Reading Papers in Undergrad

I went to school for elec­tri­cal en­g­ineer­ing, and in my cap­stone elec­tro­mag­netic the­ory course I en­coun­tered a prob­lem that I needed to read pa­pers to learn about. I asked my pro­fes­sor for his ad­vice on how to read them, ow­ing to my lack of ex­per­tise, and he recom­mended the fol­low­ing:

Don’t read the ab­stract. This is writ­ten by ex­perts for other ex­perts, so they can quickly see if it is rele­vant to their work. It will provide no use­ful in­for­ma­tion, and may in­timi­date or con­fuse you.

Do read the in­tro­duc­tion. This will con­tain an overview of ex­per­i­ment, and most im­por­tantly de­scribe some of the back­ground and mo­ti­va­tion for the ex­per­i­ment. As a non-ex­pert, this con­text is at a pre­mium. If the in­tro­duc­tion does not sug­gest there will be use­ful in­for­ma­tion, you can stop here.

Skip the body of the pa­per. The meth­ods and con­struc­tion of the work are usu­ally not the goal of stu­dents, and this part is the most dense and re­quires the great­est ex­per­tise.

Do read the con­clu­sion. This is usu­ally the meat of what a stu­dent is look­ing for—what do ex­perts think, in their own words, about their work. With this in­for­ma­tion in hand, you can de­cide whether you need to go deeper into the pa­per or have what you need. If you de­cide to go deeper:

Re­turn to the body of the pa­per. This is es­pe­cially use­ful if you are greatly sur­prised by the con­clu­sion, or if you need to be able to refer­ence images or graphs in an in­formed way. Even as a non-ex­pert you can prof­itably think about how the con­clu­sion fol­lows from the steps they have laid out. It is nat­u­rally re­quired if you want to at­tempt to du­pli­cate the ex­per­i­ment.

I have had pretty good suc­cess with this method, even in sub­jects for which I have less than un­der­grad ex­per­tise. I will add two ad­di­tional steps I found nec­es­sary, both ob­vi­ous:

Reread the pa­per. As you fol­low up on the prob­lem with ad­di­tional read­ing, you will gain im­por­tant new per­spec­tives and con­text that led to miss­ing things the first pass.

Fol­low-up the refer­ences. The as­sump­tions about how dili­gent read­ers are with fol­low­ing through vary; this can make some­thing ag­gra­vat­ingly opaque sud­denly clear be­cause it was cov­ered well in the refer­ence, or re­veal a gap­ing hole the au­thors failed to ad­dress.

Anec­do­tally, the prob­lem I was think­ing about was the elec­tri­cal sig­na­ture of can­cer. I had found a few mod­ern refer­ences to work done on the sub­ject in an­i­mals, but couldn’t lay hold of work on hu­mans for the life of me. Fi­nally on the third pass over a sur­vey pa­per cov­er­ing re­lated work I saw an ob­scure refer­ence, and fol­low­ing through with that led me to all of the work which had been done on hu­mans pre­vi­ously.