Google Analytics - 6 Share - Module 1 - Visualize Data

# Course 6, Module 1 - Visualize Data ## 0. Overview >[! quote] In this module, you’ll delve into the various types of data visualizations and explore what makes an effective visualization. You'll also learn about accessibility, design thinking, and other factors that will help you use data visualizations to effectively communicate data insights. ### Learning Objectives >[! info] Learning Outcomes for Course X, Module Y >- [ ] Explain the key concepts involved in design thinking as they relate to data visualization >- [ ] Describe the use of data visualizations to talk about data and the results of data analysis >- [ ] Discuss accessibility issues associated with data visualization >- [ ] Explain the importance of data visualization to data analysts >- [ ] Describe the key concepts involved in data visualization --- ### Glossary for Module 1 - [[Alternative text]] | [[Annotation]] | [[AVERAGEIF]] - [[Balance]] | [[Bar graph]] - [[Calculus]] | [[Causation]] | [[Channel]] | [[Chart]] | [[Cluster]] | [[CONVERT]] | [[Correlation]] | [[CREATE TABLE]] - [[Data composition]] | [[Decision tree]] | [[Design thinking]] | [[Distribution graph]] | [[DROP TABLE]] | [[Dynamic visualizations]] - [[Emphasis]] - [[HAVING]] | [[Headline]] | [[Heat map]] | [[Histogram]] - [[Inner query]] - [[Label]] | [[Legend]] | [[Line graph]] - [[Map]] | [[Mark]] | [[MAXIFS]] | [[Mental model]] | [[Movement]] | [[MINIFS]] - [[Narrative]] - [[Ordinal data]] - [[Pattern]] | [[Pie chart]] | [[Pre-attentive attributes]] | [[Proportion]] - [[R Programming|R]] | [[Ranking]] | [[Relativity]] | [[Repetition]] | [[Rhythm]] - [[Scatterplot]] | [[Sort range]] | [[Sort sheet]] | [[Static visualization]] | [[Story]] | [[Subtitle]] - [[Tableau]] - [[Unity]] - [[Variety]] - [[Visual form]] - [[X-axis]] - [[Y-axis]] <hr> ## 1. Communicate Data Insights ### 1.1 Video: _Introduction to communicating data insights_ Video introduction and overview of the entire course. ### 1.2 Reading: _Course 6 overview_ basic overview and refresher on how the certification progression is done. ### 1.3 Video: _Kevin: The power's in the data viz._ Introduction video-bio by the course presenter. I am genuinely excited to learn from Kevin. ### 1.4 Interactive Activity: Google analytics roadmap Cross posted to [[_Self Assessments]] | ==Topics== | Pre-course self-rating | | -------------------------------------------------------------------- | -------------------------- | | [[Design thinking]] | 1 | | How data analysts use visualizations to communicate about data | 2 | | Data-driven storytelling | 2 | | The benefits of Tableau for presenting data analysis findings | 1 | | Dashboards and dashboard filters | 1 (never made a dashboard) | | Strategies for creating an effective data presentation | 2 | | ==Skills== | Pre-course self-rating | | Creating visualizations and dashboards in [[Tableau]] | 1 | | Addressing accessibility issues when communicating about data | 3 | | Understanding the purpose of different business communication tools. | 2 | | Telling a data-driven story | 1 | | Presenting to others about data | 3 DSI @ Gen Assmbly | | Answering questions about data | 1 | ### 1.5 Reading: Helpful resources Coursera provided resources for course learners. No notes. --- ## 2. Understand Data Visualization ### 2.1 Video: _Why data visualization matters_ >[! cue] Def [[data visualizations|data visualization]] Refresher on the meaning of data visualization. Visualizations began a long time ago. Consider a map. This is a useful tool to help someone navigate a space. Map making is an example of data visualization. It also goes waaay back in history. The video gave an ==example== of an early map circa 1500's. But maps [go back further](https://en.wikipedia.org/wiki/Early_world_maps). I wonder what made that map significant? was it the use of "legends?" or was it just an arbitrary example. Either way, some researchers state that the visual system of the human brain represents [about 50% of the cortex](https://www.rochester.edu/pr/Review/V74N4/0402_brainscience.html#:~:text=“More%20than%2050%20percent%20of,out%20Williams%2C%20the%20William%20G.) is used to process visual information. That said, we are primed to receive information and make sense of it using our eyes. In the 1700 and 1800 visualizations using charts and graphs became more and more utilized as the need to process data became more and more intensive (industrial revolution, scientific inquiry). Now, data is ubiquitous and is encoded visually in many many many ways. >[! cue] Common visualization interactions. As an analyst working in today's data environment, will likely use visualizations in one of two ways. 1. Looking at visualizations to draw conclusions about the data it represents 2. Creating visualizations to tell a story >[! cue] 5X5 rule Your audience should know exactly what they are looking at within the first 5 seconds In the next 5 seconds your audience should understand what conclusions that your visualization is making. >[! cue] [[McCandless Method]] Using this method can help achieve a "successful data visualization" - successful I think means that the visualization achieves the 5X5 Intro: Not a deep dive into the method. 