#MRXchat: The first of many more successful MR chats!

[tweetmeme source=”lovestats” only_single=false]Since I proposed the idea one week ago, it’s been nothing but full steam ahead!  You need only look at the stats to realize just how successful our very first market research twitter chat was. (Stats gathered from wthashtag.com)

  • 266 tweets

  • 34 contributors

  • US, Canada, Indonesia, and Australia

The top contributors tonight were @lovestats, @mrxchat, @LongoMR, @BenSmithee, @joelrubinson, @bifrap1, @theelusivefish, @JeanMarie50, @spychresearch, and @JHenning.

Follow all of them on my twitter list!  @LoveStats/MRXchat

Photo credit: kakisky from morguefile.com

Some of my favourite tweets included:

LongoMR: Someone said it the other day,: there is a bias in every method know what it is and make adjustments #mrxchat

theelusivefish: Stats are the observations from which we should make our conclusions … not the conclusions themselves (imho) #mrxchat
joelrubinson: about stats. 50% ad campaigns don’t work. 80% new prods fail, yet we look for 90% confidence in stats?! crazy!! #mrxchat
theelusivefish: Get enough data points around an event and you can generally pinpoint the truth somewhere inbetween #mrxchat

If you’d like to read the entire collection of tweets, simply click here.

The plan is to continue this the first sunday of every month, 8am (there were a few folks at this time!) or pm (many more folks this time!), EST, join us! Open topic every time, but with guidance for those who would like it. If you would prefer a different day of the week, please leave your vote here. If enough people request the same day, then we can switch. Thank you everyone for making our first chat so much fun! See you next time!


3 responses

  1. Hi Annie, great initiative. I’ll join the next #mrxchat

    1. Look forward to reading you then!

  2. “joelrubinson: about stats. 50% ad campaigns don’t work. 80% new prods fail, yet we look for 90% confidence in stats?! crazy!!”

    My father, a carpenter by trade, always used to say ” a bad workman always blames his tools.”. Stats are just a tool – most of the problem with MR I feel is what I term “type III errors”, which is the wrong hypothesis to begin with. So it isn’t the stats which are to blame, we need to be more numerate, not less. It’s the thinking that is wrong…

%d bloggers like this: