4 ms·

I would love to know how a study could be conducted in such a way to prove or disprove this notion. Any ideas? Any existing data sets that could reliably mine

by gacba 16y ago

I would love to know how a study could be conducted in such a way to prove or disprove this notion. Any ideas? Any existing data sets that could reliably mined for info?

Morendil 16y ago
I wouldn't start with a data set. I would start by stating clearly enough the question we are looking for an answer to. Doing that might involve getting clear on the terms. "Productivity" is too abstract and ambiguous to be used as-is, so we first need to say "here is what we are going to measure".
- kenjackson 16y ago
  Bingo. Software engineering is filled with throwing about this term "productivity" yet its unclear as to what it means. And even what the impact is. With that said, this is not trivial even for activities where the metrics seem a lot clearer, e.g., basketball or basetball.
- narag 16y ago
  I believe that what people try to determine is not a very precise measure of productivity, but the fact that productivity measurement doesn't matter so much. The central message that I remember from Peopleware is that software teams are more about people than about numbers. Performance is going to vary, often wildly (often for the same people at different times). I loved the anecdote of the woman that wasn't very productive herself, but was a "catalyst", maybe simply a fancy way to say she was fun and nice to work with. Unless you are Google and have the right process to detect the best and resources to attract them, your best bet is to treat people well and apply common sense... as opposed to apply hard metrics and put pressure, demoralizing your people.
daviding 16y ago
It would be hard - http://www.martinfowler.com/bliki/CannotMeasureProductivity.html http://www.martinfowler.com/bliki/CannotMeasureProductivity....
stonemetal 16y ago
First you would have to define what you are measuring and develop the experiment. Then get enough HN readers to sign up to take the experiment so that you get statistically significant results. For caned results that you could look at I would think ICFP programming contest or similar would have numbers though a quick look hasn't turned anything up.

by gacba 16y ago

I would love to know how a study could be conducted in such a way to prove or disprove this notion. Any ideas? Any existing data sets that could reliably mined for info?

Morendil 16y ago

I wouldn't start with a data set. I would start by stating clearly enough the question we are looking for an answer to. Doing that might involve getting clear on the terms. "Productivity" is too abstract and ambiguous to be used as-is, so we first need to say "here is what we are going to measure".

kenjackson 16y ago
Bingo. Software engineering is filled with throwing about this term "productivity" yet its unclear as to what it means. And even what the impact is. With that said, this is not trivial even for activities where the metrics seem a lot clearer, e.g., basketball or basetball.
narag 16y ago
I believe that what people try to determine is not a very precise measure of productivity, but the fact that productivity measurement doesn't matter so much. The central message that I remember from Peopleware is that software teams are more about people than about numbers. Performance is going to vary, often wildly (often for the same people at different times). I loved the anecdote of the woman that wasn't very productive herself, but was a "catalyst", maybe simply a fancy way to say she was fun and nice to work with. Unless you are Google and have the right process to detect the best and resources to attract them, your best bet is to treat people well and apply common sense... as opposed to apply hard metrics and put pressure, demoralizing your people.

daviding 16y ago

It would be hard - http://www.martinfowler.com/bliki/CannotMeasureProductivity.html http://www.martinfowler.com/bliki/CannotMeasureProductivity....

stonemetal 16y ago

First you would have to define what you are measuring and develop the experiment. Then get enough HN readers to sign up to take the experiment so that you get statistically significant results. For caned results that you could look at I would think ICFP programming contest or similar would have numbers though a quick look hasn't turned anything up.