Visual Perception, Memory, and Dashboard Design

Data Visualization and Interpretation

10.1 Visual Perception and Why Visualization Matters

10.1.1 Why the brain is built for visuals

The whole subject starts with one fact about the human brain: it has a natural ability to process things visually. You have heard it at work many times — bosses say "guys, don't show me numbers, I'm a visual guy". That instinct is real. Visualizing things and interpreting them visually is much easier for the brain than reading numbers. The brain is designed such that after a few seconds of heavy numeric input it gives up: "it is too much for me to grasp this information, don't give me so many numbers." That is exactly why data visualization matters, and why we break information into small, small charts — so that the end user is not overwhelmed.

Visual perception — the act of interpreting what the eyes see — is what makes interpretation easy. Concretely, visual perception covers interpreting visual elements like color, size, and position. Consider color alone. We typically show anything positive in green: if you are making profit, it shows green; if you are in losses, it shows red. Our brain is so tuned to this that if you show lots of minus numbers and lots of positive numbers, it struggles. But show the same data as red and green bars and the brain immediately says "oh, I am making maximum profit here" or "I scored maximum marks in this subject." The meaning arrives without any reading.

The wiring behind this is real science. Roughly 70% of the sense receptors in the human body serve vision, and the eye plus the visual cortex of the brain form what researchers call a massively parallel processor — the highest-bandwidth channel into our thinking. That is why the professor's everyday observation is backed by measurement: perception happens in parallel, while reading numbers happens serially, one digit at a time. A screen of ten numbers takes ten careful reads; a screen of ten colored bars takes one glance.

Real-world: this is the same reason dashboards use color scales, heat maps, and traffic-light indicators — the audience reads the mood before they read a single number. Financial screens and command centers worldwide rely on exactly this trick: red, amber, and green status dots that an operator can judge from across the room.

10.1.2 The direction-giving test: words fail, arrows work

Here is a simple test of how the brain handles words versus visuals. Ask someone for directions. If the reply is "go 500 meters, take a left, a roundabout will come, then take a right", your brain gives up after a few sentences. But put arrows and symbols on a map, and the same route becomes easy to follow.

This is the logic behind Google Maps. If audio instructions alone were enough, Google would have shipped a simple audio service. Instead, Google invests heavily in maps, graphs, and visual navigation because our brain cannot process that lengthy a string of information. The moment we see a graph or an arrow, we can navigate. That is the practical summary of visual perception: our brain struggles with words and numbers, and symbols, arrows, and visuals make things easy to understand.

10.1.3 What this means for everything that follows

Two takeaways set up the rest of the topic. First, the brain is a visual processor, not a numeric one — so any report, dashboard, or chart that forces the reader to do the math in their head is fighting the audience's nature. Second, because the brain gives up quickly, the design must do the work: break the information into chunks, use color and position deliberately, and lead the eye to the important parts.

Why this lecture is ordered the way it is: first we meet the limit (short-term memory can hold only a handful of items), then the translation (visual encoding turns numbers into eye-friendly elements), then the rules (Gestalt principles describe how the brain groups what it sees), then the tool (Tableau turns the theory into working dashboards and stories). Each piece is a different answer to the same opening question.

10.2 Short-Term Memory: Why the Brain Needs Help

10.2.1 The mental notepad

Why does a screen full of numbers defeat us within seconds? The answer starts with how the brain holds information in the present moment. Ask yourself: how many items can you really juggle in your head at once, and for how long? The answers — about seven, and about eighteen seconds — shape every design rule in this subject.

Short-term memory (the ability to hold a small amount of information in mind for a short period) is nothing but a mental notepad. When we want to remember something quickly, we jot it down in a notepad and keep a summary of it — that is what short-term memory is like: a temporary mental notepad that holds information, but only for a few seconds. Within those few seconds, one of two things happens: the information is totally lost, or it moves into our long-term memory (the brain's permanent store, where information is kept for later recall). Everyone behaves this way, and we usually do not notice it. Whenever we see something, it stays in memory for about 15 to 20 seconds; after that it either goes into long-term memory or is lost forever.

