After discussing best practices to review articles with Andreas Bulling and a few other colleagues at the Wearable Computing Lab I captured below my basic 'checklist' for reviewing articles.
- Is it in the scope of the journal/conference at all?
- Is it suited as a full paper or a short paper/poster would be
- Reproducibility: are all the details given to reproduce the results?
- Literature: are relevant areas sufficiently covered by litterature?
- Benchmarks: was the system compared to a state of the art system?
- Does it make sense: is this interesting power optimization really making sense given another much more power hungry component?
- Are the conclusions overstated or understated?
- Plots: in black&white; text/label must be readable
- Complete legends for figures
- If the english is really bad: "have the paper checked by a native speaker"
- Appropriate, and enough figures/schematics? Are complex aspects
of the paper simplified by a figure?
- Contribution to the authors
- I provide a few reference that are interesting for the paper
- Style/english corrections/typos
- A few suggestions for other places to submit the work