4 element venn diagram - Information (data) - Story (concept) - Goal (function) - Visual form (metaphor) ### 2.2 Reading: _Effective data visualizations_ #### All about effectiveness of [[data visualizations|data viz]]! >[! cue] Resource: [What Makes a Good Visualization?](https://informationisbeautiful.net/visualizations/what-makes-a-good-data-visualization/) ![[McCandless Method.png]] Image credit: [McCandless](https://informationisbeautiful.net/visualizations/what-makes-a-good-data-visualization/) Though not a specific paradigm, the McCandless method **informs** elements that make a successful visualization and gives the visualization author the ability to evaluate if certain generalized criteria are being met. Key areas: 1. Information (the data) 2. Story (concept or narrative) 3. Goal (objective or function) 4. Visual form (metaphor or visual expression) #### Fung Q-D-V; A critical methodology >[! cue] Resource: [Junk Charts!](https://junkcharts.typepad.com/junk_charts/junk-charts-trifecta-checkup-the-definitive-guide.html) Fung presents a small set of questions to help us evaluate the usefulness of a visualization. Can use the questions to determine if your own visualization is effective. 1. What is the practical question? (Q) 1. Question should be well posed and interesting. 1. well-posed: search for appropriate data, 2. interesting: the question ensures an engaged audience. 2. What does the data say? (D) 1. The key is that the data be [[SMART methodology - Relevant|Relevant]] to the data. 2. Also of importance; the data's veracity and source is also scrutinized 3. What does the visual say? (V) 1. Visual representation of the data should be clear, concise, and ==directly address the question== Fung suggests the application of the framework move anti-clockwise beginning with Q - D then - V What we want is to have the three elements in agreement. if we are not in total agreement, then the chart is inevitably bad. Below is an example of what we want to result from a critique of a data visualization. Trifecta example ![[Fung Critical Framework.png]] Image credit: [Fung](https://junkcharts.typepad.com/junk_charts/junk-charts-trifecta-checkup-the-definitive-guide.html) The resource link will present a series of eight types of critiques. Reference the link to view examples of a trifecta - **Trifecta**: The trifecta means that the visualization is critically effective (possess all three elements in agreement) - **Single errors**: EG failing D, but passing Q, V etc can lead to ineffective visualizations. - **Double errors**: EG failing two of the three values even more egregious. - **Triple errors**: IE failing all three are good examples of what to avoid at all costs #### Pre-attentive attributes Visualization terminology. Leverage how the brain understands and perceives visual information. >[! cue] Def: [[Pre-attentive attributes]] Pre-attentive elements that make up a visualization that require almost no conscious processing to recognize. This reading introduces two forms: _Marks_ and _Channels_ ##### Marks >[!cue] Def [[Pre-attentive attributes, Marks]] [[Pre-attentive attributes, Marks|Marks]] are ==basic visual objects== like points (nodes), lines (edges), shapes. Each mark has four qualities: 1. Position: Where the mark is - Consider relative position - Consider scale 2. Size - how big, small, long, or tall a mark is. - viz makers need to take special attention to scaling the marks so as to not convey meaning that is not intended or doesn't exist. 3. Shape - Does the shape of the object convey information about it? ![[marks_shape_eg.png]] Image Credit: Coursera/Google - A good resource for images is [The Noun Project](https://thenounproject.com) useful to find images to help encode the shape objects in a visualization. 1. Color - color can encode simple differentiation among groups - or communicate other concepts, profit/loss, hot/cold ##### Channels >[! cue] Def: [[Pre-attentive attributes, Channels]] [[Pre-attentive attributes, Channels|Channels]] are visual aspects or variables that represent characteristics of the data in a visualization. - Specialized marks used to visualize data. >[! warning] Distinguishing between [[Pre-attentive attributes, Marks|Marks]] and [[Pre-attentive attributes, Channels|Channels]] >**Marks** indicate where something is; >- a point in a 0-degree plane >- a line in a 1-d space >- areas in 2-d space >- volumes in 3-d space > >**Channels** are the attributes that control how the marks appear. > The reading emphasizes the **effectiveness** of a channel's ability to convey information based on three aspects. 