Reflection on how to liew with statistics

Quite often, we laughed at the fact that most of us are skeptics in class. Initially, we took the class for many reasons, to fulfill a requirement in honors, to discuss with someone that keeps bringing numbers, or as in my case, to expand our knowledge in statistics. I think that the biggest lesson I learned in this class was to make a judgement out of every number I hear. Ask for the sources. Ask for the methods. Question the incentives. Simple things and rules of thumb to take into account in the situation. I am glad i took this class, not only because of the lessons I learned, but because I was able to expand my understanding and skepticism.

-Xavi

The myth of the 8 glasses of water a day

Since we were young, we were told that we needed to drink around 8 glasses of water a day will keep us healthy and well dehydrated. Yet, the estimate in many cases seems to be way above our health requirement. There's no evidence that we need to drink more water than our dictates. Researchers in relationship to the matter mention that the requirements must vary between the temperature of the place, amount of physical activity and the size of the individuals. When trying to find the source of the eight glasses of water, I found a no exit alley. It seems to be just a myth that came from someone's word and has lasted to this day. It is hard to not lie with statistics, when sometimes, we dont even know what our source is.

-Xavi

 

 

Heteroscedasticity

On the recent class presentation, the discussion about the nature of the data, specially its variation, was brought up many times. Heteroscedasticity (pronunced heterokedastic) refers to a different kind of variation in the data, while Homescedasticity refers to a constant variation.

The non-constant variation of our data is a statistical problem that can invalidate most statistical tests of significance given the harsh effects it has when trying to execute a linear regression model. Heteroscedasticity can either bias upward or downward the results, and sometimes, its not possible to see this with a simple scatter plot.  The next time you guys have to do a regression, dont forget to account for the variation on your data; but inn case you do not have an heteroscedasticity problem, dont forget to consult your professor on how to address the problem

-Xavi

40 percent ? Where does that number come from ?

Given many unfortunate events as those who have taken place in Boston and Connecticut, the craze for gun control has gone wild around the U.S. While writing a paper for my public policy class I couldn't help to notice that President Obama and Biden have been stating that  "40 percent of guns are purchased without a background check.” I am not against gun control, or more extensive background checks, but the use of the number intrigued me.

The purpose of using this number, relies on the idea that there are many ways to avoid traditional background checks, yet, the number is quite hard to estimate.
The Brady Handgun Violence prevention Act currently prevents someone who buys from a federally licensed dealer from buying a gun if he has a felony, some kinds of misdemeanors conviction or has been found mentally il, yet, a loophole in the act allows individuals to sell their guns privately, and thus, opening the possibility for background checks.

 Trying to find the source of that 40% I have gone into a crusade that gives me not reliable answer. Many different researches quote different numbers, but most importantly, the sources tend to be really biased groups from both sides of the debate.

This experience taught me, that numbers can be as dangerous as guns when dealing with people

-Xavi

TopicBlogspot3 : Multicollinearity

During the recent class presentations, I couldn't stop to notice that the topic of multicollinearity, or the high correlation between two variables, arose in many circumstances.

Multicollinearity is a common statistical  phenomenon that arises when two variables are correlated. For example. the cost of college education that someone attends will be deeply correlated to their income. If the income rises, the individual will have the possibility to attend a more expensive college.

When this circumstance arises in a regression, a high degree of multicollinearity will biase our estimates, common hypothesis testing becomes unreliable, and in some cases, the regression coefficients are more than one (which means problems).

Why  do this happen ? When two variables are related to each others, the estimated value of our dependent variable (Y) will be biased given that the relationship between the two independent multicollinear variables might  overestimate or underestimate the true values

How to avoid this ? Well, the most common way to do it is to evaluate the nature of the variables. If having more cars might be related to the amount of income of someone seems a reasonable assumption, it wouldn't be bad  to test for it. So, the next time you think about making a regression, dont forget to account for multicollinearity between variables

-Xavi

Blogging in Class

Every single one of my honors classes has had some kind of blogging. Although tedious in some cases, the discussions that tend to arise from this kind of writing have become in many cases deeply gratifying. The ability to come up with blog ideas, is a way to apply what we learn in classes, as well as a way to develop our own ideas. I personally believe that the use of blogging is a great educational tool to be used, as it fosters the same kind of sharing that of a seminar style class ...but on a bigger classroom
-Xavier

