How to Write a Literature Review Fast: Strategies and AI Tools

hey professor stuckler here today we're going to talk about writing a literature review fast and you're probably here

because maybe you've got a deadline coming up in a short period of time a month a week or even the next 48 hours

or you heard because you just feel slow I I've talked to students who have been working on their literature review for 7

years that sounds like endless torture Purgatory if I was going to construct a scene of the devil taunting me it would

be the seven-year NeverEnding lit lure review so the reason most students are not going fast is because they don't

have a system a system to work from and so they're just fumbling around trying to figure things out as they go and

maybe they've had a course at their University or even a helpful supervisor to give them pointers but eventually

they hit a stumbling block and those stumbling blocks from our Decades of experience working with students myself

as a professor at Harvard Cambridge and Oxford I found there really four let's call them four Horsemen of the lit

revieww apocalypse that you need to tackle and knock down or you're going to go slow in fact when you got a system

and strategy in place you just have a smooth ride because you kind of bat away these Horsemen they don't even arrive on

your track because you knew how to deal with them so these four are getting your topic right defining your search

strategy understanding analysis and riding up and submitting like a bro we're going to go through each of these

and I'm going to share with with you how to use AI tools in the right way to accelerate each of these notice you know

you're going to see a lot out there say use AI to write your literature review to me that's just missing the plot

because I'm seeing students get AI anxiety it's like oh I did some AI stuff my professor liked it but now now I

don't know what to do and I don't even really understand what the AI did and I feel like a fraud no the correct way to

use AI is understand the system and then you'll figure out how you can uniquely add in important ways things that

machines do better now and are going to continue to do better in the future while preserving your confidence in

those unique human things that you bring to the table so let's dive in we're going to go first to the topic getting

your topic right over and over I've said over 90% of your success comes down to getting the topic right I I remember

when I was at Oxford I had two fantastic postdoctoral researchers and they were right at the juncture where they're

trying to become independent researchers and grow Thrive maybe get some competitive funding and get their own

tenure track positions and one was really I I think objectively probably the top person I've ever worked with and

another I would put in maybe the top 20% um but what happened surprisingly is that second one got the 10e track job

faster got money competitive grant funding faster and and the difference didn't have to do with ability in fact

that that first student um he he was much stronger strongest student I've ever worked with he's now a full

professor at for the difference was the topic the second student came up on a Hot Topic everybody wanted to focus on

and really rode that wave strategically and so I encourage you guys to think about the topic where you're going to

plant a flag and Define yourself on a few Dimensions one is going to be something you're passionate about

because if you don't have the passion you're just going to feel like you're swimming against the tide against the

current and it's going to leave you feeling a bit deflated over time and it's going to be feasible so when I say

feasible means you can do it in the time that you have I I see a lot of students attracted to learning machine learning

right now and that can be a fantastic investment but it's a longtime investment when I was a student uh I got

rejected for a project that I was trying to get funding for because I said I I was going to go learn Russian and they

didn't believe I could learn Russian in the time that I had and you know looking back they were probably correct and and

to this day I can read some Russian but don't speak it um so it's got to be feasible with a skill set that you have

or you can easily acquire and for most of you right now I encourage you to think about loow hanging fruit things

that you can kind of pluck off the tree and get some big wins that are going to catapult your career because those early

successes have been shown to predict future successes and there really is what's called a Matthew effect in

science where the rich get richer and the poor fall behind so you need early quick win successes to really Propel you

on that proverbial Fast Track and the third thing you got to get on your topic right is a debate there has to be a

lively discussion I see so many students who are like I'm thinking so far ahead of the Curve it's like yeah well you're

well and good to be ahead of the curve but you're talking to yourself there's no conversation yet there and you're

going to have to you may want to get there but that can be a life's work you have to bring everybody there with you

and it's just a high-risk play and I don't recommend that now especially for many of you if you're doing a masters a

PhD early post doog that's not the highrisk play take that highrisk play when you're in a secure position don't

have to agree with me I'm sure I'm going to get slammed in the comments for that but that that's my recommendation be

pragmatic some of the students I work with talk about getting a PhD like a driver's license you can't drive until

you get the driver's license and I think that's a great analogy and a great way to think about it get to the next step

nobody wins a Nobel Prize winner here as a grab so that said your Topic's got to optimize these Dimensions well then

students say well how do I do that that's hard to do you need to find a research Gap so in that space you need

to go kind of harvest and look for gaps and what are these these are things that uh that are just missing from the field

they may be things we think we know but haven't yet been shown or or proven there may be things that have yet to be

done and there are a few sources where you can find these gaps one is by looking in research papers where they

talk about hey for future research we suggest doing this those are excellent places that you can take off the shelf

gaps it's like somebody's just say hey you over there this is a great project for the future check that out and often

I encourage students when they write a literature review to say roll out the red carpet for the next papers and say

Here's what right the lit review often Works kind of as a funnel down here and and arrives at the end it goes very

broad looks at everything and arrives at the very end by distilling and spitting out some gaps and here's what we need

here's what's missing in the field I've looked at everything here's the next steps here's low hanging fruit here's

where where we're going to go that's that's kind of the link point if you don't know where the Gap is do a lit

review that's just a good thing for you to do anyway to feel confident if you're choosing a topic and you haven't done a

literature review you could be doing something of very little marginal importance and you wouldn't even know I

mean how frustrating with that be to choose a topic invest tons of time and energy and it's like it just made a

little contribution that would be terrible I I wouldn't want that for anybody I mean most of you are in this

Bas because you want to make a difference and you want to make an impact so you know doing that leg work

up front is going to save you time now let me show now that we've got that now that you know how to comb through

research articles do that and and do your own literature review and we've got step-by-step guides on that and we're

going to take you through those next steps let's go over and let me show you how Chad GPT can help with this all

right I'm here in Chad GPT let's say please identify research gaps for a literature review topic um that I could

write in a short period of time on health inequalities just picking something random and lung cancer

screening and not I've intention Al made this a little bit unfocused to see what chat GPT comes up with while this is

pulling up a welldefined literature review topic will have Clarity on the outcome it's looking at what is like

that good or bad thing it's an it's an outcome is a result it's often an impact of something on something else is that

impact on what um and uh it might also have something that's an exposure like uh in this case might want to be

looking at I don't know the impact of poverty on access to lung cancer screening that would be a very clear

well-defined topic because it has an exposure defined and an outcome defined we have a full model called Pico which I

won't get into here which looks at the four elements of the topic the population an intervention a comparison

and outcome which you could use to help think through your topic but I I first want to emphasize with you the way to

come up with ideas for a topic because you can't really even create a constructive PECO if you're completely

lost at C so here's where chat GPT can help it can give you some ideas here uh of what you could do on your literature

review you could do a literature review on access to screening programs so you still might need to take this a step

further this can trigger some ideas say ah maybe I could look at interventions to increase access to lung cancer

screenings uh to help reduce inequalities or maybe look at what are the do a literature review on the

cultural and social barriers to people going out and getting lung cancer screening the other thing I want you to

know as I'm talking this stra to you I want your topic to be so simple that you can explain it I've had so many students

come to me and say I want to look at the role of the feminist in the Irish Wars of the 19th century and or or something

and that that's not necessarily bad but it just if you start putting in jargon and other things you can really get

yourself confused really really fast it's going to make it harder down the road so I want it to be so simple almost

you can visualize what your topic is about and what it might be showing and I know that's a bit reductionist but I

want you to reduce some of this that's the whole point of science is to reduce a complex messy world into something

that's a little more easy to access and and make sense for others so here you go you can just see Chad GPT can give you

some ideas and food for thought so let's say we want to take one of these the second thing I recommend you do

remember we talked about those three components that you you want to be passionate it's got to be feasible it's

going to be a debate we can go look for the debate so this is where Google Scholar comes in handy so let's say we

want to look at uh access interventions around access to lung cancer screening so here we go uh lung cancer screening

access inequalities interventions something like this just in Google Scholar see what's up there and lo and

behold I can see wow there's uh there's there's a lot of AC ity here there there's lots of re there's recent papers

coming out um I can see right a highle statement on this topic here here's another one disparities there's already

been some literature reviews so you do want to do something do to due diligence and make sure you're not duplicating an

existing review so I always encourage you to find something on your topic called the nearest neighborh paper what

is the closest paper to yours that's not yours and make sure you're adding just a little bit over and above that but the

