Blogs 3

Height and Dating: A Critique of Inexact Statistics

Anonymous — Tue, 08 May 2018 22:13:42 +0000

Height and Dating: A Critique of Inexact Statistics Anonymous (not verified) Tue, 05/08/2018 - 16:13

Though data-driven academics often insist that the heart of science is raw numbers, the truth is that people are influenced more by the presentation than by actual data.

By Christopher Koski
Course: Language, Gender, and Sexuality (Ling 2400)
Advisor: Prof. Kira Hall, TA Ayden Parish
LURA 2018

Such is the art of statistical analysis: Two people with two different motivations can take two identical data sets and produce facts and figures fitting two entirely different narratives. Here’s a personal favorite of mine, courtesy of Reuters. I generally consider Reuters an alright news source, but take a look at this graph:

All of the data points in this graph are correct, at least according to the Florida Department of Law Enforcement. The presentation of the data, however, is misleading. This graph seems to show a slowly increasing number of gun deaths right up until 2005, where Florida’s new Stand Your Ground law coincides with a drastic decline in firearm-related fatalities. (This law basically states that you can use lethal force when you are in a situation wherein you believe you are in danger of death or serious injury.) However, the “decline” indicated in the graph is the exact opposite of what the data actually show, which is a drastic increase in the numbers of murders committed using firearms after 2005. The author, C. Chan, apparently decided to display the graph vertically mirrored, with the values on the Y-Axis increasing as you go down, rather than up. I believe that this representation was intentional.

Unfortunately, most presentation problems are less easy to spot than this example. The one I want to talk about today is a commonly cited statistic in research on language, gender, and sexuality. In fact, Penny Eckert and McConnell-Ginet, the authors of a textbook that we used in the course Language, Gender, and Sexuality (Ling 2400) at CU Boulder, also cite this statistic: 91��Ƭ��AV half of all men and half of all women are between the height of 5’ 4” and 5’ 10”. Eckert and McConnell-Ginet use this figure to question why we do not see more heterosexual relationships in which the man is shorter than the woman. Given the statistic, with half of men and women between 5’4” and 5’10”, it seems like we should see more couples where the woman is taller.

Again, the data I have available to me (sourced from the CDC) corroborates this claim. The problem is that those numbers mean a whole lot less than it looks like they do.

The problem is that the statistic itself is vague at best. Note that there isn’t actually any overlap promised by this statement: the only two things that need to be true are that half of all men are between 5’ 4” and 5’ 10” and half of all women fulfill the same condition. Let’s look at some possibilities for height values between men (blue) and women (red), based solely on the fact that half of each must be between 5’ 4” and 5’ 10”:

This graph is not representative of reality, but is a feasible distribution of height if our only consideration is placing half of all men and women between 5’ 4” and 5’ 10” (or 64 and 70 inches, respectively). In this graph, there is basically no overlap between heights, despite our height constraint. The odds of a relationship where the woman is taller are more or less negligible on this graph. Here’s another possibility that works within the height rule:

This graph also fits the height constraint, but this time the overlap includes just about everyone. If this graph was a truthful representation of height distribution, than we would expect the proportion of relationships where the woman is taller to be around 40%!

Neither of these graphs are accurate representations of height distribution, but the fact that there is so much room for error in the 5’ 4” - 5’ 10” rule means that any claims we make based off of that statistic alone are difficult to justify. Based on the CDC’s data on height distribution, here is a more accurate graph:

The presentation suggests that there is a good amount of overlap here, but looks can be deceiving: The actual probability of a randomly selected male-female pair having a taller woman is only around 7%, based on a simulation of 40,000 random pairings. This doesn’t even take into account that people tend to date within their demographic, and the overlap within each community is even smaller. (I’ll spare you the graphs for that, but they’re easy enough to generate from the attached dataset.)

So what’s the point of all this? I don’t actually disagree with the assertion that people select for height when dating: A study by Yancey and Emerson (2014) notes that 37% of sampled men say that they will only date women shorter than them, while 55% of sampled women say that they will only date men taller than them. I am also pretty sure that Eckert and McConnell-Ginet aren’t trying to mislead anybody with their statement; this figure has been bandied about in many gender studies over the years, so it’s most likely just a part of the lexicon now, or a part of the “hall of mirrors,” as Eckert and McConnell-Ginet would say.

I do, however, think that honesty is an important part of academia, and this includes providing people with a complete picture of the truth. When we write an academic article, textbook, or any other sort of scientific journalism, we have a duty to produce a precise, fact-based account of the claims we make. We live in an age where an enormous amount of data is generated, interpreted, and presented every day, and there are people who will use their position as statistical purveyors to mislead the masses.

When you see a strange or vague claim, try to dig a bit deeper; you may be surprised at what you find.

References

Yancey, G., & Emerson, M. O. (2014). Does Height Matter? An Examination of Height Preferences in Romantic Coupling. Journal of Family Issues, 37(1), 53-73. doi:10.1177/0192513x13519256

Anthropometric Reference Data for Children and Adults: United States, 2007–2010. (2012, October). Retrieved from https://www.cdc.gov/nchs/data/series/sr_11/sr11_252.pdf

Eckert, P., & McConnell-Ginet, S. (2013). Language and gender. Cambridge, England: Cambridge University Press.

