Many people don't realise but referendums usually produce significant differences in opinion between men and women.
The pre-referendum polling and post-vote analysis from the Scottish independence referendum showed that men were disproportionately more for 'Yes' (for independence) and it was the women who voted mostly for 'No' to keep Scotland in the UK.
Likewise, if we look at the UK-wide polling regarding this June's EU referendum this pattern is repeating and our recent Northern Ireland-wide BIG100 EU polling also shows this trend.
Our 'BIG100' poll ran online over a 100 hour period from the February 8 to 12 and achieved a huge 3,023 full responses.
After data auditing - to determine only the genuine 'one-person, one-vote' responses - we considered 2,886 for the final results and analysis.
So in terms of this issue we show our BIG100 results above: Graph one shows how men voted in our BIG100 EU referendum question poll, and graph two how women voted.
We can see, only 52% of men are currently intending to vote 'Remain' compared to 64% of women.
NB the overall result was 55% Remain. Plus one in three men are intending to vote 'Leave' (33%) compared to only one in five women (20%).
These results are not surprising and mirror what is happening in polls throughout the UK.
The results also reflect the usual male-female difference of opinion when it comes to binary decision making i.e. Yes/No type decisions like the Scottish independence and EU referendums.
In terms of these binary decisions, men tend to be bigger risk takers, with females wanting to analyse the detail, and have a good long think, before coming to their decision.
That is also probably the reason why most of the current undecideds in terms of the EU referendum are women. Of course in terms of the EU referendum it's debatable what is the so-called riskier decision.
This pattern is typical in other areas of life as well.
For example most entrepreneurs tend to be men, but then again most of the bankruptcies tend to be businesses run by men.
Males tend to make quicker decisions in terms of say a house move, with the husband's wife wanting to do a bit more research, and/or have a further look around other properties.
So if the 'Leave' campaign want to win this EU referendum they are going to have to get the UK-wide women 'Remain' vote down to somewhere approaching 50/50.
This is the best realistic target for the 'Leave' campaign, as it's pretty unlikely at this stage to see a majority of women voting 'Leave'. In the end it was women that kept Scotland in the UK, and it may well end up that it's the British females that keep the UK in the EU on June 23.
Of course, this is the sort of demographic poll result that can spark discussion and a great excuse for an argument.
Polling was carried out by Belfast based polling and market research company LucidTalk. The project was carried out online for a period of 100Hours from 4pm Monday February 8 to 8pm Friday February 12 2016. The project used a combination of participants from the established LucidTalk Opinion Panel (1,200 members) which is balanced by gender, age-group, area of residence, and community background, in order to be demographically representative of Northern Ireland. In addition, a selection of respondents were either invited or volunteered to participate. In total 3,023 completed responses were received and after data auditing 2,886 completed responses were included in the data and results analysis. The data auditing process was carried out to ensure all completed surveys were genuine 'one-person, one-vote' responses.
All data results have been weighted by gender and community background to reflect the demographic composition of Northern Ireland. Because the sample is partly based on those who initially self-selected for participation rather than a probability sample, sampling error can be higher than a standard targeted sample poll. All data results produced are accurate to a margin of error of +/-4.0%, at 95% confidence. In addition to sampling error, question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of public opinion polls. In addition, all surveys and polls may be subject to sources of error, including, but not limited to sampling error, coverage error, and measurement error. All reported margins of sampling error will include the computed design effects for weighting.
LucidTalk is a member of all recognised professional Polling and Market Research organisations, including the UK Market Research Society (UK-MRS), the British Polling Council (BPC), and ESOMAR (European Society of Market Research organisations). The BPC are the primary UK professional body ensuring professional Polling and Market Research standards. All polling, research, sampling, methodologies used, market research projects and results and reports production are, and have been, carried out to the professional standards laid down by the BPC.