Simple Article Critique

Running head: ARTICLE CRITIQUE 0

Article Critique – Binary Logistic Regression

Tonia Joseph-Armstrong

Class – RSCH 8260

Professor – Craig Marker

Summer Term

Assignment 2

Binary Logistic Regression

Predicting Social Trust with Binary Logistic Regression

Article: Adwere-Boamah, J., & Hufstedler, S. (2015). Predicting Social Trust with Binary Logistic Regression. Research in Higher Education Journal, 27.

Introduction

Predicting Social Trust with Binary Logistic Regression. An article Adwere-Boamah, and Hufstedler (2015), seeks to identify the determinants of social trust on various individual adults of a particular nation sample. The study used binary logistic regression in order to evaluate the level of social trust individual adults using the five demographic variables identified as sex, race, level of education, general happiness and the importance of personally assisting people experiencing trouble. The research based its evaluation and study on sample individual adults who participated in the General Social Survey (GSS).

Summary evaluation

The study identified the role each of the predictors/variables played in determining social trust. The findings on the level of education in relation to the level of social trust in an individual revealed that the level of education is the strongest predictor of low social trust as compared to other variables/predictors. A school drop out for example, is likely to have a low social trust as compared to an individual who is a degree graduate. This finding reveals how the level of education had a profound effect on the social trust of a particular individual. The study also established that race as a predictor had a fair share of influence on the social trust of a particular individual such that the African American had a low social trust as compared to the whites. In relation to sex, females tend to be less trustful as compared to males. Happy people on the other hand are more trustful than those people who are often less happy (Adwere-Boamah, & Hufstedler, 2015). In summary, all the five variables displayed some impact on social trust as it either affected one Social trust negatively or positively. Therefore, binary logistic regression is multifaceted on various variables/predictors.

Appropriateness of the test

The binary logistic regression testing has proven in this case that it is appropriate since it applied various dimensions in evaluating social trust of various adults. It applies various means of estimating various parameters thus by application of these variables, it is then able to realize better, reliable and efficient results. The application of the five predictors in this scenario makes it appropriate as it makes it a multidimensional approach to establishing the factors affecting social trust and how these variables affect the trust of particular individuals (Adwere-Boamah, & Hufstedler, 2015). The choice to study specific individuals also makes the test achieve reliable results as it widens the scope of study for the test thus making it quantifiable and efficient tool of analysis. An application of data from the General Social Survey (GSS) also enables the tests to achieve credible and reliable results thus making the test appropriate.

Use of figures and tables

The authors of this journal did not approach tabling as one of the strategy of delivery of data though it analyzed its data in figures and ratios which also proved effective. An example is the use of the 12.7 ratio scale to estimate the number of times an individual with low educational level is going to have less social trust as compared to an individual with a higher educational level such as a degree graduate who is likely to have 12.7 times more social trust. The use of figures and ratios helped the authors develop understanding in the reader of the article and thus able to comprehend the relationship between the variables such as sex, race, general happiness, the significance of an individual helping other individuals experiencing trouble and the level of education of that individual in relation with their social trust (Harrell, 2015).

Presentation of results

The results of the test on “Predicting Social Trust with Binary Logistic Regression” stand out and appear to be more precise, clear and easy to interpret therefore effective for testing. The results clearly indicate the relationship between the distinct variables/predictors and how these variables affect the social trusts of particular individuals. By approaching a multidimensional approach to the study, the test therefore can achieve reliable and credible results (Allison & Allison, 2012).

Conclusion

After a closer view and consideration into the effectiveness of the binary logistic regression testing it is reasonable to justify that it is indeed effective and efficient in achieving better and reflexive results thus I would recommend for this kind of testing for researchers.

References

Adwere-Boamah, J., & Hufstedler, S. (2015). Predicting Social Trust with Binary Logistic Regression. Research in Higher Education Journal, 27.

Allison, P. D., Allison, P. D., & SAS Institute. (2012). Logistic regression using SAS: Theory and application. Cary, NC: SAS Institute.

Harrell, F. E. (2015). Case Study in Binary Logistic Regression, Model Selection and Approximation: Predicting Cause of Death. Regression Modeling Strategies, 275-289. doi:10.1007/978-3-319-19425-7_11