Real-world: some conditions illustrate this sharply. People with diabetes, for example, may have short-term memory loss — they cannot remember things happening right now, yet they remember things from years back. The pattern is the same for everyone, just milder: the immediate present is fragile, and only what is important or emotional survives.

This fragility is why dashboards must be visual. If we put visuals into our dashboards and keep the information refreshed, we keep people connected — we keep re-registering the information in their memory, and it becomes easy for them to interpret what they see. Every glance at a dashboard is a fresh registration; every glance at a raw data dump is another lost fight.

10.2.2 From sensory to short-term to long-term memory

Memory moves through three stages. First, sensory memory (the raw, unprocessed impression that arrives through the senses) holds that impression for a fraction of a second. For example, we see a phone number on a piece of paper; it stays in our iconic memory — the visual form of sensory memory, sometimes called the visual sensory register — for less than four seconds, until it is transferred to short-term memory, and only if we pay attention to it. Once it is in short-term memory we can do something with it: tell it to another person, or write it down, before it fades.

Then a decision is made. If the information is important enough, or if we have an emotional attachment to it, it is stored in long-term memory. That is why emotionally attached things — the birthdays of your family members, for instance — are remembered without effort. Within three or four seconds the brain decides: "is this worth it, or should I discard it?" So the flow is: senses come in (iconic memory holds the picture), a little important information moves to short-term memory to decide what to do with it, and only the genuinely important part goes to long-term memory; the rest stays away.

There are exceptions — some people can hold ten phone numbers for a few seconds without effort — but for the average human this is how it happens. And there is a good reason for the gate: if everything from sensory memory went straight into long-term memory, you would run out of space. Short-term memory gives us a buffer where we decide whether to keep information for good or to forget it, by choosing to give it attention and effort.

One technical distinction worth knowing: short-term memory refers to the ability to hold a small amount of information in mind for a short period of time, and it has two important qualities. You cannot manipulate the information you hold — that skill belongs to working memory (the part of the system that not only holds information but actively processes, rearranges, and computes with it), a different concept. And its duration is very limited. Holding a number in your head is short-term memory; adding two numbers together inside your head is working memory.

10.2.3 Capacity: the magic number $7\pm 2$

Research on short-term memory gives a good estimate of average capacity and duration. The famous result is called the magic number: the number of items you can hold in short-term memory is

$$7\pm 2$$

This result comes from a paper by American psychologist George Miller. On average, people can store somewhere between five and nine items — the "plus or minus two" accounts for individual differences. This is why people fail to remember long lists: there is a real capacity limit on short-term memory. There are ways to hack this capacity — a trick that comes a moment later — but the limit itself is well established.

10.2.4 Duration: seconds before fading, and rehearsal

Capacity is only half the story; duration is the other half. If you make no effort to remember items, they disappear in a matter of seconds. Typical estimates put the duration somewhere around 18 seconds (the spread usually quoted is about 15 to 18 seconds). With effort, you can stretch it — research commonly cites up to 30 seconds.

How do you stretch it? Rehearsal: you actively repeat the items in your head. Psychologists point out, though, that when you do this you are technically just remembering what you last said, not what you initially saw. That is why people mug: you see people repeating a number again and again to themselves — they are trying to make short-term memory hold the data for longer, repeating it so that it does not fade away. So mugging works, but it works by rehearsal, by keeping the latest echo of the number alive.

Visual intuition for the fading: picture a graph whose horizontal axis is time in seconds (0 to 30) and whose vertical axis is the strength of the memory trace. Without rehearsal the curve drops steeply and hits near-zero at about 18 seconds. With rehearsal, each repetition spikes the curve back up — like a trampoline being bounced — so the trace stays high for as long as the bouncing continues. The takeaway in one line: without a bounce, the trace decays; visualization is the bounce for data.

10.2.5 Chunking: the memory hack

Chunking (bundling multiple pieces of information into a single group, and treating the group as one item) is the way to hack the five-to-nine-item limit. The idea: bundle multiple pieces of information into a single group, and treat the group as one item. While the brain can only store five to nine individual items, research shows you can store up to four chunks — and if each chunk carries four pieces of information, you can recall:

$$4\times 4=16$$

sixteen things that you would otherwise never hold at once.

