The federal election is just a few weeks away and measure that the campaign continues, it is likely to be reading more about the surveys.
But what should you do with these surveys that predict, among other things, what party could the government and the feeling of voters form?
SBS News spoke with Simon Welsh, director of research and reputation for PollSter Redbridge Group, and Dr. Kevin Bonham, an independent election and survey analyst, about what voters should take into account.
What is the survey?
Bonham said that political opinion polls “tries to present a snapshot of what intention to voting is at the time it is taken” and is “a fairly high level vision of the voting intention contest.”
Political surveys can be presented in different ways and can use quantitative or qualitative data. Quantitative is numbers based numbers, such as “45 percent of Australians will vote [X]”Qualitative, on the other hand, is an analysis of trends that cannot be captured in clear numbers, such as” Australians are increasingly concerned about [X]”
Both are under the umbrella of political surveys and, although quantitative surveys tend to capture the headlines of the news, both types are used in politics.
“Quantitative research is often used to make decisions on campaigns and allocation of campaign resources,” Welsh said.
“Qualitative research is really used to understand voters and their priorities and their drivers deeply, and that helps to inform the messages that the parties tend to take to the campaigns.”
Surveys can be done in multiple ways such as interviews in person, telephone interviews (with a real person), ‘robopolling’ (which is usually a computer call) or online survey panels.
Australia has been more reliable in the survey, but that is changing
Compared to countries such as the United States, Australia’s mandatory voting system has historically meant that surveys were more reliable.
In a voluntary system like the United States, pollsters have to try to analyze the voting intentions of the population that probably vote.
In a mandatory voting system such as Australia, they could trust electoral data and reduce variables.
“What we have seen over the years is that the Australian electorate becomes much more fragmented,” Welsh said. “[It has] dramatically changed and accelerates in this choice. ”
In Redbridge’s research, in all of Australia there are changes particularly in the increase in generational differences, an increase in ethnic/cultural diversity, differences between regional/rural/urban voters and a change in the demography of external electoral seats.
“[20 years ago]You could run those surveys, what we call horsepower surveys, where you had work versus liberal and would get a preferred type of projection in two parts. “
“They no longer mean much due to the fragmentation of communities.”
In the survey, the preferred in two parts is a measure that shows how the votes between work and coalition would be distributed after the preferences of the minor parties are distributed.
The way in which pollsters are collecting data is also changing
These federal elections, millennials and voters of Gen-Z will overcome the Baby Boomers for the first time.
Voting groups normally collect data through telephone calls. But some surveys have suggested that younger people, particularly Gen-Z, do not like to talk on the phone.
The pollsters are aware that this risk the phone collected data of surveys that are more and, in general, more conservative.
This has led to online surveys as a more reliable and complete resource, especially with limited resources, but the increasingly common models for surveys are still in question.
Multilevel regression with prostrateification, or MRP, is a newer method for surveys. This takes a group of responses and, analyzing the demography of respondents, apply these patterns to help predict similar areas.
Bonham feels curious about the use of MRP and says: “They are not based on samples of the seats themselves and are a more complicated model.
“This is still an experiment in Australia, but it is interesting to watch and see how it is going more popular because for several reasons, including the individual seat survey per seat, it has gone badly for quite some time.”
While Bonham is curious, Welsh said that “the MRP model is probably where the future of surveys is going to lie.”
While the “human effort” is necessary to precisely interpret the data, he added that “the seat by seat survey has become so difficult because it really trusts some form of vote based on the phone”, which can be difficult, particularly among the younger voters.
“But understand that MRP is a predictive modeling and not a direct investigation with voters,” said Bonham.
What should voters take into account when reading about the surveys?
When reading the surveys data, both Bonham and Welsh want voters to pay attention to two things: methodology (the way in which the data is collected) and where the data was collected.
The methodology, such as telephone calls, can only mean certain demographic data or models such as MRP means that some seats could be predicted with little or no direct participation of voters.
Welsh wants voters to consider where the survey was conducted, warning that a broad national sample could lose important local, social and community differences.
Bonham also encourages people to consider who conducted the survey.
“Although [many] The national surveys are asked using fair questions, there are many surveys that revolve, often at the local level or by less known pollsters, who use suspicious questions and are badly informed in the media, “said Bonham.
This may include things like ‘‘, which is a tactic to persuade voters of a certain candidate through a biased interrogation style and that Wales calls “a concern” for the electoral industry.
Other experts have said that they analyze the margin of error: the range that the survey results may vary due to chance. For example, if a survey indicates that the work leads the coalition of 51 to 49 percent in a preferred two parts base with a margin of error of more or less 3 percent, the real result could be as close as 48 percent work of the coalition of 52 % or as wide as 54 percent V 46 percent of coalition.
If these details are not included in a media report, a methodology declaration may be available on the survey group website.
Ultimately, Bonham encourages voters to “read many different surveys, [and] Do not take any survey as a gospel. “
He and Welsh also recognize that voters could have to use a healthy level of skepticism with media comments, personally verifying if the information comes from voting experts or possible hidden motifs.
Bonham said “[polling] You can turn one side or another, just to get sales. “
For the latest SBS News, and .
.jpg?w=1200&resize=1200,0&ssl=1 "Industrial plays weather challenges")
