This second weekly snapshot tracked tweets from Wednesday, May 11, 2022 at 12:00 p.m. ET to Wednesday, May 18, 2022 at 11:59 a.m. ET.

 

This Snapshot was generated on May 19, 2022. The data presented is accurate at the time of posting, however as the machine learning model is updated, the toxicity measurements presented here may become outdated. Our Ontario Election Snapshot provides analysis that is consistent across the election period.

SAMbot uses a natural language processing machine learning model. This means that SAMbot is trained and tested on millions of data points so it can understand whether tweets can be considered toxic, harmful, or insulting. Machine learning algorithms are regularly updated as they learn and have access to improved data. Updates to the third-party algorithms SAMbot uses to measure toxicity can lead to slight discrepancies in toxicity measurements across different Snapshots. Our Ontario Election Snapshot analyzes all of the data collected throughout the election period using the same model in order to have consistency in measurement.

146,980 tweets analyzed

From 486 candidates and 14 official party Twitter accounts.

toxic icon

17,086 tweets identified as likely toxic*

There was a 3.1% decrease in toxicity this snapshot period compared to last period.

*Learn more about how SAMbot detects toxicity.

Key Findings

  • The second and final leaders debate was held on Monday, May 16, 2022 at 6:30pm. While SAMbot recorded a high volume of tweets during and after the debate, less than 7% of all tweets between 12:00 pm ET and 11:59 pm ET were toxic. Discussions on Twitter surrounding the leaders debate were significantly more positive or neutral in sentiment compared to the rest of the period.
  • Ontario PC Party leader Doug Ford received the most toxicity again this week, receiving over twice as many toxic tweets than Liberal Party of Ontario leader Steven Del Duca, who received the second highest volume of toxic tweets.
  • Party leaders Andrea Horwath of the Ontario NDP and Mike Schreiner of the Green Party of Ontario received an increase in total tweets this week, but both candidates didn’t see an increase in the rate of toxicity received compared to last week.
Bar graph

Toxicity Tracker

Of the 146,980 tweets analyzed:**

  • 2,391 were labelled as containing sexually explicit content
  • 4,967 were labelled as containing identity attacks
  • 6,687 were labelled as containing profanity
  • 6,784 were labelled as containing threats
  • 16,790 were labelled as containing insults

Top 10 candidates that received the most toxicity (by volume)

Doug Ford***

Leader of the PC Party of Ontario, seeking re-election in Etobicoke North, current Premier of Ontario

Received 5,760 likely toxic comments

Steven Del Duca

Leader of the Ontario Liberal Party, candidate for Vaughan—Woodbridge

Received 2,495 likely toxic comments

Andrea Horwath***

Leader of the Ontario NDP, seeking re-election in Hamilton Centre

Received 1,730 likely toxic comments

Steven Lecce***

PC Party of Ontario candidate seeking re-election in King—Vaughan, current Ontario Minister of Education

Received 1,262 likely toxic comments


Lisa MacLeod***

PC Party of Ontario candidate seeking re-election in Nepean

Received 328 likely toxic comments

Tyler Watt

Ontario Liberal Party candidate for Nepean

Received 205 likely toxic comments

Mitzie Hunter***

Ontario Liberal Party candidate seeking re-election in Scarborough—Guildwood

Received 203 likely toxic comments

Joel Harden***

Ontario NDP candidate seeking re-election in Ottawa Centre

Received 202 likely toxic comments


Mike Schreiner***

Leader of the Green Party of Ontario, seeking re-election in Guelph

Received 158 likely toxic comments

Monte McNaughton***

PC Party of Ontario candidate seeking re-election in Lambton—Kent—Middlesex, current Ontario Minister of Labour, Training and Skills Development

Received 153 likely toxic comments

SAMBOT-head-illustration

***was a Member of Provincial Parliament during the 42nd Parliament

report-icon

News from the campaign trail

  • There continues to be significant conversation on Twitter surrounding the race in the Nepean riding. Both Ontario PC Party candidate Lisa MacLeod and Liberal Party of Ontario Candidate Tyler Watt remain on SAMbot’s “top ten candidates” list for the second week in a row.
  • Mitzie Hunter, Liberal candidate for Scarborough—Guildwood and former Liberal Minister of Education, spurred online discussion this week after her announcement on Sunday, May 15, 2022. She criticized the PC Party of Ontario and its leader, Doug Ford, for what she referred to as “the Conservative gravy train at Queen’s Park.” She revealed 16 of the 41 PC Party of Ontario candidates defeated in the 2018 Ontario election were given roles within the Ontario public service shortly after losing their respective elections.
    • Hunter has received a significantly higher amount of identity attacks compared to the candidate average this week. Based on data collected thus far, approximately 3.4% of all tweets directed at tracked candidates are identity attacks, whereas for Hunter, 8.1% of tweets directed at her this week were identity attacks.
  • Green Party of Ontario leader Mike Schreiner saw an increase in tweets this week on the night of the leaders debate. Despite receiving more than 2.5 times the amount of tweets this week compared to last, Schreiner received a lower rate of toxicity; 8.9% of all tweets received by Schreiner last week were likely toxic, while this week only 4.2% of all tweets this week were likely toxic.

Due to technical difficulties that may have affected the accuracy of our analysis, we have removed findings from this snapshot related to political parties. You can view the full election report with updated political party data in the Ontario Election Snapshot.

Unless otherwise stated, tweets sent from Wednesday, May 11, 2022 at 12:00 p.m. ET to Wednesday, May 18, 2022 at 11:59 a.m. ET.

SAMbot is tracking 486 candidates and 14 official party accounts. Review the full list.

For more detailed methodology, visit our About SAMbot page.