[{"content":"This research, commissioned by Threshold and Housing Rights, examines the lived reality of renting across the island of Ireland, aiming to equip decision-makers, advocates and practitioners with robust evidence to support effective action.\nBy combining statistical findings with qualitative testimonies, the research reveals a lived experience of renters, grounded in data and hard evidence.\nWe conducted two large representative public opinion polls of renters in the private rental sector. One of those was in the Republic of Ireland (ROI), and the other was in Northern Ireland (NI), combined with in-depth qualitative interviews to produce a comparative study of renting in ROI and NI.\nTogether, these methods provide a detailed and comparative picture of renting experiences across the two jurisdictions.\nView the PDF in another tab ","permalink":"https://benh7702.github.io/articles/it/housing/","summary":"\u003cp\u003eThis research, commissioned by Threshold and Housing Rights, examines the lived reality of renting across the island of Ireland, aiming to equip decision-makers, advocates and practitioners with robust evidence to support effective action.\u003c/p\u003e\n\u003cp\u003eBy combining statistical findings with qualitative testimonies, the research reveals a lived experience of renters, grounded in data and hard evidence.\u003c/p\u003e\n\u003cp\u003eWe conducted two large representative public opinion polls of renters in the private rental sector. One of those was in the Republic of Ireland (ROI), and the other was in Northern\nIreland (NI), combined with in-depth qualitative interviews to produce a comparative study of renting in ROI and NI.\u003c/p\u003e","title":"Renting on the Island of Ireland"},{"content":"The first time the question was polled in the annual Trinity News TCDSU sabbatical elections opinion poll, students were asked their stance on whether membership of TCDSU should be mandatory or would they prefer if students had a choice to join or not. The poll was conducted between 18 and 21 February, and this specific question received responses from 835 students, indicating a margin of error of plus or minus 3.4%\nCurrently, all students are automatically enrolled in the union, with TCDSU receiving nearly €400,000 in funding from the capitation grant in the 2022/23 academic year, according to the Financial Statements provided by the union. This money is a portion of the Student Levies \u0026amp; Charges fee, which among other expenses, is shared between the four capitated bodies on campus - The Central Societies Committee, Trinity Sport Union, TCDSU, and Trinity publications. In response to being asked if students should have the choice to join TCDSU, 16.8% strongly agreed with the statement, while 19.9% indicated that they agreed. 27.7% of students disagreed with the sentiment that membership should be optional, with 12.9% strongly disagreeing. 22.8% of students neither agreed or disagreed with the statement.\nThe demographics that most supported a choice to join TCDSU were students in Multi faculty courses, Senior Freshman, and male students with 48.8%, 44.7%, and 43.9% in some sort of agreement with the proposition respectively. In contrast, only 21.4% of non binary students, 27.9% of Senior Sophister students, and 32% of Female students are either in full or partial agreement with the question.\nTo introduce an optional membership structure, Section 1.2 of the TCDSU constitution must be amended. A change to this part of the constitution would require either two-thirds of a TCDSU council meeting to vote in favour of a referendum, or a petition signed by 500 students. The referendum would then need to garner support from over 50% of students to be passed. While a combined 37% of respondents indicated a preference for an opt-in system would not be enough to enact such a measure, two findings in the data could make the unions mandate less stable. First, nearly a quarter of respondents were undecided on the hypothetical proposal; in a real referendum, a shift in the sentiment of these voters could swing the outcome. Second, there is widespread dissatisfaction: 42.5% feel unsupported by College and 41.2% feel unrepresented by TCDSU. 80% of those who feel unrepresented by the union would back an optional membership system.\n","permalink":"https://benh7702.github.io/articles/tn/su/","summary":"\u003cp\u003eThe first time the question was polled in the annual Trinity News TCDSU sabbatical elections opinion poll, students were asked their stance on whether membership of TCDSU should be mandatory or would they prefer if students had a choice to join or not. The poll was conducted between 18 and 21 February, and this specific question received responses from 835 students, indicating a margin of error of plus or minus 3.4%\u003c/p\u003e","title":"Students Divided Over Opt-In Union Membership"},{"content":"Around four in ten voters (41%) did not watch the final televised debate of the presidential election, which aired during Prime Time\u0026rsquo;s 9:35pm slot on RTÉ One last Tuesday. According to 40% of voters, Catherine Connolly enjoyed the best performance, compared to 11% for Humphreys and 9% who were unsure.