Siguro? Sigurado! Looking Beyond the Mean in Likert-Scaled Data

What do you really need to know when analyzing Likert-scaled data?

For many researchers, Likert scales are a familiar part of the research process. Strongly Disagree, Disagree, Neutral, Agree, Strongly Agree. The responses are collected, encoded, and—often almost automatically—the mean is computed.

But is getting the mean always the right approach?

This question took center stage on October 3, 2026, as researchers, educators, students, and research enthusiasts from different parts of the Philippines came together for “Siguro? Sigurado! – A Public Webinar on the Analysis of Likert-Scaled Data,” successfully conducted by Altabay Solutions.

With participants joining from different regions and representing diverse academic and professional backgrounds—from professors and teachers to students and researchers—the webinar provided an opportunity to look beyond the familiar “kuha tayo ng mean” approach and take a closer look at how Likert-scaled data can be appropriately analyzed and interpreted.

More than simply discussing statistical tests and formulas, the session encouraged participants to think critically about their data: What exactly are we measuring? What type of data do we have? What is our research objective? And what statistical approach makes sense for the question we are trying to answer?

These questions led to several important lessons that researchers can carry with them into their own studies.

Looking Beyond the Mean

One of the most important lessons from the webinar was the distinction between a Likert-type item and a Likert scale.

An individual Likert-type item—such as a question asking respondents to rate their level of agreement—is generally treated as ordinal data. However, when several items are designed to measure the same underlying construct and are combined into a validated scale, the resulting composite score may be analyzed differently, depending on the characteristics of the data and the statistical assumptions involved.

This distinction matters because it reminds researchers that not all Likert data should automatically be treated in exactly the same way.

The discussion also challenged the common practice of immediately computing the mean for individual Likert-type items. Depending on the purpose and nature of the data, frequency, percentage, mode, or median may provide more appropriate or meaningful summaries.

In other words, the lesson is not simply that “mean is wrong.” Rather, researchers should ask whether the mean is appropriate for the specific data and research objective.

Let the Research Question Guide the Analysis

Another major takeaway was the importance of choosing a statistical method based on the research question, variables, and study objectives.

As one participant aptly summarized:

“Statistical test depends on the variables and goal of the study.”

This is an important reminder for researchers who may sometimes begin their analysis by asking, “Anong statistical test ang gagamitin?”

Instead, the process can begin with a more fundamental question:

“Ano ba talaga ang gusto kong malaman from my data?”

From there, researchers can determine what variables are involved, understand the type and structure of their data, consider relevant assumptions, and select an analytical approach that addresses the research objective.

From Numbers to Meaning

The webinar also emphasized that statistical analysis does not end once a numerical result has been obtained.

For composite scores, in particular, researchers need to consider what the score represents in relation to the construct being measured. Interpretation should not simply stop at assigning labels such as agree, neutral, or disagree without considering the measurement framework and purpose of the scale.

This reinforces a broader principle in research: numbers need context.

A statistical result becomes more meaningful when researchers understand what it represents, how it was obtained, and how it relates to the question their study is trying to answer.

What Our Participants Had to Say

For many participants, the webinar also served as an opportunity to revisit assumptions and clarify concepts they had previously encountered in their own research experiences.

One attendee reflected:

“Yung likert scale na may Mean and Sd tas interpretation, mali pala yun. Okay lang pala median or mode.”

Another participant described the session as “very helpful in clarifying concepts regarding Likert scale, especially on how to interpret it properly.”

For a high school participant, the webinar provided an important foundation for future research. The session helped introduce essential concepts, including understanding when measures such as the mode and median may be considered instead of automatically relying on the mean.

These reflections highlight one of the webinar's central messages: better research begins with better statistical understanding.

Suggestions for Future Learning

The discussion did not end with the webinar. Participants also shared what they would like to learn next.

Among the topics suggested for future sessions were questionnaire validation, reliability and validity testing, Cronbach’s alpha, McDonald’s omega, Exploratory Factor Analysis (EFA), R training, research methodology, and more practical applications using real survey data.

Participants also offered suggestions on how future webinars could become even more engaging and accessible. These included incorporating more icebreakers, live polls, hands-on activities, and step-by-step demonstrations that would allow participants to apply concepts as they learn them.

These suggestions provide valuable direction for future learning sessions—particularly for researchers who want not only to understand statistical concepts but also to know how to apply them in actual research situations.

Beyond “Siguro” to “Sigurado”

At its heart, “Siguro? Sigurado!” was more than a discussion about Likert scales. It was a reminder that good statistical practice begins with curiosity, careful thinking, and a willingness to question familiar approaches.

A mean may be useful. A median may be useful. A mode may be useful. A particular statistical test may be appropriate—or it may not be.

The important question is not simply, “What can I compute?” but “What makes sense for my data and my research question?”

We at Altabay Solutions are grateful to everyone who joined the webinar, participated in the discussions, and generously shared their feedback and suggestions. With attendees from different regions and diverse academic and professional backgrounds, the session became a meaningful space for learning, exchanging ideas, and promoting better statistical practices in research.

We hope you can join us in our future webinars and learning sessions as we continue to create opportunities for researchers, educators, students, and research enthusiasts to learn and grow together.

And if you’re currently working on your own research and need guidance—from research methodology and sampling to statistical analysis, results validation, or manuscript review—Altabay Solutions is always glad to help.

Thank you for being part of “Siguro? Sigurado!” and for learning with us.

We look forward to seeing you in our next webinar! 📊