You’re checking your emails and noticed 10 new Substack posts, another 5 from Medium as well as dozens from LinkedIn — all about a new AI feature that apparently “changes everything” — again! By the time you finish reading, two more have been published. You close the tab, go back to your spreadsheet or your report queue, and wonder if you’re already behind.

Here is my honest confession: I am exhausted trying to keep up. And I say that as someone who writes about AI tools, occasionally presents on AI, and genuinely thinks this stuff is worth your attention. The pace of change right now is relentless, and pretending otherwise would be doing you a disservice. I wrote a similar article on the Higher Education User Group that describes this. The following has been altered to suit a wider audience.

In this article, I try to focus on the basics and not just for the beginners.

Domain Knowledge is the Starting Point

AI tools like ChatGPT, Copilot, and Claude are powerful but they do not know your organization. They don’t know how your office defines a metric, what your data identifiers look like, whether your headcount logic excludes certain populations, or that your reports pull from a snapshot table that refreshes overnight. You do.

This matters because AI amplifies what you give it. Hand it a vague question without organizational context and you’ll get a generic answer that sounds credible but may be entirely wrong for your environment. Before you type a single prompt, ask yourself: do I understand this data well enough to verify what AI gives me back? If the answer is no — and there is no shame in that — start there. Not with the AI tool. With the data. And we mustn’t forget that humans need to be kept in the loop. We are the ultimate decision makers. This is also what Cole Nussbaumer Knaflic refers to in her April 23 blog.

AI can generate answers exceptionally fast. But someone still needs to decide what questions are meaningful, which data is trustworthy, and which action should follow. AI might be able to generate the chart, however humans still own the interpretation.

Prompting with Context

Once you know your data, prompting becomes your most important skill despite what you read from self-proclaimed AI gurus as prompting being “dead” or whatever (pish posh I say to that and feel free to argue).

Think of a prompt the same way you would think about writing requirements for a report or a dashboard: it needs an objective, context, constraints, and an expected output. The more specificity you bring, the less the AI has to guess — and guessing is where things go sideways fast (think hallucinations).

❌ “Help me analyze enrollment trends.”

✅ “I am a data analyst at a public university. I have a flat file export with term, academic plan, and headcount for fiscal years 2020–2025. Identify the five undergraduate programs with the sharpest enrollment decline. Summarize in plain language for a non-technical director, and flag any assumptions you make about the data.”

That second prompt is in no way perfect, but it gives the AI the guardrails it needs. Notice what it does at the end — it asks the tool to flag its own assumptions. That habit alone will save you from walking into a meeting with a number that nobody can explain.

Risks Worth Knowing About

Institutional data carries obligations that general-purpose AI tools were not designed with in mind, such as data governance policies, privacy regulations, and more. An AI tool will not flag these for you. It will produce something that looks polished and plausible but ultimately the burden will be on you to guarantee compliance.

Research from the University of California, Davis confirms what many practitioners already sense: business professionals regularly fail to catch errors in AI-generated data analyses — even when they are explicitly warned the AI makes mistakes, and even when the flaws require no technical knowledge to spot (Virk & Liu, 2025). In a professional context, that is not just an inconvenience. It can mean reporting incorrect figures, making resource decisions on flawed data, or surfacing information that should have been protected. Additional research on overreliance on AI outline similar concerns (Romeo & Conti, 2025).

This is not a reason to avoid AI. It is the reason to stay in the loop — to be the person who knows the data, knows the rules, and knows enough to catch what the tool misses. That is your value.

Fundamentals Worth Knowing or Revisiting

01 Know your data before you AI it. Understand the source, the logic, and the rules before you prompt anything. AI doesn’t know your organization’s definitions, your counting conventions, or what ‘active’ means in your system. You do — and that context has to come from you.

02 Prompt with specificity. Treat your prompt like a requirements document, not a search bar. Tell the AI what system you’re working in, what the data represents, who the audience is, and what a good answer looks like. Every time.

03Verify like your director is watching. Cross-check every output against a known source of truth — a report you have run before, a count you trust. If the number feels off, it probably is. AI does not know that it is wrong. You have to be the one who notices.

Bottom Line

I am not presenting the newest features in Kentucky this summer. I am presenting the foundations — because those are what will still be true no matter what drops between now and then. How to think about a prompt. How to stay skeptical. How to be the human in the loop. Those things will always be relevant.

The professionals who will get the most out of AI are not the ones chasing every new release. They are the ones who understand their data deeply, ask sharp questions, and know enough to tell good output from plausible-sounding nonsense.

The horse is powerful. But you still need to know how to ride.

Sources

Romeo, G., & Conti, D. (2025). Exploring automation bias in human–AI collaboration: a review and implications for explainable AI. AI & Society, 41, 259–278. https://doi.org/10.1007/s00146-025-02422-7

Virk, Y., & Liu, D. (2025, August 8). Non-programmers assessing AI-generated code: A case study of business users analyzing data. arXiv. https://doi.org/10.48550/arXiv.2508.06484

I’ve been reading a lot on AI lately – most recently listening to Ethan Mollicks’ audiobook “Co-Intelligence: Living and Working with AI” and also intrigued by Luke Stanke’s LinkedIn article “The Death of the Data Analyst.” Both pieces of work made me realize that the data analyst’s job description has not been formally rewritten. The role is transforming perhaps a bit faster than job titles are catching up. As uncomfortable as it may make you (us), Artificial intelligence is already inside the tools you use every day: the BI platform, the CRM, the ERP’s reporting layer, the institution’s Microsoft 365 environment. The question is no longer whether AI changes the data analyst role. It is what kind of analyst you want to be when it does.

The good news is that the skills that make a great analyst are not being made obsolete. They are being made more valuable. Critical thinking. Institutional knowledge. The ability to translate complex data into a story that drives a real decision. These are the things AI cannot do (at least not yet). They are exactly the things your institution needs you to do more of — if you’re ready.

What’s Actually Changing — Right Now

Let’s be specific and cite some of the research that’s out there: One of the largest and comprehensive AI studies conducted by California State University showed that 74% of staff express concern about AI’s impact on job security. EDUCAUSE’s 2025 AI Landscape Study found that higher education students, faculty and staff are already using AI for at least one work-related task. The most common uses are summarizing content (74%) and brainstorming (71%). In relation to the AI-related skills that staff and faculty need to develop, 90% of respondents cite AI literacy and best practices for using AI to boost productivity as the top two.

Wharton professor Ethan Mollick’s GDPVal benchmark makes the pace of change visceral: AI completed complex professional tasks that took humans approximately seven hours — in five to ten minutes. Blind expert judges preferred the AI output 72% of the time.

What does this mean practically? Routine data cleaning, basic SQL querying, standard dashboard generation, and summary reporting are all moving toward AI-assisted or AI-automated territory. The analysts who recognize this earliest have the most time to redirect their energy toward work that genuinely cannot be automated.

“AI can make the chart. Only the human can make it matter.”

— Cole Nussbaumer Knaflic, Storytelling with Data (2025 edition)

The Real Opportunity: The Work AI Can’t Do

Here is the case for optimism, grounded in evidence: the skills at the top of the analyst’s value stack are the ones AI is least equipped to replace. LinkedIn’s Skills on the Rise: The Fastest-Growing Skills in 2026 named People Skills (Leadership and People Management) the #1 most in-demand skill across all industries. Not SQL. Not Python. People Skills — like Cross-Functional Collaboration, Team Management, and Mentorship. #3 on the list is Technical and Strategic AI (AI that goes beyond coding).

The EDUCAUSE 2026 Top 10 adds a higher-education-specific dimension that every analyst in our community should internalize: Institutional leaders are valuing human connections. A data analyst who can connect institutional evidence to student success, equity outcomes, and strategic priorities is not just doing a job. They are helping their institution rebuild trust.

Ben Jones, CEO of Data Literacy, captures the evolution plainly: the future belongs to analysts who are “keen observers”. These professionals understand the institutional terrain, they ask questions, locate the right data and share it effectively. They know what the data means in context. They ask the question behind the question. And they communicate the answer in a way that actually moves people.

Ethan Mollick refers to the human in the loop principle in his audiobook “Co-Intelligence: Living and Working with AI”. To be the human in the loop means to exercise caution with AI “hallucinations,” and to work with AI but whilst maintaining a critical perspective.

A Word of Honest Caution

Optimism is warranted. Complacency is not. Three cautions deserve your attention:

• The critical thinking risk is real. Jones cites early Carnegie Mellon University and Microsoft Research suggesting that over-reliance on AI can erode critical thinking capacity over time. If AI generates your analysis and you accept it uncritically, you may be losing the very skill that makes you irreplaceable. Stay deliberate. Question AI outputs. Maintain the habit of understanding the data yourself. Be the human in the loop.

• AI amplifies bad data. Gartner’s 2026 D&A Summit delivered a stark message: AI does not fix poor-quality data — it makes poor decisions at scale, faster. EDUCAUSE found that only 16% of institutions say their data functions operate cohesively. If your institution has data quality problems (and most do), AI is a reason to fix them urgently, not a workaround.

• The pipeline is at risk. PwC’s 2026 research shows organizations reducing entry-level analyst positions as AI handles more routine tasks. This creates a real risk: the apprenticeship pathway that has historically developed senior analysts and data leaders is being eliminated. Whether you are early-career or leading a team, advocating for that pipeline is advocating for the future of your profession. Page 1 of the Executive Briefing indicates that despite optimism around enhancing skill development and productivity gains through AI adoption, there is overriding concern around socio-economic inequity and lack of preparedness.

What You Can Do — A Practical Framework

The following table lays out the Analyst Evolution Framework in visual form. The core idea is simple: wherever you are in your career, there are concrete, doable steps you can take right now. The framework has two tracks — one for mid-to-advanced career analysts, one for early-career and aspiring analysts — because the actions are different even though the destination is the same.

