A Quantitative Ranking of Turbulent Years

Introduction — by the Teacher

History.com has asked, "Was 1968 America's Bloodiest Year in Politics?" This year's APUSH class tried to answer that question. More generally, students tried to determine, quantitatively, what years within 20th-century U.S. history were the most turbulent. It turns out that 1968 came in first, followed by 1919 and 1970. The data and an explanation of the statistical methods used to reach these conclusions appear below, along with a variety of data visualizations illustrating different representations of the turbulence.

Definition of "Turbulence" — by the Students

Before beginning our research, we first established a definition of the term "turbulence." We thought this would help all students working independently make consistent decisions about what to include in our study. We set the parameters by looking at web articles dealing with activity such as group violence and civil unrest. Based on the work of scholars at the EMBERS project at Virginia Tech, for example, we chose to exclude events motivated by personal gain and criminal activity carried out by gangs. This decision left us with civil disturbances that, according to the National Fire Prevention Association, "typically evolve from a group of people protesting against major sociopolitical issues" and include events like riots, acts ofviolence, terrorism, insurrections, assemblages, and unlawful obstructions such as widespread labor strikes. As a class, we chose to label these events as either social, political, or economic, based on the largest motivator for a group's collective action. Other than collective actions, we added notable assassinations, reasoning that the assassins were likely motivated by some type of group identification and because the murder would be considered a violent uprising against the government.

Beyond this, we thought it was also important to include nonviolent events that symbolized significant group discontent with social, political, or economic conditions. For these nonviolent events, we asked ourselves if the event had immediate or eventual nationwide significance, if the event likely involved more than 10,000 individuals, and if it was directed against state or federal government policy. While we struggled with the inclusion of nonviolent protests due to the fact that the lack of violence does not fit into the traditional definition of turbulence, when considering landmark nonviolent events such as the 1963 March on Washington for Jobs and Freedom, we concluded that they contributed to the overall character of their time and in some cases likely contributed to subsequent turbulence.

Data visualizations in this section: Bar Chart Race by Skylar; Sliding Graph by Tess.

About Measuring Turbulence — by the Teacher

To measure turbulence students searched for events within newspapers.com and recorded into a spreadsheet the amount of coverage each event received — "occurrences" in the spreadsheet. Because the number of newspapers for each year varies, students created a multiplier in order to equalize the opportunity for coverage. Regardless of the year, therefore, each event had equal opportunity for coverage because total occurrences were increased by the multiplier. Once the multiplier was applied each year and total coverage was determined, the total was then multiplied by the number of events for that year. This can can be seen on the spreadsheet, as can the terms used for searching. Whenever possible, search terms came from event titles within Stanley Kutler's Dictionary of American History. Otherwise, students used their best judgement. Search results were limited to a one-year period following each event.

More Data Visualizations — by the Students

The Number of Newspapers within Newspapers.com — When determining the turbulence for each year, due to the difference in the number of newspapers per year we had to create multipliers for each year to make the values equal. To understand the multiplier for each year, this graph precisely depicts the number of newspapers within newspapers.com for each year. Over the 20th century, the total number of newspapers decreased greatly. Some of the multipliers were extremely different as the number of newspapers decreased significantly, seen in years like 1922 as well as 1976. — by Ellie

Hierarchy Chart of Social, Political, and Economic Events — With the data that the class collected, I created a “hierarchy chart” to demonstrate a more creative way of story telling with data. I wanted to be able present a more complex representation of the data so I chose to add more components, including the three types of events and the specific events in those categories. Once I got the total points in each year, I wanted to show which specific events were found most in newspapers. I used three class periods to finish this and I enjoyed learning a new way of story telling with data on Flourish. — by Maria

Turbulence Map — To find out which states were the most and least turbulent throughout the 20th century I used the class-created spreadsheet and placed each event in the state in which the event happened. I then assigned each event the value that was calculated in the class spreadsheet. I used all of the values to form a final value for each state, which were used to create a ranking. I concluded that Oklahoma was the most turbulent state throughout the 20th century followed by California and New York. Alaska, Idaho, Iowa, Maine, Montana, New Hampshire, North Dakota, Rhode Island, Utah, Vermont, and Wyoming all tied for the least turbulent of the fifty states. — by Sean

The Students

"A Quantitative Ranking of Turbulent Years" was produced by the 2018-2019 AP U.S. History class:

Zach Belisle
Maddie Blewitt
Jordan Chappell
Quinn Fair
Ellie Frederick
Maddi Giberson
Lilliana Gurney
Skylar Hathorn
Sean Moynihan
Lizzy Nichols
Tess Palin
Maria Villarreal