Category: Data Visualization

Millennials are a generation of people born between the years 1981 and 1996. They are currently between the ages of 26 and 41. As of 2019, there are approximately 71.2 Million Millennials in the United States, making up 22% of the population, serving as the largest generation in the U.S. to date. As a Millennial, I was privileged to become a homeowner in September 2020. All-time-low interest rates coupled with Covid-19 mandates and quarantines made it difficult to find a home within my preferred county that aligned with my budget. As a result, I settled on a home two hours away from my work office and one to two hours away from my loved ones and the areas I have grown to love.

Though I was able to purchase a home in a highly unpredictable market, it was not easy at all. In our society, it has become a common trope that Millennials are unable to afford homes due to factors including rapid increases in home value, record inflation rates, and crippling student loan debt. But is there any merit to this assumption or is it a spiraling joke?

Millennial Income

The first step in answering this question is determining the income of the Millennial generation. According to the most recent U.S. Census, Millennial households made a median of $71,566 in 2020. When adjusting for inflation, this number resembles earnings of similarly aged Generation X households about 15 to 20 years in the past. Since relative income has largely stayed the same while home prices have increased, this first piece of information supports the argument that purchasing homes will be more difficult for Millennials. However, understanding how much home values have increased is also important in validating the theory. 

Home Value in the United States Today

According to the Federal Reserve Bank of St. Louis, the average sale price of houses in the United States increased from $205,000 in January 2000 to $457,000 in Jan of 2021. In 21 years, that’s an over $250,000 increase in prices. Though surprising, those numbers are not indicative of the current situation in 2022. Today, Zillow reports that the average cost of a home is $349,000 which is $100,000 less than the previous year. Those are some interesting numbers. It took 20 years to increase $250,000 in value, but one year to decrease $100,000 in value. The only other time in recent history such a large change occurred was in 2008, and it was caused by the recession. So what influenced the latest massive drop in home prices?

Interest Rates

The main factor that influenced the rise and drop in home prices was historically low interest rates. From 2019 to 2021, interest rates averaged 3.33%, which were the lowest interest rates ever recorded. This decreased the monthly mortgage of homes in the U.S. while simultaneously increasing the affordability of homes for homebuyers. As a result, homebuyers flooded the market and were able to pay more than sellers were asking to secure the homes they wanted. However, record-low rates from the Federal Reserve were intended to accommodate Covid-era policies and temporarily stimulate the economy. Thanks to vaccines and the lifting of lockdown restrictions, the economy recovered rapidly. This led to a spike in mortgage interest rates within the first six months of 2022, from 3.86% to 5.78%. This spike decreased what homebuyers could afford monthly, thus reducing offers on homes as well. 

Search US Interest Rates Over Time

What does this mean for Millennials? 

It is possible that the unusual spike in home prices was temporary and could continue to decrease over time. Unfortunately, it is impossible to predict future trends. Therefore, the best we can do is calculate the affordability of homes with the current home value average of $349,000.

So, how do you calculate the affordability of a house? Banks use a debt-to-income ratio to determine whether or not someone can afford a loan. There are two approaches to go about determining debt-to-income ratios: on the front end and the back end. Front-end ratios are computed by dividing your monthly costs by your gross income. The back end ratio is determined the same way as the front end, but adds all other recurring monthly debt into the equation. For the sake of implication, we will be using the front end ratio and utilizing an online calculator. If you bring in the median household income of $71,500, you can afford a house priced up to $263,341, with a 5.75% interest rate with a down payment of $52,668 and a loan amount of $210,673. This maximum home value falls below the today’s average home value of $349,000. Unfortunately, based on the 28/36 rule, A Millennial making the median income cannot afford the average home in 2022. 

However, this does not tell the whole story. Indeed, a Millennial most likely will not be able to afford the average home cost in 2022, however, this does not mean they cannot afford a home. The average of something naturally means that there are data sets both below and above that number. With a budget of $263,000, depending on the location, a Millennial can afford a home valued below the average of $349,000.

