|
(This post assumes that one is trying to maximize likes. There are obviously different, and arguably superior, ways of using a social network) The value of analysis like this is proving, codifying, and quantifying things that are intuitively believed to be true. Or, alternatively, disproving those things. One might believe that "more hashtags" causes "more likes", and even have a model in which "more hashtags" causes "more likes", but that doesn't mean that it's actually true. Before doing the analysis, the result could have come out either way. The author could have found that "more hashtags" did not cause (or at least, was not correlated with) "more likes". In this case, one could have taken the action of not filling in 30 hashtags, saving effort. And, for that matter, there's a lot of variance in average likes that's not accounted for by hashtag count. Further analysis might actually prove that some other factor (such as age of poster) influences both like count and hashtag count, and accounting for that other factor, hashtags might actually be useless or detrimental. ----- As a general philosophy, intuitively knowing "something is true" is not nearly as good as being able to point at data, and show general evidence that the thing is actually likely to be true. |