Take a ride with
“The scientist is not a person who gives the right answers, he’s one who asks the right questions.” Surveys are a researcher’s secret weapon – and just like a very sharp dagger, they can be both beneficial and detrimental. Receiving the right answers is strictly influenced by the way you pose the question – was it clear, was it necessary, was it relevant? When doing survey research, understanding the audience you’re addressing the survey to and speaking their language is key. Having a good view of what’s the purpose of the survey is another key. But probably the most relevant (and often overlooked) key of a brilliant survey is this – are you using the right survey scale?
Claude Levi-Strauss
There’s nothing more frustrating than having the opportunity to express your opinion on a subject matter you’re truly passionate about, only to see that your only options are “yes” and “no”. On the same token, it’s equally frustrating to have a direct, simple answer and having to expand on it for a minimum of 30 words.
Understanding survey response scales and using them appropriately is vital in the process of receiving the answers you need – and interpreting them after.
In this article, we’re tackling everything survey rating scales – types of survey scales, survey scales examples, and when you should use one or the other.
Survey scales definition
First and foremost, what are survey scales?
As Survey Anyplace puts it, “in common survey usage, a scale is an ordered series of response options, presented verbally or numerically from which the respondents select to indicate their level of feeling about the measured attribute. More properly a scale is a composite score of a number of survey questions that each measure the same attribute.”
In other words, it’s a way to simplify your respondents’ answers into measurable, manageable, and comparative increments, that you will be able to analyze, compare and learn from after the surveying period is over.
Why do we need (and actually prefer) survey scales?
As mentioned above, scales allow us to organize our respondents’ answers. This is both applicable in a technological sense – meaning it’s possible for a program like Excel or SPSS to handle the answers – and in a human sense.
Big surveys may include thousands of respondents, sometimes even more – like the population census, for instance. That’s why it’s very important to divide the survey into manageable, easy to understand values, in order to have easy-to-read survey data.
Why is the type of scales relevant to the answers?
We’ll talk about the type of scales in just a moment, but we need to understand why it’s so important to care about which one we use and why. As Neuro-Linguistic Programming (NLP) would describe it, every person has his own “mind map”. Meaning that some words, expressions, and concepts that are second nature to us may seem completely different to the person we’re addressing.
That’s why miscommunications happen, why people often get offended and why so many surveys don’t provide accurate results – because the respondents simply misunderstood (or didn’t understand at all) the question at hand.
Put simply, the proper scale will diminish the chance of this happening, assuring (as best as possible) that the question asked by the researcher is understood precisely by the one read by the respondent. Therefore, the survey scale response will be more accurate.
What types of survey scales are there and when should you use them?
Fortunately, there are a bunch of survey rating scales that we can choose from – you’ve probably already heard about some of them, like the likert scale. The most common rating scales are:
Linear Numeric Scale
Linear numeric scales are just that – linear. They offer the respondent the opportunity to choose a level of something – satisfaction, interest, curiosity, experience, or likelihood to recommend. It’s a type of semantic differential scale that will provide some parameters to indicate the answer: “very, neutral, not at all” or “high, medium, low”.
This scale should be used when we want a simple answer, an indicator of someone’s opinion, without going into details. They are most used in asking for basic feedback. The key, as with most of the scales, is to formulate the question right for the scale to be relevant.
Likert
The Likert scale is one of the most popular scales out there. As opposed to a semantic differential scale, it invites the participant to approve or disapprove (agree or disagree) with a statement at a time. The classic version uses a 5-point response scale, but it can be personalized to 7-points, 9-points, or as many as you wish (remember that too many will become detrimental to the survey, though). You can find some likert scale examples, here.
This is a more complex linear numeric scale, in a sense. It is mostly used when trying to get more in-depth in the mind of the respondent. Asking “do you agree” with something gives the question a more personal tone, making the audience more careful with their answers.
Multiple Rating Matrix
A Matrix scale is very similar to the Likert scale questions – in which it presents multiple answers for a variety of questions. Whereas Likert refers mostly to agreeing or disagreeing with a statement, the Matrix scale asks the respondents to rate several opinions, like favorability, likelihood, or loyalty for a brand. It’s a classic type of ordinal scale.
It’s basically a succession of Linear scales one after the other. Use the Matrix scale when you want to understand more in-depth the degree to which your respondent feels about some issue. Avoid asking too many questions in a single survey, since it can become overwhelming.