For example, the average cost of a home in Mississippi is $166,000 while the average in California is $465,000. The region, state, city, county, and market play a huge factor in the average cost of a house. Additionally, some local grants and programs assist in home buying, which promotes homeownership among different classes of people and makes homeownership more attainable. Therefore, we can potentially conclude that Millennials could easily afford homes within our economy.

Additional Factors to Consider

Another significant factor exists in student loan debt. Student loan debt is arguably the largest obstacle that Millennials have to overcome to reach financial independence. The average Millennial has over $38,000 in student loan debt. This number factors directly into the debt-to-income ratio of a mortgage applicant, decreasing the chances of being approved for a loan and reducing the amount Millennials qualify for. Interestingly enough, Millennials do not hold the highest amount of student loan debt on average. Gen Xers have an average of $45,000 in student loan debt, which is $7,000 more than Millennials. However, because Gen Xers are likely further along in their careers, they are already earning much more income which cancels out the difference in debt. 

In our Society, Can Millennials Afford Houses? Yes or No?

The answer is, “Maybe.” Generally, the “average” Millennial cannot afford homes priced at the national “average” home value. However, there are many other factors such as debt, government assistance, location, and more that determine an individual’s eligibility to afford a home. Millennials are indeed experiencing a much more volatile economy in a condensed amount of time, which makes it difficult to follow in our predecessor’s footsteps and invest wisely or save more aggressively. Ultimately, the eligibility for affording a house lies mainly with one’s financial responsibility, literacy, discipline, and serendipity. 

My second attempt at Dear Data covered the complaints I made within a week. This time around, out of convenience, I collected my data in a Google sheet.

The following were topic and supporting questions that helped guide my data collection:

What are my complaints?

• What annoys or frustrates me? What do I find unacceptable/unsatisfactory?
• In which settings do I complain? (Relationships, work, school, etc.)
• Is my complaint attached to a specific subject? (person, place, thing)
• When/how often do I complain? (date, day of the week, time, etc.)
• Is my complaint valid? (Am I being dramatic?)
• Who do I complain to?
• How do I manage my complaints? Do I take any further action?

Although I started collecting data with the subject of the complaint in mind, I found that I didn’t truly need this information in my final visualization. I’m already pretty aware of the things I typically complain about. Additionally, I eventually omitted the actual topics of each complaint. Though I initially thought my visualization would orbit this information, in the end, I found that the most significant trends within the data were how often I complain, who I complain to most and if I take any further action after complaining.

Here’s how it went:

My findings after collecting this data: 

• Most of my complaints were valid, although I did throw in a little drama every so often
• My husband hears the majority of my complaints
• 3/8 of my complaints were about work
• I did not take further action as much as I would like to

Now that I know this information, I can make better decisions about my complaining habits, whether I want to continue to complain, and what I’ll do with my complaints moving forward. 

Furthermore, creating my own Dear Data visualizations was a great exercise that I hope to continue. Gathering my data around life happenings and habits keeps me present and conscience, offering insight into decisions I have made or may be making. I was also able to identify and analyze which of my habits are healthy and unhealthy. Through these exercises, I practiced being introspective during a very busy life season, which I appreciate.

I struggled the most with the visualization of my data, but with time and practice, I think I’ll learn to try more things. The more I practice, the more comfortable I’ll become, and the easier it will be to track information within my daily life. 

Maps are one of the most common data visualization forms we use today. Used to highlight geographical regions in data, like “countries/regions, states, counties, or postal codes,” maps help quickly identify the area that is being reported on and sometimes incorporate specific details. 

Although once performed by hand, creating maps can be time-consuming and tedious, as they involve complex features and elements, such as scale and symbols. Therefore, to assist with processing data tied to geographical regions and account for the complexities, powerful tools such as Excel, Tableau, Datawrapper, and the like have been created. Datawrapper is a user-friendly, open-source web tool that you can use to create beginner to advanced interactive charts. 

