During the shift to using a large-n approach to research, I found complications in finding data sets to help research the question I was asking. My question shifted to ask what explains the trend in needing recognition to be considered a state. Through researching various databases, I was able to find a few sources that would help me be able to unpack my question better. I decided to use years as a state system member as my dependent variable to help explain the stability of the state. The idea behind this was to see if there was any correlation in the stability of a state compared to how long they have been considered a member of the state system.
The first dataset I located is from the Correlates of War Project database that focuses on State System Membership.[1] The data measures from 1816 year-by-year of what states have been included as part of the state system. A state is considered part of the state system before 1920 if they have a population greater than 500,000 and have diplomatic missions at or above rank with Britain and England. Post-1920 they must be a member of the League of Nations or UN or have a population of more than 500,000 and have diplomatic missions from two major powers. This can be beneficial especially when seeing large amounts of new states being entered such as in post-World War II and collapse of the Soviet Union. The limitations are that the data can be somewhat limited in the considerations of the state system as the data could be differentiated heavily if the conditions were ever so slightly tweaked.
Another resource I wanted to use was the Fragile State Index Report of 2016.[2] This report uses different analytics to a sum of scores from twelve separate indicators that can show a state’s strengths and weakness. They are rated on a scale of 1 to 10 and are divided into social, economic, and political categories. The data in 2016 covers 176 states, and I choose the 2016 report specifically because that is when the date ended for the Correlates of War Project database. The limitations of this source are that it focuses entirely on almost the negative aspects of state-building. Even the title implies that almost all states are fragile, it is limited including methods of state building and time period.
[1] “State System Membership (V2016) — Correlates of War.” n.d. Folder. Accessed October 15, 2018. http://www.correlatesofwar.org/data-sets/state-system-membership.
[2] Fragile States Index 2016 – Annual Report | Fragile States Index.” n.d. Accessed October 15, 2018. http://fundforpeace.org/fsi/2016/06/27/fragile-states-index-2016-annual-report/.
Katherine says
Griffen,
You seem to have a well-developed understanding of how your project would proceed through a large-n statistical research approach. and it seems like your 2 databases each take different approaches. However, my question would be to what extent do you think that they will combine to create one cohesive variable? When reading over your description, it appeared as though one of your data sets focused on showing if a country was or was not recognized as a state for each year. While your other data set seems to be focused on the strengths and weaknesses of states. While this seems like it could relate to your project, I’m not seeing a direct connection as to how it shows your dependent variable. For this reason, it seems like it could be beneficial as something else in your research, possibly as an independent variable? In addition, I appreciated how you turned the first dataset, which could easily be viewed as simply ordinal, into interval ratio data which would be more useful for an increased number of methods of analysis. Finally, my last question would be, how do you plan on dealing with any potential differences between what’s considered a state by different datasets? This seems like it could potentially cause difficulties in terms of your analysis if not addressed, because then one dataset would have a country for years that another does not include it. For this reason it seems like something that might be beneficial to consider as you move forward. I am interested in seeing where you go with this and if you choose to incorporate aspects of this proposal into your actual research project.
lf5995a says
Griffin,
I like how you rethought your puzzle. It seems to be well fit for a large-n research study. I think it’s very interesting that you have chosen to include the Fragile State Index Report, as this source wouldn’t immediately come to mind when thinking about why a state needs recognition. However, upon further thought, it is very clear that weak states are likely in need of global recognition the most. I am intrigued by this approach to your puzzle and look forward to reading your future posts! See you in class!
Theodora Mattei says
Griffen,
from your post it is plain to see that the large-n approach to your research puzzle is starting to take form! One question I have for you, that I have also posed to myself, is how do you plan to reconcile the different time frames from which you have found data? Similar to your experiences, I too have found sources that somewhat overlap, as both your datasets measured statehood in 2016. However, most sources measure data from longer or different periods of time, as your first dataset looked at states from 1816 onward and your second only collected data in 2016. Therefore, do you plan to find more sources to fill in the gaps from those years, or will you create your own, possibly smaller time frame? I am excited to see where your large-n analysis takes you. Good luck in your research!
joshuaoday says
Griffen,
I also found the Fragile State Index Report of 2016 a useful source. I definitely found it challenging to quantify what I am looking at, and your post is a good example of how to pull out a DV. I did not consider length of a state’s recognition. I do think the challenge with the data will be states that are not considered fragile at all, but if that state has had several splits during the 1800s. I am curious to see what your data shows for “younger” states, and what variables show stability. Hopefully we can sit down, and discuss your research because I think it will help me sort through some issues I am encountering in my research.
Josh
Dr. Boesenecker says
Both of these datasets provide you with good primary source information for your project, Griffen. I can see that you’re grappling a bit with how best to conceptualize the DV for this variant of your research, and that is fine at this stage (and something we can discuss going forward). Josh’s comments might well be helpful here as you sort out whether, in this large-n variant of your project, you are aiming to explain variation in recognition times (important, but perhaps also a bit tricky) or variation in some other aspect of a state and its membership in the system. The way that you word things in the first paragraph it is almost as if you are thinking of years as a state system member as an *independent* variable to help explain the stability of the state…so just keep thinking about which of those is the DV (the thing to be explained) as you go forward.