More Women in Education Means that More Women With B.A.’s Are Marrying Men Without (Shocking!)

In Alfred Lubrano’s article “More women with college degrees are marrying men without B.A.’s,” the author details a few heterosexual couples where the wives have bachelor’s degrees and the husbands do not.

The author uses U.S. Census data for the Education Gender Gap to show that “since 1990, the percentage of young women with a college degree has grown faster than the percentage of young men.” The data visualization appears accurate, except that it goes from 0% to 50%, rather than 0% to 100%. Because of this, the gap between women at 46% and men at 37% looks significantly larger than 9%.Screen Shot 2017-12-03 at 10.15.57 PM.png

The author then uses U.S. Census data for the Education Gender Gap in Local Counties, but I’m not from Philadelphia, so I don’t know where these counties are, and I don’t know how applicable this data is beyond those select counties.

Screen Shot 2017-12-03 at 10.16.06 PM.pngLater on in the article, the author references race/ethnicity and says that “[a]ctually, while more black women than black men aged 25 to 34 have a bachelor’s degree or higher, according to 2016 Census statistics, the nearly four-percentage-point difference (26.5 to 22.6) is half the size of the figure for all of the United States, said Temple sociologist Judith Levine. What’s different is that black women have always been more likely to get a college degree than black men, as far back as the 1940s, Levine said. Racism and high incarceration rates are among the reasons, she said.” These two sentences are the only sentences about race and it feels like this idea was just interjected and not well thought out.

After race/ethnicity is brought up (and then not evaluated), the piece immediately ends with an interviewee’s take on education: “Whether a man has a degree is not one of my first questions. At this point, life-education outweighs a degree.” I think that this conclusion for that specific person, but it is not an all-encompassing conclusion that is applicable to anyone even beyond the counties of Philadelphia.

This article has the cultural implication that while it is good that women are receiving more bachelor’s degrees than men that there is going to be even more of an intellectual disparity from now on.

Is Trump Like Manson? Or Are Republicans and Democrats Belittling Each Other Yet Again?

A couple days ago, Charles Manson, the leader of the “Manson Family” cult, died of natural causes in prison. While scrolling through the news on my phone, I came across a Fox News article “Once-reputable Newsweek compares President Trump to Charles Manson.” On a side tangent: I’ve subscribed to this source in an attempt to go against my confirmation bias, but I fear that it’s just making it worse because every time I read a Fox News article, I cringe and/or sigh. Anyway, I clicked on Brian Flood’s article because I wanted to see how Newsweek compared Trump to Manson, why this suddenly makes the new source untrustworthy, and just how badly Flood was going to condemn the author of the original source (“How Murderer Charles Manson Used Language To Gain Followers”), Melissa Matthews.

Flood claims that “Newsweek continues to embarrass the legacy of the once-reputable publication with anti-Trump hit pieces on a regular basis;” Is this why Newsweek is no longer reputable? Because they’re a politically left-leaning news source? Flood is clearly upset that “[t]he left-leaning publication literally compared a convicted mass murderer and cult leader to the President of the United States. Former President of the American Psychoanalytic Association Mark Smaller told Newsweek that part of Manson’s power lay in the type of language he used.” Flood quoted Smaller as saying “A charismatic leader knows how to speak to people in a way that will emotionally engage those people.” While comparing Trump to Manson is a dig at Trump, it seems blatantly obvious that any “charismatic leader” (even though I’m hard-pressed to say that Trump is charismatic, he is in a leadership position) would “[know] how to speak to people in a way that … emotionally [engages] … people.” This does not seem like a stretch at all. In this capacity, Smaller could have related Manson to Gandhi or Trump to the Dahli Lama, even though neither of those comparisons seems like a great idea, it is possible with those standards of “charisma.” Smaller then said that “we can look at the current position to see how language is used to form a bond with followers.”

