[RESOLVED] Iconic Failure

I started this thread over in CC, where it belongs, but it contains an interesting study that may be possible:

http://www.vbforums.com/showthread.p...eputation-Icon

I believe that there is a measurable difference in the use of the reputation icon since the forum upgrade, possibly restricted to new users, but possibly across all users. There is rarely a chance to measure the difference an icon makes, but the more I think about it, this forum now has a chance to mine its database to see something really interesting to programmers. If there has been a noticeable drop in repping coincident to the inseption of the new forum software, I would say that it would be empirical evidence of the impact that icon design has on UI design. Such a study would be pretty hard to do in most cases as the sample size would be low and could easily be biased. However, with the large number of users of this site, sample size would not be an issue. Therefore, let me add a request that somebody take a look at this in the interest of GUI design.

It reads to me like you are talking about the report post icon but are you actually referring to the "Add to this user's reputation" icon? The reputation icon looks like a black star to me.

Yes, it does, and I frankly can't much remember what it was before other than a scale....I think. However, the difference in the pattern of reps before and after the software change appears to pass the Interocular Arm Length Test of statistical significance, to me. It is entirely possible that it is just me, in which case it is meaningless, but it seems like a couple fairly simple queries against the DB could show whether the incidence of repping has dropped dramatically (either for all members or just new members), in which case the very number of users of this forum would add weight to the finding. Demonstrating an empirical difference between two icon choices should be of considerable interest to programmers.

I think it was a set of scales.

I dabbled with the templates. I added "(Rate this post)" to the thread. (You can try it out on my post here if you'd like ;) )

I think that is what we had before. I didn't need a DBA to do a test. The reality is that the number of rep ratings is down. I took a look at the numbers of reps in January of this year compared to last year. Let's see if the added text makes a difference for March.

Brad!

I just put it down to the post-Christmas Grumps.

Brad, although it might be anecdotally obvious that ratings are down it might be worth pulling some stats anyway. Particularly if you're going to start playing with the templates to see what works and by how much. I doubt the exact results are really going to directly impact my choices as a developer but as a purely academic exercise I think it would be interesting to see. Come to think of it, it'd be quite interesting to see a whole bunch of statistics like usage rates, response rates, percentage of threads being resolved etc. Those of us with an interest in data analysis would have a field day playing around with some of that stuff as I can think of all sorts of patterns we might find. E.g. has the usage rate dropped off or increased post upgrade? Is that likely to be affecting ratings? Does the qty of rating correlate to resolutions?

Of course, I'm not sure how much info you'd really like to publish from a business perspective or whether it would be worth the cost (Ialthough some of us users would enjoy playnig with the figures I doubt it would see a pay-back to you guys) but worth mulling over at least.

I looked at the stats for reputation, so I know what the difference is. I also have analytics on the forum itself. We don't share this information publicly. Sorry.

I will tell you that visits year-over-year are up. February actually saw an additional bump up for a total increase that was > 40% over last year. Prior to last month we were still greater than 10% growth, which for this size of forum is notable.

Quote:

---

We don't share this information publicly

---

Yeah, I thought that would probably be the case. Can't blame us for being nosey, though, it's in our nature:rolleyes:

but as you can see, I did share some info. Numbers are up nicely, which considering the speed and other issues we've suffered through,that is pretty big.

As was noted, the number of reputation ratings as significantly down. I pt the rate this post text back, so if you remind me in a month, we can look to see the impact.

Oh, I'll remind you. I might give it a bit longer than a month, but since you added that so early in March, the month of March might be a good comparison to the month of February.

Thanks for making that change. It should be quite interesting.

One funny side to this: It has only been a few months, but I honestly don't have a very clear memory of what the old rep symbol was. I remember some scales, and a few people have said there were words. I have no memory of words. That says something about observational skills, I suppose.

Quote:

---

Numbers are up nicely, which considering the speed and other issues we've suffered through,that is pretty big

---

Yeah, that's a big positive. With the performance issues behind us (touch wood) it bodes well.

Quote:

---

I remember some scales, and a few people have said there were words.

---

I only remember the scales too.

Quote:

---

Originally Posted by Shaggy Hiker

...but I honestly don't have a very clear memory of what the old rep symbol was. I remember some scales, and a few people have said there were words. I have no memory of words. That says something about observational skills, I suppose.

---

Well if you can stomach some major insanity for a bit, you could sign up on Stormfront. They still use the old vBulletin software for their forums. I'd do it myself as I'm curious about the old icon too but I don't want some secret intelligence agency to flag my IP address for signing up on a hate site and I can't find any other vBulletin forum that uses the old software. Everyone upgraded. In fact I think VBForums was among the last vBulletin based forums to upgrade to this one.

[EDIT]

Just discovered that vBulletin.org also still use the old software but their reputation system seems to be disabled.

It isn't even April yet, and I think the answer is in: Adding the text made a HUGE difference. Since the wording was added, I've seen enough reps to suggest that the numbers since that change are soaring.

Score 1 for text over icons. Traffic safety folks take note!