Social Media for Men since 1964
  • How can I do social network analysis projects on my own?

    This actually depends a lot on what sort of data you need, what types of analysis that you want to do, etc.

    On quora numbers, I do social media analysis about, well, Quora. For the most part, I do this with only Excel and simple statistics like mean, median, mode, percentiles, quartiles, minimum, maximum, pearson’s correlation, slope. I make dispersion graphs. Because of the limits of data collection because of 1) no Quora API and Quora not allowing bot scraping in this matter, and 2) Lack of programming skills, this normally works. The constraints are smaller data sets. The methodology clearly spells out the limits. I can do a whole lot of different tests and look to see if these smaller tests generally replicate various results to confirm findings.

    I sometimes supplement this An Open Source Platform for Complex Network Analysis and Visualization as a way of doing network analysis. I’ve also used NodeXL: Network Overview, Discovery and Exploration for Excel. In the past, I used GeoCommons I sometimes use to do fund things with geographic data. Web Scraper or DataMiner can be used to help get some social media data. (Huge limits on what both can do with Quora. There really is not a feasible way to get data out of Quora without Quora giving it to you or having programming skills to get it but not get blocked.)

    I’ve taken an educational statistics class. I understand enough of what Z Numbers, Regression Analysis, Probability, P-Value are that I can get what people say when they are doing that… but for most of what I want to do, well, it isn’t required as I’m more about using multiple methodologies to arrive at a conclusion for how things works. If you’re scared of the math, try something like Educational Research: Competencies for Analysis and Applications (10th Edition): Lorraine R. Gay, Geoffrey E. Mills, Peter W. Airasian: 9780132613170: Amazon.com: Books . I like educational statistics because I feel like the goal is more about doing research and then putting theory into practice. Educational approaches also rely more on mixed methodologies. For me, and I’ve been paid a couple of times to do social media research, having that knowledge is all you really need. Know what these numbers are and what they say. Know when to use them. Know the constraints of your data. Know your audience. The math does not need to be intimidating. Simple math can still get out valid, actionable results.

    How you do it gets a bit complex. First you tend to device a research question. Then you think about the methodology. If you’re not a programmer, you then think about how to get the data in a way that does not eat days of your life, costs you little and still has meaning. After that, you then revise your methodology or get the data. Then you analyze it. Then you write up your results, and follow it up with a conclusion. In between and possibly during every step of the way, you may want to consult with texts that explain the methodology or explain more about what you’re researching. This can be something like Page on wlv.ac.uk or Getting the Hang of Discourse Theory – PoliticsEastAsia.com to Page on sagepub.com to Wisdom in the social crowd to Research:Communicating on Wikipedia while female. It just depends on what specific methodology you’re doing and what you’re researching.

    So yeah. Don’t let the math scare you. Understand what those numbers mean. Understand how to get them. Then ask if you really need them to convey what you want to convey while you’re researching what you’re researching. (Had a fabulous discussion with my supervisor about p-value, what it means, and when we should actually get that number.)