Variable Operationalization - Essay Example

Variables for the research were operationalized using a Likert Like scale running from 1 to 10. Correspondent's perceptions therefore operationalized the dependent and independent variables. The element of operationalization raises serious validity and reliability concerns since the estimates are prone to various forms of errors.

The validity and reliability of the estimates

The fact that all variables in this research were derived from a subjective viewpoint of the respondents raised serious validity and reliability concerns. Validity and reliability are important concepts in any given research since they are used to enhance the accuracy of measurement as well as the evaluation of any given research (Tavakol and Dennick 2011). Creswell (2014) noted that they have different meanings under various types of research (either quantitative or qualitative).

Reliability is a term used to describe the stability and consistency of results of any given research (Twycross and Shields 2004). Validity, on the other hand, refers to the extent to which any given instrument measures its intended measure (Thatcher 2010). The research is valid based on several considerations, most of which rely on the researcher being able to attain all mentioned research objectives. The main cause of concern is the validity of the OLS estimates used to conduct the bivariate regression analysis. More so the issue is raised by the application of the five classical linear regression model (CLRM) assumptions (Brooks 2008). The main threat to internal validity in this study was the threat of omitted variable bias. It was however rectified by including the omitted variable. The other source of threat to internal validity is model misspecification.

The reliability of the estimated model was however moderate as indicated by the coefficient of determination R2. The R squared indicated that the chosen dependent variables could explain 48.5% to 58.5% of the variables interactions.

Relevant control variables

The relevant control variables, in this case, would involve collecting data from participants who dont belong to the civil society and hold liberal views on the role of civil societies in the democratization process. The control variable would, therefore, be people without any opinion on the role of civil societies on the enhancing or showing the legitimacy of any given government. The role of the control variable was to show the true impact of the dependent variable (s) or the absence of the independent variable. This analysis, however, was complicated by the rather subjective nature of the dependent variables. The other control variable would be to involve respondents from autocratic societies such as Saudi Arabia and ask their opinions on the roles of civil societies in the determination of legitimacy of any given government

Statistical estimators

The research depended on OLS estimators in arriving at its main conclusion. This implies that the inherent weaknesses of OLS estimators must have affected the final result. The statistical estimators used in the model specification for the regression analysis exposed the outcome to the reliability and validity issues affecting OLS estimators courtesy of the CLRM assumptions. This implies that there was a need to construct the model and choose the one with the highest coefficient of determination. The quality of the estimators was however not established in this research. Ordinary Least Squares (OLS) estimators were used for estimating the unknowns in the linear regression model used in this study. The result of the analysis indicated that despite the rather presumptuous nature of OLS estimators, they could yield high-reliability outcomes since the coefficient of determination is a rather stable estimator of the level of variable interaction in any given study.

References

Brooks. C (2008). Introductory econometrics for finance, 2ed. New York, Cambridge University Press.

Creswell, R. (2014). Research design: qualitative, quantitative, and mixed methods approach. US: Sage publications

Tavakol, M., & Dennick, r. (2011). Making sense of Cronbach's alpha. International Journal of Medical Education, 2, pp.53-55

Thatcher, R. (2010). Validity and reliability of quantitative electroencephalography (qeeg). Journal of Neurotherapy, 14, pp. 122-152.

Twycross, A., & Shields, l. (2004). Validity and reliability - what's it all about? Part 2 Reliabilityin Quantitative studies. Pediatric Nursing, 16 (10) p. 36