"Analytics provides a way to demonstrate the linkage between people and business outcomes. HR analytics (also called people analytics or talent analytics) use measurement and analysis techniques to understand, improve, and optimize the people side of business. Data are the raw numbers you track. [...] Metrics focus on counting, tracking, and presenting past data. Analytics uses statistics to help you see patterns in the data." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Data analytics is a powerful tool to increase the likelihood that you have the right problem. Both quantitative and qualitative data serve a purpose in supporting a hypothesis. They allow you to objectively measure and identify patterns and relationships." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Data mining is a common way of handling big data. It’s used to analyze big data and overcome some of the limitations of human information processing and traditional analytic techniques. This approach applies machine learning algorithms to find patterns of relationships between elements in large, messy data sets. The objective of data mining is to use the detected patterns to predict future outcomes and make better decisions." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Data scientists have extensive backgrounds in computer science, coding, machine learning, and statistics. Such an extensive background takes a long time to develop and is usually highly compensated. This has posed a tremendous barrier to many organizations. But this isn’t the only problem with relying solely on data experts. As organizations become more advanced, they are becoming more interested in using data to predict future outcomes. Merely relying on data from the past won’t suffice. Businesses must be forward thinking in how they collect their data to best serve predictive analytics. This means the employees who collect the data (e.g., those rating performance or creating and administering an engagement survey) need to understand how those data will later be analyzed." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Halo error is another common cognitive bias in employee ratings. This refers to a “halo” or aura that surrounds all ratings of an individual. If a manager has to rate an employee on several different dimensions, he or she would assign the same rating to that person on every dimension. Sometimes this may be attributable to laziness. But this also can reflect an underlying perception that performance is a singular dimension: people are either good or poor performers." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Hypotheses build the foundation for data analytics. Develop alternative hypotheses to explain the issue at hand. These hypotheses will guide your data collection." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"It’s important to look at multiple hypotheses. If your initial hypothesis isn’t supported, you’ll have little direction for moving forward. […] Thinking through viable alternative hypotheses will help you get a sense of where to find data in your organization. […] start by brainstorming. Think about the information that may be available. Is there a way it might be relevant? What sources are available to you?" (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Lagging indicators are metrics describing what happened in the past. 'Lagging' refers to the time lapse between an action and a specific outcome. […] Leading indicators are metrics that provide early indications of your progress toward an objective. […] Lagging and leading indicators work together, allowing you to make a more comprehensive evaluation." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Leniency error occurs when raters are unusually easy in their ratings, while severity error refers to the tendency to be unusually harsh in one’s ratings. Sometimes individuals commit these errors because of the language used in a rating scale. Terms like “average” and “outstanding” are relative and may lead a manager to use a personal average rather than an average that other managers may be using." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Measures of central tendency and variability work together to give you a concise summary of your data. When you don’t have any outliers, the mean is the most common indicator of central tendency. But on its own, the mean doesn’t tell you much. It isn’t until you also take the standard deviation into account that you really have a sense of your data." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Missing data can pose more or less of a problem depending on how random they are. Missing values scattered haphazardly throughout your data set may be inconvenient, but they don’t necessarily pose a threat to your analyses. When missing data follow a clearer pattern, there may be a problem. In those cases, you might need to revisit how you collect your data." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"People aren’t as good at making decisions as they think. We like to think of ourselves as rational actors, but our informational-processing limitations, emotions, and biases get in our way. The world is complex and humans have developed ways to help simplify it. So-called cognitive biases are ways our brains help us take shortcuts to deal with four primary problems: informati"n overload, lack of meaning, the need to act fast, and knowing what needs to be remembered for later." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Percentiles offer another way to understand how a data point fits into the bigger picture. […] A percentile score tells you what percentage of people fall below an individual on a given metric, so where an individual falls relative to everyone else." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"There are a few cognitive biases that commonly distort performance ratings. These inaccuracies may be intentional or due purely to human error in rating. Central tendency bias is the inclination to choose a rating somewhere in the middle of a scale, even when a more extreme score (for better or worse) is a better description. Raters often fall into this pattern when more extreme ratings require a written justification." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
"Understanding data is really about two types of expertise: data expertise and analytics expertise. Data expertise involves working directly with data - data extraction, cleaning, transformation, and management. Analytics expertise involves data analysis, data visualization, and validation." (Shonna D Watters et al, "The Practical Guide for HR Analytics: Using data to inform, transform, and empower HR decisions", 2019)
No comments:
Post a Comment