I'm... actually not sure about the legality of that. IANAL, but "You must specifically say this to use our publically-published system" seems like it could run afoul of compelled speech laws.
It at least doesn't seem to have the 'scale' problems some SE sites have, and it's done some interesting stuff to try and pre-empt some other SE problems...
It'd be interesting to know how much of that would just be closed as a duplicate on SO... Just curious, but that's one of those where codidact may not be yet plagued by scaling problems :)
@Sarov You give operational directives instead: say when, what and where, but not how ;) so 'fight the bastards in this village at break of dawn so we can get their oil reserves' without mentioning 'die' ;)
@ArthurHv universiteitleiden.nl/en/events/2021/09/… < that's the link to the webinar from yesterday. I don't know if you can access it, but if you can maybe Google scholar on the speakers names may get you somewhere?
What incentive do you offer, and how do you pick businesses without introducing bias?
How do you verify their reported findings?
I find it very unlikely that any business would willingly report the truth when it is "We treat our employees like shit and we limp along just fine, thanks."
Just because it's not your standard don't mean it wouldn't be authoritative to a degree. My colleague never ran a society or reported any number.
If you have 2-3 companies or let's be crazy and say 10-20 that show a small trend. Or not. Then you can make conclusions that at least for 10-20 companies it worked some way.
Would you better follow a guidebook that noone proved to be of any efficiency
I mean... not really. If you're trying to convince someone, maybe. But personally, I hear "My neighbour's cousin's dog found that X, so X is true" and I just think "Yyyyyeeeeeah no."
Idk what truth and dogs are doing here. Honestly it's sightly upsetting you tiptoe around the scientific rigor (which don't prove truth anyway) and the argumentative validity, when you're the first one to preach Scrum all around with fervor and be equally empty handed
If you have no idea what's like to be in X situation, and I have expertise in it, you may take my word about X situation, despite it having no scientific validity
The reason for this, is because of the point I mentioned, and where I'm at
My input and experience is utter shit data, it is still better than nothing
At that point I can only guess, but my guess is because they don't bother when they know it wouldn't be statistically significant. There's no difference between a blog post that says "Scrum is good because I said so", one that says "Scrum is good because I tried it and it worked", and one that says "Scrum is good because these 10 companies we paid said so".
If you really want/need actual studies, you may need to zoom out a bit and look to general purpose psychology. I'm sure there's been some sort of studies somewhere correlating morale to productivity.
(In fact, I know there are - I recall a study that showed that offering people more money made their mental math skills worse.)
(And yes, I'll be the first to admit that I tout agile/Scrum based on belief, rather than scientific evidence.)
There is a whole grey area between scientific evidence and belief you deny and I was aiming at. May it be poll results, survey results, any kind of authoritative expert opinion. You have all day everyday gaps in your knowledge you fill with things that are neither science or arbitrary beliefs. Read a newspaper and you're in it. I have countless examples. Decisions to believe or not are made on convincing arguments, the Scrum book for instance don't originate from your own decision alone.
Difference of terminology/perspective. In my view, there is no empty area between scientific evidence and belief, just a sliding scale/gradient. Yes, piling more anecdotal evidence on top makes belief less arbitrary.
That being said, like I mentioned, I don't personally see any difference at all between "I tried this and I like it" vs "I asked 100 people, 10 of them said they like it (90 didn't reply)".
But argumentative strength is a scale too. You could have some claims more convincing than others. I will even go one step further. Different type of people are convinced by different type of arguments. Convincing with rigor of scientific process, is one type of argumentative strength.
> I mean... not really. If you're trying to convince someone, maybe. But personally, I hear "My neighbour's cousin's dog found that X, so X is true" and I just think "Yyyyyeeeeeah no."
Personally, I find those arguments identical. Others might not.
I'll risk saying I'm sure you are less rigid in your conviction filter you pretend here because most of your knowledge today are probably not coming from scientific studies. And I do mean knowledge and not belief
Though I have thought on how one could go about it. I think the only possible valid (not feasible, but possible) approach would be to actually create new businesses. Make a third a control where you have no contact with them, a third testing your hypothesis, and a third testing your antihypothesis. Then let them run for a year, then measure.
To close the incident. You were not the one I try to convince though. There is two persons me and my boss. Personally I'd like data that I don't have to test my belief with reality. And he doesn't have his take of pie charts either. So I was frustrated. Which explain the original message on the lack of studies. That you don't find them convincing for you is a bit non-sequitur so idk
My underlying point (albeit poorly presented) is that there's no way I can help you. It's become more a question of knowing what would convince your boss specifically, rather than what would be 'objectively convincing'.
on the case of "partial studies" I have an example of book that tested software development cost and methodologies. There was a whole lot of case studies on small samples. But with very interesting findings sometimes
For example they tested TDD. They invalidated some benefited claims
Hmm I can't remember it's been too long time, but if i remember correctly there was no statistical difference with the number of defects. It's been so long I can't even remember if their test group was testing by hand or writing tests after.
Either that or they invalidated the idea it takes the same amount of time to develop than with no test. IIRC at least people not used to TDD were taking longer.
Hmm. That'd be hard to test, I'd think. My belief/intuition is that TDD makes things slower to develop at first (esp. when still learning it) but faster to maintain.
It'd be a hard sell to me that 100% test coverage is just as hard to maintain as 0%.
I mention it because I found it interesting back then. Especially the approach to try to go see what happens even if it's just partial. I remember for example we say as engineers more lines of code is generally a bad sign. They measured development costs with LOC in different companies and highlighted it was non linear. So for example 10 times bigger software took 20 times bigger costs. It's intuitive but it was nice to see it verified
> a person's opinions about the management of government Hint: Politics can be used as a singular or a plural in writing and speaking Politics has always interested me. The country's politics have changed.
