For those interested in peer reviewing other research this took me ~ minutes all in. Today I'm going to peer review this study by Ryan Jones and Sapient Nitro on Twitter and offer some opposing contradictory and better research. Background Here's the study I'm going to discuss and I'm just going to be honest it's problematic research and here's why. It didn't deal with basic data processing like removing stopwords and other common words. This means that the most common speech fragments emerge in the study not insights from keyword choices.
Although it was later argued that the safewords were the point I honestly don't see why that would ever be the case and without more effort from the authors here mobile number list I don't think this is a good justification. For theme classification stop words are useless. Anyway here at LSG we use the NLTK library to store our datapreprocessing and removing stop and other common words is a basic application of that library. Without proper processing and cleaning of the data no insight is valuable remember GIGO. The data set. BrightEdge doesn't have a very good dataset and they're not very transparent about how they get it.
Honestly I could spend all day criticizing their platform and service offerings but that's beside the point. If you are going to analyze a keyword set that will be representative at best then you need to make sure that it is as accurate a representation of the actual data as possible . So if BrightEdge has a less representative keyword corpus than say AHREFs that would mean that the insights cannot be trusted GIGO again. Fortunately here at LSG we know how to cut things like stop words and other common parts of writing when dealing with large amounts of data and I was able to get what I think is a better keyword set to use in the research.