Comments on: The Confounding Variable Of The Retweet http://brandsavant.com/the-confounding-variable-of-the-retweet/ Gaining Insight From Social Media Data Thu, 09 Feb 2012 17:40:00 +0000 hourly 1 http://wordpress.org/?v=3.3.1 By: You Got It All Wrong: The Limits of Social Media Monitoring | BrandSavant http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-1416 You Got It All Wrong: The Limits of Social Media Monitoring | BrandSavant Mon, 14 Feb 2011 12:41:59 +0000 http://brandsavant.com/?p=525#comment-1416 [...] Well, it isn’t really your fault. I’m also not going to beat up on the tool – it was built to track mentions, and I have no reason to believe that it didn’t do an admirable job (so, Meltwater folks, stand down – this isn’t about you.) The issue here is that you were asked the wrong question. Tracking mentions for predictive purposes is of dubious value – in the short term, the volume of mentions is a random walk. [...] [...] Well, it isn’t really your fault. I’m also not going to beat up on the tool – it was built to track mentions, and I have no reason to believe that it didn’t do an admirable job (so, Meltwater folks, stand down – this isn’t about you.) The issue here is that you were asked the wrong question. Tracking mentions for predictive purposes is of dubious value – in the short term, the volume of mentions is a random walk. [...]

]]> By: Understanding Klout | BrandSavant http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-1142 Understanding Klout | BrandSavant Tue, 19 Oct 2010 14:53:40 +0000 http://brandsavant.com/?p=525#comment-1142 [...] If we are interested in the realm of influence beyond the easy, immediate measures (like the confounding variable of the retweet) then the amount of data mining required presents what may currently be an intractable [...] [...] If we are interested in the realm of influence beyond the easy, immediate measures (like the confounding variable of the retweet) then the amount of data mining required presents what may currently be an intractable [...]

]]> By: Twitter’s Most Elusive Statistic | BrandSavant http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-965 Twitter’s Most Elusive Statistic | BrandSavant Mon, 13 Sep 2010 14:23:17 +0000 http://brandsavant.com/?p=525#comment-965 [...] the vast majority of those followers realize the question is rhetorical. (I admit there is also the confounding variable of the retweet to deal with here, but engagement also comes into play in retweet [...] [...] the vast majority of those followers realize the question is rhetorical. (I admit there is also the confounding variable of the retweet to deal with here, but engagement also comes into play in retweet [...]

]]> By: “Unaided Recall” in Social Media Research | BrandSavant http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-892 “Unaided Recall” in Social Media Research | BrandSavant Mon, 30 Aug 2010 15:24:50 +0000 http://brandsavant.com/?p=525#comment-892 [...] the folks monitoring for those keywords will dutifully tally this up as a “mention” (which is, as I’ve noted previously, a dubious metric anyway.) However, it’s actually a little bit more than that – it’s an unaided mention. [...] [...] the folks monitoring for those keywords will dutifully tally this up as a “mention” (which is, as I’ve noted previously, a dubious metric anyway.) However, it’s actually a little bit more than that – it’s an unaided mention. [...]

]]> By: Tom Messett http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-418 Tom Messett Sat, 08 May 2010 13:21:18 +0000 http://brandsavant.com/?p=525#comment-418 Hi Tom, nice post, I used to work for a monitoring service so I know exactly what you mean about the RT issue and couldn't agree more! Anyway, I am really interested to learn more about your thoughts on the "Eavesdrop Pitch"? Hi Tom, nice post, I used to work for a monitoring service so I know exactly what you mean about the RT issue and couldn’t agree more! Anyway, I am really interested to learn more about your thoughts on the “Eavesdrop Pitch”?