1. Accuracy 1. Does the attribute being manipulated contribute to the estimating the values being represented. 2. For example, displaying information about numbers in various colors would not help the accuracy of the data being displayed in the same way the below graphic helps convey information, context and groupings. ![[channels_accuracy_eg.png]] Image credit: Coursera/Google 2. Popout 1. control this attribute by making these marks standout visually vs the other marks. Conveys importance or some other aspect of the results of the data. 2. line length, size, width, shape, enclosure, hue, and intensity can be manipulated to Popout values in a visualization 3. Grouping 1. Grouping data marks together to show relative comparisons can aide in data visualization construction. #### Other Resources >[!cue] Other resources - [Video: McCandless - _The Beauty of Data Visualization_](https://www.ted.com/talks/david_mccandless_the_beauty_of_data_visualization?language=en#t-150183) - [McCandless Method Deep-dive](https://artscience.blog/home/the-mccandless-method-of-data-presentation) - [McCandless blog: _Information is Beautiful_](https://informationisbeautiful.net/) ### 2.3 Video: _Connect images with data_ In the video the presenter, Kevin, goes over some basic visualizations and explains how the visualization come to represent the underlying data. #### [[Bar graph]]. In general bar graphs with vertical bars are arranged with categories along X-axis and values or scale along Y-axis. A common pit fall is to arrange the Y-axis on some non-zero starting point which would alter the scale of the graph and represent the data in a less honest way. #### [[Line graph]]. Line charts are great for tracking changes over time. Here the scale remains on the y-axis. X-axis represents time The pitfalls here is you can mess with the scale of the time axis or even the values axis to under-represent or over-represent certain differences. #### [[Pie chart]] Used to convey proportion. There are some conventions used with pie charts that can vary by company or task. Such as never have more than 5 slices of data. etc. An alternative but similar chart is a donut chart that was discussed earlier in the course. A donut chart is simply a pie chart with a hole in it. They both represent parts of a whole. #### Maps Maps can be used when the countries or state ... etc are the category, colors can then be used to encode other information that is centered on the geographic. >[! cue] Resource: [DataVizCatalogue](https://datavizcatalogue.com/#google_vignette) A great resource called out in the reading is the data visualization catalog. [[data visualizations|data viz]] card has been updated ### 2.4 Reading: _The beauty of visualizing_ - https://visme.co/blog/best-data-visualizations/ - [Tableau 10-best viz blogs to follow](https://www.tableau.com/learn/articles/best-data-visualization-blogs) ### 2.5 Video: _A recipe for powerful visualization_ In this video the presenter went over some more chart types. ==The biggest consideration in your visualization design is what you want your audience to focus on.== - often less is best to avoid distractions etc. - visually represent only the data the audience needs to understand the visualizations #### Change over time - can show only the relevant time period to answer or make your arguments. - Time series charts can set this up nice. #### Data distributions - use a histogram to represent how your data is distributed. #### Ranked data - Think first about what you want to highlight. - Use a horizontal chart that shows the ranking. ASC or DESC #### Correlation charts - can show patterns in data, - but is dangerous because audience could associate the correlation as a causation relationship exists. ### 2.6 Reading: _Correlation and [[Causation]]_ This reading dives deeper into the the concepts of correlation and causation. Two resources given - [Medium article by Anthony Figueroa](https://towardsdatascience.com/correlation-is-not-causation-ae05d03c1f53) - [Khan academy](https://www.khanacademy.org/test-prep/praxis-math/praxis-math-lessons/gtp--praxis-math--lessons--statistics-and-probability/a/gtp--praxis-math--article--correlation-and-causation--lesson) ### 2.7 Video: _Dynamic visualizations_ The video introduces a few new concepts in the curriculum. 