Data Analysis Part Two

Hey! Welcome back. This is the second (and final) discussion of the data analysis.

So, the results turned out well! Here's a picture of the graph: 

When the number of risky decisions were plotted against their respective blink rates, this graph happened. What's really exciting about these results is that the p-value is equal to 0.0018, which means that there is a .0018 chance that these results are random. This means I did in fact find a correlation between blink rate and making risky decisions!

(This is what I felt like when I found out my results.)

My research consultant, Bob, and I  are really excited about these results. Since I surveyed so many different people that vary in age and gender, the next step I am going to take is to do an exploratory analysis. I am going to try and see if there are differences in blink rate when it comes to variation in gender and age. (For the purpose of this paper, though, I am just focusing on the bigger picture and the graph shown above.) 

Of course, I conducted a single study with only 50 people. More experiments need to be done to ensure that blink rate is definitely an indicator of how people will make decisions, but this is certainly a good start. 

There were some outliers, but there usually always are when it comes to collecting data from a large group of people. One person I surveyed is a wealth advisor, and he understood mathematically the risks that I was presenting him with. While most people saw the questions as bets or for sure gains, he went deeper and weighed the actual risks if one were to make the same decision many times. Even though I was presenting these options as a one time choice, it was still interesting to hear someone’s perspective on which decisions were better to make in the long run.

I also talked to someone who took no risks whatsoever. I think that this participant was under the impression that I was trying to get people to make riskier decisions, or that all of the risky options were secretly the “wrong” decisions, so he played all of them safe. However, in my survey I specified in the beginning that there were in fact no wrong answers. I also did not want everyone to answer the questions in one particular manner; I was hoping to see that they tended to prefer riskier answers if their corresponding blink rates were higher. Astonishingly, I found that result!

I would like to once again take this time to thank everybody who participated in my study. I approached many people, and a lot of them just flat out said no. I also promised everyone that I would compensate them with candy, but only three people actually wanted candy afterwards. It was really nice of everyone to help me out. Also, shoutout to my math teacher, Mrs. Bailey, for being really supportive. If you’re reading this, hi! 

Thanks for reading, and I hope you come back next week. 

Word Count: 513 

Data Analysis Part One

Welcome to my third blog post! This time, I’m going to talk about how I analyzed the data I collected. 

As I mentioned before, when people were filling out the surveys, I watched the videos to count their blink rates. I then wrote down their blink rates on their surveys once they were done with them. Once I had all 50 surveys, I went through each one of them and counted the number of times that person choose a risky answer (out of nine questions total). 

So then I had 50 blink rates with their corresponding number of risky decisions. I put all of my data in a Google spreadsheet, and then figured out what I needed to do to find the p-value, which basically would tell me the significance of my results. 

In order to do this, I had to use MATLAB, which, according to their website, is a high-level technical computing language for algorithm development, data visualization, and data analysis.  

I figure my learning how to use MATLAB is similar to how Marco Rubio learned about EDM (did you know that he said he listens to EDM - electronic dance music that is mostly played at raves?) 

Marco Rubio’s staff: Okay, we need to find something that will make younger people like you more, Marco. 

*Staff google searches something along the lines of “what do teens like”*

*Results for music show up, someone sees an abbreviation called EDM* 

Marco Rubio, after he listened to one song and was briefed on what to say: I like EDM and know what it is! 

My research consultant, Bob: Okay Emily, you should learn how to use MATLAB to analyze your data. 

*Google searches something along the lines of “how to use MATLAB”* 

*Results show up for how to enter data, make graphs, and find certain values*

Me: I like MATLAB and know what it is! 

You see, both Senator Rubio and I have preliminary understandings of what each of our respective things are. Marco needs to learn that EDM is, in fact, not a clean genre of music like he says and that its culture is not really for the GOP. On the other hand, I only have a pretty basic understanding of MATLAB and therefore can’t really compare it (and therefore prefer it) to other types of computer languages. 

MATLAB is super useful though. I had to use codes to put all of the data in a graph and then find the p-value. When I ran it, the p-value was a matrix. I turned to my math teacher, Mr. Peacher, for help, and he sat down with me and we figured out how to run the codes. We both agreed that MATLAB is pretty confusing, but definitely a great tool to know how to use. My goal is to gain a better understanding of MATLAB and other computer languages so that in the future I will be able to analyze data and find exactly what I’m looking for. 

Next week, the results will be written about. Stay tuned! As always, thanks for reading. 

Word Count: 516 

Implementation of Research

During the month of January, I filmed and surveyed 50 people in total. First, I had every person sign a consent form. After that, I filmed each participant for two minutes, and then gave him or her a survey which had nine questions. Each question presented one option with a definite outcome, and one option with a "risky" (uncertain) outcome. The questions were based on financial decisions.

Here's a question from the survey so you can see:

If you were faced with the following choice, which alternative would you choose?

_ A sure gain of $240

_ A 25 percent chance to gain $1000 and a 75 percent chance to gain nothing.

The people who chose the sure gain were deemed not risk takers, while those who decided to gamble were risky decision makers. 

While the participants were filling out the surveys, I watched the recorded video to count their blink rates. Once I was done watching the video, I deleted it to ensure confidentiality. I then wrote the number of blinks on the respective survey for future data analysis. 

Finding 50 people to participate was a bit of a challenge, but I was able to use anyone over the age of 18 for consent purposes. I used classmates and teachers at my school, but that did not amount to 50. The head of my school, Elizabeth McConaghy, sent out an email to all of the parents of students at my school asking for volunteers. A few people contacted me, and then I had 50 people. I met some people at the library, but most at school. I met with each person in a closed room with no distractions for a measure of control.

The implementation went really well, and I did not encounter anything that I didn't expect. The method I designed basically matched up perfectly with the reality of the experiment. 

The whole experience was really fun, and it reaffirmed my interest in conducting research in college and beyond. It was surprising how many people were (or at least seemed) interested in my research! I think that neuroscience and psychology interest everyone, on at least the most basic level, since they speak to who we fundamentally are. Talking with people made me realize that so many more people would be interested in studying science if the importance of the connection to themselves is clearly made. 

If I had the resources, I would use a lot of fancy equipment like fMRI and EEG to measure dopamine levels in the brain. Since I'm just a high school student, though, blink rate has been the easiest and best way to do this. If blink rate really does turn out to be a significant indicator of dopamine, this would be great for not only the scientific community, but for other places like rehabilitative services and maybe for practices of medicine. Blink rate could tell if someone is more impulsive, and if physicians or caretakers are able to see this in such a non-invasive manner, it would be much more useful when determining which treatments or medications should be prescribed.

Word Count: 518