Talk:1990 Slovenian independence referendum

Statistics

Copied from my talk page. --Eleassar ^{my talk} 13:38, 29 December 2011 (UTC)

Hello Eleassar. I have reverted this revert you made on the independenc referendum article due to your edit summary. I did not remove any information from the article except the prose about the results, and this was removed because I added in the full results in a table. Number 57 11:37, 29 December 2011 (UTC)

The data that you added are completely in opposition to STAT.SI (Slovenian Statistical Office) and other sites. --Eleassar ^{my talk} 11:56, 29 December 2011 (UTC)

The figures in the book I am using are actually from the Slovenian Yearbook and are almost all identical (Yes: 1,289,369, no: 57,800, blank/invalid: 12,412, registered voters: 1,496,860). The only figure that is different is total votes cast - the book has 1,359,581, whereas stat.si has 1,361,738. This is because stat.si include the 2,157 unused, returned ballots, which are not usually recorded in voting totals because they were not cast. Can you please reinstate the results table. Thanks, Number 57 11:59, 29 December 2011 (UTC)

The turnout was 93.2%, not 90.8%. 94.6% voted in favour of independence, not 96%. The 2,157 votes can't account for such a large difference. Please, use the STAT.SI data, as it was provided to them by the Slovenian National Electoral Commission, so it is official. --Eleassar ^{my talk} 12:07, 29 December 2011 (UTC)

Regarding the % voting in favour, the book has a footnote under the table that states "the officially released figure of the share of 'Yes' votes of entitled voters relies on a figure of 1,457,020 since 42,274 were not able to take part due to military service of work abroad". Thus the % figure they provide is not an actual percentage of votes. Both sources state there were 1,289,369 yes votes and 57,800 no votes. This totals 1,347,169. 1,289,369/1,347,169 is 95.7%, which is the % "Yes" votes given in the book. However, I cannot work out how one could possibly reach 94.6% for "Yes" votes. I assumed it might have been by including the invalid votes, but this gives 94.8% not 94.6%. Could you explain how you calculated this?

I suspect the turnout figure of 93.2 was also reached by removing the missing 42,274 non voters. However, it does not actually appear in the reference cited, so I assume you calculated it yourself using the figures on the bottom row? I'm happy to add a footnote to the table explaining the different possible figures. Number 57 12:22, 29 December 2011 (UTC)

I have reworked the table on the article to include the 2,157 returned ballots and the 42,274 non-voters, and have tweaked the wording to make it clear that the 96% of those in favour was those casting valid ballots ("Almost 96% of those casting valid ballots had voted in favour of independence, whilst at over 90%, turnout had far exceeded the 50% threshold."). I hope this is satisfactory? Number 57 12:45, 29 December 2011 (UTC)

...and the 2,434 people who voted by certificate despite not being registered... Number 57 12:52, 29 December 2011 (UTC)

I can't find any online source claiming 95.7% of support and it also doesn't make sense, as the media would then almost surely have reported the rounded 96% of support. The media have reported that 95% of voters supported the independence.[1][2][3][4][5].

To correctly calculate the % in favour, one has to count the invalid ballot papers (which differs from returned - not used). This makes then for 94.8%, reported by the political scientist Matej Tonin.[6] Slovenian Press Agency reported "almost 95%".[7]

The 93.2% turnout was reported by several publications, but taking into consideration that some voters could not vote, 1,361,738 / 1,457,020 gives 93.4%, which was also reported, for example in Delo [8] and the Archives of Slovenia,[9] and elsewhere,[10][11] so probably this is the number that should be given priority. The prominent historian Božo Repe reported 93.2% turnout (and 88.2% in support), so this can also be mentioned in the article.[12] --Eleassar ^{my talk} 13:13, 29 December 2011 (UTC)

When calculating the % in favour or against, blank or invalid votes are not typically included (see United Kingdom Alternative Vote referendum, 2011 as a typical example). There is at least one example of RTVS using the 96% figure. Anyway, as long as the table of results remains in the article, I'm happy. Number 57 13:25, 29 December 2011 (UTC)

Since you have tagged the article, I have requested a WP:Third opinion. Number 57 13:33, 29 December 2011 (UTC)

My opinion is that we should follow WP:WEIGHT. --Eleassar ^{my talk} 13:45, 29 December 2011 (UTC)

Indeed, and weight is heavily in favour of using yes and no %s that refer only to valid votes, as one can see on Election Guide. Out of interest, could you provide an example of how you would format the results table below so we can compare? Number 57 13:54, 29 December 2011 (UTC)

Hereby, one must take into consideration that in Slovenia in 1990, the invalid ballots were counted as votes against.[13] --Eleassar ^{my talk} 14:21, 29 December 2011 (UTC)

This source makes no mention of 1990, and says the order only came into effect in December 1991. Number 57 14:43, 29 December 2011 (UTC)

Thanks for pointing this out. --Eleassar ^{my talk} 15:50, 29 December 2011 (UTC)

What is the "turnout" row in your table? Turnout refers to how many people voted, yet this is not. Number 57 14:26, 29 December 2011 (UTC)

Also, several of the figures in your table do not appear in the Stat.si reference - % for (88.5% in the source), %against (4.0% in source) and the turnout % (not in source at all). Number 57 14:51, 29 December 2011 (UTC)

