Ditch 4-Point Scales: Here is Why.
Hi, Friends,
Huge thanks to The People People Group for its excellent event last week, where we, fellow HR Techies, gathered for networking and drinks.
Did you know they have a community, too? Apply to join here.
Now, let's talk about scales.
Why are we doing this?
Well, over the years, many people have been using 4-point scales without the neutral option. And it really irritated me:
What if the person truly feels neutral?
Don't we care about variance above all?
Does a 4-point scale even provide enough resolution?
In people analytics, the choice of a rating scale can significantly impact the quality and usefulness of your data. So, today, I thought, why not post about scales?!
In this issue:
Common rating scales
Why we like odd-numbered scales (and we really do)
Common Rating Scales
There are a number of different scales out there. However, most will align to 4 categories:
Binary Scale (2-point Scale)
Example: Yes/No, True/False
Use: Simple decisions, clear-cut choices.
TL;DR: Useful for straightforward questions but cannot capture any granularity common in attitudes and satisfaction. You either like things or you don't.
Likert-type Scale (5-point, 7-point, etc.)
Example: Strongly Disagree (1) to Strongly Agree (max)
Use: Measures attitudes, perceptions, and opinions.
TL;DR: Provides a balanced range of options, allowing for nuanced responses that capture the intensity of feelings. Given it's an odd scale, it has a space for neutral responses.
4-Point Scale
Example: Poor, Fair, Good, Excellent
Use: Forces a choice without a neutral option.
TL;DR: Respondents are pressured to choose a side, leading to a strong interpretation of the attitudes.
10-plus-Point Scales
Example: 1 to 10+
Use: Detailed feedback, fine-grained distinctions.
TL;DR: Allows for a truly continuous distribution of responses but may be too granular to interpret: how does 78 compare to 83, for example?
Why we like odd-numbered scales
And by "We," I mean you and I.
Balanced Options
Odd-numbered scales, such as the 5-point or 7-point Likert scale, provide a central neutral option.
This option is used when the respondent feels indifferent, and if we were to face the music, some people wouldn't have an opinion about the new policy.
In fact, researchers found that if this option is missed, respondents randomly select the option adjacent to the neutral response, be it positive or negative. In other words, they create noise.
This means your data are not accurate.
Would you want to make conclusions based on inaccurate data?
I thought so...
Improved Data Quality
Data quality is important, as evidenced by many companies investing millions, if not billions, into cleaning up, organizing, and storing their data.
Messy data has become a new big problem in people analytics.
Hence, you might as well avoid artificial noise.
If given a choice, opt for a 7-point scale, as these will provide you with neutral options, a good range to express opinions, and enough variance to run complex analytics.
Let's give 3 examples to drive my point home.
Example: "Communication within the company is effective." (1 = Strongly Disagree, 5 = Strongly Agree) - Option 3 is neutral, providing a clear picture of overall sentiment.
Key Explanation: The neutral option ensures that respondents who do not have a strong opinion are not forced to make an inaccurate choice.
Example: "Current management strategies are effective." (1 = Strongly Disagree, 5 = Strongly Agree) - A neutral option prevents respondents from feeling pressured to side with or against management.
Key Explanation: Odd-numbered scales help minimize bias by allowing respondents to safely share their opinions without siding one way or another.
Example: "Completing tasks is difficult." (1 = Strongly Disagree, 5 = Strongly Agree) - More options allow for capturing the range of difficulties and providing better data for targeted interventions.
Key Explanation: Here, the variance around the estimate is more important than the estimate itself. It tells you where we need to focus.
All in all, choose a 7-point scale when you can.
Konstantin told you so. jk! or, am i?
K
Whenever you’re ready, there are 2 ways I can help you:
#1
If you’re still looking to get started in People Analytics, I recommend starting with my affordable course:
Practical People Analytics: Build data-driven HR programs to 10x your professional effectiveness, business impact, and career. This comprehensive course will teach you everything from building an HR dashboard for business results to driving growth through more advanced analytics (i.e., regression). Join your peers today!
#2
If you are looking for support in your human capital programs, such as engagement, retention, and compensation & benefits, and want to take a more data-driven approach, contact me at Tskhay & Associates for consulting services. Or simply reply to this email!