The Ability to Share Thoughts

Through utilizing this blog as a tool for sharing knowledge and information with other classmates, it was possible to express my own thoughts and values in a public space where others could critique, teach and learn from me as well as from each other. I have never used a blog for any school course, and I do have to say I like the idea of having a "shared space" with classmates outside of the classroom. It allowed us to be able to help each other and to try to learn about each other. Having 2 blog posts do each week was fine, and the blog assignments that were required taught me how to look up sources of statistical analysis and how to understand them in a thought-provoking way. I felt that the blog became very handy, and allowed for a greater amount of discussion than what could be possible within a class time frame.

Usefulness of Blogging

Honestly, when I first started this class I dreaded the thought of having to do two weekly blog posts. After sitting down and just getting used to doing them, I found them to be a useful media for having discussions with classmates outside of class. I feel that this blog has also forced me to think about how and what I am writing. I was constantly rereading my posts prior to posting them because when you post something there is a strong likelihood that someone else will  be reading it (and judging you), and me being me, I did not want to show off my lacking grasp of the English language or be too belligerent. I think that course blogging is an effective tool at stimulating individual students to improve their writing and facilitate post class discussion (so its useful and I enjoyed it).

Blog

I think the blog was a good idea for this class.  It gave us a chance to share interesting statistics and data that we would not have seen in class, and expand on what we did cover in class. I enjoyed reading the posts and found a lot of the linked articles very interesting.  The blog is an effective way to get students more involved in real world statistics.

The Blog

I have to say that I generally agree with the comments already posted, and enjoyed using the blog as an extension for the class. It was nice to be able to link directly to an article or source material and share that with people from class in a way that I would not be able to do in class. I also appreciated the mix between what people posted for assignments and what they posted for fun. This gave me an opportunity to read other people's writing, and also just see the sorts of things that interested them outside the scope of the class. Overall, I think the blog is a simple but effective tool for furthering the topics discussed in class.

Expanding on why class blogs are cool

So...I found a blog post containing statistics about using blogs in education. (Insert Xzibit meme...) A survey was given, taken mostly by teachers, asking them which blog platform they preferred, and reasons why they liked "edu-blogs." Practically, blogs are easy to use, easy to manage, free, and free from ads. For the students, they seem to like having an online "space" to themselves, that they like to keep neat and accurate for the eyes of not only their teacher, but their peers and the rest of the world that might be reading. The audience feels more real, so the students put forth more effort to please that audience.
Although this report mainly comes from K-12 schools, I still think that a lot of these benefits apply to our blog as well. I did read every comment that was added to one of my posts, and I did put forth more effort into my posts than I probably would have in some notebook-paper homework assignment (that I would've thrown away as soon as I got it back.) In addition, reading all of your posts kept me on the same page of what everyone else was thinking about, which I could then bounce off from. So to answer again, more completely, the blog really was a great idea for this class, and I have evidence for that statement now. Thanks for a good year, everybody.

http://theedublogger.com/2012/07/26/the-state-of-educational-blogging-in-2012/

Using the Blog

Three of my Honors classes have now used blogs, and all have been effective. I think it is a great way stimulate further discussion outside of class. In addition, those that are shy, a blog gives them the opportunity to have their thoughts heard.
Something that I really liked to see was the requirement of comments. This encouraged us to actually read each other's posts and avoid repeating something others had already covered well.

Blog posts great for fostering real discussions

I thought that the blog was a really good idea for this class. It meant that we had to do less formal/essay writing that only Prof. Oyen would see, and more communication with our peers, so that we weren't just starting from scratch at the beginning of every discussion in class. It also forced me to go out and look for bad statistics in the media, on the offense for a change, as opposed to defense from the statistics that I passively hear about watching the news or reading the paper. Reading everyone else's ideas was also very helpful--being able to consistently read what your classmates are writing and thinking about throughout the semester is something that is very rare in other classes.