1. [[Static visualization]]; is one that does not change. Once it is created, it will need to be re-created if changes need to be made. 1. The key benefit here is that you, the author of the visualization, have control of the messaging the visualization is reporting. 2. [[Dynamic visualizations]]; are made that can change with some form of interactivity. 1. The benefit here is that users of the visualization will have a degree of control over what is being presented in the visualization (EG a year slider that can change the year view of a dataset) 2. The downside to this type is that as that author there is a loss of control over the messaging in the visualization. As part of the share process, out job is to make this selection fist. _What kind of visualization is needed to effectively convey the business task and the solution?_ If the data is live, then a dynamic-style dashboard is appropriate despite the lack of control the analyst will have over the data. ### 2.8 Reading: _The wonderful world of visualizations_ The reading discusses more basic chart types in greater detail than before. Just like playing a sport, to do this professionally, one must MASTER the basics. This section is by no means aiming to demonstrate mastery, but I should really understand the purpose behind some of these visualizations. #### Line chart - used to track changes over time for one or more groups. - useful to display when smaller changes exist as opposed to using a bar graph because, in a line chart, the data consumer is able to see the slope of the line to show the change rather than relative heights of bars. #### Column (bar) chart These charts use SIZE to compare / contrast two or more values. Color can be added to encode distinct groups over the same variable. #### Heatmaps The heatmap use color to compare categories in a dataset. mainly used to show relationships between two variables. I am used to showing correlation values among two variables and representing them in a heatmap. The reading shows a heatmap showing the variable `City` vs `month` encoding temperatures from `hot` to `cold` on a color gradient. ![[heatmap_eg.png]] Image credit: Coursera/Google. So the temperature of Reykjavik in June is relatively colder then the temperature of Anchorage in June. #### Pie Chart Show distribution of proportion if that is the point. The example chart is favorite movie categories. Looking at the pie, I just see that comedy is largest and most favorite of the other categories. I am not sure if this would be more or less effective than using a horizontal bar chart to show the rank order. #### Scatterplot Use a scatterplot to show the relationship between two variables. Though additional variables can be displayed. #### Distribution graph Shows the spread of values in a dataset. This is where histograms are. #### Takeaways Study your dataset look for - Change - Is there a trend? - Do observations become different over time? - line or column charts are useful - Clustering - A collection of data points with similar values - represent using a distribution graph. - Relativity - Considered in relation or in proportion to something else - Pie/donut charts used here to show relative proportions. - Ranking - position in a scale of some kind. - Use a column chart but then order it to clearly see the rankings. - Correlation - Heatmaps or scatter charts good to use here. ### 2.9 Reading: _Data grows on decision trees_ This reading focuses on building a workflow that ultimately helps you select a useful and effective chart/visualization. The reading is set up like a [[Decision tree]] to help guide you to a good starting point. To begin, start with the STORY. Evaluate the type of data you have, the analysis that has been done. - How many variables do I have ? what is needed to tell the story? - How many datasets are involved? - What is the time period being included? Am I measuring things agains time? - Are there relationships between the data that need to be shown. There are two resources provided at the end. [From data to viz](https://www.data-to-viz.com) more detailed decision tree that depends on the data type and number of variables being measured. Note: the [[data visualizations|data viz]] card is now updated with the above resources. ### 2.10 Self-reflection: _Choose your visualization type._ This was a bit tough. my gut was to select a line chart. DO NOT GLOSS over what the project is supposed to be about. ### 2.11 Quiz: _Test your knowledge on data visualizations_ FINALLY. 