The most common example is the phone number. Memorizing ten individual digits is hard. Separate the same number into three chunks — the area code, the first three digits, the last four digits — and remembering the phone number becomes entirely possible. Many of us casually remember our own numbers as "four digits, four digits, something like that" without noticing that we are chunking.

The lesson transfers directly to data. Rather than remembering big, big numbers, split the information into smaller chunks and the brain can hold it better. That is precisely what good data visualization does: less clutter, good coloring, small meaningful chunks. If we keep feeding people big tables and crosstab numbers, their brains will struggle.

10.2.6 What this means for dashboards

Put the pieces together and you get the design rules of data visualization. Because short-term memory has a limitation, we should break complex data into small, meaningful chunks — then the brain can hold it. A graph with the right color and the right positioning holds information longer, so people can act and react faster and take action. Raw data does the opposite: short-term memory gives up after a few seconds. It says, in effect: "sorry, I don't have the capacity to hold multiple-page information — I am giving up, I cannot take any decision." So do not create clutter and do not fill the screen with too much information. Logical, smaller, manageable chunks always win in data visualization, because the end user can then process the data easily.

A simple exercise shows how fragile this is. Flash a few names on a screen for a couple of minutes, then ask people to recall them: after three or four names or numbers, almost everyone struggles. Our mind's ability is such that after 10–15 seconds things start to fade away and we give up. Visualization holds on to the information before it starts to fade — and the better designed the visualization, the better we interpret the data. That is the whole reason visualization exists: it helps stakeholders get engaged and hold the information.

10.3 Visual Encoding of Data

10.3.1 What visual encoding means

How do you deliver a number without making the reader read it? That is the design problem visual encoding solves. Visual encoding (the translation of data into visual elements that the human brain can understand easily and quickly) is the bridge between the raw value and the instant understanding. The name of the topic matters less than the idea, and the idea is simple: numbers are hard for us, visuals are easy. Raw data can be very difficult and overwhelming to grasp; a visual language — icons, graphs, color, positioning — makes it easier for us to hold on to that data. Our brain is wired to process visual information faster than text, so encoding numbers into figures, graphs, and charts is the natural way to deliver a number.

10.3.2 The four benefits

Why bother encoding data visually? Four benefits came up:

1. Complex data becomes more easily accessible. A well-made chart hides the complexity and shows the pattern. A thousand sales records collapse into one rising line.
2. Analysis is faster. Remember the pre-attentive attributes: if we make something bold, our eyes immediately look at it; if we tilt one line out of a group of others, our mind goes straight to it. With these visual devices, the brain unintentionally processes information faster — no conscious reading needed.
3. Memory retention improves. As the short-term memory discussion showed, holding a number is much harder than holding a visual. Small chunks stay with us longer.
4. Communication is better. A picture communicates to a whole room at once; a table of numbers does not. Everyone sees the same pattern in the same instant — the chart, not the reader, does the talking.

10.3.3 The Nifty 500 heat map

The best illustration is the heat map (a grid of cells whose color encodes a value — the classic encoding of a number into color). Suppose you want to know how the stock market behaved on a given day. If someone shows you 100 numbers — "13 percent, 10 percent, 12 percent..." — you struggle. But show the same day as a heat map of all the Nifty 500 stocks and how they fared, and the story is instant: more greens than red means it was a good market; the mood of the day is visible in one glance.

Real-world: this is a real and commonly used example — a list of all Nifty 500 stocks color-coded by daily movement. Financial dashboards, portfolio trackers, and trading screens all rely on exactly this kind of heat map so that traders can judge a whole market without reading a single number. Beyond trading, the same encoding runs hospital bed-occupancy boards, server-monitoring dashboards, and election result maps.

10.3.4 Chart types as the encoding toolkit

The heat map is one member of a whole family of encodings. Scatter plots, bar charts, line charts, and geographic maps are the other familiar members — each maps data to position, size, color, or shape in its own way. The principle to carry forward: pick the chart that turns your data into the fewest mental steps for the viewer, because every chart type is just a different visual language for the same underlying numbers. A tool like Tableau produces these charts out of the box — the dashboard build in the coming section shows how quickly.

