Poll result display

[doire]

Title:
Poll result display

Area:
Polls

Summary:
Make 100% display width uniform between different questions in the same poll.

Description:
Now the poll results display bars where the most common response for each question is the same width for each question. If there were 3 questions each with 4 options you might get:

1 a --------------------- 100
1 b - 0
1 c - 0
1 d - 0

2 a --------------------- 25
2 b --------------------- 25
2 c --------------------- 25
2 d --------------------- 25

3 a ------- 15
3 b --------------------- 50
3 c ----------- 25
3 d ----- 10

I think it would be more visually understandable if the maximum percentage in the whole poll was used to set the display width. Like so:

1 a --------------------- 100
1 b - 0
1 c - 0
1 d - 0

2 a ------ 25
2 b ------ 25
2 c ------ 25
2 d ------ 25

3 a ---- 15
2 b ----------- 50
2 c ------ 25
2 d --- 10

It might be an option when setting up the poll if some prefer it as it is.

Sorry if this has been suggested before; I can't tell since search is down. I think it still gets my point across even with proportional fonts. My pre tags - poor pre tags.

Poll #7839 Poll result display

Open to: Registered Users, detailed results viewable to: All, participants: 49

View Respondents

This suggestion:

View Answers

Should be implemented as-is.
20 (40.8%)

Should be implemented with changes. (please comment)
5 (10.2%)

Shouldn't be implemented.
5 (10.2%)

(I have no opinion)
18 (36.7%)

(Other: please comment)
1 (2.0%)

Comments

[ratcreature]

I think this might be a problem if you have a poll where a lot of people answer to one question, but only a few answer to another question, because it only applies to part of the people. Like say one question 200 people reply to one question, because it applies to everyone, and then in the next question only seven people answer in total, because it doesn't concern the other 193 people, then in the second question all would be squished to the utmost left.

[simont]

On the other hand, if the whole point of the poll is to compare answers between the questions, it might well end up being less misleading done the new way. For instance, a common poll format I've seen goes

• Which of these is your favourite? [list of radio buttons]
• And which of them do you like at all? [exactly the same list of options, but they're checkboxes this time so you can select more than one]

It seems clear in that case that you'd like to be able to directly compare the lengths of the bars for people who liked X and people who liked it more than anything else.

I think this argues in favour of it being at the discretion of the poll creator, who can set it depending on what the poll is for.

[voldsom]

+1 to this.

(Actually, I'd also like to see my percentage bars as their actual percentage widths for the individual question, rather than the highest percentage being 100% width, but that's an unrelated aside.)

[ciaan]

+1

[archane]

+1