The research design will be classified as largely quantitative due to the analysis stemming from the data involved with the county-level voteshare information. The goal of the retrospective causal-comparative research design is to identify associations among variables. Through this experiment similar groups of voters (separated by region in Texas) would be analyzed in relation to the same independent variable (realigning elections). As opposed to a correlational study that would involve relationship, the causal-comparative study involves only comparison. This type of quasi-experimental research design is simplistic in approach and will only see the independent variable identified, but not manipulated by the experimenter. The effects of the independent variable on the dependent variable are then measured. Difficulty may arise when moving to the analyses and conclusions stage in that determining causes will have to be done carefully and take into account, as readings from the PSCI. 6350 course have suggested, variables that are both known and unknown which could have impact. It is known, for instance, that with these types of studies causation must be assessed thoroughly before affirming relationships amongst variables. Alas, due to the fact that this is a retrospective causal-comparative research study it will also need to incorporate a qualitative aspect in order to fully explain the past history that is required for understanding why a topic such as this would be investigated in the first place. It will also be necessary in order to present the political climate in Texas leading up to each election in which a voteshare will be collected. The Texas State Historical Association (an Independent Nonprofit since 1897), along with the Handbook of Texas Online, would be helpful guides for crafting the qualitative side of the research design. The threats to the causal-comparative design are brought about from the lack of randomization and inability to manipulate the independent variable in any way. The lack of randomization occurs due to the fact that the individuals (voters) are not randomly assigned into treatment groups; in fact, they are pre-selected into groups before the research begins. Another threat is the possibility of a subject selection bias. Additionally, a problem could occur if the groups that are being used are different in some significant way other than the identified independent variable. One way to possibly ameliorate this would be to break the groups down into subgroups. This is a consideration that should be taken into account by any researcher taking this topic on. Doing this would allow for the added benefit, also, of determining whether the independent variable affects the dependent variable differently in any other ways (and at different levels). Following this step, the data interpretation for this research study (should it be conducted some day) should involve both descriptive and inferential statistics. In terms of descriptive statistics, this would include the means and standard deviations. When talking about the inferential statistics, this could possibly include a t-test, analysis of variance, or chi square test.