10.4 Gestalt Principles: How the Brain Groups What It Sees

10.4.1 What the Gestalt principles are

Why does a chart with no lines still look like a complete picture? Before the theory, the phenomenon is worth feeling: show people a few scattered dots and they see a dog; show three partial circles and they see a triangle sitting on top. The brain does not just record what the eye delivers — it organizes it, fills it in, and groups it, all by itself. Gestalt principles are a set of principles that describe how our brain naturally groups and organizes visual elements. The name is German — "gestalt" simply means pattern — and in the session it was pronounced roughly as "g-slát". The work comes from the Gestalt School of Psychology, which from around 1912 studied how people perceive pattern, form, and organization in what they see. This topic is often treated as extra material, yet it is a very common concept that every UI/UX person and product designer goes through. Before the theory, two demonstrations show the phenomenon; then the explanation.

The reason these principles matter: they simplify complex scenes. Our brain does the organizing by itself — and if you know how the brain works, you can build good visualizations that work with it instead of against it.

At the most general level, the mind perceives any visual in a very simple form — a principle that will recur in every example below.

10.4.2 Opening demonstrations: the walking dog, the triangle, the panda

Three demonstrations from the session show the effect better than definitions.

Our brain has the ability to fill the gaps and provide the continuity, connecting the dots all by itself. Now the second demonstration:

There is no line, no evidence that this is what the figure shows — but the brain connects the pieces and fills the blank space. The third demonstration, and the one most people remember:

The general rules work the same way. Look at the Olympic rings: they are interlinked, one ring weaving between others — but the mind quickly sees five individual, complete rings. It does not track how they interrelate; it keeps the view simple. Or look at the dotted lines on roads: we do not need solid lines for everything. The mind can follow dotted lines and curves, which is exactly why line graphs are so easy to follow — the viewer's mind does the connecting.

Midway through the discussion a pause was taken to check the pace. One observation came back, and it is worth keeping:

10.4.3 Pragnanz: keep it simple and orderly

The first named principle is Pragnanz (a non-English word, pronounced roughly "prag-nanz"). It means: keep it simple and in order. In the demonstration, a square, a rectangle, and a circle are blended together into one figure — yet the mind notices one separate square, one triangle, and so on. The mind was not confused by which shape lies over another, or why a partial triangle is visible. It automatically differentiates, because it prefers the simpler view. That is what the rings example showed too.

Application to dashboards: keep it simple. Do not try to have too many overlapping colors. The brain is used to looking at things that are simple and orderly, and it will automatically figure out what the design says — as long as the design is logical. So a dashboard should never be filled with too many colors; the simple view wins. Overlaying too many distinct figures on one screen forces the brain into the one mode it is worst at: serial, careful untangling.

10.4.4 Similarity

The principle of similarity: whenever some shapes or colors appear in a picture, the brain treats them as part of one group. All the blue items are read as one logical group; all the squares are read as another group. Our brain treats similar shapes and similar colors as a logical unit. Color, size, shape, and orientation all act as grouping cues — this is the direct link between visual encoding (section 10.3) and grouping: the encoder chooses a channel, and similarity is the rule that turns that channel into a group.

Application to dashboards: consistency. If you show a particular value with a circle, keep circles for that value. If you use a diamond for one value in one sheet and a diamond for a different value in another, the brain gets confused. Instead, color-code and use similar shapes for things that are logically grouped — the brain will see the grouping automatically. For example, if you are showing sales or some other parameter across views, keep the same color and shape for it everywhere; interpretation becomes much easier. Same value, same look, everywhere.

10.4.5 Continuity

Continuity is the tendency of the brain to fill in and follow curves. The dog and the panda demonstrations both relied on it — the mind follows the implied line. In practice, whenever we show a trend line or a sequence of things, it becomes very easy for the brain to catch the direction. In the dashboard example shown, sales values are placed next to each other with a light continuity line: the brain immediately sees "January to February was growth, my sales are continuously going up." Wherever there is consistency and continuity, the brain reads the sequence without effort.

The design consequence is subtle but strong: the eye seeks the smoothest path. Remove the vertical axis line from a bar chart and the bars still appear lined up, because the consistent white space between labels and data carries the alignment. The same principle explains why indented text groups read as nested lists, and why a dashed trend line still reads as a single trend: the mind draws the path the designer only suggested.