Flood then attempts to give his article validity by quoting Dan Gainor, vice president of the Media Research Center: a center that documents “liberal media bias.” Gainor dubbed the piece “clickbait” and said that Newsweek “stopped being a legitimate news organization years ago.” In addition, Dan Gainor also claims that “[i]t’s almost predictable that same idiot online will write an outlandish story connecting Trump to the villain du jour, all designed to generate web traffic and ad dollars.” It’s ironic that Matthews’s article is ostracized for generating “web traffic” because it’s “outlandish” when Flood’s article generates that much more web traffic to Matthews’s article. Regardless of anyone’s political affiliation or personal feelings towards a writer, it is abhorrent to call a writer “some idiot online.”

When I went to go find out exactly how Melissa Matthews, author of the Newsweek, related Trump to Manson, the article had been changed because it “did not meet Newsweek’s editorial standards and [had] been revised accordingly.” So I have no knowledge just how bad the original article, but I do know that Mark Smaller’s claims do not appear to be outrageous or accusatory.

Significant (!) Biases

Both Brian Flood and Melissa Matthews had blatant biases that impacted what they wrote and why and how they presented the topic. I believe that each of their respective biases is significant because it (likely) impacts who their audience and whether or not their audience will agree with them. Their biases make sense to write right- and left-leaning articles for Fox News and Newsweek respectively, but it just furthers their biases. It’s also important to look at Dan Gainor’s biases because while it would appear that the Media Research Center would be non-partisan, the center only documents “liberal media bias.” It’s also painfully obvious just how much disdain Flood has for Matthews because he writes “Newsweek reporter Melissa Matthews … has recently written such other gems as “Why more orgasms can relieve sinus pressure,” and “Is PMS real or a myth?”

Sound (?) Conclusions

The only sound conclusion that I’ve found is not from the author of the attacking article, but rather a reader from Matthews’s article who stated, “Wow, both Manson and Trump used language to gain followers? Next, you’re going to tell me that they both drank water to quench thirst. You’ve cracked the case Newsweek. Now every leader will be using language to gain follows.” Perhaps I am assuming that this is a sound conclusion simply because I agree with it and because it mimics my comment that “[t]his does not seem like a stretch at all. In this capacity, Smaller could have related Manson to Gandhi or Trump to the Dahli Lama.”

Political (!) Implications

The political implications of both of these articles are ridiculous because one is a left-wing article attacking the right and the other is a right-wing article attacking a left-wing article. This constant back and forth of parties belittling each other further divides political parties.
*I converted a YouTube video of Donald Trump and Charles Manson into a GIF, but WordPress won’t let you post any kind of video without paying for it, so here’s the link in case you want to watch it.

Border Patrol unsure what to do about Trump’s executive orders on immigration

Kathryn Casteel’s article “The Border Patrol Doesn’t Know What To Do With The Thousands Of Agents Trump Wants To Hire” from FiveThirtyEight explains that while Trump has signed two executive orders regarding border patrol, the Homeland Security Department isn’t entirely sure how to handle the influx of agents that Trump wants.

Two new government analyses show that there “may be major obstacles to meeting [the] expectations [of increasing border patrol agents].” The department’s inspector general said that the department has “not establishes structure or rigorous process to determine needed staff [to] allocate them accordingly.” They even stated that they would not know what to do with “the 15,000 additional agents and officers they were directed to hire.”

Significant (?) Biases

It appears as though anyone who discusses immigration such that immigration should increase and border patrol should decrease is more liberal than conservative. And that anyone who thinks that immigration should decrease and border patrol should increase is less liberal and more conservative. However, the author does not appear to choose a side, so it is unclear whether or not she is biased or if this article is biased.

Sound (?) Conclusions

The conclusions made by the author appear to be sound because she references current events as well as historical immigration cases from the 1990s. In fact, her conclusion is more of a conclusion from an outside source rather than her own opinion, so I am unsure if this makes her conclusion more or less sound. She states that “Meissner, who now directs the U.S. immigration policy program at the Migration Policy Institute, a research organization, said that “when there is heavy political pressure for very aggressive hiring, it has forced or led agencies to cut corner, and generally it has come around to haunt them in the future with integrity issues.” The conclusion that Meissner seems sound.