4 questions, untimed, 75% minimum score required. #### Results 75% then 100%... Suggested to review glossary prior to taking future quizzes. There is a lot of definitions and concepts that could easily be confused if the precise definition is not used. ## 3. Design Data Visualizations ### 3.1 Video: _Elements of Art_ Yay. Art! Line - adding form to data visualization Shape - 2-d only. - shapes with symmetry easier to understand. Color - Use of colors can also encode data within them like in [[Heat map|heatmap]] use. - Hue is the color, Red, Green, Blue ... - Intensity how bright or dull - Value how light or dark a color is - Shades of ... color are colors with black added. - Tint of ... color as colors with white added. Space - area between in and around the visual elements, Movement - movement is used to bring attention and interesting variety - EG health and wealth of nations visualizations. - Fine line between attracting attention and distracting elements. ### 3.2 Reading: _Principles of Design_ The reading covers basic design principles as applied to data visualization. There are 9 principles of design; Checks during visualization design process/iteration: 1. [[#3.2.1 Balance|Balance]] 2. [[#3.2.2 Emphasis|Emphasis]] 3. [[#3.2.3 Movement|Movement]] 4. [[#3.2.4 Pattern|Pattern]] 5. [[#3.2.5 Repetition|Repetition]] 6. [[#3.2.6 Proportion|Proportion]] Check on visualization after design completion: 7. Rhythm 8. Variety 9. Unity #### 3.2.1 Balance >[! cue] Eg: [Balanced Column Chart](https://developers.google.com/chart/interactive/docs/gallery/columnchart) No one element or side of the visualization should distract from the other. The does not mean that one needs complete symmetry either. Example in a bar chart, each column is the same width Each space between columns is the same. This even ness provides the symmetry >[! cue] Def: [[Balance]] The design principle of creating aesthetic appeal and clarity in a data visualization by evenly distributing visual elements #### 3.2.2 Emphasis >[! cue] Eg: [The Pudding's _Where Slang Comes From_](https://pudding.cool/2017/02/new-slang/) - The visualization needs a focal point. - Emphasize the most important data - Using _color hue_ and _color value_ is effective. - This is a technical way to say, use contrasting colors for the important data >[! Cue] Def: [[Emphasis]] The design principle of arranging visual elements to focus the audience’s attention on important information in a data visualization #### 3.2.3 Movement >[! cue] Eg: [Combo chart showing movement](https://developers.google.com/chart/interactive/docs/gallery/combochart) - Either the path of the readers eye when reading - Or, an animation >[! cue] Def: [[Movement]] The design principle of arranging visual elements to guide the audience’s eyes from one part of a data visualization to another #### 3.2.4 Pattern >[! cue] Eg: [Stacked Column Chart] - The use of similar shapes and colors arranged in a meaningful way. - Also the use of consistent pattern can make complicated charts easier to read. - Using contrasting pattern can also call attention to key differences that your visualization intends to highlight. >[!Cue] Def: [[Pattern]] The design principle of using similar visual elements to demonstrate trends and relationships in a data visualization #### 3.2.5 Repetition Repetition is the extension of pattern. >[! cue] Def: [[Repetition]] This extends a particular element into a pattern and by repeating it, your audience would not need additional encoding information. #### 3.2.6 Proportion >[! cue] Eg: [Dashboard Pie chart](https://developers.google.com/chart/interactive/docs/gallery/controls) - This is also one of the most effective ways to call out data. - The first six principles should all be primary considerations. >[! cue] Def: [[Proportion]] >[! info] The design principle of using the relative size and arrangement of visual elements to demonstrate information in a data visualization. #### 3.2.7 Rhythm - Refers to creating a sense of movement >[! cue] Def: [[Rhythm]] The design principle of creating movement and flow in a data visualization to engage an audience #### 3.2.8 Variety - The visualization project should have variety of chart types, lines, shapes. - Used to keep audience engaged. - Find a good balance between minimalism and variety >[! cue] Def: [[Variety]] The design principle of using different kinds of visual elements in a data visualization to engage an audience #### 3.2.9 Unity - cohesive data visualization is more effective. All charts have unifying elements/colors/shapes >[! cue] Def [[Unity]] The design principle of using visual elements that complement each other to create aesthetic appeal and clarity in a data visualization ### 3.3 Video: _Data Visualization Impact_ The key to every visualization project is: - Which visualization can I use that will make the audience understand my conclusion in the first five seconds and then understand my argument in the next