10.4.6 Closure

Closure is the principle behind the dog, triangle, and panda demonstrations: the brain fills in missing information to perceive a complete shape. We saw it with the panda — gaps everywhere, yet the brain completes the figure.

The lesson applies beyond graphs: group points, lines, or bars into visual regions without complete borders — the reader's mind supplies the enclosure, and the screen stays free of clutter. Use closure to delete, not to draw more.

10.4.7 Focal point

Focal point is the principle you will use most frequently. If you have ten bars in a bar chart and you want one of them to stand out absolutely, color that one bar differently. Automatically, the others recede into the background and the special bar comes to the front. You never have to tell the audience "please look at the blue one" — the focal point does the telling by itself.

The dashboard example made this concrete: everything on the screen was grayed out by default, and one element was shaded — it stood out immediately, without any pointer or explanation. The audience focuses there by default, purely from this property. Use it whenever you want your users — your leaders, for instance — to focus on something. Make it a different color or slightly elevated, and the reaction comes by itself: "what is this, why is this shown differently?" That question is the sign that the focal point worked.

10.4.8 Proximity

Proximity: things that are close to each other are seen as groups. Clustered things draw our focus, so the positioning of elements helps us draw relationships. Use positioning deliberately. In the example shown, the creator placed two countries' widgets next to each other with identical parameters, so it is very easy to compare them. Had the creator put one country somewhere at the bottom and the other on the next sheet, comparison would have been very difficult. The same trick appeared with the 2020 year-to-date values: all the 2020 YTD widgets were kept next to each other, making them proximity-wise easy to interpret and compare.

Proximity is the cheapest grouping tool the designer owns: it costs nothing but white space. Ten dots become three groups by spacing alone; widgets placed side by side become a comparison pair. It can also steer reading direction — columns of closely stacked widgets encourage scanning top to bottom, rows of closely spaced widgets encourage left to right. But proximity can mislead too: if two unrelated widgets are packed together, the brain will insist they are related. The placement of every element is a statement about what belongs with what.

10.4.9 Overlap with pre-attentive attributes (Q&A)

At this point a student raised a natural worry: these ideas look a lot like the pre-attentive attributes discussed earlier — color, size, orientation, shape. The discussion that followed is the bridge between the two topics:

10.4.10 Figure and ground: foreground versus background

The last principle is figure and ground, and it may be used less in dashboards, but it shapes how any image is read. A figure-ground image can be interpreted in many ways. One classic: some viewers see a tree; shift what counts as foreground and you may see a gorilla and a tiger. Another classic: make the center of a drawing black and it looks like a vase; change the coloring and it looks like two faces. Everything depends on what we put in the foreground and what we put in the background.

For dashboards, the lesson is contrast. The contrast between foreground and background decides how your end user focuses. Just by changing the foreground and the background, the whole perception of the screen changes. The same elements can read as "content to study" or "background to ignore" purely from the contrast relationship between them.

The same idea applies to navigation. If your screen has navigation buttons — next, next, next, back — keep a consistent look and feel for all buttons. If one button is red and another is blue, the brain says "I don't know which one to choose." Consistency in figure and ground makes the choice obvious: one style for navigation, one elevated style for the single action you want taken.

One more example showed contrast used well: a chart of how much time Australians spend on social media by age. At a whole-country level, the creator removed all the spacing between the bars so that Australia reads as one continent, as a single totality — the contrast made the aggregate visible at a glance.

10.5 Pre-Attentive Attributes vs Gestalt Principles

10.5.1 One is a subset, the other is the bigger rulebook

Now that both ideas are on the table, the comparison becomes useful. Pre-attentive attributes are the visual properties that are very rapidly and automatically processed by our brain — we do not have to tell anything; the brain just does it. Gestalt principles are more like rules — the fundamental laws of how we organize what we see. Pre-attentive attributes are like a subset of visual behavior; Gestalt principles are much wider in concept and much broader in application. Where pre-attentive attributes concern specific visual properties — color, size, orientation, shape — Gestalt principles are the broader principles of visual organization of the overall theme. There are always shades of gray between the two — the distinction is useful, not absolute.

