Regression Line in Social Science

possible to give a ranking to all the members of the DAC, reaffirming the necessity to improve the methods in which the help is provided. For this reason, the study also included a survey in order to obtain recommendations from donors’ responses to provide feedback and recommendations to improve future assistance processes.

Despite criticism about the ODA processes and the quality of its assessment, Lanati and Thiele (2017) investigated the impact of foreign aid on migration. According their study, a direct connection between these two variables does not exist because it is still unexplored. The effects of foreign aid on migration have therefore not yet been determined, which opens the door to further study of the aid-migration link.

The authors expressed the expectations created on migration by ODA disbursement--under the foreign aid term--from a single pattern of study.

The data used for their research corresponded to the Development Assistance Committee (DAC) of the Organization for Economic Co-operation and Development (OECD), which proved a negative relation between the variables.

This finding is congruent to the initial proposed indication of this research, aid should raises wealth levels among the people being influenced by aid, while at the same time, other possible determinants of migration (education, poverty) are minimized and, as a result, aid reduces emigration.

2.6 Regression Line in Social Science

Quantitative approach represents measurable models for accurate analysis. The vast nature of the research fields creates debate among investigators about the most suitable method to undertake such work. However, the debates and differences in opinions cannot be definitely considered as correct or wrong after an exchange of opinions but are questionable under personal intuition. Therefore; in the particular case of linear regression application in social science, the controversy of its use is present but, at the same time, the model has been implemented in research papers.

The application of linear regression is implemented but all too often unappreciated. For Taagepera, “Most numerical results of regression analysis published in social science are dead on arrival: once printed, no one makes any further use of a single number in those tables and equations” (as cited in Taagepera, 2011, p. 73). About this kind of application, the defense of linear model applications was done by Krause (1994) who said James P. McGregor analysis about the use of linear models ignores many trustworthy developments whose strong, scathing

‧

criticism lead others to believe the absence of positive effects of linear model techniques with a skewed perspective. In addition, Krause exhorted the use of linear and even nonlinear models to embrace more advanced and more distinct approaches for political and social science complexity.

Comments in favor of linear regression are based on the idea that discarding quantitative methods is not entirely possible because of the natural importance of numerical analysis.

Regardless of the existence of other approaches in the study of social science, the selection of the methodology is valid for the exploration paper as long as a relationship of all conditions exits. Gerhan (1999, p. 166) proposed that:

Quantitative analysis plays such a role for students (and other patrons) because it is in the ascendancy in many of their fields of study, not only in the central social sciences of sociology, anthropology, political science, and economics, but also in disciplines often but not always regarded as social scientific, namely history and psychology.

Martinek (2017) elaborated on a review of a series of books aimed at teaching graduate students research methods. In her review, she included five books: Real Stats: Using Econometrics for Political Science and Public Policy by Bailey, Applied Regression Analysis and Generalized Linear Models by Fox, Econometric Analysis, 7th Edition by Greene, Basic Econometrics, 5th Edition by Gujarati and Introductory Econometrics: A Modern Approach, 6th Edition by Wooldridge. The variety of techniques correspond to the wide range of knowledge related to social science with a perspective of politics, economics, and others topics under this discipline. Linear models found in these books are part of the consideration for graduate students to contemplate once applying research. The framework of the linear models is not a limitation for broadening and exploring other models. In fact, the review explained the existence of the linear model as part of the whole compendium of possibilities.

Moreover, the expressions about linear regression or least squares are in favor of their usage in the social science field with solid observations. Coskuntuncel (2013, p. 2151) commented:

In scientific research projects, finding a relationship between two or more variables and then expressing it in a mathematical equation is an important dimension. Regression analysis has an important role in scientific research projects because it allows a researcher

‧

The interdependence of variables internationally demonstrates the correlation that exists because social interaction is not secluded. Continuous contact among actors tie one action to another in a taciturn system. As expressed by Signorino, “traditionally, international politics has been defined as the scope and extent of the relations among independent countries, thought to be the most important elements in world politics. This means that actors as well as their actions are strategically interdependent” (as cited in Hoff and Ward, 2004, p. 161).

Measures connecting democracy are part of the international affairs area. Analyzing them with linear regression to establish the relation between some variables is significant. Kedzie's paper tested “the dictator’s dilemma hypothesis by using linear regression to compare the strength of traditional predictors of democracy including economic development and education, human development and health, ethnicity and culture” (as cited in Best and Wade 2009, p. 256).

The findings of linear model cases are readily at hand. Social capital and health were studied under linear regression by Baheiraei, Bakouei, Mohammadi, Majdzadeh and Hosseni (2016, p. 8) who established that for: “the 'relationship between social capital and health status of reproductive-age women', the linear regression models considered all dimensions of social capital and socio-demographic factors as independent variables and all dimensions of health as dependent variables.”

Analysis of domestic growth in relation to unemployment in Kosovo was conducted to establish a relation between the two parameters related to the situation in the country. A paper by Misini and Badivuku-Pantina (2017 p. 5) scrutinized the dependence of social factors using scatter plot graph analysis between nominal GDP in relation to unemployment, stating "Then, the analysis of descriptive statistics will be included, and in the end the method of simple linear regression will be used comparing unemployment and nominal GDP”.

Another case of linear application conducted by Wong, Palloni and Soldo (2007) referred to the role of wealth in Mexico's old age population and international migration and using a basic regression approach, they considered that observed factors can be captured and measured with respect to migration, age, and wealth. Accounting for the aspects influencing migration, a model was developed for a complex study.

Meanwhile, for Jokela, Elovainio, Kivimaki and Keltikangas-Jarvinen (2008) international migration was found to relate to temperament in Finland. Among the dense analysis tools, a regression line was part of the research. They were able to complete an intense

‧

and extended analysis of temperament and migration in Finland while assessing different variables under different approaches. In their paper they assessed the relationship between the two variables under the linear regression model for the migration distance under the temperament scale among the participants for the corresponding research period.

Finally, in the Latin American region, immigration might have multiple influencers.

Characterized by individual and country differences, the application of linear regression is fundamental for the addition of further theories. The study of Zaman and Shamsuddin (2018, p. 617) proved this by concluding “the linear and non-linear relationships between growth, inequality, poverty, and non-poverty measures in a panel of 18 selected Latin America and the Caribbean countries by utilizing the different household available surveys for the periods of 1981–2012.”