My first impression of this tool was that it is very user-friendly with a simple interface. After a few years of experience creating and displaying charts and graphs between Microsoft, Apple, and Adobe software, and with a little experience in Tableau, I was relieved to begin utilizing a tool designed with non-technical people in mind. Using Datawrapper’s 4-step process, I was able to create visually appealing charts, graphs, and maps swiftly– and it was an enjoyable experience!

Population Over Time

My first task was to create a chart to accurately depict the change in the population of three popular African countries, Nigeria, Ghana, and Cameroon, over the past 70 years. My primary inclination was to utilize a line chart to indicate change over time. This process was fairly simple.

To experiment further, I attempted to create a table to display the same information. This quickly became more lengthy and involved. due to the way tables display information compared to line charts, creating a table required a reorganization of my data. After reorganizing the data, I learned how to use sparklines to summarize the overall trend. Next, I inserted an additional column to indicate the difference in growth from 1950 to 2020 for all three countries. This resulted in a much more informational table. 

Finally, I tried my hand at creating a map. However, the data I collected highlighted non-geographical patterns and supported analysis more than observation. In this case, I returned to the original data I collected (before cleaning) and cleaned again, seeking the 10 most populous countries in Africa in 2020, which could be displayed using a symbol map. 

A symbol map is a great map for representing specific locations within a larger region. To make the symbols vary in size, I inserted a new column containing the 2020 populations of all ten countries and sized my marker shapes based on the data in that column. My map is responsive and interactive, communicating the population information through size and leaving the numbers out. Overall, these three visualization methods helped me further understand the importance of pairing the appropriate visualization with the dataset and details you are working with.

Maximum Temperature in a Year

Next, I moved on to creating another chart detailing the maximum temperature of the month in 2021 at the National Arboretum in Washington, DC. I decided to use another table to visualize this information. Although I wanted to focus on the day with the highest temperature per month, I found that the table required me to revisit the original data and add the data covering every day of the month back in. This made the table more versatile and highly effective, as it houses a line chart showing temperature patterns over time, the maximum temperature of the month, and the specific day of the month when the maximum temperature occurred. Additionally, I was able to incorporate a heatmap to highlight the highest temperatures and the lowest temperatures in different colors. As a result, the viewer can quickly differentiate the temperatures. The result was a stunning chart that shares a vast amount of data in an easily understandable format.

For additional practice, the type of map I attempted to pair with this data was a locator map. Locator maps are perfect for displaying the places you are mentioning within an article. They don’t communicate specific data points and numbers like temperatures in °F. They simply state. “this is where it is located” or “this is where it happened.” Since the National Arboretum is located in Washington, D.C., I wanted to display the Arboretum within the context of its home district. 

Creating this map was straightforward and fun to play around with. I chose a tree as the symbol of the location marker and included an inset map of the District of Columbia so viewers could reference the surrounding area. In future locator maps, I would like to experiment with adding in multiple locations, utilizing 3D buildings, and incorporating annotations and a key. I have found that locator maps can be helpful and insightful for viewers because they provide unique context and may be easier to absorb than a paragraph of writing.

Maryland Vehicle Theft Within Counties I Have Lived In

Lastly, I created a line graph depicting the number of motor vehicles reported stolen within three Maryland counties between 2010 and 2020. Using a choropleth map, I highlighted three different counties: Montgomery County, Prince George’s County, and Baltimore County. To support the story behind my data (in a perfect world), I would be able to select a specific year within the 10-year range and observe the color change over time, However, choropleth maps do not work well for displaying the correlation between values or observing subtle numeric differences. Therefore, I revisited the possibility of a line chart to display this data.

Altogether, while using Datawrapper I gained a greater understanding that not all data related to a specific area or region works well in a map visualization. Furthermore, some forms of visualization are effective when the data is comprehensive while others get by with simplified, general information. The key takeaway here is that data visualization requires thoughtfulness and intention. And it’s okay to experiment with various visualization approaches to find the most suitable and accurate one. 

If you’re looking for a visualization tool, Datawrapper is a great place to start! It is user-friendly and there are numerous resources online to assist with any questions that may come up. Moving forward, I plan on using Datawrapper lots more.

