I have an opinion on the matter, just shared that on the custom redesign thread (I wrote it there because of another post that prompted my response): http://www.teamliquid.net/forum/viewpost.php?post_id=24472644

Heinsenzerg

Argentina2279 Posts

June 27 2015 01:53 GMT

#43

JieXian

Malaysia4677 Posts

June 27 2015 06:12 GMT

#44

if you want casual users to respond you'll need to reach out to them via mass pm and not rely on this thread

Arksun

Croatia6 Posts

June 27 2015 06:44 GMT

#45

Interesting. I'm busy at the moment, but will PM you this evening.

klipik12

United States241 Posts

June 27 2015 06:50 GMT

#46

Why make people PM and stuff instead of putting out an open survey through like google forms or whatever? You'll get more responses that way.

xM(Z

Romania5277 Posts

June 27 2015 07:57 GMT

#47

PM's give a feeling of closeness/meaning/importance/ so people will spill more bs; and, you can always spot a faker.

Dagobert

Netherlands1858 Posts

June 27 2015 09:36 GMT

#48

Consequences of using the following sampling strategies:

(A) Open letter
- nonrandom sample (<= N users who open this thread) -> analysis would only tell us something about thread-readers
- committed users only (<= n1 users who send a PM)
- very unlikely to be representative of the entire userbase (bear in mind: OPs main goal was to get representative data)
+ useful for exploring a small number of individual ideas (if you have no standardized survey ready)

(B) Random PM responder sample
+ random sample (erases most systematic selection biases)
o perhaps non-random subsample (inevitable risk of surveys, if participation is voluntary), but this can be tested and if necessary, you can bootstrap
+ large sample planning allows for sufficient N to run decent analyses
+ allows you to do (non-)response analysis (who did not reply, what are the common factors for people who do/nt, ...)
- N < registered users < users

(C) Stratified random responder sampling (added bonus)
How-to: Prepare online survey (as easy-to-use and foolproof as possible), release announcements in every corner of the website, allow unregistered users to do the same survey (preferably a unique survey button or similar on the main page), collect data, clean it up a little, categorize every datapoint by region, choose random subsample from each (can be proportional to region sample size or even).
+ widest audience
+ random subsampling (provided large enough n) allows for multivariate analyses
- no universal non-response analysis but
+ with sufficient N and n, you can now run and compare analyses (including at least response rates) taking each region into account
- may introduce noise (users replying more than once, troll behavior) if not careful, on the other hand
+ allows for trolling analysis

DaLighzic

1 Post

June 27 2015 16:04 GMT

#49

@Dagobert

Unfortunately for TL most of what he is saying is true. I have an advanced degree related to survey analysis and have studied sampling methods for quite some time. What I will say though is often-time an organization uses a convenient sample, a focus group, opt in group etc and still gathers relevant data from it so I think doing this survey would still be valuable to your team. Unfortunately the group all shares one thing in common, we have all volunteered for the survey because we read this thread.. To account for members that did not see this post, or maybe would not opt in through reading it on this forum it would be cool if you had an option to mass message the registered members who have logged on in the "X" last amount of time.

Good Luck.

Half the Sky

Germany9029 Posts

June 27 2015 16:59 GMT

#50

PM sent.

-NegativeZero-

United States2141 Posts

June 27 2015 21:00 GMT

#51

i don't want a redesign, i actually like the "classic 90s/2000s" forum style.

phantomfive

Korea (South)404 Posts

June 28 2015 00:08 GMT

#52

On June 28 2015 06:00 -NegativeZero- wrote:
i don't want a redesign, i actually like the "classic 90s/2000s" forum style.

Redesign doesn't need to change the style necessarily.....but rather makes it easier for people to find what they are looking for, and easier to use overall.

If it's done right, the visual style can be made a preference.

GunPaladin

United States1205 Posts

June 30 2015 01:23 GMT

#53

the feature is clear, LiquidCS site incoming!

Deleted User 135096

3624 Posts

June 30 2015 01:26 GMT

#54

Just as an update I'm still waiting for possible stragglers, but I wanted to let everyone know that we've received quite a few responses from this so thank you to everyone who's shown interest. I'll be shooting people a message once i've combed through everything.

Varanice

United States1517 Posts

July 01 2015 03:05 GMT

#55

On June 30 2015 10:26 wo1fwood wrote:
Just as an update I'm still waiting for possible stragglers, but I wanted to let everyone know that we've received quite a few responses from this so thank you to everyone who's shown interest. I'll be shooting people a message once i've combed through everything.

cool, hope I get picked.

Deleted User 135096

3624 Posts

July 03 2015 17:55 GMT

#56

I just sent the PMs to everyone so check your Inbox on tl.

Karis Vas Ryaar

United States4396 Posts

July 03 2015 18:51 GMT

#57

quick question. are these anonymous or not? I'm just asking because I said to pm me if you have extra questions and if its anonymous I realized that would be kind of hard for you to do

so to clarify since it never asked me for my identification I was wondering if it actually knows which specific survey I filled out. so for example if you wanted to see what I put you'd know which one was mine. not a major deal just a curiosity.