Picking a topic
Pick a topic broad enough to have literature on it but not so broad that summarizing the conversation around it would require a textbook. Pick something relevant to the field of study. Pick something you already have some familiarity with (e.g. you studied it as a special project in undergrad).
Finding sources
Make sure there are enough resources with enough similarities and differences to provide good context for the topic. It’s often better to look at topic-specific article databases versus using something general-purpose like google scholar, at least to start. Use keyword searches with wildcards and booleans as needed. It’s sometimes better to search abstracts than titles, both because titles are sometimes clickbaity and because keywords may appear in the abstract but not the title.
If you find a solid paper, check its bibliography to find more foundational works, and look at what other papers cite it Google Scholar is actually useful for this). Try to find foundational works on the topic but also new niche innovation papers.
Consider whether it makes sense to look for peer-reviewed articles vs other sources like blogs or autoethnographies. Check title, subject headings and resource type, and add entries to a backlog for evaluation. Aim for double the number of entries you actually need, since many will be rejected.
Evaluating backlog entries
- Use title and subject headings as a first pass determination for whether to add them to the backlog.
- Pick an entry from the backlog and skim the abstract and optionally the conclusion, write a brief note of the contents, and indicate whether a full read is warranted.
- Assign the paper a score (1-3, 1-5, stars, whatever works)
- Note the number of entries in the bibliography
- Note the number of times this paper is cited using Google Scholar, Web of Science, SCOPUS, or JSTOR.
- If abstract or conclusion reference papers that sound like they might be more useful, add them to the backlog.
- Once all backlog entries have notes, begin deep reading entries that scored 3, taking deeper notes and relating them to other works.
Strategies for picking the next backlog entry to evaluate
- Pick one that stood out during selection
- Pick the oldest one
- Pick the newest one
- Pick one at random using a PRNG
- Pick the one that inspires the most dread
- Pick the one with the highest citation count
- Pick the one with the longest bibliography
- Pick the one with the highest post-initial-evaluation score
Neurodivergent and other advice
- Use any tool that works for you, or combination thereof. Trello, docs, a wiki, physical notebook, zotero, jabref, etc. Start many, lose interest and switch, use multiple half-heartedly, whatever works for you in the moment.
- Stick to the evaluation process to avoid getting sucked in/going down the rabbit hole.
- Use pomodoro to evaluate material for a fixed time and then take a break.
- Try to be objective and accurate when citing and paraphrasing the work of others.
- Conversely, don’t be afraid to ruffle feathers by pitting authors against each other, just don’t do it yourself.
- It doesn’t hurt to stroke egos when possible.
- Don’t agonize over making people upset or being too aggressive due to developmental trauma. Especially if the professor is the only one who’s going to read it.
- Identify areas of consensus within the literature.
- Identify areas of contention, disagreement, or gaps in the literature.
- Identify the communities for whom the disagreement or gap is a problem.
- Use the lit review to establish credibility: you’re familiar with the existing conversation.
- Use the lit review to create context: set the stage and identify the players.
- Use the lit review to create a narrative that draws the reader into the problem.
Pages that link here: