How to write your PhD literature review WITHOUT using AI (Part 1)
ฝัง
- เผยแพร่เมื่อ 28 ก.ย. 2024
- Watch part 2 here! • How to write your PhD ...
phd.academy/bl...
For more PhD tips, ad-free versions of videos and details of my one to one coaching and my membership program, check out my website at phd.academy
0:47 The problem with using AI to write a literature review
2:04 Can you use AI to help write a literature review?
2:40 You definitely don't need to use AI
3:49 About James Hayton
4:24 The problem with summarizing papers as you go
5:59 How to build your knowledge of the literature
6:45 How to develop a general picture of the literature
8:19 Identify key influential papers
9:54 You can and should read papers multiple times
11:00 Getting to know the basics
11:55 Make sure you understand research and analytical techniques
13:14 Highly relevant papers for your specific project
13:43 Summary of the levels of basic knowledge you need
14:16 Build a collection of key, high-quality sources
14:48 Getting into details
15:23 The hard work
16:30 Any questions?
Other notes:
(2:15) The AI as an intern analogy isn't my own: I stole it from Kevin Kelly on the Tim Ferriss podcast: • Meet Your New AI Inter...
(2:40) Google scholar launched in November 2004, more than a year after I started my PhD
(8:19) L. T. Canham. Silicon quantum wire array fabrication by electrochemical and chemical dissolution of wafers. Applied Physics Letters, 57(10):1046-1048, 1990 (Cited over 7000 times as of August 2024)
Also check out this video by @writerscience on the same topic: th-cam.com/video/JRyVJKj4z7E/w-d-xo.html
Happy that TH-cam recommended this video to me. I don’t use AI in my PhD research, so my literature review is taking longer than other students who use AI, but watching this video just gave me validation that I am doing things right. Looking forward to Part 2!
Glad it was helpful!
😢oh how unfortunate that I discovered Dr. Hayton’s channel just when I’m preparing for my PhD defense in a few months(December). So many things you’ve shared on your channel that could have helped me at the start of my journey. All the same, I’m grateful for the few things I’ve learnt so far. I envy your students if you have any 😊😊😊😊. GREAT JOB SIR!!!
Thank you and best of luck with the defence!
Wooooow! That's great.
Waiting for part 2
Love from Pakistan❤
Thanks, this is very helpful. Regarding AI, I agree with everything you say about it, but disagree with creating a culture that does not use the technology in the knowledge production process. I won't debate the details here because this is clearly a complicated conversation, but just hint to the fact that we are living through the early stages of a radical transformation in epistemology. I am a sociologist and my PhD research is in human-machine (hybrid) cognition. I say this just so who reads my comment has an idea of where I come from with this argument.
I didn't argue for a culture that doesn't use technology. I argued that regardless of advances and technology, you need to build your knowledge and skills (and that if you don't have the knowledge to judge what AI is doing, then you need to fix that).
I think AI is here to stay. So we can always use it to generate ideas and then check them out in the library. Discard the framed-up sources and follow through on what is available at the varsity library.
As long as it's not the ONLY way you generate ideas. It might be here to stay, but I think the hype will fade as the problems emerge and people start destroying their own careers (it's already happening, but not getting attention).
As academics, you've got to be able to think and come up with ideas of your own (or through conversation with others). An academic who gets all their ideas from AI is like a chef who just uses a microwave.
@James_Hayton what a beautiful analogy!
@@James_Hayton Very True Sir.
Sir please please make one comprehensive video on PhD research proposal writing (particularly Literature review).
I'll add it to the list!
Thanks James. Always loved your sound advice.
James you were born to do this. Thank you so much.
Can you advise on the process of developing the research proposal?
Interestingly, in my field (psychology) the relative impact papers have does not substantially relate to the quality of the paper. I saw so many interesting, methodolically sound papers with very little impact as they might have been rather "late to the party" whereas some of the highly cited ones face severe limitations in design and counfounders.
This is why you need to understand research methods to assess the quality of individual papers. The number of citations is just one indicator of influence, and you shouldn’t just base your decisions on this one indicator.
I would hope that at least some of the highly cited papers in your field are good quality, but even if not you still need to know them if everyone else is citing them
Thank you Dr Hayton. Much valuable information
Very good video. Wonderfull insights
this is so refreshing- keep it up!
I was a policeman coming to the end of my career when I started my linguistics PhD; I had retired by the time I finished it and was able to start a new career in academia. AI didn't exist at the time. I started my lit review by choosing half a dozen seminal books and half a dozen cutting-edge journal articles and I wrote a summary of each as it related to my topic of interest. I also copied the references section for each of those sources. I pulled it all together from that, discarding 80% of what I had done up to that point. What I was left with was the gold dust for what I was researching. It also meant I had the requisite knowledge to be able to explain everything in my viva and even go beyond what I had written in my thesis.
