Read on Twitter
A mini-research question, asked and answered, on Sat night:
Does the Feenstra et al. "it& #39;s good to be first" effect hold for the tweets about WP by @nberpubs ?
Answer: Yes, weekend papers get significantly more attention. 7.3 more likes (33% of weekday) and 3.9 more RTs (32%).
Does the Feenstra et al. "it& #39;s good to be first" effect hold for the tweets about WP by @nberpubs ?
Answer: Yes, weekend papers get significantly more attention. 7.3 more likes (33% of weekday) and 3.9 more RTs (32%).
Papers appear to be smoothly posted across days of the week -- there is no significant difference in the number of posts on the weekend vs. weekdays.
While there ARE differences across the days in when things are posted ...
this doesn& #39;t appear to be driving the difference in likes and retweets.
There& #39;s no identified whether something is a working paper or a call for papers, so there is definitely noise in the dataset. However, all these results are robust to using medians instead of means, suggesting it& #39;s not just a few large CfPs driving the results.
An important fact: I& #39;m fairly certain that the order of these tweets is random (which is what they switched to with the mailing list). So, there& #39;s no obvious change that should be done. But an interesting thing to note!
A few notes on the data:
1) Dataset is last 3200 tweets
2) Dropped one tweet of a WP which skewed the sample ( https://twitter.com/nberpubs/status/1173272766151180288),">https://twitter.com/nberpubs/... for graphing purposes.
3) Goes from today back to 2017-09-12
Code is here: https://github.com/paulgp/paulgp ">https://github.com/paulgp/pa...
1) Dataset is last 3200 tweets
2) Dropped one tweet of a WP which skewed the sample ( https://twitter.com/nberpubs/status/1173272766151180288),">https://twitter.com/nberpubs/... for graphing purposes.
3) Goes from today back to 2017-09-12
Code is here: https://github.com/paulgp/paulgp ">https://github.com/paulgp/pa...
@johnjhorton you might enjoy this.