\nStrikingly, every intended Connolly voter that watched the just over hour-long showdown said that she won the debate. Connolly supporters were also the most engaged overall – reporting the highest rates of viewership among all other voting groups. This kind of unwavering and disciplined engagement points to a higher propensity among this group to have turned out on polling day.\nBy contrast, supporters of Heather Humphries were less engaged, with 42% opting not to watch at all. Even among those who did, around a third conceded that they felt Connolly had out performed her. It remains to be seen how much this deflated engagement in the final week of campaigning may have affected turnout amongst otherwise dependable voters on election day. This polling suggest that the pool of persuadable voters is narrowing for Humphreys.\nLikely Gavin voters were nearly as disengaged as Humphreys\u0026rsquo; voters, with little evidence of confidence in the communication abilities of the candidate representing Fianna Fáil\u0026rsquo;s coalition partner. This suggests Humphreys failed to cut through even to voters who are bound to be of similar political ideologies.\nVoters that were likely to have spoiled their ballot were overwhelmingly absent from the viewership, with seven in ten saying they did not watch the debate. Of those that did, the majority broke for Connolly, perhaps indicating the often shared overlap of anti-establishment sentiment in otherwise two very ideologically distinct movements.\nFinally, those who were either unsure how they would vote, if at all, were also unlikely to have viewed the debate. If they did, 90% resonated the most with Connolly\u0026rsquo;s message.\n","permalink":"https://benh7702.github.io/articles/it/debate/","summary":"\u003cp\u003eAround four in ten voters (41%) did not watch the final televised debate of the presidential election, which aired during Prime Time\u0026rsquo;s 9:35pm slot on RTÉ One last Tuesday. According to 40% of voters, Catherine Connolly enjoyed the best performance, compared to 11% for Humphreys and 9% who were unsure.\u003c/p\u003e\n\u003cp\u003eStrikingly, every intended Connolly voter that watched the just over hour-long showdown said that she won the debate. Connolly supporters were also the most engaged overall – reporting the highest rates of viewership among all other voting groups. This kind of unwavering and disciplined engagement points to a higher propensity among this group to have turned out on polling day.\u003c/p\u003e","title":"Of The Small Amount Listening, Not Enough Were Persuaded."},{"content":"A poll conducted by Trinity News indicates that the Social Democrats remain the most popular political party among Trinity students for the fifth consecutive year. Conducted alongside the Trinity News TCDSU sabbatical elections opinion poll between 18 and 21 February, this question received responses from 790 students, indicating a margin of error of plus or minus 3.4%.\nWhen asked how they would vote if a hypothetical general election were held tomorrow, 39% of respondents indicated the Social Democrats as their first preference. This support increases notably to 44.9% among female students and 43% among students in the Arts, Humanities \u0026amp; Social Sciences (AHSS) faculty. Sinn Féin (13.1%) and People Before Profit–Solidarity are practically tied (12.8%) for second place – however both parties continue to see a decline in popularity on campus. Notably, Sinn Féin have lost 40% of their support since 2022.\nFine Gael’s support remains steady at 8.6%, while Fianna Fáil continues to lag their government counterparts with 4.6% support.\nThe Labour Party and the Green Party hold support from 5.3% and 4.8% of students, respectively. According to Trinity News polling, Green Party support among Trinity students has more than halved during the term of the previous government, of which the party served as a junior coalition partner.\n5.3% of respondents indicated they would opt to not vote or spoil their ballot, whereas 5.2% favoured Independent Candidates. Aontú, far-right parties, and Independent Ireland each attracted less than 1% of first preferences in this poll.\nTrinity students’ voting intentions stand in stark contrast with broader trends among young voters in Ireland. According to the Ipsos B\u0026amp;A Exit Poll from last November’s Irish general election, as many as 29% of Irish 18-24 year olds who voted in that election gave Sinn Féin their first preference, in contrast to the 13.1% of Trinity Students who indicated that they would give that party their first preference. Additionally, younger Irish voters were three times more likely as Trinity students to rank Fianna Fáil as their top preference, while on campus the Social Democrats received support at a rate four times higher than that of the typical national cohort of young voters.