Showing posts with label data analysis. Show all posts
Showing posts with label data analysis. Show all posts

Wednesday, March 06, 2019

Sunday, October 07, 2018

GRAPHIC: What Kind Of Voter Are You Depends On How Much Government You Want: Dismantle, Streamline, Rebuild, or Expand?



An interesting analysis of partisanship in a report from the Brooking Institution. They split all voters into four categories: dismantlers, streamliners, rebuilders and expanders. They are each defined thusly:

The report then goes on to analyze how each of the major partisan groups: democrats, republicans and independents are made up of various groupings of these four types (or ideologies) and how this has changed over time. It's worth a read!

Wednesday, May 16, 2018

#NSF REPORT: U.S. Doctorate Recipients At Near All-Time High (2016 Data)



The National Science Foundation has released its annual Survey of Earned Doctorates and there are some interesting nuggets in the data (up to 2016):

  • The largest share of doctorates awarded in 2016 was in the life sciences (nearly 23 percent), followed by engineering (17 percent), and psychology and the social sciences (16.5 percent). All fields of humanities and arts made up 10 percent of doctorates awarded.
  • The time between students entering graduate school and receiving doctorates has fallen in all fields of study over the past 20 years, but, on average, it still takes years longer to earn a doctorate in non-S&E fields than in S&E fields.
  • The number of S&E doctorates awarded to U.S. citizens and permanent residents grew 2 percent since 2015 and 39 percent since 2006. The number of doctorates in S&E fields awarded to temporary visa holders grew 2 percent since 2015 and 20 percent since 2006.
  • The number of doctoral awards to temporary visa holders is highly concentrated -- 10 countries accounted for 71 percent of the doctorates awarded to temporary visa holders from 2006 to 2016. The top three countries -- China, India and South Korea -- accounted for 54 percent of the doctorates awarded to temporary visa holders.
  • Women have earned a slight majority of all doctorates awarded to U.S. citizen and permanent residents each year since 2002, and women have earned more than 30 percent of all doctorates awarded to temporary visa holders over that period.
  • From 1996 to 2016, the number of women earning degrees in S&E fields increased by 84 percent.
  • The number of doctorates awarded to blacks or African Americans who are U.S. citizens or permanent residents increased by 32 percent from 2006 to 2016. For the same period, the proportion of doctorates earned by Hispanics or Latinos increased by 67 percent.
One key take-away for me is that 1) Latino doctoral recipients now outnumber African-American and that the gap between male and female STEM doctorates is actually slightly widening, even as both numbers increase (see the graphics at the top of this page).

Hat/tip STEMPROF newsletter edited by Mark Connolly.

Wednesday, March 14, 2018

REPORT: Boys and Girls Equally Interested In STEM; Boys More Prepared

In a new report from ACT (yes, the organization that runs those other college readiness tests) there is a result that states that all though boys and girls have an equal interest in careers in science, technology, engineering and mathematics (STEM), it turns out that boys are generally more prepared for them.

Ed Week reports:
To determine readiness, students were assigned a STEM score based on combined results from their ACT math and science tests. ACT test scores range from 1 to 36. A student earning at least the benchmark score of 26 has a 50 percent chance of earning a B or higher in first-year college STEM courses, and a 75 percent chance of earning a C or higher, according to the report. The student is also more likely to stick it out with a STEM major and earn a bachelor's degree.  
The report revealed that just 18 percent of females, compared to 24 percent of males, earned a score of 26. 
The disparity is even larger among students with an interest in STEM: 22 percent of females earned a score of 26 compared to 31 percent of males. Even more surprising, females interested in STEM were less likely than all males (including those with or without an interest in STEM) to meet or surpass that benchmark score (22 percent versus 24 percent). Meanwhile, the overall level of interest in STEM between females and males is nearly equal at 47 percent versus 50 percent, respectively.
The source of the data is from the 2 million students who took the ACT in 2017.

Sunday, February 11, 2018

STUDY: U.S. Mathematics Majors Most Likely To Switch To Another Subject


The National Center for Education Statistics at the U.S. Department of Education released a report at the end of last year about students switching majors in college. "Beginning College Students Who Change Their Majors Within 3 Years of Enrollment" has a number of interesting facts about students who change majors while pursuing associate's and bachelor's degrees in the United States:

  • About half (52 percent) of students whose original declared major was mathematics switched majors within 3 years. Mathematics majors changed majors at a rate higher than that of students in all other fields, both STEM and non-STEM, except the natural sciences.
  • Within 3 years of initial enrollment, about 30 percent of undergraduates in associate’s and bachelor’s degree programs who had declared a major had changed their major at least once.
  • About one-third of students enrolled in bachelor’s degree programs changed majors, compared with 28 percent of those enrolled in associate’s degree programs.
  • About 1 in 10 students changed majors more than once: 10 percent of associate’s degree students and 9 percent of bachelor’s degree students.

Sunday, August 20, 2017

GRAPHIC: Confederate Statues Mapped In Time and Space


This post has an interesting graphic which demonstrates that statues commemorating the Confederacy (which was devoted to the maintenance of slavery and promulgation and prolongation of white supremacy) primarily happened in two waves, one around the establishment of Jim Crow laws in the 1910s and again in response to the civil rights movement in the 1960s.

This refutes the claim that these statues are primarily about "history and heritage" and supports the argument they are signals to Black people about the permanence of white supremacy.

Hat/tip to Axios.

Tuesday, March 07, 2017

POLL: 12% Of California Voters Are #LGBT

Equality California has released a poll which shows that 12% of California voters identify as LGBT. Interestingly, this number varied only slightly by geographical region in the state, from 10% in the Los Angeles area (11% in Los Angeles county) to 14% in the San Francisco Bay Area.

The poll discusses the margin of error:
The survey’s margin of error is plus or minus 4 percentage points at a 95 percent confidence level. This means that 95 times out of 100, if a response to a given question to which all respondents answered was 50%, we could be 95% confident that the true percentage would fall within +/-4% of this percentage - or between 46% and 54%. As 12% of respondents in this survey answered “yes” when asked if they were a member of the LGBT community, we can infer that the sampling error specific to that response is +/- 2.4% or between 9.6% and 14.4%.
The poll's methodology of asking the question "Are you a member of the lesbian, gay, bisexual, and transgender community, also known as the LGBT community?" online may be an explanation for why they received such larger values than the more typical 2-4% number  that we have been more used to.


Tuesday, September 13, 2016

MAP: New Data Show Same-Sex Marriage Geographic Distribution In The USA


The New York Times Upshot blog has a post today on a Working Paper (pdf) by researchers at  the U.S. Treasury department about same-sex joint married tax filers in 2013 and 2014, after the United States v. Windsor decision ended discrimination by the federal government against legally married same-sex couples (like yours truly).

By linking the tax returns of same-sex couples who filed jointly in 2014 with their Social Security records, researchers are able to give us the most accurate picture of same-sex marriages to date. And their estimate is this: In 2014 there were 183,280. same-sex marriages in America, roughly a third of 1 percent of all marriages.
Of course, implicit in this estimate is the assumption that all married couples file their returns jointly. But as a proxy for that, it’s pretty good. The Treasury Department estimates that 97.5 percent of married couples file joint returns.
One highlight of the study: Pretax household income of same-sex married couples is higher than that of heterosexual married couples. Most of that is driven by the average earnings of male same-sex couples: $176,000. On average, they make $52,000 more than married lesbian couples and $63,000 more than married straight couples.
A key factor in the reason why same-sex male couples make more than their married counterparts is where they tend to live, which is urban centers. Here's the list of locations for the top 10 places same-sex couples reside:
No surprise that San Francisco is #1 on the list (for gay couples and #3 for lesbians), but it is also interesting to note that Washington is #2 and #6 respectively. Los Angeles and Long Beach are represented well.

Friday, April 08, 2016

Mathematical Analysis Reveals TYRION Is Main Character of Game Of Thrones

This is really cool! io9 reports on a mathematical analysis of the third book in the A Song of Ice and Fire series by George R.R. Martin to demonstrate that the Tyrion Lannister is the actual main character of the series.
After ranking the characters through various methods, the three most common characters that could be perceived as the “main” stars of the series emerged: Sansa Stark, Jon Snow, and—at the top in pretty much every measured regard—Tyrion Lannister. 
It’s not really that surprising—each chapter of a book in the series is told from the perspective of a different character, and Tyrion has the most chapters to his name across the whole saga. 
The type of mathematics being deployed here is called "network science." You can read the text of the original article (titled "Network of Thrones") at Math Horizons.

Monday, April 04, 2016

Fusion Adds Data To #GayMediaSoWhite Discussion


John Walker at the Fusion website analyzed the covers of the magazines of The Advocate, Out and Attitude and produced this useful infographic (seen above) to determine the (lack of) diversity represented therein.

The summary of the data is:
Representation on the cover of Out, June 2011 – May 2016In the past five years, Out has featured 78 people on the cover of its magazine. That figure includes repeat appearances by Dustin Lance Black, Ellen Page, Neil Patrick Harris, and Zachary Quinto. The mag’s upcoming May 2016 issue will feature Troye Sivan, a gay white singer-songwriter from Australia who rose to fame on YouTube.
Sixty-six of those 78 people have been white (85%), and 31 of those white people have been queer and/or trans (40%). Twelve of the 74 individuals featured on Out‘s covers have been people of color (15%), and eight of them have been queer people of color (10%).Representation on the cover of The Advocate, June 2011 – May 2016I was unable to locate cover imagery for the following six issues of The Advocate: September 2011, December 2011, January 2012, May 2012, June 2012, and July 2012. But of the covers I was able to find—excluding the publication’s more abstractevocative cover art—at least 35 people have been featured on the cover of The Advocate over the past five years.At least 10 of those 35 individuals have been people of color (29%). At least eight of those 35 were queer and/or transgender people of color (23%). That includes the mag’s April/May 2016 cover star DeRay Mckesson, a prominent figure in the Black Lives Matter movement who is currently running for mayor of Baltimore.Representation on the cover of Attitude, June 2011 – May 2016Attitude has featured 111 people on its covers in the past five years. That tally includes repeat appearances by Adam Lambert, Dan Osborne, Gareth Thomas, Greg Rutherford, Harry Judd, James Hill, Thom Evans, and Tom Daley.
One-hundred and six of those 111 have been men (95%). The remaining five were women (5%), all of whom were white. (This gender disparity isn’t as troubling as it could be, given that—unlike The Advocate and OutAttitude is specifically marketed to gay men.) No trans or nonbinary people of any race or ethnicity have been featured on the cover of Attitude in that timeframe.
The overall data can be summarized with this final quote of the data analysis is:
By the way, of the 21 people who appeared on more than one of these magazine covers, we found that seven (33%) are straight white cisgender men. Two of the repeat cover stars are queer women (10%), Lady Gaga and Ellen Page. Just one (5%), Michael Sam—who has been seen on both Out and Attitude—is a person of color.
Discuss!

Monday, August 31, 2015

Mathematical Analysis of Women's Tennis Open-Era Challenges Serena's GOAT Status


On the eve of the start of her campaign to secure her 22nd major title, fourth of the year, Serena Williams's claim to the title of Greatest of all Time is being analyzed in an article at Five Thirty Eight titled "Serena Williams And The Difference Between All-Time Great And Greatest Of All Time." The article uses a particular mathematical technique called computing an ELO rating to assign a number (which varies over time) to the relative strength of players in head-to-head competition. The graph of all the major champions (Navratilova, Evert, Graf, Seles, Henin, Venus and Serena) are depicted.

From the graph it can be shown that Serena does not achieve the highest ELO rating of all time.

Elo definitely puts Williams in the top tier of female tennis players, but it tells a slightly more muted story than other measures. In particular: While Williams has been great, and has been doing unprecedented things for a player of her age, the relative weakness of the tier of players beneath her undermines her GOAT claim. 
The possibility that Serena has benefited from “weaker competition” is pretty conventional, and certainly debatable, but Elo gives us a useful way to examine exactly what that possibility means and what it implies.
For the unfamiliar, Elo is a rating method originally developed for chess, but eminently suitable for tennis. It’s very simple: Two players enter a match with Elo ratings based on their prior results. Elo uses their ratings to predict their head-to-head outcome, and then updates those ratings depending on the outcome.2 It’s not without limitations: Elo makes every head-to-head prediction based solely on the two players’ ratings, which, in turn, are only affected by previous match results.
However, one aspect of Serena which makes her career unique is her 2nd peak in her ELO rating, which has happened at an unprecedented stage of her career, in her 30s. It also claims that Serena was at her best in 2003 (when she won her first Serena slam).


I understand that these results will disappoint some and encourage others but I appreciate the analysis in that bring at least a measure of quantitative rigor to a discussion which is essentially Qualitiative and subject in nature: Who is the Greatest of all Time? In my mind, the figures STILL indicate hat the answer is Serena Williams.

Sunday, February 15, 2015

GRAPHIC: Age Gap In Same-Sex Couples Generally Greater Than Straight Couples'


Five Thirty Eight did an analysis of the average gap in ages between couples, and included male-male and female-female pairs in their data set (which was taken from Facebook). The results are interesting, because there is a significant difference:
Using anonymized data from U.S. users who say they are in relationships, Facebook found that the average age difference in gay couples tends to get bigger the older people get. Those in their early 20s have an average age difference in their relationships of about two to three years, but once people get into their 40s, that average age gap increases to about seven years. The age difference increases for older male-female couples, too (shown in red below), though not by as much. (Remember, this is self-reported data from people who make their relationship status public on Facebook.)
Any thoughts about why this age gap among gay and lesbian couples increases with age? The age gap between me and my Other Half is 37 months (just over 3 years), but we are in our forties so we are clearly outliers.

Happy Valentine's Day (Weekend)!

LinkWithin

Blog Widget by LinkWithin