RILEY YOUNG: Welcome back to Klaus’ Courses. In our last video, we explored the binary and the three-point scale, which are great starting points for any team.
NARRATOR: If you are concerned that having an OK option might lead to reviewers choosing it over deciding if something is bad or good, then perhaps extending the binary scale might suit you best.
RILEY YOUNG: Introducing: the four-point scale! The four-point scale is an extended version of the binary scale, meaning that you still need to grade something either as bad or good. However, now you can determine how bad or how good the interaction was. The four-point scale gives you more insight into how things are going over time compared to the two-point scale. However, removing the neutral scoring option is not overly common. It’s time to check in with Klaus’ Data-man-Dan, Daniel Figueiredo, for more.
DANIEL FIGUEIREDO: We have found out that four-point rating scales are quite a rare thing, since most teams actually prefer using a neutral rating option. Looking at the data, we can also tell that binary, three-point and five-point scales are the most common. Though some teams may actually end up using larger scales, such as the 5-point, the 7-point, or even the 11-point rating scale, we recommend those for when you have a dedicated QA team. Don’t forget frequent calibration sessions and clear scoring guidelines, as these will help you remain consistent with your grades over time.
RILEY YOUNG: Thanks, Data-man-Dan! If having a neutral scoring option is something that you need, but you want to have more options than simply good, OK and bad, then perhaps implementing the 5-, 7- or 11-point scale may suit you. These scales all have a neutral rating accompanied by an even amount of good and an even amount of bad options either side. This is great for getting deeper insights into your team’s performance over time.
However, different reviewers tend to have a different understanding over what counts as good and what counts as very good. If you have a dedicated QA team and you’re able to hold regular calibration sessions, then these larger rating scales can work wonders. But be careful – if your graders are not fully aligned on what each point in your scale means, you may end up getting inconsistent and potentially corrupted reports as an output.
NARRATOR: Calibration sessions are when reviewers are given the same conversations to grade, but they can’t see how others have scored. Once graded, the reviewers meet up and discuss the scores given. This is a highly recommended practice that will help keep your grading consistent and accurate.
RILEY YOUNG: Choosing the right rating scale for your reviews can make a big difference to your review process. Here’s what you should keep in mind: the size of your review team, how your team is prepared to receive feedback, how clear your grading guidelines are, as well as what reporting outputs you need as a result from your conversation reviews. These are all things that should help determine which rating scale you should use. We recommend to use either the binary or a three-point scale to start with, and you can adjust the size of your scale as needed. You can also download our rating scale guide to see which rating scale we would recommend for you to start with. Thanks for watching. Have an a-meow-zing day!
ANN KAER: What type of car does a cat drive? A Fur-rari.