COVID-19 Speech Sentiment – Crosstalk Dashboard

1. Monthly Sentiment Patterns

Lines show the average sentiment of COVID-19 speeches each month. Hover to see the number of speeches, dominant topic, and distinctive words. Vertical lines mark key moments in the pandemic.

2. Overall Tone by Country

Bars show average sentiment across all speeches (2020–2022). Thin lines indicate ±1 standard error, giving a sense of uncertainty around each mean.

3. How Sentiment Shifted Over Years

Lines show each country's average sentiment by year. Hover to see the change compared to the previous year.

4. Stability vs. Volatility of Rhetoric

Boxplots show distributions of monthly sentiment by country. Use the year filter to focus on a particular phase of the pandemic. The table below reports overall month-to-month variability (standard deviation).


5. Top Distinctive Words in COVID Speeches

For each country, these words have the highest TF-IDF scores, meaning they are especially distinctive in that country's speeches compared to others.

6. Speech-Level Table

This table is linked to the filters above for country and month. Download the filtered subset to quote specific speeches in your paper.

U.S.-Specific Case Study – COVID-19 Sentiment Timeline

This section focuses specifically on the United States, using a separate dataset of major national speeches and announcements from 2020–2022.

U.S. Daily Sentiment Timeline (Interactive)

Use the dropdowns on the left to filter by month, theme, or sentiment bucket.
To reset the view, either click the “All” button in the range selector or double-click inside the plot area.