@vortex_egg I'll offer a general answer:

• This is something you should be thinking about constantly.
• My own problem isn't for want of materials, but for focus in getting through a specific work.
• (I'm avoiding doing precisely that in answering your toot, BTW....)
• (No, I'm not blaming you, it's on me.)
• I try to keep tabs on the works I really want to read.
  
• I'm leaning toward having a specific index-card deck that's devoted to this, on the basis that it's 1) physically manifested and 2) can be arbitrarily re-ordered as well as 3) easily supports works being moved to the "it's been read" category.
• "#BOTI" is another related concept: Best of the Interval. Keep track of what the best books/articles/concepts of the past week/month/year/decade have been. You'll end up with a very highly-curated reading list to recommend to others.
• Scan both references and citations for promising new material. A reference with an interesting association / concept, or citations of a highly-significant work, are both promising prospects.
• Mortimer Adler's How to Read a Book, especially the later chapters, is an excellent guide.
• Consider sources as conversations or interrogations.
• Have no regrets at bailing from an unrewarding source.

I'll also look at others' recommendations, reading lists and syllabi, etc. I'm finding with time that individual recommendations and mass-market recommendations are often far less fruitful than the methods above. Subject-matter expert recommendations, on the other hand, especially for older and obscure works, are often gold.

Much new publication is not much worth reading. The back-catalogue is highly underappreciated.

#CuratingReading #Research #LiteratureSearch #StudyMethods

@vortex_egg
One cheat approach to this I have is to go through the HN daily archive, where HN is a decent prefilter. Often there's a small set of the top 30 (or up to ~100) daily items which are worth a closer look. I'll read a few days behind.

On Reddit, if you've got a good subreddit, you can set a date range (day, week, month, year) and then sort by "top" to get the most highly-rated items in that period. Reddit tends not to do a great job of quality selection, but it's not completely worthless.

The third of my two tricks is to just rely on random selection to an extent. If you've got too much material to make an informed choice on, shuffle your deck and select something at random. You'll miss some stuff, yes, but you're making an unbiased rather than a biased selection. You can also apply other filters to noise sources.

#Research #ResearchMethods #LiteratureSearch #InformationOverload #CalNewport #DavidAllen #GettingThingsDone #DeepWork #Zettlekasten #BOTI

3/end/

@vortex_egg There are two techniques in particular I'd like to suggest which ... well, they don't fully work but they seem to help:

1) Time block your information-gathering phase. Whether that's on a daily or weekly ongoing basis, or as a project phase, say "I'll scan Twitter for X minutes per day, only". And do that at the end of the day, when you've taken care of high-relevance/payoff tasks first.

2) What I call #BOTI: "Best of the Interval". On a daily, weekly, monthly, quarterly, annual ... basis, review the items you've selected as noteworthy from that period as well as the top items from the next smaller intervals, and select some n number of best items. You'll probably find that a good value is 10 <= n <= 100, but do what works for you.

BOTI draws on the 43 folders / tickler file concept, or the round-robin database. Essentially you're determining that no matter how long your research goes on, you're committing to a finite set of retained data.

(This is used in all kinds of IT systems and network monitoring, especially with long-term data history.)

You end up with higher resolution in recent / near periods, lower resolution as you go back in time. But you're constantly trying to filter up the best stuff. Since assessment can take time, you'll re-scan earlier selections to see if you'd missed something of relevance (and you can always break protocol for something especially good). But you've got a structure and have set limits on scope.

You'll also start to develop a sense with time as to what actually provides usefulness, and if you track sources, which of those are most valuable. Filter noise aggressively.

A source that sometimes generates signal but usually doesn't ... is virtually always noise. Signal tends to come through, eventually.

(This is related to my "block fuckwits" advice.)

#Zettlekasten #DeepWork #GettingThingsDone #DavidAllen #CalNewport #InformationOverload #LiteratureSearch #ResearchMethods #Research

2/

@vortex_egg This is a core challenge and failure of any academic. (Or unaffiliated researcher.) I struggle with this constantly.

That was the subtext of this toot:
https://toot.cat/@dredmorbius/106933960589722037

Cal Newport does a fair bit of writing on this, aimed at both academics and professionals. Deep Work is probably the best starting point.

A good academic programme (especially for re-entering / nontraditional students) should also address this. Talk to your advisor or department. That library-skills course you're taking is actually directly addressing this, or should (and I'd still reall like to see the course notes / outline / syllabus / readings).

David Allen's Getting Things Done is another good general time-management / goals-management guide.

Zettlekasten (or an equivalent notes-and-references-tracking system) is also very helpful.

#Research #ResearchMethods #LiteratureSearch #InformationOverload #CalNewport #DavidAllen #GettingThingsDone #DeepWork #Zettlekasten

1/

JSTOR's Text-Analyzer (beta) is fun and useful. Put any text in (upload an article, paste in portions of something you are writing) and it will analyze it and give you an adjustable search to find more articles from JSTOR's collection. Add and remove search terms and change the weighting to adjust the search.

Search is free, and JSTOR has options for independent scholars if you aren't somewhere with institutional subscriptions.

https://www.jstor.org/analyze/

#literaturesearch #researchtools

I am surprised to be impressed by Microsoft's Academic Search (http://academic.microsoft.com). Shows the citation network really, really well - cited & citing references. The AI-driven semantic search found a whole new set of vocabulary (leading to a new set of literature) for me. #literaturesearch #researchtools