10.5.2 Speed and level of processing

The two differ in speed and in level. Pre-attentive attributes are perceived within milliseconds — the moment you see the screen, you automatically differentiate what stands out. Gestalt principles work at a higher level: they describe how things will be perceived — what should be background and foreground, where should this information be placed. One works at the icon, line, and page level, in milliseconds; the other works at the theme level, organizing the broader page and picture. The examples line up with this: color, size, orientation, and shape are the pre-attentive list; proximity, similarity, and closure (the panda and the dog — the brain filling missing information to perceive complete shapes) are Gestalt examples. Gestalt principles are fundamental laws and rules; pre-attentive attributes are a quick, immediate, millisecond-level behavior.

The two columns are not rivals; the quick, automatic perception feeds the slower, organizational reading. A pre-attentive pop-out is the entry point; the Gestalt reading of the page is the arrival.

10.5.3 Where each one applies

The application split follows the theory. Pre-attentive attributes are for drawing attention to a specific data point or screen element. You can use them strategically to guide where the user focuses — "I will make all of this gray, and I want this one red." You can apply them directly to any screen or any graph, straight away. They are the fast, surgical tool: one change, one element, instant effect.

Gestalt principles guide the design of the visual composition at a broader level. Before you start on any screen design or poster, you sit down with the team and decide what exactly you want to achieve: should I give a gap here, should I place this button here or at the top, should I keep all buttons together, should submit come before back, or back before submit? These are layout decisions, and they influence how people move through the screen. This is why designers — UI/UX people — take care of these things very religiously: the mind-behavior questions are their daily work. When it comes to Gestalt principles, you are influencing layout and composition, not just color and size. A product designer deciding where the "submit" button lives is running the same mind-behavior calculus a dashboard builder runs when deciding where the profit chart lives — the layout is a conversation with the viewer's visual system.

10.6 Tableau Dashboards: Sheets First, Assembly Second

10.6.1 Individual sheets are the building blocks

The dashboard work follows a strict order: individual sheets first, assembly second. In the demos so far, each demo was built as an individual sheet — one chart that answers one question, complete in itself. Keep creating whatever sheets you have in mind. When the data changes, the sheet changes too, so the sheet stays useful at any moment.

The key learning: dashboards are simple — the real effort goes into building the individual sheets. A dashboard is simple placing of the sheets; after that it is just linking and cross-connecting them to each other. Learn to make individual sheets, because the building blocks of a dashboard are these sheets. Once you are clear on your data points, you bring them in, drag and drop, and it is done. Tools these days are so user-friendly that there is no complexity in the mechanics; the whole effort goes into sheets — how do I make a running total, how do I show this graph, how do I show a scatter — while dashboards and stories are more about placing your information in a logical manner.

10.6.2 Profit by category: rows, measures, and instant charts

The first sheet in the demo was profit by category. The setup: when you say row, Tableau shows what you want as rows. Here, category and subcategory were placed at the row level — so the sheet shows category and subcategory. Then the profit measure was dragged from the data pane and placed as the measure, and Tableau automatically made the chart.

Real-world: the dataset used is the default Sample Superstore that ships with every Tableau version. Nothing was custom-built — you can replicate every example at your end the moment you install Tableau.

10.6.3 Profit by state and profit by segment

The second sheet was profit by state. Whenever you drag geo-spatial fields — here, region and state — and place a sum on top, Tableau produces a geographic chart. The state-wise profit makes California instantly visible as the state with all the good profit. Labels, coloring, and formatting are easily changed through the properties — you can change the background color, the font color, and so on, the way you want.

The third sheet was profit by segment: three segments, and which segment gives you profit. It is a minimal sheet — it could have added more — and you can always keep adding: one more sheet and your fourth sheet will come. Each sheet is complete by itself.

10.6.4 A dashboard is a collection of sheets

Once the individual sheets exist, the dashboard is a collection of individual sheets. You give a bigger picture by placing them together. To create one: click the dashboard icon (or use "new dashboard"). Tableau shows all the sheets you have built. If you had ten sheets, you might decide to use only four or five, or two or three — it is absolutely up to you and depends on the dashboard's purpose. But it will show you everything available.

