Using Visuals to Enhance Scientific Presentations

Choosing the right chart is half the battle in data storytelling. This one visual helped me go from “𝐖𝐡𝐢𝐜𝐡 𝐜𝐡𝐚𝐫𝐭 𝐝𝐨 𝐈 𝐮𝐬𝐞?” → “𝐆𝐨𝐭 𝐢𝐭 𝐢𝐧 10 𝐬𝐞𝐜𝐨𝐧𝐝𝐬.”👇 𝐇𝐞𝐫𝐞’𝐬 𝐚 𝐪𝐮𝐢𝐜𝐤 𝐛𝐫𝐞𝐚𝐤𝐝𝐨𝐰𝐧 𝐨𝐟 𝐡𝐨𝐰 𝐭𝐨 𝐜𝐡𝐨𝐨𝐬𝐞 𝐭𝐡𝐞 𝐫𝐢𝐠𝐡𝐭 𝐜𝐡𝐚𝐫𝐭 𝐛𝐚𝐬𝐞𝐝 𝐨𝐧 𝐲𝐨𝐮𝐫 𝐝𝐚𝐭𝐚: 🔹 𝐂𝐨𝐦𝐩𝐚𝐫𝐢𝐬𝐨𝐧? • Few categories → Bar Chart • Over time → Line Chart • Multivariate → Spider Chart • Non-cyclical → Vertical Bar Chart 🔹 𝐑𝐞𝐥𝐚𝐭𝐢𝐨𝐧𝐬𝐡𝐢𝐩? • 2 variables → Scatterplot • 3+ variables → Bubble Chart 🔹 𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐢𝐨𝐧? • Single variable → Histogram • Many points → Line Histogram • 2 variables → Violin Plot 🔹 𝐂𝐨𝐦𝐩𝐨𝐬𝐢𝐭𝐢𝐨𝐧? • Show part of a total → Pie Chart / Tree Map • Over time → Stacked Bar / Area Chart • Add/Subtract → Waterfall Chart 𝐐𝐮𝐢𝐜𝐤 𝐓𝐢𝐩𝐬: • Don’t overload charts; less is more. • Always label axes clearly. • Use color intentionally, not decoratively. • 𝐀𝐬𝐤: What insight should this chart unlock in 5 seconds or less? 𝐑𝐞𝐦𝐞𝐦𝐛𝐞𝐫: • Charts don’t just show data, they tell a story • In storytelling, clarity beats complexity • Don’t aim to impress with fancy visuals, aim to express the insight simply, that’s where the real impact is 💡 ♻️ Save it for later or share it with someone who might find it helpful! 𝐏.𝐒. I share job search tips and insights on data analytics & data science in my free newsletter. Join 14,000+ readers here → https://lnkd.in/dUfe4Ac6

11 plots in data science that are used 90% of the time (with precise usage👇) Visualizations are critical in understanding complex data patterns and relationships. They offer a concise way to understand the: - intricacies of statistical models - validate model assumptions - evaluate model performance, and much more. The visual below depicts the 11 most important and must-know plots in data science: 1) KS Plot: - It is used to assess the distributional differences. - The core idea is to measure the maximum distance between the cumulative distribution functions (CDF) of two distributions. - The lower the maximum distance, the more likely they belong to the same distribution. - Thus, instead of a “plot”, it is mainly interpreted as a “statistical test” to determine distributional differences. 2) SHAP Plot: - It summarizes feature importance to a model’s predictions by considering interactions/dependencies between them. - It is useful in determining how different values (low or high) of a feature affect the overall output. 3) ROC Curve: - It depicts the tradeoff between the true positive rate (good performance) and the false positive rate (bad performance) across different classification thresholds. 4) Precision-Recall Curve: - It depicts the tradeoff between Precision and Recall across different classification thresholds. 5) QQ Plot: - It assesses the distributional similarity between observed data and theoretical distribution. - It plots the quantiles of the two distributions against each other. - Deviations from the straight line indicate a departure from the assumed distribution. 6) Cumulative Explained Variance Plot: - It is useful in determining the number of dimensions we can reduce our data to while preserving max variance during PCA. 7) Elbow Curve: - The plot helps identify the optimal number of clusters for the k-means algorithm. - The point of the elbow depicts the ideal number of clusters. 8) Silhouette Curve: - The Elbow curve is often ineffective when you have plenty of clusters. - Silhouette Curve is a better alternative, as depicted above. 9) Gini-Impurity and Entropy: - They are used to measure the impurity or disorder of a node or split in a decision tree. - The plot compares Gini impurity and Entropy across different splits. - This provides insights into the tradeoff between these measures. 10) Bias-Variance Tradeoff: - It is used to find the right balance between the bias and the variance of a model against complexity. 11) PDP: - Depicts the dependence between target and features. - A plot between the target and one feature forms → 1-way PDP. - A plot between the target and two feature forms → 2-way PDP. 👉 Over to you: Do you find it easier to interpret plots over numbers? ____ If you want to learn AI/ML engineering, get this free PDF (530+ pages) with 150+ core DS/ML lessons. Get here: https://lnkd.in/gi6xKmDc ____ Find me → Avi Chawla Every day, I share tutorials and insights on DS, ML, LLMs, and RAGs.

