The two datasets needed to reproduce this article are included in the data
folder.
table1_respondentsRCI
: contains RCI of each party for 9135 respondents surveyed through monthly surveys conducted from January to August 2022. The electoral riding of each respondent is also included.table2_duringCampaign
: contains the predicted vote shares of the parties for each electoral riding during the campaign as found on Qc125.com.
These two datasets are then agregated into a third dataset, table3_aggregatedData
which is generated by the script located at codes/generate_table3.R
.
Each observation of this dataset represents one respondent. It contains 9135 observations and 9 variables.
- respondent_id: unique number identifying respondents across the different surveys.
- month: the month in which the respondent was surveyed.
- riding_name: the name of the respondent's electoral riding.
- riding_id: the id of the respondent's electoral riding. Each riding has its own numbered id.
- rci_CAQ: the respondent's RCI score for the CAQ, which falls within a range of -1 to 1.
- rci_PLQ: the respondent's RCI score for the PLQ, which falls within a range of -1 to 1.
- rci_QS: the respondent's RCI score for the QS, which falls within a range of -1 to 1.
- rci_PQ: the respondent's RCI score for the PQ, which falls within a range of -1 to 1.
- rci_PCQ: the respondent's RCI score for the PCQ, which falls within a range of -1 to 1.
- voteInt_CAQ: vote intent of the respondent for the CAQ. 1 if the respondent indicated the intent to vote for the CAQ, 0 for another party.
- voteInt_PLQ: vote intent of the respondent for the PLQ. 1 if the respondent indicated the intent to vote for the PLQ, 0 for another party.
- voteInt_QS: vote intent of the respondent for QS. 1 if the respondent indicated the intent to vote for the QS, 0 for another party.
- voteInt_PQ: vote intent of the respondent for the PQ. 1 if the respondent indicated the intent to vote for the PQ, 0 for another party.
- voteInt_PCQ: vote intent of the respondent for the PCQ. 1 if the respondent indicated the intent to vote for the PCQ, 0 for another party.
Each observation of this dataset represents a single day's Qc125.com prediction of a riding's vote share of a political party. It contains 52388 observations and 6 variables.
- date: the date of the observation.
- riding_name: the name of the respondent's electoral riding.
- riding_id: the id of the respondent's electoral riding. Each riding has its own numbered id.
- party: the political party associated to this prediction. Options are CAQ, PLQ, QS, PQ and PCQ.
- pred: the predicted vote share of the party.
- moes: the margin of error of this prediction (as reported by Qc125.com).
This dataset is generated by the file located at codes/generate_table3.R
. Each observation of this dataset represents a riding with its vote fragility index (before the campaign) and its campaign volatility (during the campaign). It contains 125 observations and 4 variables.
- riding_id: the numerical id of the riding.
- riding_name: the name of the riding.
- n_riding: number of respondents in the riding.
- fragility_index: vote fragility of the riding before the campaign generated from survey data. From 0 - not fragile to 1 - fragile.
- fragility_index_mrp: vote fragility of the riding before the campaign generated by MRP (in the
mrp
folder). From 0 - not fragile to 1 - fragile. - volatility: campaign volatility of the riding generated from Qc125.com data. From 0 - not volatile to 1 - volatile.