The Bottom Line

The data analyst role in higher education is not disappearing. It is expanding in scope and contracting in routine. The analysts who thrive in the next five, ten, fifteen or so years will be the ones who embraced AI as a collaborator, protected their critical thinking, deepened their institutional knowledge, and learned to communicate data in ways that rebuild — not erode — the trust our institutions desperately need to earn back.

None of this requires waiting for your institution to have a strategy. It requires you to start, right now, with the tools you have and the time you can protect.

The views and opinions expressed here are personal and do not necessarily reflect the official policy or position of my employer.

In this series, we have explored how applying user interface and user experience principles to data visualizations can enhance clarity, engagement, and decision-making. Beginning with foundational ideas—such as treating visualizations as interfaces, understanding the business context, and recognizing the benefits of user-centered design—we then examined how to ground practice in core design considerations, from context and constraints to knowing the audience and choosing appropriate graphs.

Building on that base, we turned to the experience layer: reducing clutter, making purposeful use of visuals and interactive features, and considering how users interact with dashboards in real settings. We also emphasized the importance of gathering user feedback—through conversations, usability sessions, surveys, and embedded feedback mechanisms—to iteratively refine dashboards so they better align with real-world needs and workflows. From there, we delved more deeply into accessibility, looking at how color and contrast, alternative text, labels, legends, navigation, and device compatibility can make visualizations more perceivable, operable, understandable, and robust for a wider range of users.

Taken together, these perspectives highlight that effective data visualization is not only about accurate data and attractive charts, but about designing interfaces that support people—each with their own goals, contexts, and abilities—in making sense of information. Even a basic understanding of user design principles can spark new approaches to layout, interaction, accessibility, and feedback that improve how dashboards are experienced across an institution. As a practical next step, reaching out to colleagues in accessibility, user experience, or related areas can open up valuable collaborations and help bring these ideas into everyday visualization work.

Need a refresher on the previous posts? Click on each one below:

Introduction

Part 2: Core Design Considerations

Part 3: The Experience Layer

Part 4: On Accessibility

Part 5: User Feedback

After considering how accessibility features such as color and contrast, alternative text, labels, legends, navigation, and device compatibility can make dashboards more inclusive and effective, an important next step is to understand how these designs work in practice for real users. Accessibility guidelines provide a strong foundation, but they cannot anticipate every context, workflow, or level of data literacy found across an institution.

For that reason, the series now turns to the role of gathering user feedback in dashboard design. By engaging stakeholders directly—observing how they interact with visualizations, listening to their questions, and incorporating their suggestions—designers can iteratively refine dashboards, so they are not only theoretically accessible, but genuinely usable and valuable in everyday decision-making.

Gathering Feedback

Gathering user feedback is a critical step in designing dashboards that truly meet the needs of their audience. It closes the loop between design intentions and how people actually experience and use the visualization in their day-to-day work.

Why Feedback Matters

User feedback helps identify misunderstandings, pain points, and gaps that are not visible from the designer’s perspective. It also surfaces opportunities to simplify navigation, clarify labels, reduce clutter, and strengthen accessibility for a wider range of users.

Ways to gather feedback

• Informal conversations and demos

Walk key stakeholders (e.g., advisors, deans, department chairs) through a draft dashboard and ask them to “think aloud” as they interpret the charts, filters, and navigation. Listen for confusion, extra clicks, or repeated questions.

• Structured usability sessions

Give users realistic tasks (for example, “Find the retention rate for first-year students this year” or “Identify which program has the highest growth”) and observe how easily they complete them. Note where they hesitate, what they miss, and which elements they ignore.

• Surveys and quick polls

After sharing a dashboard, send a short survey asking about clarity, ease of use, and perceived usefulness (e.g., Likert-scale items plus one or two open-ended questions). This can be especially helpful when the audience is large or distributed.

• Embedded feedback mechanisms

Include a small “Feedback” link or icon directly on the dashboard, inviting users to report issues, suggest enhancements, or ask for additional views. This makes it easy for insights to accumulate over time without organizing formal sessions for every iteration.

Making feedback actionable

• Look for patterns, not one-off comments

Prioritize changes that address recurring pain points, especially those that affect core tasks or critical decisions.

• Tie changes back to design principles

When feedback points to confusion, revisit UI and UX principles related to clutter, labels, navigation, and accessibility—for example, simplifying filters, improving title wording, or adjusting color and contrast.

• Iterate and communicate

Treat dashboards as living products rather than static reports. Share release notes or brief updates with users when changes are made, and explain how their feedback informed those revisions to reinforce engagement and trust.

By intentionally incorporating user feedback into each iteration, dashboard designers can better align their visualizations with audience needs, improve usability and accessibility, and help foster a more collaborative, user-centered data culture.

Conclusion

Gathering user feedback is essential for designing dashboards that genuinely support people’s day-to-day work. It bridges the gap between the designer’s intentions and how users actually interpret charts, navigate filters, and apply the information to decisions. Feedback reveals misunderstandings, pain points, and usability issues that might otherwise go unnoticed, while also highlighting opportunities to simplify navigation, improve labels, reduce clutter, and strengthen accessibility. Feedback can be collected in multiple ways: informal demos and “think-aloud” sessions with stakeholders, structured usability tests based on realistic tasks, short surveys or polls after dashboards are released, and embedded feedback links that let users share comments as they work.

Making this feedback actionable requires looking for patterns rather than reacting to one-off comments, and prioritizing changes that affect core tasks or critical decisions. When users report confusion, designers can return to core UI/UX principles—clarity of labels and titles, intuitive navigation, appropriate use of color and contrast, and reduction of visual noise—to guide revisions. Treating dashboards as living products, not static reports, means iterating regularly and communicating updates back to users, including how their input shaped each new version. This iterative cycle helps align visualizations more closely with audience needs and supports a more collaborative, user-centered data culture.

In the previous part of this series, we focused on how to refine the experience of using data visualizations by managing clutter and making intentional choices about visuals and interaction. An equally important dimension of that experience is ensuring that dashboards and charts are usable by people with a wide range of abilities, preferences, and technologies.

The next part of this series will examine accessible features in more depth, including how color and contrast, text alternatives, labels, legends, navigation, and device compatibility can be designed to support perceivable, operable, understandable, and robust visualizations. By embedding accessibility into the design process rather than treating it as an afterthought, practitioners can create data experiences that are both more inclusive and more effective for all users.

Providing Accessibility

Accessibility in terms of human interaction with a computer interface is defined as a set of features “designed to help people with disabilities use technology more easily.”¹⁸ When designing user interfaces, one must consider the variety of accessibility challenges and look at the appropriate guidelines to incorporate practices in the design strategy. The Web Content Accessibility Guidelines (WCAG) 2.0 list four principles that guide the development of accessible web content for people with varying types and levels of impairments. Information needs to be perceivable, operable, understandable, and robust in order to make it easier for users to consume the content as intended.¹⁹ Concepts such as screen size, colors, typeface and font are some of the factors that UI professionals consider when developing accessible interfaces such as mobile apps or web pages. With data dashboards becoming more and more interactive in nature, there is a growing need for embedding accessibility features in order to accommodate for a wider and diverse audience with differences in vision, hearing, cognition, or language. A non-exhaustive list of recommended accessibility design concepts follows.

Color and Contrast

“Color is one of the most important things to understand in data visualization and frequently is misused,” as Wexler et al. state in The Big Book of Dashboards.²⁰ It must be used with intention and consistency. Color is an important UI element that contributes to the audience being able to make sense of what they see as they visually navigate a data dashboard. Color should not be the only way to convey a message and functionality should never be compromised for aesthetics. The following example includes two different line charts. While the first applies color to each of the lines, the second applies color and contrast in a way that emphasizes the more important element in the chart (purple trend line), while the less important elements are shown in gray. Gray may have a calming effect on the overall visualization.

INSTEAD OF THIS

TRY THIS

A general rule of thumb when using colors in a chart is to limit to no more than seven. The more colors the more confusing or busy the chart. If all categories of data need to be included in a visualization perhaps a simpler type of chart needs to be considered using a single color, such as in the example below.

INSTEAD OF THIS

TRY THIS

There are numerous guides to accessible color palettes that designers can reference online such as those found on the W3C or one’s own corporate branding guidelines. To summarize, palettes are designed to make it easier for designers to choose web- and user-friendly colors in their data visualizations.

Alternative text description (alt text)

Alt text is a written description added to an image in order to convey meaning and explain the content. Formatting a visual in this manner enhances understandability of a UI element. Visually impaired users may use screen readers to understand the visualizations.¹³

The Alt text would include: The Undergraduate School Applications Surpass the Graduate and Professional for the first time Graph, seen in figure 1 above, shows a Line chart of the total number of applicants who submitted applications to each of the school categories of Graduate, Professional, and Undergraduate in the last five years. The Undergraduate School applications surpassed both the Graduate and Professional applications for the first time in 2023 with a total number of 1,200. Data was sourced from the College.

Another way to convey the information accessibly is to create a text only version and save it on a separate page that is linked in the image caption. For example, one would caption the previous graph as a “text-only description” for the Undergraduate School Applications Surpass the Graduate and Professional for the first time Graph.

Labels

Adding labels will add clarity to the visualization. Such text objects provide instructions or additional context about the data elements being displayed. Labels can be incorporated in the chart itself and/or reside in a separate area next to the chart. If done right, embedding labels in the context of the chart helps users to not have to guess at the numbers or their meaning. Looking at the following two bar graphs, both convey the same information. Which one, however, does a better job at instantly communicating the information quicker?

INSTEAD OF THIS

TRY THIS

Legends

Legends clarify and categorize the meaning of different elements on a data visualization and can often be used in lieu of data labels on a chart itself, such as seen in the previous figure. It is common to include legends when there is more than one category. They often help to reduce visual clutter and are popular in maps. Legends are typically placed below or parallel to a chart and may or may not include a title; this will depend on how obvious the categories being defined are. A good practice when including a legend title is to give a descriptive name instead of simply the obvious and default “Legend.” Looking at two versions of the same map below, one can see how a) replacing labels with an appropriate legend, and b) including a descriptive legend title, result in a more effective visualization. Note that the data points are being illustrated using sequential colors, whereby the percentages are ordered continuously from high to low.

INSTEAD OF THIS

TRY THIS

One more way this map can be made even more effective is better is by using a three-color diverging palette that segments on three percentage categories or points in order to highlight in which states female applicants account for over 50% of the total population, which is a noteworthy break point in the data. The title now also highlights the message even more. Please note some modifications to the fictitious data were made for emphasis.

Navigation

Navigation pertains to the use and exploration of a data visualization or dashboard. Navigation needs to be functional. The user needs to instinctively understand how to navigate without thinking excessively about it. When designing a data visualization, one common recommendation is to place the most relevant information at the very top (left aligned). Another design decision is to customize the default alignment of the Axis title to one that is read easier and more naturally by the audience, such as illustrated on the y axis of the second figure below.

Another recommendation is to include the filters panel (if applicable) in a consistent place.²¹ Additionally, including instructions on how to navigate are also helpful in guiding the user’s navigation process.

Device Compatibility

Dashboard dimensions need to be set according to the device that the user may use to interact with the information. Screen size does impact the way a user perceives the information digitally and if not accounted for may contribute to a negative overall experience. Optimizing the data visualization for different devices such as tablets, lap tops, or mobile phones will increase usability. Embedding responsive design methods, which include the layout dynamically shifting to accommodate the dimensions of a specific device, will create a smoother user experience.²²

Conclusion

Accessibility in data visualization means designing dashboards and interfaces so people with a wide range of disabilities can use them effectively. Grounded in the WCAG principles—perceivable, operable, understandable, and robust—accessibility requires attention to screen size, color, contrast, typeface, and font, especially as dashboards grow more interactive and complex. Color should be used intentionally and sparingly, avoiding reliance on color alone to convey meaning and limiting the total number of colors to reduce confusion. Designers can draw on established, accessible palettes from sources like W3C or institutional branding guidelines. Other key practices include adding alt text that clearly describes the chart type, data, and key message; or providing a separate, linked text-only description so screen reader users can fully access the information.

Text-based supports such as labels and legends also play a crucial role in accessibility. Clear labels embedded in or near charts help users quickly understand what values represent without guesswork, while well-designed legends can replace dense labels, reduce clutter, and clarify categories—especially when paired with descriptive titles instead of generic “Legend” headings. Effective navigation further supports accessibility by making it intuitive to explore a dashboard: placing key information at the top left, aligning axis titles for natural reading, keeping filters in a consistent location, and adding simple navigation instructions. Finally, designing for device compatibility through responsive layouts ensures that dashboards remain usable and legible across desktops, laptops, tablets, and mobile phones, preventing screen size from becoming a barrier to understanding the data.

The next part of the series will explore the role of user feedback in dashboard design. By observing how stakeholders use visualizations, listening to their questions, and incorporating their suggestions, designers can iteratively refine dashboards so they are not just compliant with guidelines, but truly usable and valuable in everyday decision-making.

Share

Leave a comment

Bibliography

1. Galitz WO. The essential guide to user interface design: an introduction to GUI design principles and techniques: John Wiley & Sons, 2007.

2. Ward MO, Grinstein G and Keim D. Interactive data visualization: foundations, techniques, and applications: CRC press, 2010.

3. Zhu Y. Measuring effective data visualization. In: International symposium on visual computing, Anonymous, pp.652-661: Springer.

4. Anonymous Alberto Cairo’s tips for more effective data visualization​, https://ijnet.org/en/story/alberto-cairos-tips-more-effective-data-visualization%E2%80%8B (2018, accessed Apr 21, 2023).

5. Anonymous ISO 9241-11:2018, https://www.iso.org/standard/63500.html (2018, accessed Nov 22, 2023).

6. Sridevi S. User interface design. International Journal of Computer Science and Information Technology Research 2014.

7. Roth SF, Lucas P, Senn JA, et al. Visage: A user interface environment for exploring information. In: Proceedings IEEE Symposium on Information Visualization’96, pp.3-12: IEEE.

8. Ezera A. Boost engagement with interactive data visualizations 2022.

9. Experience WLiRU. Analytics and User Experience | Nielsen Norman Group Full-Day Training, https://www.nngroup.com/courses/analytics-and-user-experience/ (2023, accessed Nov 22, 2023).

10. Salehi B. UX Researchers and Data Scientists: Work Together | Built In, https://builtin.com/design-ux/ux-data-science-work-together (2023, accessed Apr 22, 2023).

11. Mais V. Four User-Centered Strategies for Designing Useful Data Visualizations | by Valerie Mais | Nightingale | Medium, https://medium.com/nightingale/four-user-centred-strategies-for-designing-useful-data-visualizations-5c6daeb91c8f (2020, accessed Apr 21, 2023).

12. Pappas L and Whitman L. Riding the technology wave: Effective dashboard data visualization. In: Human Interface and the Management of Information. Interacting with Information: Symposium on Human Interface 2011, Held as Part of HCI International 2011, Orlando, FL, USA, July 9-14, 2011, Proceedings, Part I, Anonymous , pp.249-258: Springer.

13. Deakin J. The ultimate guide to designing accessible dashboards, https://bootcamp.uxdesign.cc/the-ultimate-guide-to-designing-accessible-dashboards-b140afe0029f (2023, accessed Jun 5, 2023).

14. Ognjanovic S, Thüring M, Murphy RO, et al. Display clutter and its effects on visual attention distribution and financial risk judgment. Appl Ergon 2019.

15. Paas F and Sweller J. Implications of cognitive load theory for multimedia learning. The Cambridge handbook of multimedia learning 2014.

16. Dillon A. User interface design. London: Macmillan, 2003.

17. Wong K. This is how design principles help you declutter visualizations, https://uxdesign.cc/this-is-how-design-principles-help-you-declutter-visualizations-25ba1f14bd0a (2021, accessed Apr 21, 2023).

18. Anonymous Computer Basics: Using Accessibility Features, https://edu.gcfglobal.org/en/computerbasics/using-accessibility-features/1/ (accessed Nov 22, 2023).

19. Anonymous Web Content Accessibility Guidelines (WCAG) 2.0, https://www.w3.org/TR/WCAG20/#guidelines (, accessed Nov 22, 2023).

20. Wexler S, Shaffer J and Cotgreave A. The big book of dashboards: visualizing your data using real-world business scenarios: John Wiley & Sons, 2017.

21. Anonymous Best Practices for Effective Dashboards, https://help.tableau.com/current/pro/desktop/en-us/dashboards_best_practices.htm (2008, accessed Sep 22, 2023).

22. Deakin J. The ultimate guide to designing accessible dashboards, https://bootcamp.uxdesign.cc/the-ultimate-guide-to-designing-accessible-dashboards-b140afe0029f (2023, accessed Jun 5, 2023).

23. Bin Habib M. The Intersection of Data Visualization and UX Design :: UXmatters, https://www.uxmatters.com/mt/archives/2023/03/the-intersection-of-data-visualization-and-ux-design.php (2023, accessed Apr 21, 2023).

Considering Clutter

In the area of data visualization, clutter is defined as an abundance of visual stimuli. Ognjanovic et al. conducted a study that focused on financial data visualizations and how subject matter experts use visual cues in a variety of displays with variations in visual clutter.¹⁴ According to the authors, clutter was deemed to have a negative impact on judgment, visual attention, and performance. This “display clutter” impacts user ability to interpret, understand, read, and assess information. An activity such as reading graphs and charts with numbers and text is considered a “conscious cognitive activity,” according to Paas and Sweller.¹⁵ When a user is presented with an excessive amount of information – related or not – in a small amount of space such as a data dashboard or mobile application, the user may experience a high cognitive load that may result in diminished human-computer interaction. Minimizing the number of steps that a user needs to take to understand a data visualization or providing clear explanations of how the data was compiled are both good techniques in reducing clutter.¹⁶ Removing unnecessary borders around or gridlines within charts additionally help to reduce visual noise. Also coming up with more effective ways of adding value labels or possibly excluding them completely is an additional technique.¹⁷

INSTEAD OF THIS

TRY THIS

Using Effective Visuals and Interaction Features

Apart from the choice of charts or graphs, a visual may be enhanced and become more impactful when one or more of the following UI-influenced design choices are applied:

• Include a meaningful title: this increases dashboard usability, enhancing the ability of users to read the content and minimizing confusion about what the chart is illustrating. The following chart seeks to emphasize the factor that powered the record number of early admissions applicants and does so by using an active title and subtitle for further context. The second version excludes the redundant “Figure 1.” text in the beginning of the title. The purple color is also added to the categorical variable of “in-state” so as to make the visual connection to the bars being highlighted.

INSTEAD OF THIS

TRY THIS

• Include the values that the user has filtered on: this customizes the view inside the title of the graph and serves as additional context and a point of reference. In this way, the user instantly knows how the view has been filtered and what data values are being included. This is a useful tip especially when sharing a screenshot of the visualization with another person who may not know how the data is being filtered in the first place.

• Add data labels on select areas of a graph: this allows the user to easily understand and locate the most critical information being shown in a series or position.

INSTEAD OF THIS

TRY THIS

• Add graphics such as photos, icons, or pictures: do this only if their inclusion adds value or relevance to the message or story being told through the data. If the photos are not clear or perhaps pixelated or if they do not support what a graph is showing then they should be excluded from the visual all together. In the case of the following example, one may argue that both visuals deliver the same message and emphasize the percentage distribution. However, the version with the human icons may have a humanizing effect on the graphical representation. In the end, it is a stylistic decision on the part of the designer.

  INSTEAD OF THIS

  

  TRY THIS

  

• Highlight the most important point using an “at-a-glance” chart: when needing to highlight a single point or number in order to focus the audience’s attention on that, opt for a chart that centers on that point. For instance, look at how much more effectively the waffle chart on the right delineates the percentage of applicants from North Carolina. This chart type is very useful in providing an at-a-glance insight.