Then you choose the sequence. Drag "profit by category" into the dashboard area. Drag the next sheet below it if you want it below — or delete and re-drop it side by side. The moment you move your cursor, the sheet snaps into place, and Tableau automatically adjusts and fits each graph into whatever space is available.

10.6.5 Layout, containers, and fitting the screen

For a quick layout with three sheets, the automatic fit is enough. If you have many small widgets to place, you may choose a horizontal or vertical container (pane) to make the group more logical. A container is exactly what its name says: a box that holds a set of widgets together, arranging them in a row (horizontal) or a column (vertical), so the group moves and resizes as one unit — proximity made structural. And every sheet or dashboard has two view modes: standard and entire view. Click a chart and say "use my entire space" — it will take whatever space is given to it. The same is available for every sheet on the dashboard.

10.6.6 Interactivity with "use as filter"

Interactivity is the very basic thing of dashboards, and in Tableau it is easy to link. Initially, clicking a state in the state chart has no effect on the other charts — there is no linkage. Then the use as filter icon was clicked: it connects all the sheets with each other as filters. Now the interactivity appears.

That is how you filter and see what is going on, and the filter can always be removed. Besides linking charts, you can also expose the filters to the end user directly — right-click and hand over the controls; the full details come in the next session.

Real-world: this is the core of modern dashboard interactivity — a leadership dashboard where clicking a region instantly re-aggregates every other chart for that region, without any new report being built.

10.6.7 Extra objects: web pages, logos, and text

A dashboard can carry objects beyond charts. You can add a web page object: drop it on the canvas, and it renders a live web page — in the demo, the Microsoft page. You can add an image: a logo, for example, of Microsoft. You can add a text box for commentary — "this is a demo Microsoft web page" — and center it, give it a border, or make it bigger. Dragging the handles automatically makes space on the canvas.

10.6.8 Device preview and export options

Dashboards are not screen-bound. You can decide to show it on a phone — the layout changes automatically to the phone layout — or switch to screen, desktop browser, or other devices, and decide the filter behavior. When you are ready, everything can be exported: copy it as an image, export the image, show it as a PDF, send it as a report, or export it to PowerPoint. One dashboard, many delivery channels: the same sheet-first design renders for the boardroom screen, the pocket phone, and the emailed report without being rebuilt.

10.6.9 Color consistency across sheets

The visualization theory from earlier sections shows up in the build. In the demo dashboard, the profit and loss coloring was kept consistent: the same scale here, the same scale there — so the end user can compare colors across charts easily. If one chart had a different scale than another, the user would go for a toss. Consistency rules: if you show something in blue in one view, keep blue for the same thing elsewhere; use red only for a deliberately different meaning. Then the layout decisions follow — what should be bigger, where should the story start. Should you start with the state map and then go two levels down? For this data, the subsets suggest starting with the bigger picture. Experience and hierarchy decide what to place first.

10.6.10 The Power BI comparison (Q&A)

The dashboard demo invited a natural comparison with the tool many students already use:

The practical takeaway of the comparison: the charting mechanics are near-identical, and the real difference is architectural. Tableau's building blocks are sheets that get assembled; Power BI's building block is the single-pane dashboard itself. If you are comfortable with one tool, the other is mostly a rearrangement — and the sheet-first discipline is the transferable skill.

10.7 Stories and Storyboards: Dashboards That Narrate

10.7.1 Story points are like slides

A story (or storyboard) is one level above the dashboard. If a sheet is one chart and a dashboard is a collection of sheets, a story is a collection of dashboard states arranged for presentation — like a slide deck. You add a caption and a commentary to each screen: "my screen number one will be profit by category", then a caption, then commentary. Then you add the next story point. Each story point can carry a dashboard or another sheet, and you can run the whole thing like a presentation: navigate from one point to the next and play it back. With a mixture of screens and dashboards you can build a beautiful story that actually plays.

10.7.2 Building the narrative from big picture to detail

Stories are built the way good business narratives go: start from the big picture and narrow down.