good thing here is I can see that here's here's one here inter prevention designed to increase the uptake of lung

cancer screening Equity or endoscop review looks like this one's already been done but what I I and that's

helpful to know so I don't go then invest all this time and energy come out the other side like oops somebody

already did it well that was kind of a waste of time so I really encourage you here to go through some of these and

identify a topic where you see that discussion and debate from the citations in Google Scholar uh for your lit review

paper to know that there's enough papers know their citations but also to see if you can find a gap if in the back of

your mind you're hoping to publish this down the road to say you're not just duplicating an existing lit review okay

so now you've got your topic defined we come to the next fork in the road and this is where we encounter our next

Horsemen of the lit review apocalypse and that's a search strategy and I know some of you maybe thinking well what's a

search strategy and that's the problem a lot of people are just kind of willy-nilly hunted around on Google

looking for stuff here there putting things together ending up with a mountain of papers maybe dropping them

into a reference manager if they have a reference manager you need a reference manager Where Have You Been get zoto now

if you don't have it all right I digress so um yeah you you you need a search strategy and and what is that got my

topic great well and good where am I going to find articles for this and what articles am I going to look at what am I

going to include and what am I also going to exclude you see if you don't have this you're going to take forever

you're going to get overwhelmed there's just too much mass information out there and you need to have some filtration

system to to kick out the stuff you don't want it's like if you go fishing right I'm throwing out a net and I'm

going to get a whole bunch of stuff in there I'm going to get some Seabass some tuna maybe some sardines and anchovies

you got to know am I looking for a sardine or an anchovy what am I going to throw back so I got a student who is

doing a search and getting way too many articles and I was like well what what's your search strategy what your incl

inclusion exclusion she's like I don't know like oh okay well her her topic was on uh the impact of digital feedback on

uh English learning English as a foreign language learning it's like okay well um maybe you want to include only certain

types of studies like only those that looked at an intervention or maybe you only want to include certain types of

digital feedback tools maybe you specifically don't want AI tools maybe right maybe you want those that are

completely autonomous from the teacher and so you really have to think hard about the nuts and bolts of your review

to make sure you're getting exactly what you want because if not you're going to feel like you're on the back foot and

you're always finding another article and like oh look here here's another one and you just can C down a rabbit hole

and waste so much time and have to restart things and it changes your analysis and thinking about it and this

idea you know right maybe you've done this before just kind of working it out as you go along that works in a

contained project but something of this magnitude summarizing an entire body knowledge you can't just put together

this hodg Podge in this way it's just a recipe for frustration and taking too long and I'd say the main reason people

aren't going fast is not because they can't go fast it's because they're taking too long and I know that might

sound kind of well isn't that the same thing it it's not because you got these roadblocks that are stopping you from

just continuing to have a smooth easy ride so you need to Define your search strategy and you're usually going to

fall into one of two camps you're going to go down the path of Google Scholar and you're going to search there in

which case you still need inclusion and exclusion criteria and we've got dedicated training on setting those up

for your topic you have to have the clear topic right before you can do that and knowledge of what you want and where

you're focusing your review and that incision you're making into the world to say something new um the other type that

I strongly recommend as you know I'm big Advocate is a systematic review and sometimes called a scoping review

integrated review lots of different names for it I just call it systematic review and I aim to do that because

that's kind of the highest most well-respected form of a lit review and that takes a more formal approach and

doesn't search things like Google Scholar which put you in a filter bubble you know if you're schooling Instagram

or YouTube the algorithm showing you things it thinks you want to see that's not great that's not scientific because

as scientist we want to do things that if I put in Search terms then divage will get the same search with those same

terms or Jessica will get the same thing with those terms that's reproducible that's scientific and that that's what

you want to Aspire for so there you're going to look at databases so if you search with these keywords on this date

no matter who in the world does that you're going to get the same thing now those search is you'll often build up

from keywords iteratively and again there's there's some training around that I think it does take a few seconds

to learn how to do that but once you've got it and master it it's a valuable skill to have in your kit and it forces

you to be very clear about what you're looking at so it's harder to go off and get der railed on the road to finishing

your literature review on that systematic path but you you then either way need to have a search strategy know

which databases you're going to search know what's in or out and kind of Define and articulate something you're going to

want to write in your literature review anyway what keywords you want to put in all right now notice I didn't inject an

