Member-only story
Focusing on the polarization: Designing effective charts for the Likert scale
#26: 100% Stacked Bars or Diverging Stacked Bars
When it comes to visualizing the Likert scale, we have quite a few options. We can use 100% stacked bars, a panel bar chart (or its grouped version), or some variant of the diverging stacked bars depending on the placement of neutral answers. Each one of those charts focuses on slightly different aspects of the data, but the diverging stacked bar with the most antagonist values in the center is the best. The short answer is that they are the most insightful, empower quick data scan, introduce flexibility with placing the neutral responses, and show both total agree/disagree shares and degree of polarization. For a more comprehensive explanation, I recommend Daniel Zvinca’s article. He thoroughly analyzed the diverging stacked bars usage, and I agree with every word there.
This time I redesigned another PEW Research Center chart showing the survey result conducted among adult Americans. The chats show the distributions of respondents who agree and disagree with two statements about how Chinese and American social media companies handle users’ personal data. The first is compliance with privacy policies, and the second is using users’ data in a way that is comfortable for them. I like that both questions are…
