Central tendency bias refers to the tendency of an individual to rate most items on a survey in the middle of a rating scale.
This type of bias occurs most often with internal surveys – when managers have to rate their employees. For example, on a 10-point scale managers may be likely to rate employees between 6 and 8 for most categories.
Managers may do this because they want to avoid showing preferential treatment to certain employees or to simply avoid backlash from employees for making ratings on the extreme ends of the scale.
This type of bias occurs less frequently when customers are filling out surveys for a company because there are lower stakes.
Customers aren’t directly rating people they know and interact with on a daily basis, so they’re more willing to provide extremely high or extremely low ratings when they feel the need to do so.
The Problem with Central Tendency Bias
There are two problems caused by central tendency bias:
1. The data may be inaccurate.
If managers rate each employee in the middle of a rating scale simply because they’re afraid of providing extreme ratings, this means the data collected in the survey will be inaccurate and not reflect true employee performance.
2. The data will be unhelpful.
If managers rate each employee in the middle of a rating scale for each category, then the performance of all of the employees will appear mostly equal. This makes it difficult to identify the individuals who should be promoted or given a bonus.
How to Avoid Central Tendency Bias
There are three ways to avoid central tendency bias in surveys:
1. Don’t require managers to provide justification for their ratings.
Sometimes managers don’t want to provide ratings on the low end or high end of a scale simply because they’re required to provide justification for unusually low or high ratings.
By not requiring managers to provide justification for their ratings, they’re more likely to be honest about whether or not an employee should be rated high or lower because they don’t have to do additional work in providing a justification.
The benefit of this approach is that you’re likely to get more accurate data, but the downside is that you don’t actually get the reasoning behind particularly low or high ratings.
2. Allow for rank ordering.
Instead of asking a manager to provide a 1-10 rating for each individual employee on overall productivity, you could instead ask them to rank the employees from least productive to most productive.
This will force managers to identify low-productivity and high-productivity employees simply because not each employee can be the least productive or most productive.
3. Ensure that questions are clear.
Another obvious way to avoid central tendency bias is to make sure that the survey questions are clear. Often when questions are unclear, managers simply assign a rating in the middle because they’re unsure of the question being asked.
For example, consider the following survey with unclear questions:
- How responsible is employee X on a scale of 1-10?
- How would you rate the leadership of employee X on a scale of 1-10?
Now, consider the revised survey with clear questions:
- Rate the responsibility of employee X on a scale of 1-10, with 1 indicating that they’re not responsible in any manner and 10 indicating that they’re completely responsible for their actions and their work.
- Rate the leadership of employee X on a scale of 1-10, with 1 indicating that they have never taken a leadership role on a project and do not exhibit any leadership traits and 10 indicating that they always take on leadership roles and exhibit leadership traits on all projects when needed.
The revised survey is more likely to produce accurate data because they’re much more clear about what the manager is supposed to provide ratings for.