AI tool here and that's because AI is not helpful here AI is partial and the way AI searches can kind of work is

there's a tool I recommended just recently in another video called scace it does have a l review Search tool but

it kind of snowballs and it optimizes and gives you top sources and it can be useful for getting the sense of the lay

of the land but it's searching a partial set of the literature and it's like well why would you want to go search for

scientific knowledge with only one eye open not really a good way to go I don't recommend it but AI is going to enter

and make it come back in our next stage which is the next Horseman of the Apocalypse is uh knowing how to analyze

the data and write up and I think a lot of students struggle here because they just try to write up okay I got the I'm

I'm going to vomit on page and write up that could be useful but most people don't really vomit very

effectively I mean I'm I'm only partially joking you you need to have a sense of what you want to say before you

write rather than I'm going to figure it out as I write which is what many of you may have done in assignments before

there's a level of thinking where you could say okay I found these papers and I'm just going to summarize what East

paper did block by block well that really is something like a four or five-year-old could do that's not higher

order creative thinking the real power of thinking in a literature review comes through synthesis when you have say a

set of Articles and suddenly you can see things that nobody else could even imagine before because you start seeing

patterns emerge across those articles you start seeing Trends and currents and that's really helpful that's the real

value because that's where you start to spot gaps you say hey all these guys maybe looked at depression with this

index that's not that great and they're not getting good findings but these few studies that did use a good measure of

depression founds a completely different result if you're just kind of looking at the trees one to one each individual

article you won't see that b but when you zoom out you see something different uh personal experience one of the bigger

findings that um came in my own personal professional work uh was when a lot of people were having to debate it was time

of recession is unemployment bad for your health and some studies were done and they were find no unemployment is

actually good for your health paradoxically how can that be well people were at home they're going to the

gym or they're cooking at home and and they're getting healthier other people are finding no unemployment is terrible

people are are drinking too much and they're killing themselves and they're stressed out of their minds and they're

on the couch eating potato chips and it's like well how do you reconcile this well as I look closer at each individual

study said well these are just contradictory but when I took the whole map of studies out there I found wait a

second the studies in Sweden and Finland and places with strong social safety nets and support systems they're getting

completely different patterns than those where people they lose their job and they're just poor and completely exposed

to their life falling apart like in some parts of the US for example uh where I'm from um and that's about literature

review synthesis and so I want you to get to those results and this means really getting deep intimate and dirty

with your papers I recommend going so far as extracting knowledge if you're doing systematic review you might

extract the core features about your research question into an Excel data sheet and I'll got some summary

templates for you to check out on how you can do that and you can see that in our live full systematic review

step-by-step training course that you're going to find if you go to playlists on my channel it's about 12 12 hours of

training uh but it will it will change your life I'm kind of biased I think it's great training but I've seen so

many students get great results and publish a literature review in a very short period of time is awesome and uh

I'm I'm really proud of that and uh you're going to love it so you need to get that that Clarity what that involves

is something called a strip method you need to strip information out of these papers the key points key things that

found in relation to your topic and and that big overarching theme and question the impact of what what of your paper

and start lining those up togethers and you know twist and spin and perturb the data and look at at different angles

until you start seeing some meaty thick key insights now when you write up you want to focus the literature review and

create a structure around those big themes okay again notice I haven't talked about Chad GPT in any of this and

none of this now what you can do that can be quite powerful now we can come back to scace which can be quite helpful

I'm going share my screen again so here we go scace fastest reach search platform ever has a function where you

can come here and you don't want citation generators you don't want to paraphrase you don't want to have it

right for you you don't want to do an AI lit review this is all going to get you in trouble and it's just going to make

you feel like a fraud don't do it but go to extract data and uh drop in your PDFs in your inclusion here and this can help

analyze and cluster some themes I would rather have you identify them on your own when you drop all these in there it

will give you some ideas about main themes that you can explore let me show you an example this how how you can use

these main themes I've got a set of the articles I use for a literature review on uh the impact of commercial factors

on sexual and reproductive health and so just come here upload your literature review files again this is not a

substitute for you doing this yourself but it can be an extra check and here you go you can engage you can chat with