\nAs of the 2022 census, 14% of 18-24 year olds usually resident in Ireland are ineligible to cast a vote in a General Election, as only those with either Irish or British citizenship have the right to vote. According to the World University Rankings 2025, 39% of Trinity students are international students, so it is likely that a significant proportion of Trinity students are also ineligible to vote in Irish general elections. Therefore, while these polling are a statistically significant capture of the political preferences of Trinity’s student body, they may not directly translate into voting outcomes in national elections.\n","permalink":"https://benh7702.github.io/articles/tn/ge/","summary":"\u003cp\u003eA poll conducted by Trinity News indicates that the Social Democrats remain the most popular political party among Trinity students for the fifth consecutive year. Conducted alongside the Trinity News TCDSU sabbatical elections opinion poll between 18 and 21 February, this question received responses from 790 students, indicating a margin of error of plus or minus 3.4%.\u003c/p\u003e\n\u003cp\u003eWhen asked how they would vote if a hypothetical general election were held tomorrow, 39% of respondents indicated the Social Democrats as their first preference. This support increases notably to 44.9% among female students and 43% among students in the Arts, Humanities \u0026amp; Social Sciences (AHSS) faculty. Sinn Féin (13.1%) and People Before Profit–Solidarity are practically tied (12.8%) for second place – however both parties continue to see a decline in popularity on campus. Notably, Sinn Féin have lost 40% of their support since 2022.\u003c/p\u003e","title":"Poll: Social Democrats maintain huge lead as most popular party among Trinity students"},{"content":"A Trinity News’ snap poll indicates that 54% of voters in the presidential race intend to vote to re-open nominations (RON).\nThe poll was conducted over a 24-hour period from 4pm on Monday, February 24, to 4pm Tuesday, February 25, with a sample size of 453 voters.\nWith most respondents saying they were decided on how they would vote, the margin of error is plus or minus 4.7%. The poll broadly reflects demographic proportions in terms of year group, faculty, and Union involvement, consistent with last week’s poll. Seán Thim O’Leary comes in second with 25.6% of the vote, followed by Patrick Keegan at 11.3%, while Giovanni Li stands at 9%.\nRON leads in all but one demographic captured by this snap poll, with over 50% support across these categories. The highest proportion of RON votes comes from senior sophister (fourth year) students, with 72.7% of that cohort indicating they would give RON their first preference. Arts, Humanities and Social Sciences students follow with 58% intending to vote RON, as do 57.8% of students who have previously held a position in TCDSU.\nIf these results were borne out, it would mark the first time in TCDSU history that a RON outcome is returned in a contested sabbatical officership election. The only RON result in recent memory occurred in the uncontested race for University Times Editor in 2022.\nIn last week’s larger opinion poll, RON had 19.82% of the vote – a figure that was at the time the largest RON vote ever recorded in a contested race in Trinity News polling. After responses that came in before the launch of the RON campaign were excluded, Trinity News’ analysis showed that 30% of students who decided their preferences later in the week gave RON their first preference.\nThe surge in support for RON has come after the launch of a “RON on pres” campaign page on Instagram last Tuesday. At the time of writing, the page has amassed over 440 followers, surpassing any of the official campaign pages of the other candidates in the race.\nThis snap poll was conducted alongside the voting opening on Tuesday, February 25. At the close of voting, 1,320 students had cast their ballots across the six sabbatical races. At the time, 609 students had cast their vote in the arts block, 340 in the business building, 296 in the Hamilton Building, 64 in St James Hospital, and 11 in Trinity Hall.\n","permalink":"https://benh7702.github.io/articles/tn/ron/","summary":"\u003cp\u003eA Trinity News’ snap poll indicates that 54% of voters in the presidential race intend to vote to re-open nominations (RON).\u003c/p\u003e\n\u003cp\u003eThe poll was conducted over a 24-hour period from 4pm on Monday, February 24, to 4pm Tuesday, February 25, with a sample size of 453 voters.\u003c/p\u003e\n\u003cp\u003eWith most respondents saying they were decided on how they would vote, the margin of error is plus or minus 4.7%. The poll broadly reflects demographic proportions in terms of year group, faculty, and Union involvement, consistent with last week’s poll. Seán Thim O’Leary comes in second with 25.6% of the vote, followed by Patrick Keegan at 11.3%, while Giovanni Li stands at 9%.