Political Implications

This article clearly has a political implication given that it is discussing a salient political topic right now. If border patrol increase in such a way that it cuts corners, it is likely that what some may consider the “issue” of immigration may not actually get “better.” It just doesn’t appear that those who want to decrease immigration are enacting anything that will actually decrease immigration in the way that they want.

The Unreliable Statistics of Gang Violence

The article “Gang Stats Aren’t Remotely Reliable, But Voters Keep Hearing About Them Anyway” from FiveThirtyEight discusses the issues with gang violence data.

Gang Violence Data and How its Specifically Being Used

Politicians like Ed Gillespie, a Republican candidate for the governor of Virginia, are making outrageous claims about gang violence and its role in policy positions. Gillespie has connected the gang MS-13’s rising crime to the policy positions regarding sanctuary cities of the Democratic candidate for governor of the same state. Gillespie has “asserted that there are over 2,000 MS-13 gang members in Fairfax County, Virginia and that the group’s membership is growing.” The author argues that there is a significant issue with this claim: Gang membership and levels of gang violence are impossible to quantify with any certainty.

Various fact-checks have linked Gillespie’s claim to other relative statistics such as a Fairfax County government presentation on gang violence, but ultimately the other statistics while relative are not relevant enough to claim as the exact same.

Why There are Discrepancies with Gang Violence Data

According to Meena Harris, the director of the National Gang Center, trying to quantify gang membership is incredibly difficult because there is “no universal definition of ‘gang,’ and the debate still continues over what construes a gang and a gang member.”How can data quantify something when statisticians can’t even qualify that something? This lack of consistency means that each agency in each state is qualifying “gang” and “gang membership” differently. It’s been shown that “[d]eterming gang involvement … is up to the individual police departments, and most either underreport gang killings or do not report them at all.” This is best exemplified by the fact that in 2015, New Orleans reported zero gang killings to the FBI enough a city report “found that gang members were involved in 49 murders that year.”


Based on the data and its inconsistencies, the author stated that there may be a couple of reasons as to the issues with reporting gang violence such as “increased awareness of gangs,” “more resources being dedicated to countering gangs,” or “better training of officers to identify gang members.” These conclusions feel sound because the author provides more than one option as to why there are issues with data on gang violence, meaning that the author has a higher likelihood of being correct.

Social/Cultural Implications

Even though this gang violence data is being used by politicians, the article has more of a social and cultural implication than a political one. It’s more social and cultural because it’s not so much that the data is being used by politicians, but that it’s being used politically to impact the ways in which residents of the United States live their lives socially and culturally.

Significant (?) Biases

I did not find any significant biases in this analysis other than small biases such as the author being a crime analyst based in New Orleans, so it makes sense that he would reference the 49 murders from New Orleans in 2015 and that he would even be writing about this issue. The article feels more liberal than conservative, but ultimately more moderate than either. I think that it’s important to note that while there are some salient biases in this article, there don’t appear to be any significant biases.

(More) D(ata) of D& D*

*Yes, I’m very aware that I’m a nerd.

Cheyli’s most recent blog post referenced an article that I’ve read recently: “Is Your D&D Character Rare?” When I read this, I realized just how common all the characters the people in my group are playing.

Common Combinations of Race/Class

The most common combination of race and class to play, as Cheyli mentioned, is a human fighter. Your “race” is essentially the species you’re playing such as human or gnome and your “class” is the type of job you have such as wizard or fighter. Of elves, the most common is ranger, with the second most common being wizard and third, rogue. Of dragonborns, the most common class is paladin. Of tieflings, the most common is warlock. In our group, there is a human fighter, an elf wizard, an elf rogue, a dragonborn paladin, and a tiefling warlock.
I found this particularly odd because none of the players in my campaign, including myself, think that our characters are especially common. Ever since I read this article two-ish weeks ago, I was confused as to why these seemingly unique characters would actually be the least unique or most common. Is it because, like Cheyli said, “people are more naturally drawn to things [they know] because they are easy to understand”? Is it because the less creative the person the less likely they are to make a creative character? Or is it because of something else? I’ve been hard-pressed to think that a character such as a dragonborn paladin is easy to relate to and therefore, unfamiliar and that a less creative person is going to be drawn to this type of character. While I completely agree with Cheyli that human fighters require less creativity, I don’t think that the other characters are common because of a lack of creativity. Perhaps, this frequency is more related to the nature of the game or race/class synergy than a player’s creativity. 

