This assignment started off by reading an article that explains what “proportion of variance explained” means. After reading this, we had to examine a document full of tables that showcases the proportion of variance in internet use explained by personality traits. Next, we had to read 2 articles about Cambridge Analytica, about their rise and fall, and why they likely did not impact the 2016 election. Finally, we had to create a discussion board post, in which we explained what we learned about “proportion of variance explained”, what we learned from examining the tables, and if we think Cambridge Analytica could have done what they claimed to have done.
Honestly, for this assignment, everything I learned is very well summarized in my discussion board post, so I am including everything I wrote for that below.
From the first article of the assignment, by Miles, I learned that variance is essentially how much people differ on a dependent variable, such as how much someone uses Facebook. Then, you can try to predict a dependent variable value based on an independent variable value. So, building off of my other example, this could be like using a person’s extraversion score to try to predict how often they use Facebook.
So, moving into what “proportion of variance explained” means, you can compare your predicted dependent variable value to the actual dependent variable value, and see how close they are to one another. The closer these two numbers are to one another, the greater the proportion of variance explained. For example, if the extraversion scores led to a perfect prediction of Facebook use scores, all of the variance would be explained by that variable.
From the tables in the second document, the main thing I learned was that most personality and demographic variables do not explain much of the variance in variables related to internet use. For example, the big 5 personality traits (besides extraversion) often did not explain a substantial amount of the variance in internet use. Additionally, many other variables such as income, education, self-esteem, and many others often had negligible percentages when it came to amount of variance explained.
However, it is important to note that physical, verbal, vicarious, and overall sadism can explain a solid chunk of enjoyment in trolling, Additionally, factors such as age, gender, and extraversion can sometimes explain a slim to moderate proportion of internet use.
Based on the last two articles we read, I do not think Cambridge Analytica could have done what they claimed to have done. According to the Kris-Stella Trump (2018) article, many low opinions of the company were expressed, they recruited new clients through tricks, and their claims “quickly fall apart” when inspected.
Additionally, based on the Andy Kroll (2018) article, Cambridge Analytica used “unethical methods to obtain a massive trove of Facebook data to fuel its psychographic tactics”. Their methods included injecting propaganda, bribing politicians, and trying to entrap candidates.
So, reflecting on these two articles, I am confident in saying that the company likely could not have done what they claimed to have done. They did not win Donald Trump the 2016 election, and if they had any influence, it was minor, and unethical as well.
Reflecting on this assignment, I think the first article was helpful, because I was not quite sure what proportion of variance explained meant beforehand. I knew what variance was, but had not really applied that and taken it to the next step. So, after reading the article on that, I feel a lot more confident about what it is and will be able to utilize the concept in the future if need be.
Additionally, I think the graphs were a great thing to include after the article. After learning about the concept, seeing the graphs gave me a visual representation of the concept, as well as related it back to topics we have been discussing, such as personality traits and internet use. It was a great transition, allowed me to ensure that I understood what proportion of variance explained was, and could get ready to explain the findings afterwards.
It was interesting to learn about Cambridge Analytica, because I actually had never heard of them before. I was completely unaware that any company was thought to have influenced the 2016 election. It was interesting to learn about their ideas, but when I learned about their flaws, I was glad to see they had failed.
Finally, the discussion board post was good because it essentially touched on every single part of the assignment, making sure I really understood the materials I had learned from. We are not always required to write about all the sources we read from, but this required it, and even though it takes longer, I am more confident in my understanding of the material now!
Nicky's Blog 