Useful templates (LaTeX, Python, ROOT) and general guidance here: JTS Public Dropbox

• Use a citation manager:
  • Readcube Papers ($$)
  • Endnote ($$)
  • Zotero ($0+)
  • Mendeley ($0+)
  • Jabref ($0)
• Download the citation file directly from the journal and into your citation manager
• Start a literature review using LaTeX where you can store notes on this paper and, crucially, related papers:
  • MikTeX (Windows)
  • TeXShop (Mac)
  • Overleaf - note: I pay for all the features for Overleaf so ask me to join as a project owner to provide you with access to all the features
• The content of the paper is just as important as how that paper fits into the greater scientific literature on that topic
• Do not read the paper word for word on your first pass (unless it is very short and does not follow the conventional format)!
  • Pass 1: Read the title, author list, abstract, introduction, section headings, figure captions, and conclusions
  • Pass 2: Scan the references for any papers that you've already read and scan the list of papers that have cited this paper
  • Pass 3: Determine the three things that you want to extract from the paper and then skim the paper looking for those three things
  • Add this paper to your literature review by (1) paying close attention to how this paper fits into the bigger picture, (2) noting what you learned, and (3) noting three things that still don't make sense
  • Pass 4+: if necessary, read the paper word for word - but - have an agenda for the information that you are looking for

• Guidance on writing abstracts is here
• Sample presentations:
  • Bad example (PDF)
  • Good example (PDF)
  • EDM talk (PPTX)
  • SAM poster (PPTX)
  • SAM talk (PPTX)
• How to start writing a talk/presentation
  • General guidance on the structure of the narrative is here: 2022-03-01-hints-how-to-scientific-presentations-jts
  • Presentations should be 1 slide per minute (roughly and this includes the title slide and conclusion slide)
  • Pass 1: The title of each slide should be a complete sentence (with proper grammar and punctuation) that indicates the main take away message for that slide that is suitable for the intended audience
  • Pass 2: Think about the figure that best illustrates that sentence and put in a placeholder
  • Pass 3: Add supporting text
  • Pass 4+: iterate with constructive feedback and revise for legibility and visibility

• Learn a tool for making publication quality figures (diagrams and plots)
• A diagram shows how the experiment was carried out
  • Adobe Illustrator ($$)
  • Inkscape ($0)
  • Save the diagram in EPS or SVG format so that it does not look pixelated when changing sizes
  • Some things to consider when designing a diagram: 2019-10-02-conceptual-diagrams-phy-451-jts
  • Designing a diagram is an iterative process which requires constructive feedback: 2017-08-08 Blog Post
  • Another example of the iterative process from the PHY 451 Advanced Lab Class is below:
    • pictures of Diode Laser Spectroscopy Experiment:
    • first pass student diagram:
    • first pass JTS diagram (page 2 of PDF): 2019-10-04-first-pass-conceptual-diagrams-jts
  • An example of a professional publication quality diagram is Fig. 1 from PRL 113, 013001 (2014) (see below, made with Illustrator):
• A plot shows the data extracted from the experiment
  • ROOT and Python+matplotlib are both free and excellent tools
  • Save the diagram in EPS or SVG format so that it does not look pixelated when changing sizes and can be manipulated in an illustrattion software (see above)
  • Basic required elements of a 2D plot:
    • x-axis title label and units in parentheses, units should be chosen such that the x-values are between -100 and +100
    • x-axis labels values are large and legible, should be the same size
    • y-axis title label and units in parentheses, units should be chosen such that the y-values are between -100 and +100
    • y-axis labels values are large and legible, should be the same size
    • title of plot should describe the relevant conditions under which the data was taken and possibly include the timestamp of the dataset
    • data markers should be closed red circles
      • if the data density is high (the points overlap), then the data markers should be open red circles
      • if there are two data sets, then the second data set should be blue crosses
      • if there are multiple data sets, then the color of each data set should encode information about the variable that was varied between data sets
    • data markers should include uncertainties in the y-values and possibly the x-values if relevant
    • data markers should not be connected with lines just to guide the eye
    • if you use curve fitting to model the data, then you must plot both the (data vs fit) as well as the residual (data-fit)
    • example Python code is here: Python Curve Fitting
    • examples ROOT code are here: ROOT example