INSTEAD OF THIS

TRY THIS

Conclusion

We have seen how “Using Interface Design Methods in Data Visualization” shifts from foundational decisions—such as context, business needs, audience, and graph selection—to the experience layer: how a visualization actually feels and functions for users. A central focus is the management of clutter. Research shows that excessive visual stimuli, or “display clutter,” can hinder judgment, attention, and performance, especially in cognitively demanding tasks like reading charts and dashboards. Overloaded screens increase cognitive load and can diminish the quality of human-computer interaction. Practical strategies to reduce clutter include minimizing the steps needed to understand a visualization, clearly explaining how data was compiled, removing unnecessary borders and gridlines, and rethinking or selectively applying value labels to reduce visual noise.

Beyond decluttering, this part of the series highlighted UI-influenced techniques for making visuals and interactions more effective and meaningful. Designers can improve impact by using clear, descriptive titles (and subtitles) that emphasize key insights and reflect any filters applied, so users instantly know what subset of data they’re viewing. Adding data labels only to the most important points helps users quickly locate critical information without overwhelming them. Carefully chosen graphics—such as icons or photos—can humanize the data when they are relevant and high quality, but should be omitted if they distract from the message. Finally, “at-a-glance” charts, like a focused waffle chart, are recommended for highlighting a single, important number or percentage. Collectively, these practices help transform well-chosen charts into clear, engaging, and purposefully designed visual experiences.

Next up in this series, we will examine accessible features in more depth, including how color and contrast, text alternatives, labels, legends, navigation, and device compatibility can be designed to support perceivable, operable, understandable, and robust visualizations.

Share

Bibliography

1. Galitz WO. The essential guide to user interface design: an introduction to GUI design principles and techniques: John Wiley & Sons, 2007.

2. Ward MO, Grinstein G and Keim D. Interactive data visualization: foundations, techniques, and applications: CRC press, 2010.

3. Zhu Y. Measuring effective data visualization. In: International symposium on visual computing, Anonymous, pp.652-661: Springer.

4. Anonymous Alberto Cairo’s tips for more effective data visualization​, https://ijnet.org/en/story/alberto-cairos-tips-more-effective-data-visualization%E2%80%8B (2018, accessed Apr 21, 2023).

5. Anonymous ISO 9241-11:2018, https://www.iso.org/standard/63500.html (2018, accessed Nov 22, 2023).

6. Sridevi S. User interface design. International Journal of Computer Science and Information Technology Research 2014.

7. Roth SF, Lucas P, Senn JA, et al. Visage: A user interface environment for exploring information. In: Proceedings IEEE Symposium on Information Visualization’96, pp.3-12: IEEE.

8. Ezera A. Boost engagement with interactive data visualizations 2022.

9. Experience WLiRU. Analytics and User Experience | Nielsen Norman Group Full-Day Training, https://www.nngroup.com/courses/analytics-and-user-experience/ (2023, accessed Nov 22, 2023).

10. Salehi B. UX Researchers and Data Scientists: Work Together | Built In, https://builtin.com/design-ux/ux-data-science-work-together (2023, accessed Apr 22, 2023).

11. Mais V. Four User-Centered Strategies for Designing Useful Data Visualizations | by Valerie Mais | Nightingale | Medium, https://medium.com/nightingale/four-user-centred-strategies-for-designing-useful-data-visualizations-5c6daeb91c8f (2020, accessed Apr 21, 2023).

12. Pappas L and Whitman L. Riding the technology wave: Effective dashboard data visualization. In: Human Interface and the Management of Information. Interacting with Information: Symposium on Human Interface 2011, Held as Part of HCI International 2011, Orlando, FL, USA, July 9-14, 2011, Proceedings, Part I, Anonymous , pp.249-258: Springer.

13. Deakin J. The ultimate guide to designing accessible dashboards, https://bootcamp.uxdesign.cc/the-ultimate-guide-to-designing-accessible-dashboards-b140afe0029f (2023, accessed Jun 5, 2023).

14. Ognjanovic S, Thüring M, Murphy RO, et al. Display clutter and its effects on visual attention distribution and financial risk judgment. Appl Ergon 2019.

15. Paas F and Sweller J. Implications of cognitive load theory for multimedia learning. The Cambridge handbook of multimedia learning 2014.

16. Dillon A. User interface design. London: Macmillan, 2003.

17. Wong K. This is how design principles help you declutter visualizations, https://uxdesign.cc/this-is-how-design-principles-help-you-declutter-visualizations-25ba1f14bd0a (2021, accessed Apr 21, 2023).

18. Anonymous Computer Basics: Using Accessibility Features, https://edu.gcfglobal.org/en/computerbasics/using-accessibility-features/1/ (accessed Nov 22, 2023).

19. Anonymous Web Content Accessibility Guidelines (WCAG) 2.0, https://www.w3.org/TR/WCAG20/#guidelines (, accessed Nov 22, 2023).

20. Wexler S, Shaffer J and Cotgreave A. The big book of dashboards: visualizing your data using real-world business scenarios: John Wiley & Sons, 2017.

21. Anonymous Best Practices for Effective Dashboards, https://help.tableau.com/current/pro/desktop/en-us/dashboards_best_practices.htm (2008, accessed Sep 22, 2023).

22. Deakin J. The ultimate guide to designing accessible dashboards, https://bootcamp.uxdesign.cc/the-ultimate-guide-to-designing-accessible-dashboards-b140afe0029f (2023, accessed Jun 5, 2023).

23. Bin Habib M. The Intersection of Data Visualization and UX Design :: UXmatters, https://www.uxmatters.com/mt/archives/2023/03/the-intersection-of-data-visualization-and-ux-design.php (2023, accessed Apr 21, 2023).

Let’s set the scene. You’ve been asked to build a dashboard. The deadline is two weeks out. You open your laptop, pull up your data, and think: I’ve got time. I’ll figure it out as I go.

Two weeks later, you’re still waiting on data access, your stakeholder has changed the requirements twice, and you’re submitting a half-finished version at 11:47 PM the night it’s due.

Sound familiar? You’re not alone — and it’s not because you’re bad at your job. It’s because most analysts are never taught to treat a dashboard build the way it actually deserves to be treated: as a project.

The Hidden Complexity of ‘Just a Dashboard’

Here’s a question worth considering: When was the last time a dashboard took exactly as long as you thought it would?

Dashboards look deceptively simple from the outside. A few charts, some filters, maybe a summary table because we all know how much users love their ability to download to Excel! But behind the scenes, there’s stakeholder alignment, data sourcing, cleaning, modeling, design decisions, feedback loops, and quality control/assurance (QA). That’s a project — with moving parts, dependencies, and real risk of delay.

Yet most analysts approach a dashboard build the way most of us approach assembling IKEA furniture: dive in, reference the instructions only when something goes wrong, and hope for the best. (We’ve all been there with the leftover screws – I know I have been rather creative with a couple of bookshelves).

💡 Ask yourself: Do you start building before you’ve fully defined what you’re building?

What Project Management Actually Means for Analysts

Before anyone checks out — no, this is not about Gantt charts and resource allocation matrices. Nobody’s asking you to become a project manager.

But borrowing a few basic principles from project management can transform how reliably and confidently you deliver analytical work. Specifically: breaking work into defined phases with clear milestones before you start building.

This matters for three reasons.

1. First, it forces you to surface problems early. Data access issues, unclear requirements, scope creep — these are much easier to handle on Day 2 than Day 12.

2. Second, it gives you something to communicate. When a stakeholder asks where things stand, you have a real answer instead of a vague ‘it’s coming along.’

3. Third, it protects your time. When new requests come in half way through your build, a milestone plan lets you say ‘yes, and here’s what that means for the timeline’ instead of silently absorbing scope without anyone knowing.

A Practical Framework: The Dashboard Milestone Map

Here’s a simple five-phase approach you can apply to almost any dashboard build. The example below assumes a two-week delivery window, but the structure scales up or down with the timeline.

A few things worth noting about this framework.

• The data audit phase (Days 3–4) is the one most people skip — and the one that causes the most delays. Always validate your data before building anything. Finding out that a key field is inconsistently formatted on Day 9 is a bad time.

• Day 10 as buffer is not a luxury; it’s a requirement. If everything goes perfectly, use it for polish. If something slipped, use it to recover. Never plan a dashboard build with zero buffer.

• The milestones are your early warning system. If you hit Day 5 and you don’t have a working draft yet, that’s not a problem to fix on Day 14. That’s a flag to raise now.

💡 Think of milestones not as checkboxes — but as questions you’re answering along the way: Am I on track? Do I know something now that changes the plan? Does my stakeholder need to know anything?

The Habit That Changes Everything: Estimate Before You Build

Before starting any dashboard, try this: write down every sub-task involved in building it. Not the broad phases — the actual tasks. ‘Pull data from source system.’ ‘Check for nulls and duplicates.’ ‘Build date filter logic.’ ‘Draft layout.’ ‘Get stakeholder sign-off on v1.’

Estimate the time for each. Then add 20–30% as a buffer. If that total doesn’t fit before your deadline, you’ve just discovered something important — and you’ve discovered it before the build, not after.

This exercise also has a useful side effect: it makes the work feel manageable. A two-week dashboard build can feel overwhelming as a monolith. As twelve discrete tasks with time estimates, it becomes a plan.

💡 When is the right time to flag a potential delay? Before it becomes a certainty. Early communication about a risk is a sign of professionalism. Silence followed by a missed deadline is not.

A Note for Project Managers Reading This

If you work with analysts and you’re reading this thinking ‘yes, I wish they’d all do this’ — fair. But it’s worth asking: are your analysts set up to work this way?

Do they have enough lead time before a deadline to actually plan the work? Are requirements clear and stable when they start building? Is there a low-friction way for them to flag scope changes or data issues without it feeling like an escalation?

The milestone framework works best when the environment supports it. Analysts who are expected to absorb last-minute requests without timeline adjustments will always look like they’re bad at managing time — even when the real issue is something else entirely. Good project management is a two-way street.

