Indeed, the people who use Input are not representative of Firefox users in general. Heartbeat is a much more accurate tool for general sentiment since it's not as obviously self-selected.
Those who remember Gregg's speech from Whistler know that Input denotes passion and not necessarily sentiment of the general population. Additional context from Matt Grimes who heads up the User Advocacy team: "We do know that when we first launched Pocket there were complaints on Input and the press picked up on this. So we ran a Heartbeat study to find out if this was the general sentiment or indeed showing a very passionate subset of users. You can see the results here: https://useradvocacy.mozilla.org/reports/Desktop/PocketSurvey/ Basically the general population finds Pocket to be neutral to positive. Many users had not heard of it and actually rated it MORE positive after we had informed them of the existence of Pocket." Thanks, David On Dec 18, 2015 4:26 PM, "Luis Villa" <[email protected]> wrote: > I'd be careful about trying to claim any sort of statistical validity for > the feedback button. It is hidden in a menu most users will never see, much > less use; by definition the people who find and use it are not > representative. The only valid data you'll get on Pocket is by actually > measuring use. > > Luis (nightly user since '99 who loves the Pocket integration) > > On Fri, Dec 18, 2015, 12:04 AM Angly Cat <[email protected]> wrote: > > > On Friday, December 18, 2015 at 1:37:22 PM UTC+6, Francesco Lodolo > [:flod] > > wrote: > > > I don't think looking at 90 days is fair. Let's look at 7 days, at > least > > > there's something manageable > > > > > > https://input.mozilla.org/en-US/?q=pocket&date_start=2015-12-10&selected=7d > > This is not how statistics works. To get a representative statistical > > sampling, you should be widening time range, not narrowing it. Pocket was > > shipped with Firefox 38.0.5 at 2nd of June, more than 6 month ago. So > > basically we should look at all feedbacks within this time (minus month > or > > two in order to filter 0-day rage feedbacks). > > > > > 4 out of almost 4k people who took the time to use input to give > > > feedback (and we all know that people rarely take the time to give > > > positive comments). Would you define this as 'overwhelmingly negative'? > > > I don't. > > I'm sorry, but you're really not good in statistics. Regarding this > matter > > we don't and shouldn't care about feedbacks that are not about Pocket. > E.g. > > if there would be a hundred negative feedbacks about Pocket and a billion > > positive feedbacks about anything but Pocket - that's 'overwhelmingly > > negative' Pocket feedback. > > _______________________________________________ > > governance mailing list > > [email protected] > > https://lists.mozilla.org/listinfo/governance > > > _______________________________________________ > governance mailing list > [email protected] > https://lists.mozilla.org/listinfo/governance > _______________________________________________ governance mailing list [email protected] https://lists.mozilla.org/listinfo/governance