Screen Shot 2017-10-28 at 11.00.16 PM.png

Examples of Race/Class Synergy

To look at this theory, I decided to look at each of the players in my campaign’s characters:
Elves get a boost to both dexterity and intelligence (depending on the elf you choose) and while dexterity is helpful for wizards, intelligence is required for them. Specifically, in the Player’s Handbook (PHB), elves get a +2 to dexterity and, depending on how you do it, a +1 to intelligence. The +2 means that anytime this player rolls for a dexterity based activity, they’ll receive a bonus in their favor for this activity. Similarly, the +1 means that anytime this player rolls for an intelligence-based activity, they’ll receive a bonus for in their favor. For wizards, spells are an intelligence-based activity, so the higher a wizard’s intelligence, the better they are at spell casting. Ultimately, elves are inherently more dexterous and intelligent, so they make better wizards. This helps to explain why so many players create this race and class combination.
Rogues are incredibly dexterous, but there’s not a ton of requirement for intelligence so it’s understandable why they would then be the third most common class for the race of elves. While this pairing isn’t the most helpful, it makes sense why players would pair them together.
Dragonborns get a boost to both strength and charisma, and paladins are strength and charisma based. Specifically, in the PHB, dragonborns immediately get a +2 to strength and a +1 to charisma. This means that dragonborns are going to be inherently stronger and more socially capable when they are paladins.
Tieflings get a boost to both charisma and intelligence and warlocks are charisma based. While there are other reasons as to why this race and class are paired, it helps when their characteristics are paired together.
While I only looked at four race/class combinations, I feel like it gives a pretty good sense that the frequency at which people play as certain characters is more related to the nature of the game than creativity alone.

Limitations of the Data and How I Interpreted it

I looked at the data as “of race x, the most common class is y.” However, the data can be looked at as “of class y, the most common race is x.” If you were to look at it this way, it would then be: Of fighters, the most common race is human; of rogue, the most common race is again human; of wizard, the most common race is elf; of paladins, the most common race is human; of warlocks, the most common race is tiefling.
I also slightly altered what data I looked at to fit what characters the players in my campaign play as. Specifically with elves, the most common race is ranger, but instead, I looked at both wizards and rogues.
The researchers collected 109,188 race and class combinations between August 15th and September 15th, 2017 on the website D&D Beyond. This website is for players who don’t necessarily have access to an in-person campaign, meaning that this data completely excludes tabletop players (or the nerds who, let’s say, gather in a basement every Saturday).
The data included the most frequent used races and classes, but not all of them. It allows for a general idea of the most common combinations, but fails to show the most detailed picture that it could. It did not include sub-races or sub-classes.
It’s also important to look at how the data is visualized. While the colors help show the modal occurrences and help emphasize the more popular race/class choices, the data is shown only in raw scores. The frequencies are not relative to each other, so it is difficult to only look at this and have a good idea for the actual percentage occurrences.

Accurate Data Visualization used Inaccurately: Domestic Terrorist Threat Edition

In the article, “White American men are a bigger domestic terrorist threat than Muslim foreigners,” the author, Jennifer Williams, explains that “[s]ince Trump took office, more Americans have been killed by white American men with no connection to Islam than by Muslim terrorists or foreigners.” According to a study conducted by New America, a nonpartisan think tank in Washington, DC, “between 2001 and 2015, more Americans were killed by homegrown right-wing extremists than by Islamist terrorists.”