Dear Data is a project by two information designers, Giorgia Lupi and Stefanie Posavec, involving the collection of data from their daily lives on postcards and mailing them to one another. On the front of each postcard is their chosen data visualization, and the back features a key detailing how to read the data. Much like popular mood and habit trackers, the postcards feature unique visualizations. These visualizations are comprised of custom symbols, color variation, shape, thickness and length of lines, and the use of left and right, all defined by the designer behind the data. 

To kick off my own Dear Data adventure, I started by identifying a topic and supporting subquestions to guide my data collection:

• What are my eating habits?
  • What have I eaten? (category: snack, fruits, meal, etc.)
  • What did I eat? (specifics)
  • When did I eat? (date, time of day, etc.)
  • What was I doing while eating? (working, watching a show, at dinner, etc.)
  • What were the portions? (mini, small, mid, large, surplus, etc.)
  • Was it home-cooked? If so, who cooked?
  • Was it fast food or carry-out? If so, from where? How much did it cost?
  • Did I visit a restaurant? If so, what restaurant? With whom? How much?
  • Why did I eat? (hungry, craving, bored, snacking, offered to me, just because, etc.)
  • What was my hunger level? (low, average, high, starving)
  • Was it satisfying? (satisfied, neutral, dissatisfied)

Next, I created a simple table in my sketchbook to track everything I ate for an entire week. My subquestions became each column within the table, and the rows were the meals, organized by date and time.

As I collected, I started noticing trends within the data. First off, the week in which I tracked my eating habits was a peculiar week to do so. It was an extremely busy week that consisted of completing design work all day, having late-night work sessions, attending multiple meetings and class sessions, and commuting for various appointments and errands. The main trend I noticed was that I ate out every single weekday. My reasons for purchasing food instead of eating at home or cooking were primarily out of convenience, and to save time. I also think eating things I knew I would be satisfied with offered a sense of comfort and peace of mind in contrast with the chaos of my work week. Another trend I noticed– I tend to purchase decent portions of food and span my consumption of those meals across a couple of days. I also do not have a routine for my meals; I just squeeze in food where I can and often wait until I’m hungry (with an average or high hunger level) to eat.

Once all of my data was collected for the week, I began exploring possible ways to refine and visualize all my findings. I started by reviewing each column and deciding what was most important to display in my final visualization. I eliminated insignificant columns and details to simplify things, including dish name, fast food/carryout, restaurant, and why I ate. Some of this information resembled other existing information closely, while other information was irrelevant to the story I wanted to focus on. This left me with simplified versions of the columns that would eventually define my key: the day, time of day, level of activity, portion size, hunger level, satisfaction level, meal category, the reason for eating, and whether or not the food was home-cooked.

Here’s what I ended up with:

Overall, I learned a lot about myself through this exercise. For starters, the majority of the times that I ate in one week, I was trying to multitask and accomplish something else while eating. The frequency at which I ate each day was inconsistent from day to day because I squeezed in meals where I could. I also only ate when I had an average or high hunger level. Finally, 1/9 of the times I ate was a home-cooked meal– and I was not the chef!

When visualizing information, selecting a decent chart or graph type to display your data may seem like a no-brainer. We often throw things together quickly, choosing a chart type on impulse. It might surprise you that this is not the ideal approach to visualizing information. In fact, a better practice involves asking yourself a couple of questions about your data. 

The Nature and Purpose Questions

According to Scott Berinato, you should consider the following questions about the nature and purpose of your visualization in order to start thinking visually:

1. Is the information conceptual or data-driven?
2. Am I declaring something or exploring something?

These two questions combine in a 2×2, creating four possible types of visualizations: Conceptual-Declarative, Conceptual-Exploratory, Data-Driven-Exploratory, and Data-Driven-Declarative.

The Four Types

Conceptual-Declarative

In the quadrant between conceptual and declarative lies Idea Illustration. The goal of idea illustration is to simplify complex concepts. This kind of visualization illustrates processes and frameworks, commonly resulting in a simple and metaphorical visualization. Idea illustration is often utilized in presentations and teaching and can often involve trees, bridges, and circles, showing up as decision trees, org charts, tree diagrams, cycle diagrams, and the like. 