10 reasons why your dashboard lacks clarity. A - Don't put everything in one dashboard. => A dashboard made for everyone, is a dashboard used by no one. B - Help users see, not read. => "Good data visualization takes the burden of effort off the brain and puts it on the eyes." Stephen Few's C - Don’t use maps if they’re not relevant. => Even if your colleague worked so hard to get these ZIP codes. Ask yourself : Does the map add value to the business? D - Zoom in when necessary. => Sometimes (for specific reason) you'll need to truncate your axis. Because Usain Bolt has no intention of running the 100m in under 7 seconds. E - Declutter your charts. => It's a constant balance between space optimization and chart comprehension. F - Use double encoding on purpose. => Displaying the same KPI twice in the same chart may raise questions you don't want to hear during the kick-off meeting. Keep it clear. G - Rotate your charts to see full labels. => "My neck has been hurting lately, but I'm not sure why." H - Clean your pie chart. => Pie charts are hard enough to understand quickly, so let's not make them even trickier. I - Use aggregation to your advantage. => If your message is clear with 36 bars, why use 156? J - Use color to your advantage => The purpose of color is not to make your dashboard funky, but to attract the eye, to alert and to assist readability... Find this High Resolution visual + 50 other in the Dataviz Clarity Gallery here : https://lnkd.in/eThSWtWv #Businessintelligence #Datavisualization #DataAnalytics

How to create great MBB slides as a PRO (even if you’re not a designer) I’ve probably made over 10,000 slides in my career. According to Malcolm Gladwell’s 10,000-Hour Rule, that technically makes me a slide master God. (Or maybe just someone who needs new hobbies.) But here’s the twist: I’m not a designer. I’m a lawyer by training. My natural design instincts are zero. And yet, I’ve learned to make slides that work, not because of fancy design skills, but because of a few simple principles that anyone can apply. Here’s how to make slides that actually work: 1. Always start with the message. Before adding any charts, graphs, or visuals, ask yourself: • What is the key takeaway? • How do I want the audience to react? • What elements best support this message? Tattoo this on your brain: A slide where the elements don’t support the message is a bad slide. Period. 2. Use a strong title and subtitle. Every slide needs a clear title that tells you what it’s about and a subtitle that provides the key insight. Example: • Title: “Global sales performance” • Subtitle: “North America drives 60% of growth” 3. Stick to simple and consistent layouts. A good slide should be easy to read at a glance. My go-to layouts: • One chart with commentary: Chart on the left, key takeaways on the right. • Two simple charts side by side: For comparing metrics or trends. • Three columns: When comparing options or showing steps, use three aligned boxes with short text. • Image and text pairing: Image on one side, the message on the other. Keep it simple. The art belongs in the museum, not on your slide. 4. Less is more with text • Bullet points, not paragraphs. • Short phrases, not long sentences. • If your slide looks like an essay, start over. 5. Alignment and precision matter. Nothing makes a slide look messier than poor alignment. • Align elements consistently. • Use symmetry wherever possible. • White space is not your enemy, clutter is. 6. Keep charts simple and actionable • Bar charts, line charts, scatter plots, stick to what works. • Always label axes and show units. • Highlight key data points. 7. The 5-second rule • Can you tell what the slide is about in 5 seconds? • Is the key insight crystal clear? • Would a stranger understand it instantly? The bottom line: If your slide doesn’t support your message, it’s just a distraction. And if your message isn’t clear, neither is your impact.

