Welcome back. Step two in performance management is to find proper ways of evaluating progress. Traditional performance reviews usually take place on an annual basis and that is often insufficient. A lot of things may change in a year and the feedback, even if provided properly, may be too late to actually improve performance. There is extensive scholarly research on why traditional performance reviews do not work. Of the few factors that influence employee performance rating, the actual job performance is not the most important. Can you believe it? Many scholars claim that rater biases account for 50 to 60% of all the variance in ratings. A global consulting company, Deloitte, chose to alter their performance evaluation as they learned that over half of their executives believed that the traditional system drove neither performance nor engagement. Besides, it took too much employee time and consequently money. So they chose a simple metric aimed at avoiding biases. Raters may rate other people's skills inconsistently, but they are highly consistent when rating their own feelings and intentions. To see performance at the individual level, Deloitte chose to ask team leaders, not about the skills of each team member, but about their own future actions with respect to that person. As soon as we have proper indicators to track and measure, we need to start collecting them so they can actually be analyzed over time. Although there are specialized applications, for a small organization with a limited HR technology budget, Excel spreadsheets may be the most appropriate technology option. As long as you have correct metrics that accurately represent the data, that measures important metrics, it is sufficient tool for [INAUDIBLE] and analyzing performance information. Presenting data may be almost as important as collecting and analyzing it. In a situation when you need to promote people analytics in an organization, the second thing you'll probably have to do after collecting performance data is present it to other team members. Even the most correctly presented and analyzed data can lead to different managerial decisions. But before we switch to that, please complete a two-question on-screen quiz devoted to analytical decision making. As a matter of fact, many people will reply to the same question differently, depending on how the same data is presented. We tend to be more approving of an option if its description focuses on positive outcomes. That brings us to perception biases that may be still with us even if we use data analytics. First, framing. Depending on how information is structured and presented, humans would draw different conclusions even though the facts are exactly the same. Confirmation bias. We tend to disregard information that contradicts our preconceived notions. Anchoring. Anchoring means focusing on some or just one data points that we consider especially telling. Loss aversion. It is a result of humans being generally risk averse. We tend to consider losses of great importance, even if they have the same amount as gains. And finally, status quo. All else equal, we tend to support the status quo, or default version of any decision. We have to be consistently aware of these biases when working with data.