File talk:UK popular vote.svg

There are no numbers on this graphic and it's very hard to understand. Isn't it possible to find a simple line-graph to better represent the results? —Preceding unsigned comment added by 193.225.200.93 (talk • contribs) 11:03, 11 May 2010 (UTC)

The aim is to show the shares of the vote at each election, rather than just the fluctuating level of each party as in this one. The vertical bars facilitate comparison of the performance of different parties in the same election, and also show election timing. The level of votes at each election is the real data; the lines between elections are purely cosmetic.

Regarding numbers, I initially had 0%, 20%, 40%, 60%, 80% and 100% on the left side, but worried that it might create an appearance of bias in one direction or other. Kanguole 15:16, 11 May 2010 (UTC)

Deceptive data?

I assume each bold, vertical line represents an election at that particular year, correct? In that case, I think the lines between each election is deceptive, because it implies that changes happened between those election points as well.

For example, if one party in 1980 had--for example--50 members, and the next point for 1984 had--again, for example--58 members, the way the data is apparently represented, it implies that, in every year in between 1980 and 1984, that party gained two members... which I don't think is the case (that is, from 1980-1984, there were 50 members, and then in the 1984 election, that party gained eight members at that one moment). Unless, perhaps, there's something I don't know about British politics and that they have elections every year for seats?

I recognize that this discussion page isn't for requesting corrections, and that corrections be requested on the page this graphic is on, but because this graphic is on several pages and this point I'm making isn't about if this graphic is appropriate for the page(s) its on BUT about the data the graphic is representing, I felt it appropriate to post something about it here. -- 66.92.0.62 (talk) 22:19, 29 March 2011 (UTC)

Your interpretation of the vertical lines is correct. It's true that the only real data here is the proportion of votes in each election, so it would be more accurate to just show bars for the elections and nothing in between. It wouldn't be accurate to show steps, because the graph represents proportions of the vote at elections rather than proportions of representatives. Presumably public voting intention does change between elections, and the thin lines and light shading here are a naive interpolation. But I think the only more accurate alternative would be nothing at all between elections, and that would be hard to read, especially when there are big swings. Kanguole 22:50, 29 March 2011 (UTC)