I also spent time looking specifically at how a literature review should be structured because this was the section I was most at risk of turning into a list of papers.
My first instinct was to work through the sources one at a time: explain what one author says, then explain the next. The guidance I read made it clear that this would describe the literature but would not necessarily demonstrate a strong critical understanding of it.
The literature review needs to be organised around ideas, themes and debates, not around the order in which I found the sources.
This was probably the most useful structural change.
Instead of creating a paragraph for Doyle, a paragraph for Dwyer and a paragraph for Bhattacharya, I need to decide what argument I am trying to make and then bring the relevant sources into that discussion.
For example, a paragraph about independence can combine research on autonomy, self-determination and lived experience. A paragraph on sensory accessibility can compare evidence around sensory reactivity, predictability, perceived control and opportunities for recovery.
The paragraph should belong to the argument, not to the author.
I also noted that stronger paragraphs often bring two or three sources into conversation with one another.
This gives me a way of showing whether researchers agree, whether one study extends another or whether there are tensions between their conclusions.
That is much more useful than repeatedly summarising individual papers because it lets me demonstrate that I understand the wider field rather than only the contents of each source.
A literature review should not create the impression that every source agrees.
I need to identify where the evidence is inconsistent or where different methods produce different conclusions. I also need to ask what questions remain unanswered.
This is particularly relevant to physiological sensing. There is evidence that signals such as HRV can provide information about autonomic activity, but there is not a simple or universal relationship between one biosignal and sensory overload.
That unresolved area is not a weakness in my report. It is part of the research gap and helps explain why the project needs further investigation.
My notes also reminded me to look beyond the authors' conclusions.