Perhaps. I've removed the disputed tag now as I agree with your data. --Eleassar ^{my talk} 15:50, 29 December 2011 (UTC)

Thanks. I've removed the dubious tag though - the 90.8% turnout figure is based on the 1,359,581 votes cast divided by registered voter number of 1,496,860 (hence why it is in the row with the registered voters, not the entitled voters). This is not calculated by myself, but is the figure presented in the results table in the book. I can scan it if you want. Number 57 16:38, 29 December 2011 (UTC)

Ok. Some numbers seem to be missing. Actually, the number of support votes should be compared to the total number of eligible voters. It was stated per law that the majority (that is 50% and one voter of all eligible) should support the independence, 88.5% did (per stat.si). These numbers are the ones actually relevant.[14][15] --Eleassar ^{my talk} 16:52, 29 December 2011 (UTC)

I thought the issue had been cleared up now, and you had inserted your preferred way of recording the percentages in the text. However, I see you also changed the table to reflect your preferred %s (and in a totally different manner to the table you suggested below). Can we agree that as your preferred %s appear in the text, the results table remains in the standard format of presenting %s for yes and no? Number 57 11:05, 30 December 2011 (UTC)

Why not present in the table the results that actually influenced the events? This makes them more easily readable. If you think your data should be presented too, put them in the table along them in a way. Presenting in the table, which currently takes up one third of article, only the data that are less important, because they did not influence anything and have been reported only by a single source, means giving excessive prominence and therefore undue weight (the data should be presented in the proportion to the prominence of each viewpoint). See WP:UNDUE (a policy) and WP:RSUW (which is not a policy, but nonetheless). --Eleassar ^{my talk} 13:10, 30 December 2011 (UTC)

I don't understand what you mean. The results are the results. It is not a question of presenting "my data", it is a question of presenting the results as one would present the results of any other referendum. The Undue tag is rather bizarre, but I will seek a third opinion again. Number 57 16:52, 30 December 2011 (UTC)

That's ok. To be more precise, the tag has been added because "an article should not give undue weight to any aspects of the subject but should strive to treat each aspect with a weight appropriate to its significance to the subject... Note that undue weight can be given in several ways, including, but not limited to, depth of detail, quantity of text, prominence of placement, and juxtaposition of statements." --Eleassar ^{my talk} 17:19, 30 December 2011 (UTC)

What is being given undue weight though? Number 57 21:28, 30 December 2011 (UTC)

The % of support votes, of opposing votes etc. compared to the total votes cast. More significant is the % of support votes (88.5), of opposing votes (4.0) etc. compared to the total number of eligible voters, as this was what actually had to be determined and was found out. Also, the turnout of registered voters (1,496,860), more significant is the turnout of the voters that were counted in the calculation (1,457,020), because it was actually reported then. --Eleassar ^{my talk} 22:05, 30 December 2011 (UTC)

But as I have tried to explain several times already, this is not the standard for how the results of referendums are presented. When reporting results in a table, the %s yes and no should be the %s of valid votes cast. The special rules for this referendum are explained in the text, which is appropriate, but they should not be in the results table. Number 57 22:13, 30 December 2011 (UTC)

This "standard" seems not to be suitable here, and using it means a too prominent placement of results that are less significant. --Eleassar ^{my talk} 22:37, 30 December 2011 (UTC)

Response to third opinion request:

You should stick with the official result, as they served the basis for all subsequent actions. The differences in calculations can be described below the table, but I still don't see the point of reporting the slight deviation.—Dmitrij D. Czarkoff (talk) 08:59, 1 January 2012 (UTC)

Format clarification (on Number 57 request) until the consensus for another format exists, keep in table only the data that is specifically stated in the official source. That means the following table should be installed:

Entitled to vote
Total	1499294
Included in the electoral registers	1496860
Voted with certificate	2434
Voted
Total	1361738
At polling stations	1329523
Before polling day	19233
Not in place of residence	12982
Number of voters total	1457020
Choice	Votes	%
Yes	1289369	88,5
No	57800	4,0
Invalid ballot papers	12412	0,9
Returned ballot papers	2157	0,1
Did not vote	95282	6,5

All the further modifications should be the result of consensus. — Dmitrij D. Czarkoff (talk) 13:39, 1 January 2012 (UTC)

Results tables

Number 57 version

Choice	Votes	%
For	1,289,369	95.7
Against	57,800	4.3
Invalid/blank votes	12,412	–
Total votes cast	1,359,581	100
Returned ballots	2,157	–
Unable to vote[a]	42,274	–
Registered voters/turnout	1,496,860	90.8
Entitled voters[b]	1,499,294	–
Source: Nohlen & Stöver, Stat.si

^a Those unable to vote were working abroad or involved in military service or military exercises. If they are excluded, voter turnout was 93.3%.

^b 2,434 people were not registered to vote, but were able to do so using certificates issued.

References

Choice	Votes	%
For	1,289,369	94.8
Against	57,800	4.3
Invalid/blank votes	12,412	0.9
Total votes cast	1,359,581	100
Returned ballots	2,157	–
Unable to vote[a]	42,274	–
Registered voters /Turnout	1,496,860	93.4
Entitled voters[b]	1,499,294	–
Source: Stat.si