Bringing It All Together

The shift being proposed here isn’t big. It doesn’t require new tools, new training programs, or a complete change in how you work. It’s really just three habits:

1. Plan before you build. Map the phases, estimate the tasks, identify the dependencies. Spend an hour at the start of a project doing this — it will save you many hours at the end.

2. Use milestones as checkpoints. Not just to track progress, but to prompt the question: do I need to flag anything? Is anything going to affect the deadline?

3. Communicate early and proactively. A heads-up on Day 5 is manageable. Silence until Day 13 is a problem.

Dashboards are analytical work, yes. But they’re also delivered work — which means they have stakeholders, deadlines, and real consequences when they’re late or incomplete. Treating the build process with the same rigor you bring to the analysis itself isn’t overhead. It’s professionalism.

A Personal Note: Deadlines Are Not Universal — But That Doesn’t Make Them Optional

Here’s something I’ve learned from working across very different organizational cultures: not everyone thinks about time the same way.

Early in my career, I worked in solution development for a global 4PL company — a world where RFP responses ruled the calendar. If a client submission was due at 5:00 PM on a Thursday, it was due at 5:00 PM on a Thursday. Not Thursday-ish. Not end of week. There was no negotiating with a procurement deadline, and there was no sympathy from a client evaluating twelve competing bids if you asked for an extension. That environment drilled into me a very specific relationship with deadlines: they are fixed, they are respected, and you build your entire work plan around honoring them.

It probably helped that the company was German-founded. And if you’ve ever worked closely with German business culture, you’ll know that timelines are not a suggestion — they’re closer to a moral commitment. There’s a reason Germany has a reputation for precision and reliability in industry. The clock is taken seriously.

Contrast that with — and I say this with genuine affection — the average Greek approach to time in business context. As someone who has navigated both, I can tell you that ‘I’ll be there in five minutes’ and ‘the report will be ready soon’ can mean rather different things depending on which cultural lens you’re looking through. Time in many cultures is relational and contextual, not mechanical. A deadline shifts because relationships matter more than schedules, or because flexibility is seen as a sign of trust rather than a failure of planning. Here is an interesting read for anyone who is interested in learning more about cultural differences with regard to the concept of time: Edward Hall’s “The Dance of Life: The Other Dimension of Time” talks about monochronic vs. polychronic time — where polychronic cultures treat time as subordinate to relationships and context, not as a fixed, mechanical resource.

Neither worldview is wrong, exactly. But they create real friction when they collide — especially in cross-functional teams, global organizations, or any environment where one person’s ‘flexible’ is another person’s ‘late.’

💡 Have you ever missed a deadline and genuinely not understood why it was a problem? Or held firm to a timeline that your colleagues found unnecessarily rigid? Chances are, you were experiencing a difference in how deadlines are culturally understood — not just a difference in work ethic.

I’ve also had to recalibrate my own expectations when moving into higher education. Coming from a fast-paced logistics and solutions environment — where a delayed deliverable could cost a contract worth millions — stepping into the rhythms of an academic institution required real adjustment. Timelines in higher ed often move at the speed of committees, academic calendars, and shared governance. What felt like urgency to me wasn’t always read the same way by others. And that’s okay — but it does mean you have to be more deliberate about building shared expectations, not just assuming everyone in the room shares your definition of ‘on time.’

This is exactly why the milestone framework in this post matters beyond just personal productivity. It creates a shared reference point. When you document phases, assign dates, and agree on milestones at the start of a project, you’re not just organizing your own work — you’re establishing a common language around time and accountability that cuts across cultural and organizational differences. The milestone plan becomes the neutral ground.

Whatever industry you’re in, whatever culture you work within: the goal isn’t to impose one group’s relationship with time onto everyone else. It’s to make expectations visible, explicit, and agreed upon — so that when something shifts, everyone knows it, and the conversation happens early rather than after the fact.

And on the bright side: once you’ve done this a few times, scoping a dashboard build becomes second nature. The leftover screws become a distant memory.

Learner analysis as data visualization

Learner analysis is essentially structured data about humans: demographics, prior knowledge, experience, motivations, and constraints (McDonald & West). When we treat that as data, we can start to visualize it in ways that inform design decisions instead of relying on vague intuition.

In my upcoming presentation for HEUG Alliance, “Communicating Tech to Non-Techies: The Sequel,” my audience spans higher ed professionals from admins to IT managers, both technical and non-technical. Rather than think of them as a blur of job titles, I mentally bucket them into those five audience types—Novice, Specialist, Peer, Manager, Executive—and imagine each group as a slice of a stacked bar chart. Then I design problem-based scenarios and a small role-play that hit those slices in different ways, so each “segment” can see themselves and their communication challenges in the examples.

Think:

• A bar chart of experience levels across your audience (novice, specialist, peer, manager, executive).

  

• A sunburst of comfort levels with key concepts or tools.

  

• A simple matrix that crosswalks “learner type” to “preferred explanation style” or “examples that resonate.”

  

Project management as the frame for those visuals

If learner analysis is the data, project management is the frame and the blue tack that keep the picture straight (and stuck) on the wall. Communication plans, needs assessments, and timelines are all, in a sense, visualization tools for complexity (Wiley, 2018). They help us answer questions like:

• Who needs what information, in what format, and when?

• Where are the decision points, handoffs, and risks?

• How do the needs of learners and stakeholders show up over the life of the project?

Section 7.5 of Wiley’s Communications Planning chapter emphasizes understanding “the information needs of the various members and stakeholders of the project,” which is just another way of saying: map out who needs to see which part of the data about learners, at what time, and in which medium (Wiley, 2018). A communication matrix is basically a table-based visualization of that flow.

For example, if you’re designing an online class on the Basics of Student Aid for first-level business analysts in a financial aid office, you can visualize the project like this:

• A timeline that shows when you’ll meet with SMEs, when drafts are due, and when pilot testing happens.

• A stakeholder diagram that clarifies who gives input on learning objectives vs. who approves them.

• A simple gap chart that compares “current skills” vs. “required skills” for those analysts.

All of those visuals make it easier to see where learner data needs to show up and who is responsible for acting on it.

Visualizing gaps and pace layering

Dirksen’s discussion of gap identification and “pace layering” naturally lends itself to visual thinking (Dirksen, 2015).

When she talks about aligning gaps with goals and then proposing solutions, I immediately think of:

• A before/after bar chart: current performance vs. desired performance.

• A layered diagram: fast-changing elements (tools, tactics, specific procedures) on the outer rings, and slow-changing elements (core mental models, deep domain understanding) in the center.

Pace layering—understanding that some aspects of learning change quickly while others evolve slowly—is especially powerful when visualized. You might sketch:

• Outer layer: quick-win microlearning, job aids, FAQs.

• Middle layer: course modules, practice activities, coaching.

• Inner layer: long-term mindset shifts, conceptual frameworks.

From a project management standpoint, this kind of diagram helps teams decide where to invest time and resources. Fast-changing layers might call for flexible, easily updated assets, while slow layers might justify deeper, more intensive design up front. It also supports risk management by making it clear which layers are most vulnerable to change over the life of the project.

Milestones, risks, and resources as charts and matrices

Key project management principles—setting milestones, allocating resources, and managing risk—can all be made more actionable when we treat them as visual stories of the work (Wiley, 2018). In instructional design projects, this might look like:

• A milestone timeline that shows when learner research happens relative to content development, prototyping, and evaluation.

• A resource allocation chart that indicates SME availability, ID workload, and review windows.

• A risk matrix that includes communication breakdowns, late SME input, or shifting goals as explicit, visible risks.

Here is a short section of a project management plan that really helped me plan, design, and execute a data storytelling course I delivered at the College Business Management Institute at the University of Kentucky:

The scope document itself becomes the central artifact where these visuals live and evolve, capturing any changes, disruptions, or new constraints (Wiley, 2018). It’s also where Conway’s Law becomes painfully visible: if communication between teams is fragmented, you can practically see the system fracturing in your diagrams. Learning Loop’s summary of Conway’s Law underscores how misaligned structures and communication patterns result in disjointed products—something we can literally trace in our project maps and dependency charts (Learning Loop, n.d.).

When communication breaks down, you often see it first in the visuals: milestones slip, dependency arrows multiply, risk levels shift from yellow to red. That’s a signal that the instructional design process is at risk of producing something misaligned with learner needs.

Bringing it all together: data, visuals, and lived experience

In my current role as a data analyst, I straddle the worlds of data, story, and learning. I build learning experiences for my own audiences, but I don’t always work within a traditional project team or against hard deadlines. The upside is that I can experiment freely with how I visualize both the content and the learners themselves.

Behind the scenes, I think about:

• Who is enrolling (roles, departments, experience levels).

• What their “before” vs. “after” behavior looks like by comparing their pre-course survey responses to their post-course. How did their initial expectations about the course compare with their overall experience?

• How to represent those shifts in simple visuals—completion charts, quiz performance distributions, or even informal “profile” sketches of common learner types.

When I look back to my time as a Solution Developer in logistics, I see similar patterns. Responding to RFPs and RFQs required us to use a solid project management (PM) framework and to truly understand clients as “audiences” and then coordinate across five departments to tell one coherent story. That was one of the most challenging jobs I ever had but what made it slightly easier was teamwork and the fact that every single individual worked around out a specific PM model. Project management was essentially the art of visualizing that complexity: tracking requirements, workloads, constraints, and timelines so we could tailor each proposal to the client’s specific context. It’s striking how similar that is to learner analysis in instructional design.

Both in RFP work and in instructional design:

• You start by understanding your audience in a structured, often visual way.

• You use PM tools to map the work of responding to that audience.

• You rely on continuous feedback—client reactions or learner performance—to refine your next iteration.

From a data visualization lens, better project management practices don’t just improve outcomes because they are “more organized.” They improve outcomes because they make invisible dynamics visible: who the learners are, what gaps matter most, which risks are emerging, and how well the project is tracking toward its goals. When we can literally see those things—on charts, in matrices, in timelines—we’re far better equipped to design learning that resonates with real humans, not just with our initial assumptions.

