Thread

1. socializer · 1790604534 · focus · HN ↗

   My impression is that this doesn't show anything. It essentially implies that brand loyalty is equivalent to paid astroturfing, but if you've been around hobby communities on the internet, you know that a lot of people are just brand blowhards. And in any case, by its own admission, none of the results in this article are statistically significant, so it fails to find the effect it implies to have found.

   I strongly believe that there's a lot of paid product placement and paid political talking points on Reddit, so it's actually remarkable that this (LLM-generated) article fails to find it. I suspect it's just bad design.

   If you're already spending money on tokens, there's a very simple experiment you could do instead: just pay one of these companies to position a made-up product and see what they do.

   1. edelbitter · 1790774265 · focus · HN ↗

      > I suspect it's just bad design.

      I don't have good suggestions on what would improve the design while staying with the general approach, but I imagine reporting a "likely not significant" could have at least be improved to a "proven to be not significant" null result by adding the "obvious" control group. Just check how "unlikely" mentions of common memes and catchphrases compare to those brands. Are there non-brand mentions that are distributed similar to the brand mentions, showcasing that these numbers really do not mean much either way?

Unofficial Hacker News client; not affiliated with Y Combinator.