Design and Critique

(link) What Do Men Think It Means To Be A Man?
Source Visualization Critique:
Our source visualizations are from a FiveThirtyEight article titled “What Do Men Think It Means To Be A Man?” by Ella Koeze and Anna Maria Barry-Jester. This set of visualizations pulled its data from a survey of 1,615 adult men conducted by FiveThirtyEight and WNYC Studios in May 2018. Below are three of the six visualizations included in this article. These visualizations were intended to accompany an article’s paragraphs of both written statistics and direct quotes collected from the 2018 survey. Thus, instead of creating an engaging, complex visualization that takes time to explore, the designer’s goal was to give readers quick, concise takeaways as they read through blocks of text. They only utilized static horizontal aligned bar graphs, only a handful of contrasting colors, and keeps surrounding information to a minimum. Only one of the six visualizations is segmented by age, which adds one extra dimension of information. As a result, insights can be easily pulled by readers of all levels of visualization expertise.

Design Rationale
Complete set of survey questions: While exploring the source dataset, we discovered that there was much more useful data beyond what the FiveThirtyEight article captured. While the authors only visualized answers to six questions, they had asked a total 25 usable questions in their survey—thus, we decided to create a dynamic bar chart that changes data based on the selected survey question, which can be browsed through via two arrow buttons.

The left axis of the bar chart displays the range of responses to each question, and the bottom axis represents the percentage of each response. We decided to make this a horizontal bar chart due to the lengthy left axis labels.

Categorized data: During our data exploration, we also noticed that the authors’ survey questions were already categorized into four broad categories: masculinity, lifestyle, work, and relationships. To ease the process of cycling through questions, we decided to allow users to filter the data down to a specific category using playful icons. Each time an icon is clicked, users have access to a new set of questions. Categorized data: During our data exploration, we also noticed that the authors’ survey questions were already categorized into four broad categories: masculinity, lifestyle, work, and relationships. To ease the process of cycling through questions, we decided to allow users to filter the data down to a specific category using playful icons. Each time an icon is clicked, users have access to a new set of questions.

We ran into a challenge with this second chart due to the way that demographic data was collected. Ideally, we would have wanted each bar to represent an answer choice (similar to our first bar chart), and each bar would be comprised of stacked bars representing each demographic group (ex. ages 18-34, ages 35-64, and ages 65+ all stacked into a singular bar). However, the data was not structured this way: demographic buckets did not sum to 100% for a single answer choice, only multiple answer choices for a single bucket (ex. age 18-34) summed to 100%. As a result, our left axis displays all of the possible demographic buckets, and the stacked bars correspond to answer choices.

Another decision made during this process was to remove the demographic breakdown view on the right for questions that were originally presented as multi-select (select all that apply). As a result, the breakdown bars would’ve summed up to more than 100%. Another way to iterate on this project in the future would be to provide alternate views to the stacked bar chart that provide the user with different views of the data.

To facilitate comprehension of demographic breakdowns, we added hover interactivity within our visualization: when a user hovers over a specific answer choice on the left bar chart, the corresponding stacked bar (on the right) is emphasized so that a user can see how much of each demographic group answered in a specific way. [Example: A user is interested in the percentage of gay/bisexual respondents that answered “very masculine” to the question “In general, how masculine or ‘manly’ do you feel?”. They hover over the “very masculine” bar on the left chart, and can look to the right chart to see that 20% of gay/bisexual respondents answered “very masculine.”]

Podcast audios: The authors also embedded WNYC’s 40-minute “Death Sex & Money” podcast within their article, in which men were interviewed on their thoughts on masculinity. We thought that this podcast provided valuable context to the quantitative data, so we decided to insert audio clips from the podcast for each of the four interview categories. These audios change based on the selected category.