References

Dirksen, J. (2015). Design for how people learn [Kindle version]. New Riders. https://a.co/d/gHIPUdx

Larson, M. B., & Lockee, B. B. (n.d.). Project management defined. In Project management for instructional designers. EdTech Books. Retrieved February 8, 2025, from https://edtechbooks.org/pm4id/1-1-project-management-defined

Learning Loop. (n.d.). Conway’s Law. Learning Loop. Retrieved February 9, 2025, from https://learningloop.io/glossary/conways-law

McDonald, J., & West, R. (n.d.). Conducting a learner analysis. In Design for learning: Principles, processes, and praxis. LibreTexts. Retrieved February 8, 2025, from https://socialsci.libretexts.org/Bookshelves/Education_and_Professional_Development/Design_for_Learning_-_Principles_Processes_and_Praxis_%28McDonald_and_West%29/01%3A_Instructional_Design_Practice/01%3A_Understanding/1.05%3A_Conducting_A_Learner_Analysis​

Wiley, D. (2018). Communications planning. Project management for instructional designers. https://edtechbooks.org/pm4id/7-5-communications-planning​

Wiley, D. (2018). Project definition and context. Project management for instructional designers. https://edtechbooks.org/pm4id/1-2-project-definition-and-context​

Wiley, D. (2018). Project scope. Project management for instructional designers. https://edtechbooks.org/pm4id/7-2-project-scope​

In the first part of this series, we explored how incorporating user design principles into data visualizations can enhance clarity, engagement, and decision-making. Building on that foundation, the next installment will introduce core design considerations—such as context, constraints, and delivery environment—that shape every visualization, and then move into two closely related elements of practice: understanding who the audience is, what they need from the data, and how to choose appropriate graphs that effectively communicate those insights.

Knowing the Audience

Discovering the audience is step one in the iterative design process, as most research shows.¹¹ This strategy is used to identify the users, their needs, expectations, experience, possible biases, and background, all of which can help in organizing and categorizing the data to use, the navigational flow, the graphical and statistical elements to include, the tool to use, and other visual elements. Getting to know the audience by way of interviews, workshops, surveys, or brainstorming sessions can minimize flawed drafting of initial requirements, optimize workflows, create a solid prototype, simplify the design of the data visuals, and save time and money. If the audience is not well versed in correlation coefficients or statistical regression analysis, the designer needs to consider how to compile the data and what charts to use to visualize information in a way that makes sense and is easy to understand, yet without losing focus of the issue or problem that needs solving.

Understanding the Business and Outlining Objectives

It is crucial to know what information needs to be conveyed to address the topic of the data being explored. Understanding the business function and any relevant operations are key. This includes gathering business requirements, outlining specifications, listing the data that needs to be gathered and analyzed, and determining what training and documentation will need to be created. A foundational principle in UI is the setting of learning objectives or building of action maps, which are both methods for streamlining the design process. A user’s experience with a visualization will also be largely influenced by the setting of specific objectives by the designer. Only when considering these objectives will the designer be in a capable position of building a dashboard or visualization that caters to the needs of the intended users who ideally will be able to make sense in full of the data visualized.

Choosing Appropriate Graphs

Data visualization at its core is the collection of appropriate graphs or charts, along with any necessary contextual information such as text or graphics, to convey specific information to end users. In their 2011 conference paper, Pappas and Whitman refer to a wave of technology in regards to creating effective data dashboards.¹² They anecdotally see people as being flooded with data dashboards as institutions work toward providing ample information using effective and intuitive tools to help their customers make data-driven decisions. Choosing the right charts is important to the user’s interpretation and understanding of the data presented. There are plenty of resources online that aid the practitioner in selecting the right chart to relay the desired message to the audience. The following is a non-exhaustive list of typical applications of graphs and charts, accompanied by examples that demonstrate their use in addressing questions.

Question: How do you tell a story with a single number or data point?

Question: How do you compare data?

Question: How do you show survey data?

Question: How do you depict change over time?

Question: How do you show relative performance?

Conclusion

This part of the series built on the idea that user-centered design principles can make data visualizations clearer, more engaging, and more useful for decision-making. It introduced key design considerations—such as context, constraints, and delivery environment—that underpin every visualization, then focused on two closely linked practices: identifying who the audience is and what they need from the data, and choosing graph types that communicate those insights effectively.

Knowing the audience is the first step in an iterative design process. By understanding users’ needs, expectations, experience, biases, and backgrounds—often through interviews, workshops, surveys, or brainstorming—designers can better determine what data to include, how to structure navigation, which tools and visuals to use, and how to simplify complex concepts without losing sight of the core problem. Equally important is understanding the business context and clearly defining objectives: gathering requirements, specifying data, and planning training and documentation enable designers to build dashboards that truly support users’ goals. Within this framework, selecting appropriate charts becomes critical, especially as institutions and their stakeholders are increasingly inundated with dashboards. Using resources and best practices to match chart types to specific questions helps ensure that visualizations remain interpretable, actionable, and aligned with the needs of their intended users. Next up in this series, we will concentrate on how to reduce clutter, and make purposeful use of visuals and interactive features.

Data visualizations are now central to how colleges and universities surface insights, communicate performance, and support decision-making. Yet even well-prepared data can fall flat if the way it is presented overwhelms, confuses, or fails to meet the needs of its intended audience. As institutional leaders, faculty, and staff are increasingly asked to engage with dashboards and analytics tools, the overall experience of interacting with data has become just as important as the numbers themselves.

User Interface (UI) design and User Experience (UX) principles offer a powerful lens for improving that experience. UI focuses on the visual and interactive elements of a system—layout, navigation, labels, colors, and controls—while UX considers how effectively and efficiently people can use a product to accomplish their goals. When data visualization practitioners structure charts and dashboards to minimize clutter and optimize visibility, accessibility, and understandability, they are already drawing on UI and UX techniques, whether they name them as such or not.

This article series explores how foundational UI principles, drawn from human-computer interaction and usability research, can be applied to data visualizations in higher education. It highlights factors such as audience, business context, appropriate graph selection, clutter reduction, visual and interactive features, accessibility, and feedback as key levers for more effective, user-centered dashboards and reports. By connecting practices from UI/UX and data visualization, the goal is to offer practical guidance that helps analysts and data professionals design interfaces that not only look good, but meaningfully support users in exploring information, understanding patterns, and making informed decisions. Let’s start, shall we?

About Data Visualization

Data visualizations include a variety of methods to present complex data in an understandable way with the ultimate goal of providing insights and supporting decision-making whether at an operational or executive level.² The effectiveness of a data visualization is measured in several ways and there are numerous best practices outlined by practitioners and researchers alike.

In Zhu’s conference paper, the author states, “existing definitions of effectiveness are incomplete and often inconsistent,” which makes evaluating the efficacy of a visualization challenging.³ Nonetheless, definitions and measures such as the extent to which a visualization improves a user’s tasks, do exist and provide a foundation for information visualization assessment.

For visualizations to be effective, multiple factors such as data extraction, wrangling, chart selection and others need to be carefully measured, structured, and executed. When considering the front-end, user-facing experience, Alberto Cairo provides the following tips:⁴

• Add context to increase understanding of the data and subject being visualized.

• Find patterns in the data to add clarity.

• Focus on the facts when adding creativity.

• Emphasize the need of the audience and deliver the solution using the appropriate tool.

About User Interface Design

The history of user interface design goes back to the development of the first computer, According to Galitz, a user interface is defined as a system that users interact with. User interface design is “a subset of a field of study called human-computer interaction (HCI)” that focuses on how people work effectively with computers.¹

A big component of UI is usability, which pertains to how effectively and efficiently users are able to interact with a product and the outcome of that interaction.⁵ Developers of user interfaces are required to study technical and non-technical factors when building user products. Some steadfast rules of user interface design include the following:

• Give the user control by providing opportunities for interaction as intuitively as possible.

• Reduce the user’s cognitive load so as to minimize errors and not burden the user’s memory.

• Maintain consistency by providing meaningful context displayed in an organized, instinctual way on the interface, which will set expectations for continual use across interrelated applications.⁶

Using Interface Design Methods in Data Visualization

Using interface design methods in data visualization means treating dashboards and charts as interactive products rather than static outputs. By borrowing techniques from user interface design—such as structuring layouts, guiding navigation, and intentionally shaping interactions—practitioners can better align visualizations with user needs and behaviors. This approach provides the foundation for the benefits outlined in the following section.

There are several benefits to incorporating user design principles in the creation of data visualizations, and this is becoming increasingly relevant as users seek more from the data they see and use. Applying core design principles can help improve the clarity and impact of the message delivered in data visualizations.

Benefits of incorporating UI in Data Visualizations

Existing research on information-focused user interfaces examines how these interfaces can best support users in exploring the data they are presented. The collaborative work of S.F. Roth et al. investigates Visage, a prototype for information analysis that allows users to explore visualizations using an information-centric approach, giving them control to manipulate visualizations for their own needs.⁷ The takeaway from this study is that giving users the ability to explore data may, in turn, lead to increased user engagement.⁷

Numerous corporations highlight that the key to connecting with audiences is to create opportunities for dynamic and useful engagement with content.⁸ This can improve users’ understanding of the content and expedite decision-making processes.⁸

Currently, significant efforts are being made to incorporate user design techniques more smoothly and iteratively into data visualizations. Researchers and professionals are publishing more guides and best practices on the topic and providing perspectives for further consideration. Consulting companies such as Nielsen Norman Group offer professional courses on analytics and user experience.⁹ Efforts are also being made to blend the knowledge and expertise of data scientists and UI designers to implement better quality products by using analytics in research to discover what user experience is really like.¹⁰

Conclusion

Data visualizations have become central tools in higher education for surfacing insights, tracking performance, and guiding decisions. Yet even high-quality data can lose impact if dashboards and charts are confusing, cluttered, or poorly aligned with user needs. As more institutional leaders, faculty, and staff interact with analytics tools, the experience of using data—how intuitive, efficient, and meaningful it feels—has become just as important as the underlying numbers.