When you present it to your stakeholders, you narrate over each point — "here is the overall summary, the top states are these" — then you navigate: "now let us go into this and see what happened in Texas." The audience follows the narrowing-down and you end with a conclusion. Stories are mainly for presentations: the presenter does the narrating, and the story does the sequencing. Whatever the flow of your business case, that is how you build the story.

10.7.3 Stories are interactive, not static pictures (Q&A)

A cluster of questions came up about whether stories are just slides, and how they relate to Power BI:

The correction underneath the cluster is the essential one: a story only looks like a static picture. Because each story point is a live sheet or dashboard, the interactivity — filters, drill-downs, URL actions — survives inside the story. The presentation frame does not freeze the data; it sequences it.

10.7.4 Practice before the next session

The closing advice was practical: try your hands before the next session. If you have other commitments, at least come with your own trial — "I got stuck here, I was not able to drag that." It is okay to struggle. Dragging and dropping looks easy when someone else does it, but you will not get that feel until you do it yourself. Remember short-term memory: the demo stays in your memory for a few seconds and then it goes away — the only way to keep it is to rehearse it, exactly as the memory research says. With your background, it will not take more than two or three iterations to become absolutely professional. Nobody masters any tool completely — even the demonstrator has many features unexplored — but the basic features needed to present something are not difficult. In five minutes, the demo built a graph; the effort, again, goes into the sheets.

Real-world: the planned next session is a hands-on dashboard and storyboard build from scratch — the fastest way to convert this session's theory into working memory.

Exam Guidance Summary

No formal exam information — mark distributions, question patterns, or expected problem types — was given in this session. What the session did give are strong signals about what matters for assessment-style work, and they are worth treating as guidance:

• Master the individual sheets. Every repeated emphasis in the session pointed the same way: sheets are the building blocks, the real effort goes into them, and a dashboard is a simple collection of them. Expect the ability to build a sheet from a question to be the skill being tested.
• Expect the next session to be hands-on dashboarding and storyboarding, built from scratch. Practice dragging fields, building the three demo sheets (profit by category, profit by state, profit by segment), assembling a dashboard, and wiring up the "use as filter" interactivity.
• Apply the visualization theory deliberately: use focal points instead of explaining, keep color scales consistent across sheets, group related widgets by proximity, and keep the design simple (Pragnanz).
• Stories are dashboards plus narrative — know how story points, captions, and navigation turn a dashboard collection into a presentation.

Key Industry Applications

The real-world connections from this session, collected in one place:

• Memory research in dashboard design. George Miller's magic number $7\pm 2$ and the chunking result ($4\times 4=16$) explain why dashboards use small meaningful chunks, less clutter, and refreshed visuals: the audience's short-term memory simply cannot hold anything bigger.
• Google Maps is the standing proof that visual navigation beats verbal instructions — arrows and maps are chosen over audio because the brain cannot process lengthy spoken directions.
• The Google homepage is a deliberate figure-and-ground design: white screen, elevated search bar, and nothing else — the visual says "I'm good at searching, use me for that."
• The Nifty 500 heat map — a real stock-market screen where color encodes the mood of the whole market in one glance; the same color encoding powers hospital occupancy boards, server-monitoring dashboards, and election maps.
• Tableau and the Sample Superstore dataset — the industry-standard dashboarding tool and its built-in demo data, used to build profit-by-category, profit-by-state, and profit-by-segment sheets, then combine them into an interactive dashboard.
• Power BI comparison — Power BI places widgets directly on a single pane; Tableau's sheet-first workflow and its storytelling feature are the differentiators.
• Dynamic web actions — Tableau dashboard actions can pass a clicked value (for example, a state) as a URL parameter, opening the state's Wikipedia page on the fly inside the dashboard.
• UI/UX practice — product designers apply Gestalt principles religiously to button placement, grouping, and layout; the same rules govern dashboard composition.
• Emerging technology watch — the session opened with a note on how fast this industry changes: quantum computing, if it becomes a reality, could crack today's password defenses in seconds, where classical computers would take years; artificial intelligence today is estimated to be only five to ten percent of what it is capable of — voice recognition and generative AI are still evolving very fast. The advice for staying current: keep closer to the industry and follow a few good newsletters.