any individual paper um and you can pull out specific features which is doing now the method used the results this could

be useful for extracting data um but you can ask some questions here and again you can unlock high quality if you pay

for it um you get more tokens if you pay for it we've got a discount below David S 40 40% off David S2 for 20% off uh

annual or monthly again I'm not sponsored by them I find this is a helpful tool we've brokered that

discount for our FastTrack students but say here um how new commercial factors impact on

sexual and reproductive health and it will give you some ideas it will give you some things that maybe these

become subheadings for your outline in the structure of describing your literature review and so I can see here

there's some stuff about conflict of interest so maybe I'm going to summarize some things about conflict of interests

maybe here there's a problem with uh the or the role of the pharmaceutical industry now ultimately the big point is

you're going to need a framework for summarizing your key points so that they can really Bloom to the four because

when you do a literature review there's going to be so many details so much information your job is to highlight the

big points the big messages and there's no simple magic formula for that but um by breaking down and analyzing the data

you will see these themes and Trends start toer of the four and you can use AI tools to help give you this check so

here you can see almost as well an idea of how an extraction can help you to see patterns now you can see this is methods

used it's not really doing a good job of extracting the methods the methods might be the method here is case studies well

we did case study on May Mar's Community Based midwives so it doesn't really extract that well let me show you

another example of what a formalized extraction table looks like so so you can understand and I'll show you one

that one of our students has been working on so here's an example of an extraction sheet where you might strip

out specific components well what was the study design here was a paper looking at inequalities what racial

comparisons were being made what was the inequality measure what did they find and again it's just forensically going

through and pulling out key elements so that when you look across the papers you this student could start seeing patterns

oh well the East Coast had higher inequalities than the West Coast the inequalities between black and white

were worse than those between Hispanic and white other insights that you just couldn't see by looking at one paper

alone a more formal way of doing this just for your own knowledge as a side note um is called meta analysis when you

might go so far as using these articles as data and turn them into a quantitative data set that you fully

reanalyze most of you that's more advanced you could definitely do it but get the review done and then let's move

on to some of these more Advanced Techniques where we integrate quantitative tools or other Advanced uh

analytical instruments sometimes you don't even need them to get a big payoff um in terms of what you find no again

that's the principle AI can help but is it necessary no can it help you go faster if you don't use it in the right

way you could even feel more lost and lose more time finally with this writing up you need to have a system for rioting

and so many students maybe they got taught some rioting in high school but by now they've long forgotten it and

they're just riding they have no idea what they're riding why they're riding what goes where so fundamentals you need

structure you need outline of course we talked about that but we recommend a pier system and the pier system just

breaks down the most basic unit of an academic paragraph where each paragraph makes one point and is broken into pier

and we call this like a mini hamburger model because it's kind of like hamburger with a bun and the meat in

between and then the other bun at the end and so the P of Pier is the one big Point each paragraph should make one big

point then stuffed inside your burger you're going to have evidence and examples your Tu appear um evidence

examples so um right uh evidence examples pack that in there and you need to explain it and then you come to an R

which is your repeating or linking point to flow into the X paragraph I've got other videos where I go through lots of

academic text showing this format showing the system and how it's common again and again and again makes it

really easy for your readers to follow because your big point of the paragraph is going to be right up at the top so

for example you might have a paragraph that says you know the war in Ukraine has been devastating okay well you kind

of know what that paragraph is going to be about what's coming next well you can even foreshadow you know well I'm going

to give you some evidence and examples of that over 10 million people have lost their lives are making up a number no

offense to anybody wherever you stand on this side of the war this is not a political Channel I know it's impossible

to be completely a political but I don't want to go through a long discussion about Fuko all right so uh yeah so

evidence examples 10 million people have died maybe the explanation is the deaths of come not so much from direct you know