Water cycle diagrams are a good example of idea illustration because most of them highlight the main stages (evaporation, condensation, precipitation, and collection) in a cycle format. Most of these diagrams do not visualize the complexities of the water cycle. They often omit additional phases that may make the concept more challenging to understand, including surface run-off, percolation, transpiration, and more. Instead of throwing all aspects of the cycle into one diagram, they are grouped into four basic categories. Through teaching, these categories are further explained. The simplification that exists within water cycle diagrams is a key feature of conceptual-declarative visualizations. 

Conceptual-Exploratory

Also known as Idea Generation, conceptual-exploratory visualization is “typically used in group brainstorming sessions to gather as many diverse points of view as possible.” Idea generation is used for discovery, learning, concept simplification, collaboration, and design thinking. This form of visualization often takes place informally, involving whiteboards, sketchbooks, butcher paper, napkins, etc. This approach is a “go-to for thinking through complexity.”

Design thinking involves a lot of idea generation, including mind mapping. Mindmapping is a graphical technique in which participants build a web of relationships. By branching out from a central main idea, this strategy creates a visual representation of related terms and ideas. Mindmaps are successful conceptual-exploratory visualizations because they support the discovery and dissection of complex ideas. 

Data-Driven-Exploratory

The third visualization type exists between data-driven and exploratory. Visual Discovery seeks to identify trends, make sense, and achieve deep analysis. Within the 2×2, this quadrant is the most complicated, as it incorporates big, complex, and dynamic data and covers two subcategories: confirmation and exploration. Exploration is open-ended, while confirmation is more focused exploration. 

The $Billion Dollar o-Gram, explained further by its creator, David McCandless, is an example of a data-driven exploratory visualization. It displays American spending, earning, potential giving, fighting, giving, owing, and accumulating. McCandless describes this visualization of complex data as a “landscape that you can explore with your eyes, a kind of map, a sort of information map.” Over the years, there have been many iterations of the $Billion Dollar o-Gram, including interactive versions, that allow users to explore and make sense of the vague “billions” of dollars often mentioned by American news outlets. This evolving data promotes discovery and analysis, characteristics of data-driven exploratory visualizations.

Data-Driven-Declarative

Finally, data-driven-declarative visualizations are also called Everyday Data Viz. This form of visualization is the one most people choose on impulse. Everyday data viz looks like conventional charts and typical graphs including, bar charts, pie charts, line charts, and scatter plots. It is often utilized in formal presentations and serves the purpose of providing context and affirming information. The data behind these charts is often simple and of low volume. Everyday charts and graphs support a simple narrative and make clear and distinct points based on established data. 

Much like numerous pie charts we observe and create regularly, Lumina’s opportunity pie chart clearly depicts how people in the United States understand opportunity based on their 2019 National Racial Justice and Equity Framing Survey. This chart makes it clear that the majority, 56% of the participants, agree that “opportunity isn’t equal in the United States. This is followed by 35% who agree that “everyone has equal opportunity in America.” Finally, the remaining 9% agree with neither of the preceding statements. The data behind this chart is simple, separated into three categories. The chart itself is also simple, straightforward, and declarative, making clear and distinct points. 

Conclusion

We often default to everyday data viz when seeking a way to visualize important information. However, it is imperative to identify the visualization that suits the specific data you are working with. The four types 2×2 should serve as a guide, informing the decisions you make about the visualization of your information. These decisions will determine “the skills you’ll call on, the tools you’ll use, and the media you’ll visualize.” Furthermore, asking the necessary questions will ensure that your visualization is successful and effective.

The Significance of Data Visualization

Over thousands of years, humans have developed and perfected the practice of collecting and analyzing data. Over time, we have found that data is a valuable asset that empowers people and organizations to identify problems, gather accurate information, and make informed decisions. Data collection and analysis are highly effective. Even so, data visualization takes things a step further, allowing people to dissect and display complex information in a comprehensible and user-friendly way. 