The article included this stacked bar graph: 

My Interpretation of this Visualization

From my interpretation of this visualization, it shows that between 2008 and 2016, there were a little less than 20 attempted domestic terrorist incidents by those who are left-wing, there were a little more than 60 attempted domestic terrorist incidents by those who are Muslim, and there were maybe 115 attempted domestic terrorist incidents by those who are right-wing. The issue I have with this data is that it shows the “type” of person who attempted a domestic terrorist incident, but it does not depict how they tried to act out this incident (what kind of weapon they used, where they tried to do this, etc.). It’s also pretty difficult to see just how many incidents were attempted; since the stacked bar graph starts with domestic terrorist incidents that were acted out, it’s hard to judge the exact number of the blue data, or the foiled terrorists attacks. I feel like this visualization could still show the same data, but perhaps it’d be easier if the orange bars (acts carried out) were next to the blue bars (acts foiled), instead of the blue bars being pushed up against the orange bars.


Where did this visualization come from?

I searched this info graphic and found that it was included in another article, “Home Is Where the Hate Is” by David Neiwert back in June. Then I found another article, “How We Analyzed Domestic Terror Incidents,” that breaks down exactly where the researchers obtained data for this visualization, who they recruited to analyze this data, how they categorized people as either “right wing,” “Islamist,” or “left wing.”


Does the data from this visualization make sense?

After reading this article, it appears that the data was obtained ethically and in a sound manner—it’s from public databases and Freedom Information Act requests. And that the people they recruited to judge the data are well-informed and knowledgeable about the topic—these include people such as the director of the Institute for National Security and Counterterrorism at the Syracuse University College of Law and the former senior domestic terrorism analyst at the Department of Homeland Security’s Office of Intelligence and Analysis, now the owner and CEO of the consulting firm DT Analytics. And that the terminology they chose was loosely specific to certain ideologies—“left wing” includes animal rights activists, environmental activists, and anti-racist extremists; “right wing” includes militia movements, white supremacists, anti-government activists, anti-Muslim extremists, anti-immigration extremists, and anti-abortion extremists, including radical Christians; “Islamist” includes theocratic extremists inspired by groups such as the Taliban, al-Qaida, and the Islamic State.


Why did the author use this visualization?

After looking in to this specific data visualization, I understand it a bit better and feel as though it was chosen for the article “White American men are a bigger domestic terrorist threat than Muslim foreigners” as a means to get the author’s point across in a more succinct way, but not necessarily in a way that accurately depicts what the data is saying. For example, the author claims “between 2001 and 2015, more Americans were killed by homegrown right-wing extremists than by Islamist terrorists,” but this data visualization refers to the years between 2008 and 2016. Ultimately, I don’t think that this accurate data was used in the most accurate way possible.

Educating girls is not about just social justice, but capitalism, too

In the article “Educating girls is not just about social justice,” Mansoor Qaisar explains that education is “no longer simply a matter of social justice—they are intrinsically linked to economic growth, social well-being, and stability. This places girls’ education at the core of all human development endeavors;” education for girls is not solely about ethics or morality, but rather the larger impact that it has for the economy.

The article focuses on the fact that girls, specifically in Pakistan, are not receiving much, if any, education because they’re not getting equal rights. It explains that while there are larger policies being passed in the public sphere, there are sexist cultural values that are persistent in the private sphere (homes, for example), which perpetuate the lack of education for girls.

The author examines urbanicity versus rurality and along the same lines of the public sphere versus the private sphere, they conclude that, “only in the most rurally isolated areas do these notions of strict gender roles apply.”

According to the Annual Status of Education Report for 2016 in Pakistan, “48 percent of the poorest girls in the 5-16 age bracket are enrolled in school as compared to 68 percent of the poorest boys from the same age group.”

This article definitely has a liberal or social justice bias simply because of its focus, but it also has a capitalistic bias because the main argument as to why girls need to be educated is so that they can financially contribute to Pakistan and increase its GDP: “[e]ducation represents critical input in human resource development and is essential for the country’s economic growth. It increases the productivity and efficiency of individuals, while achieving a skilled [labor] force capable of leading the economy towards sustainable growth and prosperity.” The author has a similar liberal or social justice bias because he is the Senior Information Officer at the Population Council office in Pakistan and has been for more than fifteen years.

I think that the conclusions are sound because, ultimately, the conclusion is that education needs not to be gendered, even though it places a significant emphasis on means and production. So, I think that the conclusion is as sounds as it could be given that it has an obvious bias.