War casualties but indirect consequences like damage to sanitation sewer systems and Associated knock on consequences

like infectious diseases and if this now you've got your repeating or linking sentence and if this war continues the

consequences will likely grow larger if right I don't know NATO doesn't take urgent steps to intervene and that might

flow to your next paragraph then talking about the global response to the war but notice that keeps things simple I see so

many hamburgers that are just like stuff I got some bacon and egg and beetroot and mushroom and like just too much

stuff in the burger it's not tasty anymore um so keep it focused keep it lean and mean one big Point each

paragraph you need a writing system this is again where chat GPT can help you and in the background I'm working on

creating a custom GPT for all of you that you can use that embeds our Pier system but take your paper and your

draft and uploaded chat GPT and ask it to follow Professor stuckler Pier system to make sure it's fing good habits of

academic writing have it edit for style not substance so you need to get that prompt right in chat GPT style formal

academic writing not substance you don't want it to change the content that's what you need to preserve yourself when

you changing substance that's when AI detectors like ding ding ding ding ding going off and you get in trouble by the

wayspace has an awesome AI detector go check it out a l them have false positives I got a video where I review

that below all right then we are coming to our I know thanks for sticking with me guys this has been a long video but I

hope it's a helpful video submitting those of you who want to submit to a journal or even you're submitting for a

grade submit like a pro don't make rookie mistakes and what are rookie mistakes when professors like me are

reviewing your paper whether it's for prer review or as an editor sending it to a journal or for a grade we're busy

we're going to skim it and we're looking for indicators quickly of quality so if you got a spelling error in your

title probably have done a cursory not very thorough job and it's already primed me to make a judgment the reality

of the psychology is of how people engage with these things this has been studied elsewhere I won't go into that

is they read quickly and make a relatively snap judgment about where it's at so for peer review is it I'm

going to send this for a review or not am I going to accept this or reject it is this an a paper is this a c paper and

then they go look for reasons to justify that decision so especially make sure your front matter does not have howlers

what I call howlers it's like oh man it's so painful it's just screaming off the page how bad it is chat GPT can

really help with that run your paper through it and say please identify any major grammatical errors or typos get

rid of those there's no reason this day and age even if English is not your first language you can't have a grammar

checker I also highly recommend grammarly it's not an AI tool but it is like bread and butter tool that you

should be using especially if English is not your first language free version is more than good enough for what you need

and will catch those grammar mistakes the second thing if you're going to submit to a a journal is to know how to

write a cover letter and know how to choose reviewers to recommend and not recommend why is this important I get a

lot of students saying that I need somebody famous and actually I don't don't recommend that that can saddle you

with problems of a different kind having somebody famous on your paper because they have a reputation for better for

worse and you're just latching yourself onto it so really the issue comes in submitting like a pro like you've done

it in a million times before and the first thing that the editor is going to look at and really notice well it's

going to be the title it's going to be the abstract uh of course but it's going to be your cover letter cuz that's when

you're speaking directly to the editor and explaining the value ad of your paper if you've never written a cover

letter um check out our dedicated training on submitting which will help you identify the right journal to submit

to how to write a cover letter we've even got a dedicated video with the executive editor of the quite possibly

the most important medical journal in the world the Lancet I'm sharing her tips and then those tips apply across

social sciences across Natural Sciences across Humanities across the piece really fantastic tips on how to stand

out and get picked up and increase your chances for success I mean what a waste to spend all this time and effort to get

to the Finish Line only to like run out of gas and fall flat and get kicked out because you made rookie mistakes don't

let that happen to you also don't forget use your reference manager get your references in line it's just something

that you shouldn't mess up it's just too easy to let that Horseman of the Apocalypse get you guys like gotcha at

the very end because you didn't get your references right click of a button with a reference manager you're going to

format everything to looking beautiful and adjust for whatever you're submitting many programs will require uh

APA MLA Harvard whatever Vancouver formatting click a button you're going to be doing everything just the way it's

supposed to be done with without wasting endless time and energy formatting guys all right that was a lot of ground we

covered in a very short period of time I am sure you're going to have a bunch of questions and lately in my Facebook

group FasTrack grad we've got lots more master classes that go in dedicated topic with live student examples on each

of these topics on how to conquer these four horsemen of the literature review apocalypse and you can submit video

questions be in direct touch with me right to our communities of practice these are professors researchers grads

all across the board collectively working to try to help each other produce the best quality research

possible it's 100% free and valuable and again we can be in direct touch in the DM so if you are trying to get a paper

done fast that is a tremendous resource for you click the link below and check out my next videos If you are at the

writing stage that go in detail of how to write the discussion how to write the introduction hints save the introduction

for last more than that later um how to write the methods and will look forward to seeing you there