The choice of whether to use an odd or even number of options in your response scales for questions on opinion mostly comes down to whether a “neutral” or “middle” response is desired or required.
… and we are going to go ahead and state right out front that we avoid odd scales (with a neutral response) on questions around opinions at all costs.
Why? (read on)
The topic for this particular post is the number of response options you should have in your “Likert” or “contiguous” scales.
Likert or contiguous scales are ones that run in even increments from one extreme to the other. An example would be “Strongly agree” on one end through to “Strongly disagree” on the other, with versions of lesser agreement or disagreement in the middle. We ourselves almost always use response scales of:
What we are NOT addressing here is whether to allow for “N/A” and “Don’t know” in your options – and we will be talking about that in the next post “Question Scales Part 2: Should I allow for “N/A” or “I don’t Know” in the response options?“. (Stay tuned.)
This is the first item to consider and will make a big different on which way to go:
In the first (on opinion), one could theoretically be completely neutral on each – though honestly we just don’t believe you.
In the second though (on fact), “neutral” is meaningless as there is no “central point”. So for fact-related questions the choice of an odd or even number of options is immaterial. Re the actual number of options you’ll want though, we’ll be addressing that later in “Question Scales Part 3: How many response options should I have, and what is a Just Noticeable Difference? (JND)“.
Of all of these, only the last (#6) is really valid for our discussion, and even then, when is someone’s opinion TRULY neutral?
Obviously an odd number of responses allows from a middle or “neutral” option; whereas an even number results in a “forced choice” in one direction or the other, even if it’s to a very small degree.
And the later is what we prefer – no matter how small that increment between positive and negative is.
The reason? As mentioned above the only valid use of “Neutral” would be when a respondent truly is exactly neutral in their opinion – which in our experience pretty much never happens.
The other side of that coin is that by having a “neutral” you are now providing an “easy out” for the people in #3, where they just don’t want to take the time to think about it.
In other words the response is almost always and incorrect reflection of their actual opinion.
In my experience of human nature, and after running surveys for 30 plus years, (supported luckily for me by a substantial amount of research) – here’s the thing …
If provided with the “forced choice” of having to select one or the other side of neutral, positive or negative, people are generally less willing to straight out “lie” about their opinion (by then randomly choosing one side or the other), and are more likely to take the time to provide a more accurate and true response.
And while you’re thinking it; yes they could just skip the question altogether (if this is allowed) – which we would FAR prefer to an incorrect answer.
(Our thoughts on making questions “required” FYI are in Pro Tip Series 4 – Should I make some or all of the questions “required”? )
Hope this was of help, but as always give us a shout if you would like to discuss!
Next up – the second in our 3-part series on question scales: Question Scales Part 2: Should I allow for “N/A” or “I don’t Know” in the response options?
Thanks for reading, and if you’re interested in discussing a survey for your organization call us at 1-604-219-7876, email us at [email protected], or just book a discovery call for a one-on-one chat.
Let's have a chat
Maureen Simons is a senior human resources and communication consultant with over 25 years of experience helping clients achieve their business and organizational objectives through their people. (Linkedin)
Adam Hunter has a Bachelors degree in Mechanical Engineering, an MBA, and 35+ years of technical and programming experience, resulting in a broad mix of analytical, statistical, project-related and business skills. (Linkedin)