Everyone loves statistics they agree with! Otherwise, they become methodologists.

Hello, hello!

Man, it has been a long time! I checked my last post and it was about 6 years ago at this point... I knew it had been a while, but I didn't realize exactly how long it was. Craziness!

Why am I even picking up the (metaphorical) pen again? Jonny and I had recently met up with Claire and we were going through a nostalgia trip. One of the items mentioned was my blog... I've just been thinking about it in the back of my mind, but it's intrigued me little by little. I feel like I finally have an idea today of something to possibly write about. I don't know if I'll keep going (seeing my consistency before, not likely!), but it feels nice typing something out.

For a quick life update, I'm currently working as data entry and analysis for an oil company. It's been a great learning experience, and I've developed a love for Excel and VBA. Macros are amazing! Well, programming itself is amazing, but I digress... If anyone ever needs any Excel help, I'm always down to lend a digital hand.

Hopping back to analysis, I've seen so many statistics getting thrown around these days... Whether it be talking about the health benefits/detriments of substances or the percentage of support for a candidate, it's hard not to see some percentages pop up these days. It reminds me of an old phrase,

"90% of statistics can be used to say anything 50% of the time."

While quite silly, it points out that statistics can be misleading... The facts themselves may be true, but there are two main points of contention:

1. The population of the data
2. The usage of the data

Take, for example, the statistic that 4 out of 5 dentists prefer Orbit gum, something they touted in old commercials (I don't have cable anymore, so I may be dating myself). Let's say they were given the options of Juicy Fruit, Orbit, and Take 5. To make things simple, let's say all scientists...

1. Chewed Orbit for 21 seconds
2. Chewed Take 5 and Juicy Fruit for 20 seconds
3. Had to put a :) or :( next to each brand; 4 dentists put :) near Orbit, with one dissenting.
4. Wrote an adjective next to each brand of gum, with two dentists putting "flavorful" for Orbit. Only Juicy Fruit received one other "flavorful" description

Some statistics you can make are:

1. Orbit lasts longer than all other competitors
2. More dentists preferred Orbit over other leading brands
3. 4/5 dentists prefer Orbit
4. Twice as many people think Orbit is more flavorful than Juicy Fruit!

Ok, hopefully my point is coming across. While all these factors themselves are true, they don't mean much in terms of absolutes. If you looked at the base data, you can see that there is not that much of a difference between all measures of this study. Yet, saying "Twice as many people..." carries some weight behind it if you don't look at the data.

Also, even if these stats are true, how applicable is to different populations? Are these dentists all people who have grown up near an Orbit gum factory and chewing Orbit gives them the nostalgia of home? Are these dentists all over a certain age? There's dozens of points you can come up with, questioning how applicable these studies are to everyone.


I remember for psych classes and experiments, there are certain guidelines and reviews that are followed. I'm really rusty at remembering what (considering it's been over 5 years since I've been out), but there were controls to help avoid some common issues. Yet, even then, there was talk of pressure to find something. Typically, with statistical experiments and analysis, you are trying to find something. If I remember my high school stats (Mrs. Smith, please don't hate me) right, it's specifically trying to "reject the null hypothesis." If you fail to find statistical evidence going against the hypothesis in place, you "fail to reject the null hypothesis." Part of the issue is that if you receive grant money to study something, yet your experiments don't turn up anything new, it doesn't look the best for you...  If your experiments aren't fruitful, it's hard to receive more money to conduct other experiments. This results in a lot of pressure in trying to find something, anything to help reject the null, which in turn can be misleading..


My point is, in order to prevent being misled, you would have to go down to the basic data and analyze it there. Yet, honestly, who has the time to examine the methodology for every single statistic that pops up? Even just reading the above takes some time and it's meant to be simple... While that does raise questions about my verbosity, the point still stands; it's unrealistic to analyze every single stat and fact that comes our way. Yet, in not doing so, we make ourselves susceptible to false information.

Why does it even matter?

We tend to disagree with statistics that question our world view, but we readily accept any statistics that further our own. Humans are naturally cognitive misers, which means that we want to know everything with the least amount of effort. If something comes up that's new/haven't heard before, normally we accept it. If it's new information that conflicts with our current worldview, only then will we break it down to see where the fault lies within. With the amazing world wide web, you can find information about so many things. Unfortunately, it also means that you can easily find statistics and information supporting your own views, whether they be right or wrong.

My point is, I think it's wise to take a step back every now and then to question what we take as fact. It's fine to have an opinion or take a stand for something, but it's important to recognize how you got there. Opinions, like stress, can affect you in small dosages. If you're hit by a great stress, your body can recognize and work against it. On the other hand, if there's a multitude of small stresses, some that you may not even be aware of, it may do more damage/change in the long run. It's all too easy when reading or browsing to find other facts or opinions shaping and agreeing with your worldview without your recognition *cough cough Reddit*. As long as you're aware and take a step back to recognize your own growth and change, though, you may help to eliminate your own biases.

Anyways, I feel like my hands can't fully translate what's floating around in my brain completely just yet. I may have to take another crack at writing something else in the future...

If you made it this far, thank you for reading!

Dusty