Quick challenge: Say the color of each word aloud as quickly as possible. Surprisingly difficult, isn't it? That’s because you’re not reading the words themselves. You’re identifying the color they're printed in first, then reading the words. That's the Stroop Effect. Your brain handles text and visuals through two distinct pathways — one for words and another for colors. Typically, these systems collaborate. But when they conflict, it slows down processing. Consider the implications for data visualization: • When text and visuals are misaligned, your audience experiences the same kind of mental conflict as in the Stroop test. • When labels contradict the data, comprehension is hindered. • When a legend requires viewers to interpret colors separately, insights become tougher to grasp. The most effective data visualizations ensure that visual and textual elements are synchronized. • Titles should clearly convey to the audience what they're viewing. • Labels should be integrated directly into the visualization to avoid forcing viewers to switch focus. • Visual contrast should enhance the message, not compete with it. When text and visuals work in unison, insights become instinctive. When they don't, understanding is delayed. Are your charts making understanding easy or difficult? Art+Science Analytics Institute | University of Notre Dame | University of Notre Dame - Mendoza College of Business | University of Illinois Urbana-Champaign | University of Chicago | D'Amore-McKim School of Business at Northeastern University | ELVTR | Grow with Google - Data Analytics #Analytics #DataStorytelling

🚩 How To Flag Misleading and Dishonest Charts (https://lnkd.in/e9cB8r4E), a practical guide on how to spot misleading charts to communicate insights more accurately and more reliably — with plenty of examples and design guidelines to create honest charts. Kindly put together by Nathan Yau. 🚫 Charts aren’t merely a visual representation of data. ✅ Charts are visuals that have a specific job to do. ✅ Don’t cut bar chart baselines — always start at 0. ✅ Don’t expand the y-axis beyond the max value. ✅ Don’t choose narrow segments to highlight a point. 🤔 Beware of smooth operator as it often hides real data. 🚫 Correlation doesn’t mean causation: validate and verify. ✅ Don’t add time gaps in the timeline: it hides what happened. ✅ Avoid leading titles, as people use them to interpret data. We often think of charts as visual representation of data. But as Nick Desbarats says, charts are visuals that have a job to do — e.g. make people aware, take an action, find an answer, filter or look up values. To do that job well, they need to be honest. And if they don’t, they spread skewed and biased messages, fast. Charts combine visual encodings (e.g. color, area, position, direction, length, angle) with scales. If the data is scarce, visual encodings fill a space based on available data — against the scales we choose to use for it. If the scales are chosen unfairly, or the data is cherry-picked, charts tell a wrong story. Here are some of the common attributes of dishonest charts: 🎢 Slopes → Artificial steepness of lines suggests notable changes. 🚢 Damper → Values appear smaller if y-axis expands beyond max. 🍒 Cherrypicker → Choosing narrow segments to highlight a point. 🌊 Smooth operator → Avgs show patterns, but hide bumps in reality. 🗑️ Overbinner → Clumping data into general groups to hide diversity. 👀 Base Stealer → Shortened y-axis makes tiny differences seem large. 🦋 Probable Cause → Showing 2 things follow similar/opposing patterns. ⏰ Time Gap → Points in time are purposely selected, others left out. 🔥 Storyteller → Leads with narratives, then squeezes data to support. 📇 Descriptor → Words chosen to deflect or invite misinterpretations. Different design choices lead to different charts, along with different interpretations attached to them. And that interpretation is often linked to what a reader already knows, what they expect, or what they choose to believe. The purpose of a good chart is to make wrong interpretations less likely. Unfortunately, there are plenty of charts that intentionally invite wrong interpretations. So be careful in choosing the data set to rely on, check sources, and explore not only what is there, but also what is missing. As Nathan suggests, a single data set can represent infinite narratives, depending on the angle you look from. So be cautious about the story you are telling, and avoid common but dishonest attributes that always invite wrong conclusions. #ux #design

