford, 1990). Second, our results point out that an instrumental motive also drives feedback seeking across performance dimensions. Participants preferred to receive feedback about the most important dimensions as opposed to the least important dimensions. This finding blends well with a similar within-person study of information-seeking behaviour in organisations. Morrison and Vancouver (2000) found that employees seek more of a given type of information to the extent that they perceive that information of that type is important. So, organisations should try to convince employees about the importance of perfor-mance dimensions that are crucial for organisational effectiveness in order to encourage feedback seeking in the direction of organisational goals.
The role of learning goal orientation in influencing the instrumental motive was not supported. While learning goal orientation qualified the effect of importance on feedback seeking as expected, results showed that participants with a high learning goal orientation sought more feedback about the least important performance dimensions than participants with a low learning goal orientation. One possible explanation for this unexpected finding is that importance makes the instrumental motive so salient in the feedback-process that it overrides the effect of learning goal orientation. For both people with a low and a high learning goal orientation, obtaining feedback about their most important competencies might be crucial for their future development and performance. A learning goal orientation adds little to the salience of the instrumental motive when seeking feedback about important dimensions, as the importance of these dimensions already activates the instrumental motive to a high degree. In contrast, when seeking feedback about their least important competencies, it might be that especially learning goal oriented individuals are interested to obtain feedback for further development. People with a low learning goal orientation are less interested in obtaining feedback about their least important competencies as they are less focused on developing competence by acquiring new skills and mastering new situations. A second explanation might be that we did not take performance goals into account in the current study. Individuals with a performance goal orientation seek to demonstrate and validate the adequacy of their ability by seeking favourable judgments and avoiding negative judgments about their ability (VandeWalle & Cummings, 1997). Performanceoriented individuals can be motivated either to "outperform" others and to demonstrate their competence and superiority, or to avoid failure and to avoid looking incompetent. One typical finding is that especially a performanceapproach (also called "prove") goal orientation is positively related to performance (for reviews, see Elliot,1999;Pintrich, 2000) suggesting that individuals with a performance-approach orientation tend to exert much effort and to work hard in order to accomplish their goal of performing better than others. When seeking feedback about their most important goals, participants might have also been driven by performance-approach goals to demonstrate their competence on the assessment task and outperform others. Recent research suggests that striving to outperform others (performance goal orientation) is not necessarily inconsistent with trying to attain task mastery (learning goal orientation) and these goal orientations may interact in predicting individual responses in achievement situations (Van Yperen & Janssen, 2002). Of course, these are tentative explanations. In this study, we focused exclusively on the concept of learning goal orientation as it more closely parallels the feedback-seeking framework of Ashford and Cummings (VandeWalle et al., 2000). Future research should take both learning and performance goal orientations into account to further test the different explanations proposed to account for the unexpected interaction effect. Future research might also examine whether performance goals interact with individuals' self-assessed standing as research suggests that individuals with a performance-avoidance goal orientation might be more driven by the egobased motive when seeking feedback (Janssen & Prins, in print).
This study is not without limitations. A first concern deals with generalisability. In this study, we undertook substantial efforts to create a situation and a task that were as realistic as possible for the participants. For instance, the design of our task corresponds closely to organisational practice where people are provided with feedback about various performance dimensions on one single occasion, for instance upon completion of developmental assessment centres, 360-degree feedback, and self-assessment tools. However, it should be acknowledged that some organisational realities could not be simulated. Furthermore, in the current study, participants sought feedback about individual performance. However, organisations are increasingly using teams to organise the work environment. Therefore, an important question for future research is whether organisations can encourage employees to seek feedback about team performance instead of individual performance by emphasizing the importance of team work (e.g., by rewarding team performance).
A second drawback was that we measured individuals' intentions to seek feedback and not actual feedback-seeking behaviour. Previous research has shown that employees' feedback-seeking intentions are sometimes reconsidered and modified in a public environment because of fear for face-loss (e.g., Levy et al., 1995). In the current study, participants expressed their interest in obtaining feedback in a private computerized environment and thus, experienced very little feedback-seeking costs. Future research might want to examine within-person feedback decisions in a more realistic public environment. It would be interesting to see whether initial feedback-seeking intentions towards specific performance dimensions will be modified to other performance dimensions when the actual feedback-seeking act takes place in the presence of significant others (colleagues, supervisor, etc.).
In conclusion, recently several calls have been made stressing that more research using within-person designs is needed to gain a better insight in human behaviour (e.g., Fleeson, 2004). The results of the present study indicate that previous findings obtained in between-person studies of feedback seeking hold relatively well at a within-person level of analysis. This further highlights the relevance of the current study as these results offer additional evidence for the complementarity of the between-person and within-person approaches in studying human cognition and behaviour.