User Interface (UI) and User Experience (UX) principles offer a powerful lens for improving that experience. Connecting UI/UX practices with data visualization helps analysts design dashboards and reports that look polished and, more importantly, help users explore information, recognize patterns, and make informed decisions. By structuring layouts, guiding navigation, and intentionally shaping interactions, practitioners can better align visualizations with user needs and behaviors.

Research on information-focused user interfaces underscores the benefits of this approach. The work of S.F. Roth et al. on Visage, a prototype for information analysis, shows that giving users control to manipulate and explore visualizations can increase engagement and deepen understanding. Many corporations similarly emphasize creating opportunities for dynamic, meaningful interaction with content to strengthen comprehension and accelerate decision-making. Among numerous UI design factors, key considerations for more effective data visualizations include knowing the audience, understanding the business context, choosing appropriate graph types, managing clutter, leveraging visual and interactive features, and ensuring accessibility. The field is moving toward smoother, iterative integration of these techniques, supported by growing bodies of best practices, professional training from groups such as Nielsen Norman Group, and closer collaboration between data scientists and UI designers. Subsequent parts of this series will explore these elements in more depth, with practical examples for designing and implementing user-centered data visualizations in higher education. Stay tuned for more!

Sources

1. Galitz WO. The essential guide to user interface design: an introduction to GUI design principles and techniques: John Wiley & Sons, 2007.

2. Ward MO, Grinstein G and Keim D. Interactive data visualization: foundations, techniques, and applications: CRC press, 2010.

3. Zhu Y. Measuring effective data visualization. In: International symposium on visual computing, Anonymous, pp.652-661: Springer.

4. Anonymous Alberto Cairo’s tips for more effective data visualization​, https://ijnet.org/en/story/alberto-cairos-tips-more-effective-data-visualization%E2%80%8B (2018, accessed Apr 21, 2023).

5. Anonymous ISO 9241-11:2018, https://www.iso.org/standard/63500.html (2018, accessed Nov 22, 2023).

6. Sridevi S. User interface design. International Journal of Computer Science and Information Technology Research 2014.

7. Roth SF, Lucas P, Senn JA, et al. Visage: A user interface environment for exploring information. In: Proceedings IEEE Symposium on Information Visualization’96, pp.3-12: IEEE.

8. Ezera A. Boost engagement with interactive data visualizations 2022.

9. Experience WLiRU. Analytics and User Experience | Nielsen Norman Group Full-Day Training, https://www.nngroup.com/courses/analytics-and-user-experience/ (2023, accessed Nov 22, 2023).

10. Salehi B. UX Researchers and Data Scientists: Work Together | Built In, https://builtin.com/design-ux/ux-data-science-work-together (2023, accessed Apr 22, 2023).

My take on learning styles: helpful, but handle with care

When I took the VARK inventory, I came out as multimodal, which felt spot-on: I like a mix of visual, auditory, reading/writing, and hands-on experiences.

VARK focuses on how we prefer to receive information across those four modalities, and multimodal learners tend to draw on several of them rather than just one. Honey and Mumford, in contrast, helped me think more about how I engage with learning—coming out strong as a Reflector and Pragmatist meant I like to observe, think things through, and then apply concepts in concrete ways.

Those two profiles together give me a kind of “learning fingerprint”: I gather information in multiple ways, then slow down to reflect and look for practical application. That doesn’t mean instruction must always match those preferences to be effective, but it does help me understand why certain activities feel more natural to me than others.

Why learning styles aren’t the whole story

Dirksen (2015) reminds us that “our learners are not necessarily like us,” which is a critical guardrail against designing only for our own preferences (Location 501). Learning style inventories can nudge us to ask better questions about learners—Who are they? What do they need? What motivates them?—but they shouldn’t be the sole basis for instructional decisions (Dirksen, 2015; Understanding learning styles, 2024).

Research on learning styles in higher education has repeatedly failed to confirm that matching teaching to a labeled style (for example, “only visual” or “only auditory”) reliably improves outcomes, even though the idea remains popular with many educators. What seems more promising is using learning preferences as one data point among many, and then building rich, multimodal experiences that give learners multiple ways to engage with the same core ideas.

So what are learning styles good for?

Looking across models like VARK and the Felder–Soloman index, a few practical themes stand out for educators and instructional designers. The University of Waterloo’s guidance on learning styles, for example, emphasizes using them to diversify instruction and foster inclusion, not to pigeonhole students into fixed categories. Felder and Soloman’s work highlights that some learners lean toward active vs. reflective, visual vs. verbal, or global vs. sequential approaches—differences that can inspire varied activities rather than rigid labels.

That translates into design moves like:

• Adding interaction, discussion, and hands-on tasks for active learners.

• Including diagrams and visual frameworks for visual learners.

• Framing content with “big picture first” for more global learners.
  These are not separate courses for separate “types,” but different entry points into the same concepts.

How this shows up in my data training

As a multimodal, reflector–pragmatist learner, I naturally gravitate toward inclusive, mixed-format design—something I’ve tried to bring into my PS Query teaching (PeopleSoft Query (PS Query) is a user-friendly, web-based reporting tool that allows users to extract real-time data from the PeopleSoft database without writing complex SQL code). In this online course that has looked like combining visuals (infographics summarizing SQL commands and their PS Query equivalents), written guides, and hands-on labs in a test database environment so learners can immediately apply what they see and read.

One artifact I’ve created is a short PDF rapid guide that’s been described as “easy to use and pleasant to scroll through,” which intentionally blends visual layout, stepwise structure, and prompts for hands-on practice.

That kind of design can speak to multiple preferences at once, even when the primary medium (say, a PDF or video) pushes you more heavily toward one modality. The real challenge—and opportunity—is deciding what to emphasize when time is short or the format is constrained, without defaulting to just one style and ignoring everyone else.

Learner personas: a better complement to learning styles

Learner personas helped me move from abstract preferences to real people. Connie Malamed (2024) frames personas as well-developed fictional profiles grounded in audience analysis—interviews, conversations, and stakeholder input—rather than wishful thinking. Wallace (2023) warns about the “assumption abyss,” where we over-rely on imagination instead of data when defining our “typical” learner, which is easy to do if we’re working ahead of an actual cohort.

To avoid that trap, I drew on feedback from five former learners in my online course, then used their responses to build a persona for a future data storytelling course. That process surfaced motivations, pain points, and expectations that go far beyond “visual vs. auditory,” and it aligns with constructivist, active-learning approaches described by Lewis and Bryan (2021) and Rutherford-Hemming, where learners build knowledge by connecting new ideas to prior experience.

Why this matters for data analysts and trainers

For data analysts in higher education, learning styles and personas both point to the same big idea: use data, not just instincts, to design better learning. Learning analytics and related approaches are already being used to identify which instructional methods support student success, personalize learning opportunities, and flag learners who may need extra support, which is only possible when we treat preferences as variables to study rather than fixed categories to obey. Analysts who understand these nuances can partner more effectively with instructional designers, helping them test which multimodal strategies genuinely move the needle on engagement, performance, and equity.

For trainers and instructional designers, especially in higher ed, the stakes are high: you are designing for adults with diverse roles, backgrounds, and constraints. Blending learning style awareness with learner personas, analytics, and active learning theory helps avoid oversimplified “visual vs. auditory” debates and pushes design toward inclusive, evidence-informed experiences that can scale.

Wrapping it up

Pulling this all together, my view is that learning styles are worth engaging with—as conversation starters, design prompts, and self-reflection tools—but not as strict blueprints for how we must teach. They become most powerful when combined with learner personas, real audience data, and learning analytics, especially in data-focused contexts where we can actually test whether certain design choices work. For data analysts and trainers in higher education, that blended approach offers something better than chasing the perfect “match”: it offers a way to design learning that is inclusive, adaptable, and grounded in both human insight and hard evidence.

References

Davidson, C. N., & Katopodis, C. (2022). The new college classroom [Kindle version]. Harvard University Press. https://a.co/d/1uB9aq4

Dirksen, J. (2015). Design for how people learn [Kindle version]. New Riders. https://a.co/d/gHIPUdx

Felder, R., & Soloman, B. (2002). Learning styles and strategies. North Carolina State University. http://www4.ncsu.edu/unity/lockers/users/f/felder/public/Learning_Styles.html

Lewis, N., & Bryan, V. (2021). Andragogy and teaching techniques to enhance adult learners’ experience. Journal of Nursing Education and Practice, 11(11), 31–40. https://www.researchgate.net/publication/353272841_Andragogy_and_teaching_techniques_to_enhance_adult_learners’_experience

Malamed, C. (2024, November 3). Learner personas for instructional design. The eLearning Coach. https://theelearningcoach.com/elearning_design/audience/learner-personas-for-elearning/

Understanding learning styles: Centre for Teaching Excellence: University of Waterloo. (2024, May 21). Centre for Teaching Excellence | University of Waterloo. https://uwaterloo.ca/centre-for-teaching-excellence/catalogs/tip-sheets/understanding-your-learning-style

Wallace, D. (2023, August 15). 4 pitfalls in crafting effective learner personas for training. eLearning Industry. https://elearningindustry.com/pitfalls-in-crafting-effective-learner-personas-for-training

Why Start with Power Query in Excel?

Power Query in Excel is a friendly, visual way to step into real data preparation without learning a programming language or buying a new BI tool. Power Query is built into modern versions of Excel under the Data tab, so you can start where you already live instead of adopting an entirely new platform. It lets you:

• Import from many sources (CSV exports from PeopleSoft, SQL Server reporting tables, Access databases, text files, web sources, etc.).

  

• Clean and reshape the data using a point‑and‑click interface instead of complex formulas or code.

  

• Automate repeatable prep so next term’s file is ready with a single refresh instead of another afternoon of copy‑paste.

  

If you are being nudged toward more data wrangling, Power Query is often the easiest “first real data prep tool” because it feels like an extension of the Excel skills you already have.