I often stress this in workshops: if you want people to actually read the explanation of your visual, place the information next to the visual, NOT separate from it. And this morning (I woke up way too early) I stumbled onto a great eye-tracking study that shows some evidence for this. The researchers compared two layouts of the same figure: Separated – text far away from the visuals. Readers read the title, skipped the text, and jumped straight to the figure without context. Integrated – text and visuals placed together. Readers were far more likely to read the explanation and connect it with the visual. The results show that integrating text with visuals helps the reader also read the accompanying text. Yet, in peer-reviewed papers, subsidy proposals and reports, we often do the opposite [e.g. "See figure 3, three pages down"]. So when you're designing a figure, infographic, or diagram, make sure that the explanation of the visual is integrated into the visual, and not presented separately. Otherwise, your explanation might be ignored, or worse misunderstood. Reference: Holsanova, J., Holmberg, N., & Holmqvist, K. (2009). Reading information graphics: The role of spatial contiguity and dual attentional guidance. Applied Cognitive Psychology, 23(9), 1215–1226. doi.org/10.1002/acp.1525 #infographics #makingsciencesexy

Data science or data analytics without storytelling is void. You can do all the SQL, all the Python, all the modeling — but if the final insight is not communicated in the right visual form, the value is lost. This cheat sheet is a perfect reminder that choosing the right chart is not decoration — it is part of analysis. It breaks the decision down by purpose of insight: 1) Composition Waterfall, Progress bar, Pie, Gauge — great when you want to show parts contributing to a whole or target progress. 2) Comparison Bar charts, Row charts, Line charts, Combo charts — useful when comparing categories or trends over time. 3) Distribution & Relationship Histogram and Scatter plot — when you want to show how values are spread or how two variables interact. 4) Stage Analysis Sankey and Funnel — ideal for visualizing drop-offs or flow across process stages. 5) Single Value KPIs Number & Trend cards — best for dashboards where one metric needs to stand out with context. The skill is not in plotting a chart — the skill is in selecting the correct one for the question being asked. Your analysis is only as powerful as the clarity of how you present it. cc Metabase #DataAnalytics #DataScience #DataVisualization #StorytellingWithData #BI #Metabase #DashboardDesign #DecisionMaking

Why showing text and graphics simultaneously is like trying to watch two movies at once - and the better alternative backed by research. Your brain has limits. Let's use them wisely. Most eLearning overloads learners with: ↳ Dense text blocks ↳ Complex graphics ↳ Information overload Here's the science-backed solution: 1️⃣ Split Processing Power • Your brain has two channels • Visual for graphics/images • Auditory for spoken words • Don't max out either one 2️⃣ The Power of Voice • Narration > on-screen text • Frees up visual processing • Reduces cognitive strain • Better retention rates 3️⃣ Strategic Implementation • Use audio for explanations • Keep visuals clean and focused • Sync narration with graphics • Let each channel do its job Real-world application: ☑️ Replace text walls with narration ☑️ Sync audio/visual timing perfectly ☑️ Save text for key terms only ☑️ Design for dual-channel processing The results? ↳ Reduced cognitive load ↳ Improved engagement ↳ Faster learning curves The secret isn't more content. It's smarter delivery. Your learners' brains will thank you. What small change could you make in your next course to ease your learners’ cognitive load?

Ever explained something perfectly… only to have learners still look confused?😵💫 You gave them the words. But did you show them? 👉🏼Enter Dual-Coding Theory — a simple but powerful principle: Learners absorb more when we combine words (spoken or written) with visuals (images, diagrams, videos). 🤔Why is that? Because our brains process verbal and visual info through two separate channels. Use both—and you double the chances of understanding and retention. 💡Why Dual Coding works in learning: ••• Learners store content in two mental “folders” — verbal and visual — making recall easier. ••• Visuals support complex or abstract text by making it more concrete. ••• But too much of either = cognitive overload. So balance is key! ✏️Practical examples: Teaching a new expense reporting tool? ✔️ Provide a step-by-step guide (verbal). ✔️ Add screenshots of the tool and a process diagram of the steps (visual). Now learners see and read what to do—making it much more likely they’ll get it right the first time. 📌Bottom line: Dual-Coding isn’t about making content prettier—it’s about making it stick. So the next time you design a course, job aid, or workshop, ask yourself: “Where can I show, not just tell?” Your learners’ brains will thank you. #InstructionalDesign #LearningAndDevelopment #AdultLearning #TheLnDAcademy