five seconds? #### Changes over time; - [[Line graph]] (BP) - [[Bar graph]] - [Stacked bar](https://datavizcatalogue.com/methods/stacked_bar_graph.html) - [Line/area chart](https://datavizcatalogue.com/methods/area_graph.html) #### Between Objects; - Ordered bar chart - [[Bar graph]] which each bar is ordered. Arranged horizontally with bar on y-axis - Ordered column chart - Grouped bar chart, [Multi-set Bar](https://datavizcatalogue.com/methods/multiset_barchart.html) #### Data composition >[! cue] Def: [[Data composition]] The process of combining the individual parts in a visualization and displaying them together as a whole - [[Pie chart]] or [[Donut chart]] - a [Stacked Bar chart](https://datavizcatalogue.com/methods/stacked_bar_graph.html) - [Treemap](https://datavizcatalogue.com/methods/treemap.html) - [stacked area chart](https://datavizcatalogue.com/methods/stacked_area_graph.html) #### Showing relationships The charts compare the relationship of two or more variables with one another. - [[Scatterplot]] - [Bubble chart](https://datavizcatalogue.com/methods/bubble_chart.html) - [[Column chart]] with trendlines - [[Heat map|Heatmap]] #### 3 key elements of good viz 1. The visualization has a CLEAR MEANING 2. The visualization makes excellent use of contrast 3. The visualization has refined execution. Basically uses [[Principles of Design]] and ART Theory (position, size, shape, color (Hue, Value, Tint, Shade), movement, balance, emphasis ...) ### 3.4 Reading: _Data is Beautiful_ Breaking down the [[McCandless Method]] ![[McCandless Method.png]] Image credit: [Knowledge is Beautiful] #### Information: - What is the data you're attempting to convey. #### Story - Story drives the inspiration to action. The visualization should drive the narrative. - Avoid: "informative, but not inspiring" #### Goal - goal of the visualization is to make the data useful / useable. - What are you trying to achieve with the visualization, what is the point you're making? #### Visual Form - the elements that give the visual elements structure - what makes it beautiful The reading reviews two visualizations created by McCandless. The first is one compiled from AKC data for their best in show. The next is about sea levels rising. ### 3.5 Video: _Design Thinking and Visualizations_ This video introduces design thinking >[! cue] Def: [[Design thinking]] A process used to solve complex problems in a user-centric way Kevin gave an example of AirBnB using design based thinking (user centric thinking). The presenter introduces the concept of [[Design thinking]] as it applies to data visualization. There are five phases, but not necessarily in any order. 1. Empathize 2. Define 3. Ideate 4. Prototype 5. Test The details discussed in the following reading. ### 3.6 Reading: _Design thinking for visualization improvement_ Case study: banking analyst. The current setup at a bank has a budget view as a donut chart. Each section of the donut is defined by a spending limit and a user can see their spending relative to the limit of each of their budget categories previously defined. The exercise is that we go through each of the five aspects of the design thinking process and see what to do. ![[design_think_ex_bank.png]] #### 1. Empathize for empathy (terrible name) we need to think from the customer perspective. From a customer perspective; this is nearly unusable. I am unable to tell how much is left in each category. The only think I know is how full each category is relative to each spending limit. There are no dollar values. and I have to keep going back and forth from the legend to the donut. So, from the perspective of an existing user.. I wouldn't use it. too confusing. Sample questions: **Do the colors and labels make sense?** - No, too many blue colors, Education and groceries nearly indistinguishable. I don't know which one is which. **How easy is it to set or change the budget?"** - I don't see a button for that in the UI given. **When you click on a spending category in the donut chart, are the transactions in the category displayed?** - No idea. #### 2. Define Are there other visualizations that could be used to help customers? The reading has the idea of including an "income" stream. I disagree with that. **Can you think of anything else?