According to TechTarget, “Data visualization is the practice of translating information into a visual context, such as a map or graph, to make data easier for the human brain to understand and pull insights from. The main goal of data visualization is to make it easier to identify patterns, trends, and outliers in large data sets. The term is often used interchangeably with others, including information graphics, information visualization, and statistical graphics.”

Data visualization is powerful and impactful. Visual representations of data allow people to relay a specific message utilizing numbers and records to support their case. However, data visualization can also misconstrue messages, aid with misinformation, and result in the use of data to tell a desired story as opposed to the truth.

Technological Advances and Data Visualization

With time, humans have discovered and invented more efficient ways to collect, process, analyze, and display data. Today, “software tools for a wide range of visualization methods and data types are available for every desktop computer.” With the impact of the Information Age, people’s practices have generally evolved, resulting in the replacement of physical objects with digital representations. This makes capturing and displaying data much more accessible and cost-effective. 

However, now that data visualization exists within anyone and everyone’s reach, it is much more difficult to track progress within the world of data visualization, especially historical progress. Michael Friendly writes, “it is hard to provide a succinct overview of the most recent developments in data visualization, because they are so varied, have occurred at an accelerated pace, and across a wider range of disciplines. It is also more difficult to highlight the most significant developments, that may be seen as such in a subsequent history focusing on this recent period.”

Furthermore, data visualization, like everything else, requires context that communicates our reality. Without this context, an audience will not receive correct, important, and meaningful information. Although technological advances make it easier to automate data visualization, these advances can also result in missed opportunities in which we are displaying and sharing data that is useless, confusing, or misleading instead of sharing information with its significance and relevance in mind. To incorporate our reality into data visualization, Giorgia Lupi suggests that, when dealing with data, people should prioritize and practice data humanism. With the implementation of technological advances, we begin losing our human touch. According to Lupi, achieving data humanism occurs by embracing complexity, moving beyond standards, incorporating context, and remembering that data is imperfect. She states, “to make data faithfully representative of our human nature, and to make sure they will not mislead us anymore, we need to start designing ways to include empathy, imperfection, and human qualities in how we collect, process, analyze, and display them.”

Though technology has benefited us in many ways, it is imperative to ensure that what we publish and share is genuine, true, and captures the essence of human nature. 

What Makes a “Good Chart”?

A good chart must be easily understandable. Most audiences include individuals from varied backgrounds who have distinct ways of processing information. Therefore, a good chart should be simple, clear, and self-explanatory. Elements that support a comprehensible chart include keys/legends, explicit titles, stated units of measurement, and labels.

From a designer’s standpoint, a good chart must involve the elements and principles of design. Some examples of elements are line, shape, color, texture, type, space, etc. Principles include alignment, balance, hierarchy, emphasis, contrast, movement, proportion, white space, and more. Maintaining the elements and principles of design help with creating charts that are widely understood. 

I respond well to charts and graphs that maintain a distinct use of color balanced with decent whitespace. Additionally, I am easily overwhelmed by graphics that contain a sizable amount of complex information, so I appreciate charts that are simple and easily absorbable. Furthermore, I love work that pushes the boundaries of how data can be visualized, displaying data in unique and engaging ways. I also love charts that include visual elements such as photos, illustrations, and icons. Finally, I value useful, functional, and interactive data visualizations. In recent years, I have become familiar with bullet journaling and mood tracking, which has been very helpful for me. 

Here are some examples of good charts I respond to:

Distinct Color Balanced with Decent Whitespace

^{Source 1, Source 2, Source 3}

Simple and Easily Absorbable

^{Source 1, Source 2, Source 3}

Charts that Exemplify Divergent Thinking and/or Include Visual Elements

^{Source 1, Source 2, Source 3, Source 4: Photos I took of the menu of a restaurant I visited in Orlando last year}

Functional, Useful, and Interactive Data Visualizations

^{Source 1: Functional information to navigate the DMV subway system, Source 2: A quick reference guide for photography, Source 3: Useful information to support color theory in design, Source 4: Informative information about typography for designers, Interactive Data Viz}