undefined | Better HN

0 pointsserge2k10y ago0 comments

> For a data driven company, it shows they either failed to collect related data and/or that data shows the article was accurate despite the poor choice of sources.

They do an annual survey of SDEs, and they do release that data internally (aggregated/anonymously). SO they do have data. The problem is if they release data taht says "Oh well X% of our engineers are happy with the company" Then when happens when someone says why not (X+Y)?

It's data without comparisons to give better context.

0 comments

fweespeech10y ago

> They do an annual survey of SDEs, and they do release that data internally (aggregated/anonymously). SO they do have data. The problem is if they release data taht says "Oh well X% of our engineers are happy with the company" Then when happens when someone says why not (X+Y)?

If they can show a clear majority are happy with their jobs, that is the strongest argument they have. It also shows the articles claiming otherwise are in error.

There is always going to be someone who complains but data is more valuable than "well, all these people are malcontents or misquoted".

srtjstjsj10y ago

It's not an election. "Clear majority" is an extremely low bar for "job satisfaction", especially when you consider that over the years, the happy portion vote every year, while many of the unhappy people depart.

fweespeech10y ago

Then what bar would you set?

I'd be happy with a low bar. My guess is it isn't even hitting that low bar.

vkou10y ago

If they can show that a clear majority are happy with their jobs, all that demonstrates is survivorship bias. Given their churn rate, this is not a good thing.

j / k navigate · click thread line to collapse

0 pointsserge2k10y ago0 comments

> For a data driven company, it shows they either failed to collect related data and/or that data shows the article was accurate despite the poor choice of sources.

It's data without comparisons to give better context.

0 comments

fweespeech10y ago

If they can show a clear majority are happy with their jobs, that is the strongest argument they have. It also shows the articles claiming otherwise are in error.

There is always going to be someone who complains but data is more valuable than "well, all these people are malcontents or misquoted".

srtjstjsj10y ago

fweespeech10y ago

Then what bar would you set?

I'd be happy with a low bar. My guess is it isn't even hitting that low bar.

vkou10y ago

If they can show that a clear majority are happy with their jobs, all that demonstrates is survivorship bias. Given their churn rate, this is not a good thing.

j / k navigate · click thread line to collapse