Monday, March 27, 2017

Shameless Self Promotion: I wrote a chapter in a book about Open Educational Resources!

Let's make the academy better for science and better for our students, and let's make it better for free.

Want to learn how? I recommend a Open: The Philosophy and Practices that are Revolutionizing Education and Science, edited by Rajiv Jhangiani and Robert Biswas-Diener.



In the spirit of open resources, it is totally free.

In the spirit of open pedagogy and quick sharing of teaching ideas, I wrote a chapter for the book about how I've gone about sustaining a blog dedicated to teaching for the last four years. The basic message of my chapter: I blog about teaching, and you can, too!  Here are all the chapters from the book:


Johnson's "The reasons we don’t study gun violence the same way we study infections"

This article from The Washington Post is a summary of an article from the Journal of the American Medical Association. Both are simple, short articles that demonstrates how to use statistics to make an argument. Here, that argument is made via regression in order to demonstrate the paucity of funding and publications for research studying gun related deaths.

What did the researchers do? Regression. A regression line was generated in order to predict how much money is spent studying common causes of death. We see that deaths by fire arms aren't receiving proportional funding relative to the deaths they cause. See the graph below.

https://img.washingtonpost.com/wp-apps/imrs.php?src=https://img.washingtonpost.com/blogs/wonkblog/files/2016/12/Capture7.png&w=1484

How to use in class:

1) How is funding meted out by our government in order to better understand problems that plague our country? Well, it isn't being given to researchers studying gun violence because of the Dickey Amendment. I grew up in a very hunting friendly/gun friendly part of Pennsylvania. I've been to the shooting range. And it upsets me that we can't better understand and study best practices for safe gun ownership.

2) Another issue: We don't talk about suicide enough. Half of the gun deaths were suicides.

3) There seems to be under-funding of possible accidents, as opposed to diseases, that cause death (shooting, motor vehicle, falls, and asphyxia). Why might this be?

4) The above image demonstrates correlation/linear relationships as well as gun violence as an influential observation.

5) Regression, y'all. 

The WP article states, 

"If public health issues were funded based on their death toll, gun violence injuries would have been expected to receive about $1.4 billion in federal research funding over about a decade — compared with the $22 million that it actually got, the study found." 

They predicted Y (research funding) based on X (death toll) and found a discrepancy, and the discrepancy is used to make an argument about the funding short fall. If you go to the JAMA article, they describe the research article publication shortfall as well. According to that regression equation, there should be over 38K articles published about gun deaths. Instead, there are 1,738.


Monday, March 20, 2017

Retracton Watch's "Study linking vaccines to autism pulled following heavy criticism"

This example from Retraction Watch illustrates how NOT to do research. It is a study that was accepted and retracted from Frontiers in Public Health. It purported to find a link between childhood vaccination and a variety of childhood illnesses. This would be a good case study for Research Methods. In particular, this example illustrates:

1) Retraction of scientific studies
2) The problems with self-report surveys
3) Sampling and trying to generalized from a biased samples
4) What constitutes a small sample size depending on the research you are conducting
5) Conflict of interest

This study, since retracted, studied unvaccinated, partially vaccinated, and fully vaccinated children.

And the study found "Vaccinated children were significantly less likely than the unvaccinated to have been diagnosed with chickenpox and pertussis, but significantly more likely to have been diagnosed with pneumonia, otitis media, allergies and NDDs (defined as Autism Spectrum Disorder, Attention Deficit Hyperactivity Disorder, and/or a learning disability)."


But the study surveyed moms who home school their children. A group that is historically, but not exclusively, anti-vaccination. From the study:


"Homeschool organizations in four states (Florida, Louisiana, Mississippi, and Oregon) were asked to forward an email to their members, requesting mothers to complete an anonymous online questionnaire on the vaccination status and health outcomes of their biological children ages 6 to 12."


And money to conduct this study was crowd sourced via a pro Autism:Vaccination website. There are other problems with this study, as noticed by Retraction Watch as well as the summary piece below. The sample sized used was relatively small for this kind of research, no one verified the various diagnoses via medical records, etc.


So, there is plenty for your students to consider with this study. Maybe you could create a methodology for this study that would fix the current, flawed methodology. Or you could just give your student the summary of the study and ask them to find the problems.

Further treatment (and deconstruction) of the study can be found here.

Monday, March 13, 2017

I've tracked all my son's first words since birth [OC]

Reddit user jonjiv conducted a case study in human language development. He carefully monitored his son's speaking ability, and here is what he found:
https://imgur.com/gallery/KwZ6C#qLwsn9S...go to this link for a clearer picture of the chart!



How to use in class:
1) Good for Developmental Psychology. Look at that naming explosion!
2) Good to demonstrate how nerdy data collection can happen in our own lives.
3) Within versus between subject design. Instead of sampling separate 10, 11, 12, etc. month old children, we have real-time data collected from one child. AND this isn't retrospective data, either.
4) Jonjiv even briefly describes his "research methodology" in the original post. The word had to be used in a contextually appropriate manner AND observed by both him and his wife (inter-rater reliability!). He also stored his data in a Google sheet because of convenience/ease of tracking via cell phone.

Monday, March 6, 2017

Annenberg Learner's "Against All Odds"

Holy smokes. How am I just learning about this amazing resource (thanks, Amy Hogan, for the lead) now?

The folks over at Annenberg, famous for Zimbardo's Discovering Psychology series, also have an amazing video collection about statistics, called "Against All Odds".

Each video couches a statistical lesson in a story.

1) In addition to the videos, there are student and faculty guides to go along with every video/chapter. I think that using these guides, and instructor could go textbook free.
2) The topics listed approximate an Introduction to Statistics course.

https://www.learner.org/courses/againstallodds/guides/faculty.html