\u003c/p\u003e","title":"Poll projects RON result in TCDSU presidential election"},{"content":"A poll conducted by Trinity News predicts victories for Aoife Ní Bhriain and Charles Hastings in the races for Oifigeach na Gaeilge and Editor of The University Times, respectively. Meanwhile, Seán Thim O’Leary currently leads the presidential race with 36.4% of decided voters giving them their first preference. Buster Whaley holds a narrow six point advantage in the Education race, whereas Channing Kehoe appears destined for a comfortable victory in the Communications \u0026amp; Marketing election. In the fiercely contested welfare \u0026amp; equality and Ents races, John Garvey and Yuv Garg lead with 39% and 35% of the votes, respectively.\nThe poll was conducted among a sample of 867 students between 18 and 21 February, with an estimated margin of error of plus or minus 4.7%. It gathered responses from a representative cross section of intended voters, and included voting likelihood and ranked data, offering an insightful look at the dynamics of the closely contested elections. In single seat races, a candidate must secure at least half of the votes, ensuring no combination of votes allows their opponents to overtake them.\nPresident In the presidential race, Seán Thim O’Leary leads with 36.4% of the vote, followed by Patrick Keegan at 26%. Giovanni Li stands at 17.7% among the decided voters in this category, behind the 19.8% who would prefer to re-open nominations (RON) in this race – the highest RON figure ever recorded by a Trinity News poll in a contested race.\nThe unusually high proportion of respondents in favour of RON follows the launch of an online campaign in the early hours of the second day of polling. Since the launch of the campaign’s Instagram page, the Trinity News poll has seen an increasing number of students deciding to give their highest preference to RON. Excluding responses collected before the campaign launch, our analysis shows that up to 30% of students deciding their preference later in the week opted for RON in this race. Although this statistic carries a higher margin of error of 6% due to the smaller sample size, it points to a possible momentum shift, especially considering 50% of likely voters remain undecided in the presidential race. Strikingly, senior sophister students and those who have held positions in the union, elected to vote to reopen nominations at 23.9% and 25.8% respectively throughout the whole polling period.\nO’Leary leads with a decent majority amongst most demographic groups, with 42.3% support among those who would vote for a centre left party in a hypothetical Irish general election. Keegan is the most popular with students who would like the union to take part in more direct action and left wing voters, garnering 35.5% and 37.1% of the votes from each group, respectively. Li enjoys 47.6% of the health sciences vote, while being most preferred by students identifying with centre right or right wing parties.\nEducation As the race with the least amount of decided voters this election cycle, the outcome of the Education race remains firmly unknown. Buster Whaley holds a small lead over Conchúr Ó Cathasaigh, with each candidate polling at 51% and 45.3% respectively, while 3.7% of voters would prefer to re-open nominations in this race.\nWhaley has a 14 point edge over Ó Cathasaigh amongst Arts, Humanities and Social Science students, whereas Ó Cathasaigh has an 8 point lead amongst STEM students, a 27 point lead with Health Sciences students, and a 12 point lead with those who have held a position in TCDSU. A strong turnout in any of these respective demographics could swing the race in either candidate’s favour – as 50% of voters remain undecided.\nWelfare Alongside this year’s Ents race, the welfare \u0026amp; equality race is the most contested TCDSU election since 2018. John Garvey tops the poll at 39%, followed by Nina Crofts on 26%, Deirdre Leahy follows closely behind at 24.8%, and Bhagarvi Magadi is polling at 7%. A further 3.1% of voters opted to reopen nominations.\nCrofts leads amongst junior fresh and senior fresh students with 40% and 36.23% support from these cohorts respectively. Leahy has the support of 41.5% of health science students, whereas Garvey leads in nearly every other polled demographic, securing 54.1% of the senior sophister vote and 40.7% of students who align with centre left parties. He is also the most transfer friendly candidate, attracting the most transfers from two out of his three competitors, as well as those who initially ranked RON first. Although this highly contested race will most definitely take a few counts to determine a victor, Garvey enters the final week of campaigning in the strongest position to become next year’s welfare \u0026amp; equality officer. That said, with 47% of voters still undecided, a robust showing by any of the candidates in the final week could shift the dynamics of this race.\nComms Channing Kehoe is set to become the next communications \u0026amp; marketing officer of TCDSU, garnering 87.8% of the votes in the only uncontested race this year. As is typical of an uncontested election, the percentage of votes to re-open nominations is slightly higher at 12.2%.\nKehoe has assembled a strong backing across a wide range of demographics, with 91.7% of senior sophister and 90.7% of female voters backing her candidacy. The largest cohort of voters who are opting to re-open nominations are males, but even if 17.3% of this group ultimately vote as such, Kehoes path to victory will still be a very straightforward one.