Thank you for sharing!
Insightful and waiting for part 2, thank you for such helpful content in a world of distraction which is known as Ai
part 2 is out next week!
This is very helpful-I’m really looking forward to part 2! I have a question: would it be possible to make a future video reflecting on writing routines? I have a part-time job and aim to complete my work before the day begins and other obligations come up, which means an early rise for an undisturbed 2.5 to 3-hour work session. Would you advise starting each day by reviewing (and, if necessary, rewriting) yesterday's work? I’m asking because I know you emphasize intentional ("slow") writing, aiming for each version to be the best draft at that moment. I’m curious whether it’s better to trust yesterday’s edits and just continue to move forward without looking back, or to start each day by reviewing yesterday's work, or perhaps to create one or two moments during the week for that second review of what has been written?
I'd read over yesterday's work (and make some quick edits), then carry on where you left off. I also tend to read over what I have (or the last paragraph or two) if I get stuck. I read my work a lot, so by the time I finish, say, a 1000 word script, I may have read it 10 times through.
@@James_Hayton Thank you so much for the information! One last question: do you have a system for archiving previous drafts? Do you work directly in the main (final) document, and does that document contain comments, or do you have a separate system for gathering comments?
I tend to have one working draft and an offcuts document (so if I remove a section I don't have to delete it). I tend not to keep multiple prior drafts, though an easy way to do this would be to email copies to yourself.
These days I'm generally giving comments rather than receiving them. I do this by making notes on the pdf and then talking them through with students so it's a 2 way process and I can ask them questions rather than just leaving comments.
@@James_Hayton Much appreciated!
Dear Dr. Hayton,
Thank you for being one of the most honest and sincere youtubers teaching PhD students to do the hard work instead of looking for shortcuts that will only result in a mediocre thesis. Most of the youtubers online that recommend AI tools to PhD students do so because they get a share of every subscription. Interestingly, in this case no one talks about conflict of interests hahaha.
What would you say is a good way to identify quality in a paper?
Obviously the journal will play a big role, would you also look at number of citations and then at how robust the methodology followed was?
Many thanks for all your work!
Indeed... they should declare that it's sponsored content. I'm considering doing a reaction video to one of the worst offenders, but I'm not sure if I want to go down that route.
Identifying quality isn't easy, especially when you first start. The number of times a paper has been cited is a good place to start as it indicates impact (and that others have deemed it to be quality). For newer papers, though, you can't rely on this (and there may be excellent papers that have gone unnoticed).
You can also judge based on the writing (does it make sense? have they clearly described their methods? is it logically consistent with itself?). The very best papers tend to be more clearly written because the authors are confident in the content. If they're trying to hide something, they might make it impossible to understand.
Beyond that, you've got to develop some strong understanding of specific research techniques. Without that, it's really impossible to judge what they've done
Thank You Sir, for such a useful video. I am preparing my literature review for PhD. The video gives me a direction to start with. Thanks again Sir for such a great video for PhD students.
Very well timed! thank you .. Waiting for part 2
Insightful 💯
Your content has helped me immensely when starting to do deep dives into literature. Thank you so much!
You're welcome!
Thank you
You're welcome!
@@James_Hayton
I have a question. As a non-native speaker, I sometimes write my ideas in my native language first, then translate them into English. However, the translation often doesn't sound natural, especially for formal writing. Do you think it's a good idea to use AI to polish my original ideas and make them smoother? Here's an example of how I used AI to make my writing sound more natural.
Maybe, but I wouldn't rely on it exclusively. I think it's important to practice speaking and writing in a second language (not just translating from your first language). Also try reading good non-fiction (not just academic text, which is often very poorly written) in English (and read it out loud to get used to the rhythm of the language). It's a longer-term investment, but worthwhile, I think.
@@James_Hayton Thank you for your suggestion. I now know what steps to take next. Have a great day!
What do you suggest should be included in a 15-minute master thesis presentation of an eighty- pages work?
Stick to the most important points and don't try to summarize everything. The audience or panel can ask questions on anything you leave out.
Most important points would be
- what problem/ question/ need you're addressing
- what you did
- What you discovered
- Why this matters/ consequences or next steps
How much background you give will depend on the audience, but with only 15 minutes don't spend too long on it.
Also, with limited time, it's essential to rehearse and make sure you can run to time. Maybe even write a script so you don't lose time improvising. My last live presentation was entirely scripted: th-cam.com/video/D9RGfRVHMKQ/w-d-xo.html
@@James_Hayton Many thanks prof. This is quite helpful. Fifteen minutes is such a short period and it is overwhelming to think of what to include. Again, it's so tricky going into details beyond what is on the PowerPoint slide.