Forced Ranking Scale
The Forced ranking scale is the method by which you ask the respondent to categorize a number of items (usually up to 10) in order of preference. This is considered better than simply rating the items individually because it brings in a new condition – no two items can be on the same level. So, the respondent must prioritize what they think is truly at the top.
Use this scale when you want to understand the real priorities of your respondents. Make sure to always shuffle the order in which the items are presented, so that the audience doesn’t subconsciously pick the top answers first.
Paired Comparison Scale
The Paired Comparison scale is similar to the Forced Ranking in the sense that it forces the audience to prioritize their favorite pick. However, this one limits the choice to only two answers, so it’s simpler for the respondent and provides a clearer answer for the researcher.
Use this scale when you managed to boil down the items (be it features of the product, options for upgrades, or brand communication strategies) and are not sure about which one would work best. Only opt for this scale after you’ve already used the Forced Ranking scale in a previous survey – you may overlook some items that are crucial for the respondents.
Adjective Checklist
Adjective Checklist scale is ideal when trying to “pick the brain” of the respondents – to receive some customer feedback. It’s usually used to understand the general view of a brand, a product, or a service, in a customer satisfaction survey.
It works by giving the audience a number of adjectives and asking them to choose the ones that they associate with the brand/product/service. It’s important to have a mix of negative and positive adjectives.
Use this scale when you want to accurately see what the consensus is when it comes to the subject matter. Using adjectives that maybe you wouldn’t personally associate with the topic could be relevant and yield surprising results. As with the Forced Ranking scale, don’t forget to randomize the answers for each respondent.
Fixed Sum
The Fixed Sum scale is a more in-depth approach to the Forced Ranking scale. Instead of choosing between several items in order of importance, the respondents have to allocate a specific cut of a whole (100 hypothetical dollars or 100 percentage points).
Use this scale when you want to get a more accurate representation of where your audience’s priorities really lie.
Pictorial/Graphic Scales
Pictorial/Graphic scales are the ones you’ll mostly see in hospitals, malls, and other public spaces. They are accompanied by representative pictures/graphics, in order to convey the emotion to the respondent. Very pleased – green happy face; very displeased – red sad face.
Use these scales when working with respondents that might not speak the survey’s language very well. They are ideal, not surprisingly, for public venues – since people will only be in passing and most of them will not be willing to stop and take a proper survey. This makes the answers more intuitive, easier and faster to answer, increasing the response rate.
General recommendations, regardless of the scale used
- Make the questions clear, using basic vocabulary, without metaphors, play on words or puns;
- Never ask two scale questions at the same time;
- Don’t try to manipulate the answer in one direction or the other with the survey question;
- Use a healthy balance of positive and negative feedback – allow the respondent to speak his mind accurately about the subject matter;
- Don’t make the survey longer than it needs to be – try to restrict it to what’s really relevant for the research.
You can choose a survey template from a massive list, here.
Conclusions
Surveys can only be accurate and useful for the researchers when done right. There are plenty of survey scales examples out there – in this article we’ve covered the most important. Using the right ones can make the difference between valuable data and a frustrated respondent.
The research method may vary, but one thing’s for sure – your scales should always be used correctly!
Frequently asked questions about survey scale
Choosing the right survey scale is important because it directly impacts the quality and reliability of the survey data collected. The scale should accurately capture the respondent’s perspective and allow for meaningful analysis and interpretation of the results. Selecting an appropriate scale ensures that the survey responses accurately reflect the respondent’s opinions or behaviors.
Yes, you can customize a survey scale to fit your specific research needs. While it’s common to use established and validated scales, you may modify or adapt them as necessary. However, it’s important to ensure that the modified scale maintains reliability and validity.
To choose the right survey scale, consider the following factors: Research objectives: Clarify the specific research objectives and the type of data you want to collect. Determine the constructs or variables you want to measure, such as attitudes, opinions, behaviors, or perceptions.Response format: Consider the response format that best captures the intended meaning or level of the construct being measured. Different constructs may require different response options, such as agreement/disagreement, frequency, importance, or satisfaction ratings. Level of granularity: Determine the level of detail or granularity required for your research. Choose a scale that provides enough discrimination between response options without overwhelming respondents.