How Power Query Helps with Higher Ed Work

Here are concrete ways it can help with the kinds of tasks data professionals in higher education see every day:

• Automate recurring term jobs
  Build a query once to filter to active students, standardize student program codes, and merge in a census snapshot. Next term, point to the new extract and hit refresh instead of redoing all the steps manually.

• Combine files from multiple sources
  Stack multiple CSVs (e.g., one per college, campus, or term) into a single table in seconds, rather than copying and pasting tabs together.

• Fix common data quality issues
  Remove duplicates, split concatenated fields, standardize values (such as “FT”, “Full Time”, “Full‑Time” to a single value), and safely change data types for dates and numbers.

• Reduce Excel formula sprawl
  Logic that might have lived in layers of VLOOKUPs and nested IF statements can be captured as clear, step‑by‑step transformations that are easier to review and reproduce.

The big win is that Power Query records your steps. Each filter, split, merge, or transformation becomes a documented step that can be replayed and modified later, which is a huge help for auditability and collaboration with other analysts.

You Don’t Need to Know R, Python, or SAS

Many higher ed colleagues feel pressure to “catch up” with programming tools, but don’t have the time, support, or comfort level to dive into them right away. Power Query offers:

• A no‑code or low‑code environment where most work is done via menus and right‑clicks.

• The ability to grow into the underlying M language later, but only if and when you’re ready.

• A way to build good data prep habits—documented, repeatable, query‑driven workflows—that will transfer nicely if you eventually move into R, Python, or BI tools.

Think of Power Query as a bridge between “I clean everything manually in Excel” and “I script my entire pipeline.”

Where to Start: A Simple First Project

A good first project is something you already do repeatedly in Excel and that annoys you. For example: preparing an enrollment file each term, cleaning up a financial aid disbursement report, or combining multiple queries into one student‑level dataset.

A minimal starter path:

1. Export your usual SQL query result to CSV or Excel.

2. In Excel, go to Data → Get Data → From File and load it into Power Query.

3. Use the interface to:

   • Remove columns you don’t need.

   • Combine values from multiple columns to create concatenations.

   • Filter out rows you always delete.

   • Fix obvious issues (trim spaces, change data types, rename fields).

4. Click Close & Load to bring the cleaned table back into Excel as a new sheet.

5. Next time you get the same report, point Power Query to the new file and hit Refresh.

Helpful Resources to Learn More

Here are some approachable ways for you to build Power Query skills.

Introductory Power Query Documentation

• Power Query documentation (Microsoft Learn) – https://learn.microsoft.com/en-us/power-query/​

• Power Query for Excel help (Microsoft Support) – https://support.microsoft.com/en-us/office/power-query-for-excel-help-2b433a85-ddfb-420b-9cda-fe0e60b82a94​

• Learn to use Power Query and Power Pivot in Excel – https://support.microsoft.com/en-us/office/learn-to-use-power-query-and-power-pivot-in-excel-42d895c2-d1d7-41d0-88da-d1ed7ecc1025​

Beginner-Friendly Video Tutorials

• “How to use Power Query – Microsoft Excel Tutorial” (YouTube) –

• ​“How to use Microsoft Power Query” (YouTube) –

• ​“Excel Power Query Tutorial – First Steps” (YouTube) –

• ​“Power Query – Beginner to PRO Masterclass in 30 minutes” (YouTube) –

• ​“Transform Excel Data in Minutes (2025 Tutorial Part I)” (YouTube) –

Conclusion

Power Query in Excel gives higher ed analysts a way to move from “I just pull the data” to “I own the whole pipeline” without jumping straight into a programming language. It lives in a tool you already use, lets you document each step of your cleanup, and turns those painful, repetitive term‑by‑term tasks into a refreshable workflow instead of a weekend project.

If you’ve been feeling intimidated by data prep, start small: pick one routine report you always clean up manually, pull it into Power Query, and see how many of your usual steps you can capture as repeatable transformations. Then share what you learn—show a colleague in your office, or propose a short brown‑bag where you rebuild a common process together. The sooner you take that first, low‑risk step, the sooner data wrangling stops being something to fear and starts being a core part of your analytical superpower. Good luck! And if you manage to pull something together, comment here!

Leave a comment

A proud moment as a learning experience designer recently reminded me why this work matters so much.

I created an intermediate, self-paced PeopleSoft Query course for university staff called “Query Inside Your Box!” The title and some of the course design were inspired by George Couros’ ideas in The Innovator’s Mindset about innovating inside the box—the belief that people can create meaningful, innovative teaching and learning opportunities within their existing constraints, tools, and roles.

For many, PeopleSoft Query can feel like a purely technical skill, something you “have to learn” to pull reports. I wanted to flip that script. My goal was to design an experience where staff could see themselves not just as report runners, but as creative problem-solvers who use data to improve their work and the experiences of students, faculty, and colleagues.

Moving beyond step-by-step instructions

A key design choice in this course was to offer multiple, diverse, and regular assessment opportunities that go beyond “click here, then click there.” One of my favorite activities is a problem-based challenge where learners receive a realistic reporting problem and are not given the step-by-step instructions. Instead, they must decide themselves how to construct the query and apply aggregate functions.

Example reporting problem:

You work in the Institutional Research office at a university. The Registrar has asked you to provide a headcount and credit-hour summary for undergraduate students for the current fall term. The report needs to:

• Include only degree-seeking undergraduates.

• Include only students who are enrolled as of a specific date and not withdrawn.

• Be summarized by college and student level (e.g., freshman, sophomore, junior, senior).

Show, for each college and level:

• Number of distinct students.

• Total enrolled credit hours.

• Average enrolled credit hours per student.

Using Query Manager, construct a query that:

• Pulls enrollments from the term’s class enrollment table(s).

• Joins to student career/term data to filter to active, degree-seeking undergraduates as of census.

• Joins to the academic organization/college table to roll students up to the correct college.

• Uses aggregate functions to produce one row per college and student level with the required counts and sums.

Learners are not given any navigation or step-by-step instructions; they must decide:

• Which records/tables to use (e.g., term enrollment, student term, academic organization).

• How to join them correctly.

• Which fields to group by.

• Which aggregate functions (COUNT, SUM, AVG) to apply to produce the final summary.

Instead of walking them through a recipe, the exercise asks them to think for themselves, test their assumptions, and reflect on their choices. In other words, they practice innovating inside the box of PeopleSoft Query. Reflection is a crucial part of this activity because it helps learners surface their decision-making process, consolidate what they’ve learned, and become more intentional and confident in their future querying work.

Here are some of the reflective prompts from this activity:

Knowing what you know now about applying aggregate functions on fields in PS Query, how often (if ever) would you need to double-check the results of a function? How would you double-check a query that includes one or several aggregations?

How comfortable were you in attempting to assign two different aggregate functions on two fields?

What was the most challenging part of this activity?

What did you enjoy about this activity?

Can you provide an example of how you could use aggregate functions to support your work?

These questions are designed to move learners from “Did I do it right?” to “How do I know it’s right, and how will I use this skill in my real work?”

A learner reflection that made my day

One learner’s reflection captured exactly the kind of mindset shift I hoped this course would create. Here’s what they shared (quoted with permission):

On double-checking aggregate functions:
“I will most likely run reports before and after creating the aggregate to spot check some of the math.”

On using multiple aggregate functions:
“After completing this assignment, I am more comfortable with multiple aggregate functions. It was intimidating to attempt at first 😊”

On the most challenging part:
“This was a very interesting activity, because it was presented as a problem and I had to find the solution. It was challenging to decide what the best plan of action was, without being provided explicit steps to work through. It put what I thought I knew to the test!”

On what they enjoyed:
“I enjoyed being able to build a query from a problem, without the explicit steps, even though that was challenging 😊”

On applying this to their work:
“My team plans to use aggregate functions to build a report showing total number of missing course grades across the College, for end-of-term audits.”

This is everything a learning experience designer hopes to see:

The learner acknowledges initial intimidation, then grows into greater confidence.

They develop a strategy for validating their work (“run reports before and after…to spot check”).

They enjoy the challenge of solving a realistic problem without being spoon-fed.

Most importantly, they see a concrete way to apply the skill in their real context—supporting end-of-term grade audits across a college.

That is learning that sticks.

Why this moment matters

This reflection was a proud moment—not just for the learner, but for me as a designer.

It affirmed that:

Challenging learners respectfully—by removing some scaffolding and inviting them to struggle productively—can build real confidence.

Even “technical” tools like PeopleSoft Query can be vehicles for creativity, judgment, and innovation.

When learners see a direct line from a course activity to a real problem in their day-to-day work, they are far more likely to adopt and sustain new skills.

It also reminded me that innovating inside the box is not about flashy tools or unlimited freedom. It is about designing thoughtful experiences within existing systems that empower people to think, experiment, and own their learning.

Universal design in action

This course also reflects key principles of Universal Design for Learning. By offering multiple ways for learners to engage with the material—such as realistic problems, reflection prompts, and self-paced practice—it supports different motivations and comfort levels. By presenting concepts through written scenarios, hands-on querying, and metacognitive questions, it provides multiple means of representation and helps learners connect abstract ideas to concrete tasks. And by allowing learners to demonstrate their understanding through how they design queries, validate results, and describe real-world applications, it offers multiple means of action and expression that honor different strengths and working styles.

A note to fellow learning experience designers

To anyone designing learning—whether it’s your full-time role or a side passion—moments like this are why the work is worth it.

When a learner says, “It was intimidating at first, but now I’m more comfortable,” or “It put what I thought I knew to the test,” you are witnessing growth in real time. When they go on to describe how they’ll use that skill to improve a process, support colleagues, or better serve students, you’re seeing the ripple effect of your design.

So here’s to:

Building courses that trust learners to think.

Embracing productive challenge over step-by-step handholding.

Creating spaces where people can innovate inside their own “box”—their tools, their roles, and their everyday work.

This experience with “Query Inside Your Box!” has been one of those energizing reminders for me. Proud of my learners. Proud of the work. And excited to keep designing experiences that stretch, support, and celebrate people as they grow.

Subscribe now

Share