360-degree Feedback - Results

Results

Several studies indicate that the use of 360-degree feedback helps to improve employee performance because it helps the evaluated see different perspectives of their performance. In a 5-year study, no improvement in overall rater scores was found between the 1st and 2nd year, but higher scores were noted between 2nd and 3rd and 3rd and 4th years. Reilly et al. (1996) found that performance increased between the 1st and 2nd administrations, and sustained this improvement 2 years later. Additional studies show that 360-degree feedback may be predictive of future performance.

Another study including 33 workplace leaders showed that when feedback was given to these individuals the majority of participants (84%) were rated as demonstrating change (54% had major change and 30% had some change). The top trends for areas of strength that emerged from the data were competent (97%), likeable (48%), respected (36%), trustworthy (27%), and smart (15%). The top areas for development were communication (63%), assertiveness (24%), engaging one's team/creating a team approach (18%), and improving work/life balance and stress (18%). Although some may criticize this kind of feedback, some studies have shown that it can be efficient and reliable.

Some authors maintain, however, that there are too many lurking variables related to 360-degree evaluations to reliably generalize their effectiveness. Bracken et al. (2001b) and Bracken and Timmreck (2001) focus on process features that are likely to also have major effects on creating behavior change. Greguras and Robie (1998) tracked how the number of raters used in each particular category (direct report, peer, manager) affects the reliability of the feedback. Their research showed that direct reports are the least reliable and, therefore, more participation is required to produce a reliable result. Multiple pieces of research have demonstrated that the scale of responses can have a major effect on the results, and some response scales are better than others. Goldsmith and Underhill (2001) report the powerful influence of the evaluated individual following up with raters to discuss their results. Other potentially powerful factors affecting behavior change include how raters are selected, manager approval, instrument quality (reliability and validity), rater training and orientation, participant training, supervisor training, coaching, integration with HR systems, and accountability.

Some researchers claim that the use of multi-rater assessment does not improve company performance. One 2001 study found that 360-degree feedback was associated with a 10.6 percent decrease in market value, and concludes that "there is no data showing that actually improves productivity, increases retention, decreases grievances, or is superior to forced ranking and standard performance appraisal systems."

Maury Peiperl of Stanford's General Management Department, proposed four paradoxes that explain why 360 evaluations do not elicit accurate data: The Paradox of Roles, in which an evaluator is conflicted by being both peer and the judge; the Paradox of Group Performance, which admits that the vast majority of work done in a corporate setting is done in groups, not individually; the Measurement Paradox, which shows that qualitative, or in-person techniques are much more effective in facilitating change; and the Paradox of Rewards, which shows that individuals evaluating their peers care more about the rewards associated with finishing the task than the actual content of the evaluation itself.

Additional studies found no correlation between an employee's multi-rater assessment scores and his or her top-down performance appraisal scores (provided by the person's supervisor), and advised that although multi-rater feedback can be effectively used for appraisal, care should be taken in its implementation. This research suggests that 360-degree feedback and performance appraisals get at different outcomes, and that both 360-degree feedback and traditional performance appraisals should be used in evaluating overall performance.