What are the most useful types of employee performance data? What’s the difference between qualitative and quantitative data?
In one step of Aubrey C. Daniels’s management approach, you collect baseline data about your employees’ current performances. You use this data as a reference point for evaluating the effectiveness of your behavioral intervention.
Check out the two types of employee performance data you should collect.
Assess Performance to Date
Daniels suggests clearly communicating the purpose of the evaluation process—he notes that many employees are uncomfortable with performance assessments because they fear being judged or punished based on the results. This discomfort often stems from past experiences where evaluations were used solely for disciplinary purposes. To mitigate this fear and gain accurate employee performance data, emphasize that evaluations are intended to identify areas for growth and improvement, not to find fault or assign blame.
(Shortform note: Framing performance assessments in this way helps you create what Amy C. Edmonson refers to as psychological safety in the workplace. In The Fearless Organization, Edmonson explains that psychological safety is the belief that one can speak up, take risks, and risk failure without fear of judgment or reprisal. By assuring employees that the purpose of performance assessments is growth and improvement, you foster a culture where they feel safe to engage in open communication and experimentation. This enhances employee engagement (the ultimate goal of Daniels’s management approach) as well as performance and innovation.)
What Kinds of Data Should You Collect?
Daniels says managers should collect two kinds of baseline data about employee behaviors: quantitative and qualitative data.
Quantitative data is more objective, so Daniels recommends focusing on that whenever possible. He also recommends using raw data—like simple frequency counts, response times, and numerical scores—over processed data like percentages and averages. Raw data provides a more accurate picture of behavior as it captures specific, concrete actions rather than potentially distorting the results through calculations or transformations.
For example, instead of averaging the time employees take to respond to customer inquiries, you might track individual response times for each inquiry. This allows you to see how consistently employees meet response time targets and identify any outliers or patterns that could signal specific problems.
(Shortform note: Daniels’s emphasis on using raw data aligns with Darrell Huff’s warnings about processed data in How to Lie with Statistics. Huff cautions that processed data can easily distort the reality of a situation. For example, averages can obscure outliers or significant variations in the data, giving a skewed impression of overall performance. As a manager, you probably wouldn’t intentionally distort the facts, but it’s easy to make statistical mistakes that lead to inaccurate conclusions. By focusing on raw data, managers can gain a more transparent and accurate view of employee behavior, allowing for clearer insights and better decision-making.)
Qualitative data describe your subjective assessments of aspects of performance that can’t be counted, like an employee’s acumen for customer service. Because qualitative data are subjective, many people consider them less reliable than quantitative data.
However, Daniels says there are two ways to ensure your qualitative data are fair and reliable: First, develop standardized rubrics that assign ratings based on certain behavioral criteria. For example, suppose you’re evaluating an employee’s customer service skills. In that case, your rubric might include categories such as communication clarity, problem-solving ability, and empathy, with each category rated on a scale of 1 (poor) to 5 (excellent). Second, ask someone else to review your qualitative data; this reduces the risk of personal bias and makes evaluations more consistent.
(Shortform note: Another way to enhance the reliability of your qualitative data is to ground your assessments in well-established theories, frameworks, or scientific evidence. For example, suppose you want to evaluate an employee’s customer service skills. In that case, you might start by researching the important components of stellar customer service as well as related concepts like emotional intelligence. Then, you can use what you learn to build a standardized rubric that captures these elements, ensuring that your assessments are objective and aligned with recognized standards. This is similar to what researchers do when they conduct a literature review, which helps them place their studies in the context of existing knowledge and methodologies.)