]]> By: Dennis McDonald http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-416 Dennis McDonald Wed, 05 May 2010 16:17:25 +0000 http://brandsavant.com/?p=525#comment-416 Here are some additional concern about social media metrics, or rather, not with the metrics themselves but how they might be used in some cases: "OMB's New Guidance on Social Media is an Improvement - But There's a Catch" http://www.ddmcd.com/guidance.html Here are some additional concern about social media metrics, or rather, not with the metrics themselves but how they might be used in some cases: “OMB’s New Guidance on Social Media is an Improvement – But There’s a Catch” http://www.ddmcd.com/guidance.html

]]> By: Dennis McDonald http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-415 Dennis McDonald Wed, 05 May 2010 16:06:11 +0000 http://brandsavant.com/?p=525#comment-415 I'm not dismissing social media monitoring statistics; they can be of great use. My concern has to do with how easy or hard it is to relate the different categories they use to other (e.g., known or documented) population statistics. Having recently evaluated several different monitoring tools I thought they could be doing a better job on things like making their population categorization rules and processes more transparent. I’m not dismissing social media monitoring statistics; they can be of great use. My concern has to do with how easy or hard it is to relate the different categories they use to other (e.g., known or documented) population statistics. Having recently evaluated several different monitoring tools I thought they could be doing a better job on things like making their population categorization rules and processes more transparent.

]]> By: Tom Webster http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-414 Tom Webster Wed, 05 May 2010 14:48:21 +0000 http://brandsavant.com/?p=525#comment-414 Dennis - I'm with you, especially because I am currently a sample-survey-and-statistics guy myself. Yet I am reluctant to dismiss social media monitoring metrics, especially when half the US population is using social networks. Yes, the non-response bias is troubling, but I think the answer is the continued application of sound sampling techniques (properly weighting sources and samples, for instance) within the framework of a structured monitoring program. There's a there there, though yes, it needs some work :) Dennis – I’m with you, especially because I am currently a sample-survey-and-statistics guy myself. Yet I am reluctant to dismiss social media monitoring metrics, especially when half the US population is using social networks. Yes, the non-response bias is troubling, but I think the answer is the continued application of sound sampling techniques (properly weighting sources and samples, for instance) within the framework of a structured monitoring program. There’s a there there, though yes, it needs some work :)

]]> By: Dennis McDonald http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-413 Dennis McDonald Wed, 05 May 2010 14:42:28 +0000 http://brandsavant.com/?p=525#comment-413 Excellent post. I'm less concerned about the inconsistency of "sentiment analysis" than I am about the underlying population numbers and what they represent. As a former sample-survey-and-statistics guy I'm especially sensitive to what populations the measures of mentions and conversations are referring to and whether it's possible to relate the conversation metrics to known populations. Sometimes this is not a problem if you are primarily looking at an entry point into conversations. But if you're tracking, say, how well a government program or a new product introduction is performing with a particular population or market segment, the lack of a definable and measurable connection between conversation and population can be a big problem. Excellent post.

I’m less concerned about the inconsistency of “sentiment analysis” than I am about the underlying population numbers and what they represent. As a former sample-survey-and-statistics guy I’m especially sensitive to what populations the measures of mentions and conversations are referring to and whether it’s possible to relate the conversation metrics to known populations.

Sometimes this is not a problem if you are primarily looking at an entry point into conversations. But if you’re tracking, say, how well a government program or a new product introduction is performing with a particular population or market segment, the lack of a definable and measurable connection between conversation and population can be a big problem.

]]> By: Hugo from Zeta http://brandsavant.com/the-confounding-variable-of-the-retweet/comment-page-1/#comment-410 Hugo from Zeta Tue, 04 May 2010 13:46:53 +0000 http://brandsavant.com/?p=525#comment-410 Great post! Modern monitoring tools are simply not sophisticated enough to deliver quality sentiment scoring. That's why it's crucial to use some good, old-fashioned elbow grease when analyzing social media data. Great post! Modern monitoring tools are simply not sophisticated enough to deliver quality sentiment scoring. That’s why it’s crucial to use some good, old-fashioned elbow grease when analyzing social media data.

]]>