Stephen D. Krakie
12/13/17
Techniques in Computer Analysis
Reflective Paper
Help Received: None

When I first started this class, I was skeptical that I was going to learn anything new or interesting. However, throughout the past semester I found myself learning much more than I thought I would. Not so much about statistics and how to manipulate variables and whatnot in SPSS, but about how to actually conduct a good political science experiment. That was by far the most interesting and relevant thing that I took from the course.
When I started this class, I was honestly dreading it due to the focus on statistics and manipulating numbers and variables. All things that I thought I had left behind in statistics rat year. Therefore, I was worried that I would even be able to pass the class thanks to my barely passing those classes years ago. I would do the reading and not understand what they were talking about, and I was very stressed in the first few weeks about how I would be able to do in the class, especially when my first quiz grade came back as a D.
One thing that I did find useful in learning the concepts was SPSS. Even though while I would be doing the assignments I would become very frustrated almost every time, if I sat for long enough with it and tinkered with the functions enough I found that 95% of the time I could get it figured out. I found that simply sitting and messing around with different functions was more useful for me to understand what was going on then reading the book itself. I would read the book to get the general gist of what the functions we were doing were, and then sit and work until I eventually figured it out. And I did fairly well on most of the homework because of this.
Another thing I learned that is important is what really goes into setting up a political experiment, such as the difference between a theory and a hypothesis. I even commented in one of my blogs about how much more difficult it was to build a theory and hypothesis than I originally thought. Also, I learned about what independent and dependent variables actually look like in a political experiment. Although I’ve learned about these before in other statistical classes, it was helpful to actually see what they looked like in practice and having to build our own.
The best learning experience I have had throughout the semester though has been working through the group project. We have not even finished our project yet and I have already learned a lot from the mistakes we’ve made already. One of the biggest lessons I have learned that I mentioned earlier is that, if you are giving a survey, and you want a very specific kind of answer to your questions, specificity is everything.
When we built our survey, we assumed at times that people taking it would assume to answer how we wanted them to. This was not the case for several of our respondents. Several questions, such as our question about what ROTC you are in and if you are commissioning or not, was one that was misunderstood by almost everyone. Therefore, we had to adapt how we were going to collect the data based on the kind of responses we got. And then of course there were the people who just didn’t take it seriously and gave us goofy answers, that we had to somehow find a way to record.
So, after collecting our data we created our scale, and we were very careful and methodical in how we set up the scale in order to maximize our variables. However, when it came to applying our variables to our scale we ran into even more problems that I mentioned in my blog. Often the responses did not fit cleanly into the scale. For example, and response for how many hours a week that someone spends working about might be 12-15, but our scale only measured in increments of two hours. Therefore, we have to adapt and chose to record whatever was the lowest end of the responses in cases like this.
The final lesson I have learned from this experience is how to actually apply the SPSS functions to our data. I have to say it was very satisfying after spending hours scaling our data and entering it all by hand into SPSS and researching which functions we should use, to have SPSS do the analysis for us in a half second. We were able to see where exactly the correlations were, and were surprised by what we found, learning that you’ll often find things you did not expect in these experiments.