We retrieved all tweets posted in the English language during 12 consecutive weeks, from September to December 2021, that contained any of the following six hashtags: #phdlife, #phdspeaks, #phdvoice, #phdchat, #phdtips, #phdstudent. We then measured the sentiment (positive, negative or neutral) associated with each original tweet (excluding retweets). Of the 91 229 tweets we retrieved, 43,941 were positive, 12,298 were negative, and 34,990 were neutral. Mann-Whitney U tests were performed to compare the average number of likes and retweets of positive versus negative tweets. Negative tweets received significantly more likes than positive tweets (14.5 vs 12.3; P<0.001); negative tweets were also retweeted more than positive tweets but the difference was not significant (1.7 vs 1.5; P=0.383). The Twitter API and the “rtweet” R package (cran.r-project.org/web/packages/rtweet/vignettes/intro.html) were used to retrieve the tweets; the “syuzhet” R package (rdrr.io/cran/syuzhet/) and the Bing lexicon (Liu, 2012) were used for the sentiment analysis; all analyses were performed with R software (R Development Core Team, 2021).