\nEnts In the joint-most contested election this year, Yuv Garg leads the race with 34.6% of respondents giving him their first preference. However, his advantage over Orla Norton is within the margin of error, with Norton polling at 30%. John O’Hara has the support of 21.7% of voters, while Finn Hallwood received 10% of first preferences. 3.6% of voters would prefer to re-open nominations.\nGarg enjoys strong support across various demographics, particularly among multi faculty, health sciences, and students who would vote for a government party in an election. He is also the most preferred candidate among senior sophister students, securing 43.8% of their support. Norton enjoys 35.9% and 34.3% of the centre left and student union vote, respectively. O’Hara has the backing of 36.6% of junior fresh students, making him this year groups most favoured candidate.\nOn further analysis of the second and third preferences, it appears that Norton is the most transfer friendly candidate, securing between 40% and 52% of the available transfers from her competitors. With around 35% of voters still yet to decide who they are voting for in this close race, it is plausible that any of the leading candidates emerge victorious.\nGaeilge In only the second ever election for the role of Oifigeach na Gaeilge sabbatical role, and the first contested one, Aoife Ní Bhriain holds a decisive 16 point lead over her competitor. Ní Bhriain is polling at 56.8%, while Niamh Leddy has secured 40.2% of the first preferences, with 3% of deciding respondents choosing to reopen nominations.\nLeddy would need to convince two thirds of the undecided voters to flip the race in her favour. Ní Bhriain’s lead amongst certain demographics is the strongest seen in this year’s polling – she is on course to win 70% of senior fresh students and 63.9% of the vote from her peers in the STEM faculty. However, her lead amongst senior sophister students and those who have had a role in the student’s union are well within the margin of error – leaving a potential opening in the electorate for Leddy if she can encourage these highly engaged voting groups to give her their preference.\nEditor In what is shaping up to be an uneventful race for editor of the University Times, Charles Hastings leads with 61.4% of the vote, while Sajal Singh has secured 31.1%. 7.5% of the voters in this race would prefer to re-open nominations. Hastings has a particularly strong support base amongst students who would support centre right or right wing parties, male voters, and those who have held a position in the students union. Singh would need to convince nearly three quarters of the undecided votes in this category to support her if she is within a chance of securing the position.\n","permalink":"https://benh7702.github.io/articles/tn/main_poll/","summary":"\u003cp\u003eA poll conducted by Trinity News predicts victories for Aoife Ní Bhriain and Charles Hastings in the races for Oifigeach na Gaeilge and Editor of The University Times, respectively.  Meanwhile, Seán Thim O’Leary currently leads the presidential race with 36.4% of decided voters giving them their first preference. Buster Whaley holds a narrow six point advantage in the Education race, whereas Channing Kehoe appears destined for a comfortable victory in the Communications \u0026amp; Marketing election. In the fiercely contested welfare \u0026amp; equality and Ents races, John Garvey and Yuv Garg lead with 39% and 35% of the votes, respectively.\u003c/p\u003e","title":"Poll: RON campaign sees major upward momentum even as O’Leary leads"},{"content":"Abstract This paper explores the use of copula models for analysing competing financial risks, with a particular focus on modelling the dependence between loan prepayment and default events. Beginning with a theoretical overview of copula functions and their application within a competing risks framework, we evaluate the performance of different copula families through simulation studies. These simulations demonstrate that copula-based models can effectively recover both marginal and dependence parameters. The methodology is then applied to a real-world dataset of over two million loans from the Lending Club. Marginal distributions are selected using AIC criteria, with a Weibull distribution fitted for prepayment times and a three-component Weibull mixture for defaults, due to evident multimodality. The results highlight the value of flexible marginal modelling and copula-based approaches in capturing complex interdependencies between financial risks.