[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/sci/ - Science & Math

View post   
File: 239 KB, 819x1428, 1574646725199.jpg [View same] [iqdb] [saucenao] [google]
11422194 No.11422194 [Reply] [Original]

What is the definition of good science? Are there any tell-tale signs of garbage tier research?

>> No.11422216
File: 377 KB, 400x521, yudkowsky bayes.png [View same] [iqdb] [saucenao] [google]
11422216

https://www.readthesequences.com/The-Dilemma-Science-Or-Bayes

>> No.11422482

>>11422194
you can call all bio research garbage. leading in number of paper retractions

>> No.11422523

>>11422194
Depends on the level you are talking about, could be anything from poorly sourced and argumented to lacking sufficient attempts of alternative explantions

>> No.11422571

>>11422194
>Studies with a sample size <100
>Sponsored by a corporation, or even worse a government agency
>Conclusion doesn't fit the research, e.g. "Based on our findings that humans come in multiple distinct phenotypes, we conclude that race is a social construct with no basis in biology"
>Anything from Harvard in the past decade

>> No.11422596

If it looks like every other paper and is accepted by another careerist researcher for publication in a journal, it's good

If it contains original ideas, it's bad

>> No.11422717

If it's hard science then its ok (not good). If it's not be wary of the results and assume they are wrong until a replication study is performed or some similar result is obtained. Also a good tell mark for individual researches is to check their publication history. If it looks like they churn out a shit ton of papers on what seems to be a highly iterated experiment, their research is probably garbage. Gauss' advice is best, Few but Ripe.

>> No.11422880

>>11422571
>>>/pol/

>> No.11422893

>>11422880
Bonus: purposefully nonsensical conclusion that doesn't fit the data and the graphs to get past the skimming peers

>> No.11424002

>>11422596
based

>> No.11424030

>>11422216
Well that was a pile of garbage.

>> No.11424042

>>11422571
>mfw majority of oil & gas research has been done by major corporations and the USGS

>> No.11424043

>>11424042
>and the USGS
Rwally?

>> No.11424723

>>11424042
With that sort of situation you have to take into account the bias and the nature of the paper - for a example an oil company funding a study on "New Methods to Estimate Remaining Oil Amounts" is probably good science; accurate info means more profits for the company. A study on how new drilling methods effect ground water quality is going to be hit or miss though.