Blog Post

Overall, I thought the blog posts were a great way to enhance our knowledge, especially in this class. Because many people in this class did not come from an advanced statistical background, I felt it helped many of us get through this class with a better understanding of the fundamental concepts. Also, I feel that the posts made us think outside of the box sometimes, which is a great thing to have in any class.

Blog interpolation

I thought that the blog was an excellent format for extra discussion in within the group and encouraged a more "lively" format than would have been possible otherwise.  One of the things that I enjoyed the most about the blog was the ability to easily link to original source material and develop an opinion based upon another person's conducted research. If we have learned anything from this course it is that interpretation of research and statistics can be broad and it should be the job of the consumer to utilize information at hand in the most efficient way possible. Although I probably would have chosen a different format for assignments I still thought that it was accessible enough that, as students, we could utilize its features in a way that expanded the salience of the course.

Blog

I think the blog was a great way to encourage additional discussion for the class. It provided a medium by which we could bring up topics or elaborate further on topics that were brought up in class. In terms of being able to use outside sources, and cite additional research studies, the blog was very useful. Overall I think the blog was a great way to enhance the class.

What do you think of the blog?

As this is the last week of class, I have one last blog question for you to answer. What did you think of using the blog throughout the semester? Please, post your honest thoughts about using this medium for written discussions and assignments.

Autism on the rise?

I came across some interesting statistics showing the increasing rate of autism in the last decade. In 2012, the CDC estimated that 1 in every 88 children and 1 in every 54 boys has autism. The latest survey done in 2008 showed that autism is up 23% and since 2006 and 78% since 2002. I did not realize how prevalent this disorder had become. I can now see why people are starting to scapegoat and grasp on to any answer they receive. This disorder is truly an epidemic in the United States.

http://www.webmd.com/brain/autism/news/20120329/autism-rates-cdc-2012

Red Brain, Blue Brain

A study was published in the journal PLOS ONE on February 13, 2013 concluding that both Democrats and Republicans use different parts of the brain during the decision-making process. This study was completed by Dr. Darren Schreiber and his colleagues at the University of California, where 82 people (35 males, 47 females) were studied during a simple gambling experiment. Party affiliations were recorded at voter registration records. The results for both Democrats (60 people) and Republicans (22 people) were conclusively about the same when looking at gambling choices, but different parts of the brains were activated when thinking through the decision. The researchers were able to predict an individual's political affiliation 82.9% of the time by looking at the brain scans.
Mean age for participant Democrats=22    Republicans=28

Republican Results:

Use the right amygdala- Associated with the "Flight-or-Flight" response
More fearful dispositions
Greater sensitivity to "threatening stimuli"
Reward, fear, risky decisions
More squeamishness

Democrat Results:

Use the left Posterior insula- Associated with self and social awareness and subjective feeling states. 
The left insula is linked to the "Theory of mind," or the "ability to understand what others might be thinking."
Less attention to "negative stimuli or threats"

These differences are linked to social, psychological and logical constraints.

 

Are these differences linked to pre-determined biology, or by external environmental stimuli which affects us, thus impacting affecting how we think? More than likely, it is a bit of both. I am leaning towards the ideology that it is primarily based on how we are affected by our environment. As Dr. Schreiber stated, "We are hardwired not to be hardwired."

Concerns:

The sample size is too low for such a generalization to be statistically backed. 

The number of Democrats and Republicans that were tested should be somewhat the same amount, and the mean ages are also relatively different from one another to be pronounced. Usually younger adults are Democrats, and older adults are sometimes more conservative. 

Thirdly, there are definite environmental and other biological factors which play a role in how we cognitively process events or information. These factors such as one's gender and if they have a family, "race," socioeconomic status, parents' views, education and where they live all play an important role on how we look at other people, as well as political, economic and social policies. These factors are NOT brought up within the media coverage of this study, so many people think that this study represents the difference between black and white, which it does not!

Schreiber D, Fonzo G, Simmons AN, Dawes CT, Flagan T, et al. (2013) Red Brain, Blue Brain: Evaluative Processes Differ in Democrats and Republicans. PLoS ONE 8(2): e52970. doi:10.1371/journal.pone.0052970

 http://www.livescience.com/27213-brain-scans-predict-political-party.html