\nView the pdf in another tab here PDF ","permalink":"https://benh7702.github.io/projects/copulas/","summary":"\u003ch3 id=\"abstract\"\u003eAbstract\u003c/h3\u003e\n\u003cp\u003eThis paper explores the use of copula models for analysing competing financial risks, with a particular focus on modelling the dependence between loan prepayment and default events. Beginning with a theoretical overview of copula functions and their application within a competing risks framework, we evaluate the performance of different copula families through simulation studies. These simulations demonstrate that copula-based models can effectively recover both marginal and dependence parameters. The methodology is then applied to a real-world dataset of over two million loans from the Lending Club. Marginal distributions are selected using AIC criteria, with a Weibull distribution fitted for prepayment times and a three-component Weibull mixture for defaults, due to evident multimodality. The results highlight the value of flexible marginal modelling and copula-based approaches in capturing complex interdependencies between financial risks.\u003c/p\u003e","title":"Using Copulas to Model Prepayment and Default in Competing Risks Analysis"},{"content":"Abstract Ranked Data arises when judges are asked to order a finite set of items, and it is present in various fields, such as health economics, information retrieval, and election studies. Often, judges only provide partial rankings of the finite set of items, be it by survey design or indifference to certain items after a particular preference level. It is often also true that judges can be subdivided into different clusters based on the both their ranking and social covariates via Mixture Models for ranked data. This project proposes a unified approach at the intersection of both of these realities, namely a Bayesian Mixture of Experts Model for Partially Ranked Data. The proposed model shows encouraging initial applications to election data from the 1997 Irish Presidential Election.\nView the pdf in another tab here PDF ","permalink":"https://benh7702.github.io/projects/capstone/","summary":"\u003ch3 id=\"abstract\"\u003eAbstract\u003c/h3\u003e\n\u003cp\u003eRanked Data arises when judges are asked to order a finite set of items, and it is present in various fields, such as health economics, information retrieval, and election studies. Often, judges only provide partial rankings of the finite set of items, be it by survey design or indifference to certain items after a particular preference level. It is often also true that judges can be subdivided into different clusters based on the both their ranking and social covariates via Mixture Models for ranked data. This project proposes a unified approach at the intersection of both of these realities, namely a Bayesian Mixture of Experts Model for Partially Ranked Data. The proposed model shows encouraging initial applications to election data from the 1997 Irish Presidential Election.\u003c/p\u003e","title":"An Extended Mixture of Experts Model for Rank Data: Insights from Irish Elections"},{"content":"Abstract This report models the incidence of Condorcet paradoxes in elections that use ranked-choice voting. It is known that the occurrence of such paradoxes becomes more likely as the number of voters increases, assuming each candidate is equally preferred by the voters. However, such a scenario is not plausible in the real-world, as some candidates will inevitably be preferred more or less by the electorate than others. In the absence of real data, ballots were generated using a function that samples the Plackett-Luce distribution, which in this context considers the voters propensity to vote for a candidate using weights. Then, an algorithm was run that tests the frequency of Condorcet paradoxes for various amounts of voters and candidates, inferring the probability of its occurrence for various sets of these parameters. The programming language used to generate the ballots and implement the algorithm was R. Interestingly, the probability of a Condorcet paradox tends to zero very quickly for a set of candidates with randomised weights.\nView the pdf in another tab here PDF ","permalink":"https://benh7702.github.io/projects/hamilton/","summary":"\u003ch3 id=\"abstract\"\u003eAbstract\u003c/h3\u003e\n\u003cp\u003eThis report models the incidence of Condorcet paradoxes in elections that use ranked-choice voting. It is known that the occurrence of such paradoxes becomes more likely as the number of voters increases, assuming each candidate is equally preferred by the voters. However, such a scenario is not plausible in the real-world, as some candidates will inevitably be preferred more or less by the electorate than others. In the absence of real data, ballots were generated using a function that samples the Plackett-Luce distribution, which in this context considers the voters propensity to vote for a candidate using weights. Then, an algorithm was run that tests the frequency of Condorcet paradoxes for various amounts of voters and candidates, inferring the probability of its occurrence for various sets of these parameters. The programming language used to generate the ballots and implement the algorithm was R. Interestingly, the probability of a Condorcet paradox tends to zero very quickly for a set of candidates with randomised weights.\u003c/p\u003e","title":"Sampling Plackett-Luce Distributions to Study Condorcet Paradoxes in Ranked-Choice Voting"}]