** - I think annotations of actual spend or remaining budget would be helpful. But this approach to budgeting in general may need to be rethought out. #### 3. Ideate - I would like to see a way for a user to adjust the categories. etc. - Let's see buttons that allow editing of budget - How about a metric that that "projected budget spend" or something. Displaying the projected end of month spend vs budget. Sucessful budgeting means changing budget, and smart resource allocation. If we are projected to be over budget, what would we plan to do about it? #### 4. Prototype. - at this stage, developers would take over. #### 5. Test. Testing the finished visualization/dashboard with other members of the team. Iterative development would be most beneficial to the customer. I think in a second round we would think along the lines of the business and really get strong justifications for any changes that we would be making to the existing product. Change management is pretty costly so that would need to also be factored in. ### 3.7 Quiz: Test your knowledge on designing data visualizations 4 questions. untimed, 75% minimum score required. #### Results 100% Tricky question around Color terms from the beginning of the section. it was a _select all that apply type_ of question that really only had just one correct response. Be careful. ;) ## 4. Visualization Consideration ### 4.1 Reading: _Pro tips for highlighting key information_ invite audience, keep engaged >[!cue] Rem: 5x5 rule We want audience to **==process and understand==** the information I am trying to share in the first five seconds. The reading will teach you how to engage your audience immediately. Beginning with this visualization, the reading will define concepts and iteratively improve the _base visualization_: ![[visualization_base.png]] Image Credit: Coursera/Google. Looking at the base visualization, nothing is labeled. I have no context to understand what the data are showing. The only thing I can see is that yellow ~~did better~~ is higher than than green or blue relative to their starting position. I can infer that there are three distinct things being compared here because they are all different colors. Line chart assumes change in time, and I don't know what the y-axis is, exactly. #### 4.1.1 [[Headline|Headlines]] that pop >[! cue] Def: [[Headline]] Text at the top of a visualization that communicates the data being presented The headline should immediately communicate what the visualization is about. ![[visualization_base+headline.png]] Image credit: Coursera/Google Adding headline is critical to understanding the intent, content, and context of the visualization. With the added information about the content, I now know that yellow is not "better" I can now even draw some conclusions or further assumptions in the chart. Rents in tri-city are trending down. This means that rent level in currency unit is inferred on the Y-axis. The line plot used infers that this is change over some unit of time. The precise unit of time, days, months, or years, are still unclear. Given the color's used I am assuming that each line represents a city in the tri-city area. >[! note] No use of acronyms >EG: _Average_ is used and not _Avg._ >[! warning] Best Practices >**Guidelines:** > 1. Be brief > 2. No longer than the chart length. Single line > 3. Always above the data > >**Style:** >1. No italics >2. No acronyms >3. No abbreviations >4. No humor/sarcasm #### 4.1.2 [[Subtitle|Subtitles]] that clarify >[! cue] Def: [[Subtitle]] Text that supports a headline by adding context and description The subtitle adds the context of which "Tri-City Area" we are talking about. ![[visualization_base+headline+subtitle.png]] Image Credit: Coursera/Google Some of us may also be wondering why not identify that these towns are in California or part of the San Diego area? I believe we skip that because .... perhaps this audience is already aware OR this is intended for a live presentation which explaining the area being near San Diego could be done verbally. >[! warning] Best Practices >**Guidelines** >1. Use to ==clarify== context >2. usually fewer than 30 characters >3. Directly below headline > **Style** >1. Smaller font than headline >2. no undefined words >3. no styling (bold, italic, caps) >4. No acronyms or abbreviations #### 4.1.3 [[Label|Labels]] that identify >[! cue] Def: [[Label]] Text in a visualization that identifies a value or describes a scale. Our working concept is that a label identifies data in relation to other data. - typically the X and Y axes get labels ==**Always label your axes**== ![[visualization_base+headline+subtitle+labels.png]] Image Credit: Coursera/Google Now we know what our unit of time is (month) and what is being measured in the y-axis and the unit of measurement (Average in $) We can also label data directly to avoid using legends. ![[visualization_base+headline+subtitle+labels+direct_Labels.png]] Image credit: Coursera/Google Now each category is labeled next to the corresponding line, with a matching color to encode it. Interesting that each month is not labeled. I like that choice since I don't really need to know it. It might be