Off

Traditional

White

Racial Stereotypes in my Childhood Film

Anonymous — Tue, 08 May 2018 21:20:10 +0000

Racial Stereotypes in my Childhood Film Anonymous (not verified) Tue, 05/08/2018 - 15:20

As children, we often do not notice how we are socialized into ideologies of race through cartoon movies.

By Shawg Ahmed
Course: Language in US Society (Ling 1000)
Advisor: GPTI Maureen Kosse, TA Irina Wagner
LURA 2018

When children go out into society and become adults, they may believe that one racial identity speaks only one way and another speaks only another. While there is nothing wrong with categorizing how speakers of different racial identities speak, the problem arises when people also associate language with socioeconomic class status and job potential.

One of the biggest socialization factors is television. Television teaches kids about how the world works, but often what kids learn are stereotypes (Lippi-Green 2012).

I chose to focus on the Disney film Princess and the Frog to demonstrate and exfoliate the stereotypical views displayed in characters based on their social identities and their language that they speak. In this movie, all the African American characters speak African American Vernacular English (AAVE), while the White American characters speak Standard American English (SAE). Being able to learn that different social identities speak different varieties of English is not a bad thing. However, in this film, all the African American characters that speak AAVE occupy working/poor class status and live in smaller houses and work in service-based positions, as seen in the character of the waiter.

This kind of representation, which presents AAVE speakers only as laborers, leads people to believe that those who speak AAVE are not smart enough or do not have the potential to succeed in other higher-level occupations. The result of this stereotypical thinking is very harmful to the African American community. Based on sociological research, an example could be this: An African American goes out to apply to similar positions as their White Americans peers, but they are less likely to be hired to higher-paid positions because there is an unconscious bias that those who speak AAVE are not capable enough. Others might speak in SAE, which is perceived as more prestigious and the more “normal” and “correct” way of speaking English.

I chose to research this topic in Disney’s Princess and the Frog because as a child, I grew up believing that speaking AAVE shows that a person is uneducated. I always tried to stay away from speaking it, whereas my African American friends spoke it freely. I was socialized to believe that in order for me to get into good universities and earn high-paying positions, I needed to speak more Standard American English.

I decided to explore the roots of where I first was socialized to believe that way, and the first thing I could think of was Disney movies, particularly my favorite movie, Princess and the Frog.

My goal with this research project was to explore how socialization into language may influence society in more ways than one might expect. I hope to eventually lead people in society away from continuing to think in the ways they were socialized, from the time they are children watching Disney movies to when they are adults watching news outlets. Speaking AAVE does not equate to your level of knowledge, education, or potential.

Off

Traditional

White

Orange Caligula and Euphemism

Anonymous — Tue, 08 May 2018 21:15:00 +0000

Orange Caligula and Euphemism Anonymous (not verified) Tue, 05/08/2018 - 15:15

Euphemism is a linguistic tool used to communicate with indirect means, generally around a topic that causes psychosocial discomfort.

By Lauren M Hoos
Course: The Study of Words (Linguistics 1010)
Advisor: Prof. Laura Michaelis; TA Annebeth Buis
LURA 2018

Historically, euphemism has been correlated to fear or superstition. For example, often times we hear death referred to as passing away; while death is inevitable, avoiding direct reference to it both acknowledges the sensitivities of the hearer and comforts the speaker. Another linguistic example includes the word devil, which comes from the Greek word diabolos, or ‘thrower’, meaning one who throws insults. This euphemistic strategy is grounded in superstition, in particular the fear that speaking its name may cause it to appear. Other common euphemistic terms for devil are prince of darkness or fallen angel. Additionally, the German word brun, which translates to ‘brown one’, is used to refer to a bear, while in Russian the creature is called medved, which translates to ‘honey-knower’.

Since the election in 2016, and the inauguration of #45, speaking of the President seems to be a sensitive subject that brings a shift in mood of the conversation, whether it shows itself through fear, passion, or vehement opposition. Since noticing this linguistic shift, I have taken interest in the use of euphemism in reference to the current president of the United States. All over the internet, media, and in every day communication, we hear nicknames for him, ever-evolving, and increasingly humorous. I am interested in analyzing cross-cultural perspectives and the use of euphemism by speakers discussing the President.

An important property of euphemism, which will figure in this study, is the fact that euphemistic expressions lose their efficacy relatively quickly. Dr. Steven Pinker coined this phenomenon as a ‘euphemism treadmill’: When a euphemistic term becomes the conventional way of referring to something, the term loses its euphemistic value, requiring the creation of a new indirect form. The etymology of toilet (from French toilettes ‘towels’) illustrates this. The term has come to refer to the commode itself, and the euphemism restroom is more commonly used in American English.

In this proposed project, I expect to find evidence of the ephemerality of euphemism—that is, expressions will lose their euphemistic force over time and therefore must be frequently replaced. This explains the rapid evolution of nicknames. I expect to hear passionate discourse regarding our President. Along with performing discourse analysis on the use of euphemism around his name, and looking for discourse markers that are abnormal to the subjects’ typical phonological patterns when discussing the POTUS, I will analyze definite descriptors used as a workaround for his name, expecting to find a connection between the use of euphemism around a topic or person which creates a sense of unease, and the sociolinguistic function of euphemism.

With the help of my advisors, Dr. Laura Michaelis-Cummings and Dr. Kira Hall, I will use this research to defend my honors thesis.

Off

Traditional

White