R: Ecological inference data sets of the 2002 New Zealand...

ei_NZ_2002 {ei.Datasets}

R Documentation

Ecological inference data sets of the 2002 New Zealand General Election.

Description

This tibble contains 69 data sets corresponding to the 2002 New Zealand General Election. Each data set includes party and candidate vote results by voting unit as well as their associate cross-distributions (for votes and percentages) at the district (electorate) level.

Usage

data(ei_NZ_2002)

Format

A tibble containing 69 observations and 6 variables:

Number_of_district: Number assigned to the district/electorate by the New Zealand Electoral Commission.
District: Name of the district/electorate.
Votes_to_parties: A tibble for each electorate/district with the party votes recorded in each voting unit of the district.
Votes_to_candidates: A tibble for each electorate/district with the candidate votes recorded in each voting unit of the district.
District_cross_votes: A tibble for each electorate/district with the parties-candidates cross-distribution of votes in the entire electorate/district.
District_cross_percentages: A tibble for each electorate/district, with the parties to candidates voter transition probabilities (in percentages) in the entire electorate/district.

Details

Description of the Votes_to_parties, Votes_to_candidates, District_cross_votes and District_cross_percentages variables in more detail, where N(i), R(i) and C(i) denote, respectively, the number of voting units, party voting options and candidate voting options in district i:

Votes_to_parties: A list of 69 tibbles/data.frames, with each data.frame containing N(i) observations and 2+R(i) variables. The two first variables, City and Address inform, respectively, about the place in the district where the voting unit is located and the voting unit address. The rest of the columns correspond to the votes gained by the different party voting options competing in the district. The orders of the voting units in Votes_to_parties and Votes_to_candidates coincide.
Votes_to_candidates: A list of 69 tibbles/data.frames, with each data.frame containing N(i) observations and 2+C(i) variables. The two first variables, City and Address inform, respectively, about the place in the district where the voting unit is located and the voting unit address. The rest of the columns correspond to the votes gained by the different candidate voting options competing in the district. The orders of the voting units in Votes_to_candidates and Votes_to_parties coincide.
District_cross_votes: A list of 69 tibbles/data.frames, with each data.frame containing R(i) rows and 1+C(i) columns (variables). The first variable, which is labelled after the name of the district, contains the names of the parties in the same order than in corresponding Votes_to_parties tibble, the rest of the variables (columns), ordered as in the corresponding Votes_to_candidates tibble, are labelled as the candidate voting options.
District_cross_percentages: A list of 69 tibbles/data.frames, with each data.frame containing R(i) rows and 1+C(i) columns (variables). The first variable, which is labelled after the name of the district, contains the names of the parties in the same order than in corresponding Votes_to_parties tibble, the rest of the variables (columns), ordered as in the corresponding Votes_to_candidates tibble, are labelled as the candidate voting options.

Note

The New Zealand Electoral Commission had no involvement in preparing these data sets. The raw data has been pre-processed in order to guarantee their straighforward usefulness in ecological inference procedures. Some small discrepancies exist among the figures in District_cross_percentages and District_cross_votes. The percentages are a direct translation of the published data, whereas the vote counts have been adjusted using integer linear programming to make them congruents with the figures in Votes_to_parties and Votes_to_candidates. More details in Pavia (2021). For the official results, visit https://www.electionresults.govt.nz.

Author(s)

Jose M. Pavia, pavia@uv.es

Source

Own elaboration from data available in https://www.electionresults.govt.nz, retrieved 19 January 2019.

References

Pavia, JM (2021). ei.Datasets: Real datasets for assessing ecological inference algorithms, Social Science Computer Review, forthcoming.