more critical if the time period spanned multiple years. ??? no idea. might be a good design discussion with the team prior to finalizing the visualization. >[! warning] Best Practices >**Guidelines:** >1. Direct label to replace need for legends >2. fewer than 30 characters >3. Next to data (direct labeling) or below/beside axes > >**Style** >1. Be thoughtful on color coding. >2. Callouts to point to the data >3. no styling. #### 4.1.4 [[Annotation|Annotations]] that focus >[! cue] Def: [[Annotation]] Text that briefly explains data or helps focus the audience on a particular aspect of the data in a visualization ==Not all datapoints need to be annotated== See that only the highest average rent was called out on the chart. Now I have an anchor value to make more informed conclusions about the data underlying the chart. ![[visualization_base+headline+subtitle+labels+direct_Labels+Annot.png]] Image credit: Coursera/Google >[! warning] Best practices >**Guidelines:** >1. Draw attention to ==certain== data >2. length can vary, limited to whitespace on plot >3. Immediately next to data being annotated > >**Style:** >1. No styling >2. No text rotations >3. Do not distract viewers from the data. ### 4.2 Video: _Accessible visualizations_ The presenter reviewed a few ways to make data visualizations more accessible - **Labeling**... labeling and making good use of color-encoding do away with the need for a legend. - **Text alternatives**- these convert the visualization into other forms and formats (like braille) - **Text-based format** - convert the visual content into a text description - **Distinguishing** using high contrast foreground or background. Could do away with color encoding all together and just go with alternating grayscale textures and patterns - Simplify - LESS is more ... too often the curse of expertise tells us that we need to info dump on everyone so that they are on the same page... not true. try displaying only the needed data. ### 4.3 Video: Andrew: Making data accessible become more empathetic to the audience. Making the visualization more inclusive, you are making your data clearer and more impactful for everyone. - **Prioritize inclusivity!** - **Clarity without context**... that is, you will not be around to guide every audience member after the visualization is published. So, have the data be self-explanatory as much as possible. - **Empathy** ... by having empathy for your broad audience, being inclusive makes you a better analyst and data storyteller. - **Strive for action and IMPACT**: clear and inclusive visualizations are more effective in conveying insights, motivating action, and ensuring broader understanding. ### 4.4 Reading: _Design a chart in 60 minutes_ Here the reading goes through how to allocate time in designing a chart. There are 4 stages: - prep (5 min) - This is a brainstorming session. Expansive comprehensive thinking. - Talk and listen (15 min) - Talk and listen to stakeholders, Identify the question behind the question - Sketch & Design (20 min) - Draft the approach to the problem define the timing and output - Prototype and improve (20 min) - Here the visual solution is generated. - Iterate on ideas/feedback. - This might mean making different versions on the same data. That's ok. ### 4.5 Hands-on: _create your own viz_ review of [[Design thinking]]; empathize, define, ideate, prototype, test ### 4.6 Quiz: _Test your knowledge on exploring data visualizations_ 4 questions, Untimed, 75% minimum required #### Results 100% ## 5. Module 1 challenge ### 5.1 Glossary review - [[Alternative text]] | [[Annotation]] | [[AVERAGEIF]] - [[Balance]] | [[Bar graph]] - [[Calculus]] | [[Causation]] | [[Channel]] | [[Chart]] | [[Cluster]] | [[CONVERT]] | [[Correlation]] | [[CREATE TABLE]] - [[Data composition]] | [[Decision tree]] | [[Design thinking]] | [[Distribution graph]] | [[DROP TABLE]] | [[Dynamic visualizations]] - [[Emphasis]] - [[HAVING]] | [[Headline]] | [[Heat map]] | [[Histogram]] - [[Inner query]] - [[Label]] | [[Legend]] | [[Line graph]] - [[Map]] | [[Mark]] | [[MAXIFS]] | [[Mental model]] | [[Movement]] | [[MINIFS]] - [[Narrative]] - [[Ordinal data]] - [[Pattern]] | [[Pie chart]] | [[Pre-attentive attributes]] | [[Proportion]] - [[R Programming|R]] | [[Ranking]] | [[Relativity]] | [[Repetition]] | [[Rhythm]] - [[Scatterplot]] | [[Sort range]] | [[Sort sheet]] | [[Static visualization]] | [[Story]] | [[Subtitle]] - [[Tableau]] - [[Unity]] - [[Variety]] - [[Visual form]] - [[X-axis]] - [[Y-axis]] ### 5.2 Quiz: _Module 1 Challenge_ 8 questions, Timed, 80% minimum to proceed #### Results 100% ## Summary and takeaways >[! summary-top] Module ___ Key Take Aways >Design thinking and accessibility are the most important.