3 ms·
It’s a fair comparison … just because your own suspicions are confirmed doesn’t mean it’s not one. To be honest, I find this criticism baffling and confusing.
by arrrg 11y ago
It’s a fair comparison … just because your own suspicions are confirmed doesn’t mean it’s not one.
To be honest, I find this criticism baffling and confusing. I would agree that I would want to dig deeper now and figure out the exact variables that explain the difference – but what we have here is an extremely compelling starting point and an extremely interesting difference.
- morgante 11y agoHow is it a fair comparison though? In a study, you should clearly express what variable(s) you're changing with each variation and only change those. In this case, the author varied both the number of messages and the medium. That makes it impossible to say which is actually the cause of the effect, and I speculate that the number of messages is the better explanation. It's like comparing the reliability of USPS and UPS by sending 4 different packages on the former and 1 on the latter. That's obviously not a valid study.
- arrrg 11y agoThere are many hundreds of variables that are different between the variants … you chose to focus on this one for some weird reason. Such is comm research. There is no other way, there really isn’t. This makes sense as a starting point.
- ncallaway 11y agoThe author did not vary the number of messages. The author sent the same number of messages in both cases. Facebook's "Full Conversation" view automatically collapses multiple messages together, but the author was explicit in that they sent the messages in exactly the same fashion: "As well, there have been a few comments on how each messaged as typed. Please note that both Facebook Chat and Facebook Full Conversation allow you to send a new message on pressing enter, which I did in both cases. The input was identical, the visual output was not."