The Variables Associated With Health Promotion Behaviors Among Urban Black Women (Hepburn, 2018). https://doi.10.1111/nu.12387

            Research questions (RQ) determine the type of statistical test to use. The table below shows the four types of RQ, types of statistical tests, and the key or essential words.

Source: Walden University Academic Skills Center (2021).

 Correlation is a statistical tool used to measure the existence and strength relationship between two variables, dependent and independent variables, while in linear regression, the independent variable predicts the outcome of the dependent variable (Dendukuri & Reinhold, 2005; Frankfort-Nachmias et al., 2020). The continuous variables in correlation are interchangeable (A is related to B, same as B is related to A) but fixed in linear regression (A predicts B not same as B predicts A) (Walden University Academic Skills Center, 2021).

            The researcher in the above article tries to establish a relationship between health promotion predictor (HPP, independent variable) and health promotion behavior (HPB, dependent variable or outcome) among urban black women (N=132) ages 30 to 64 years in a United States metropolitan region in 2015. The HPP consists of health literacy measured with Newest Vital Sign (NVS), self-efficacy measured with New General Self-Efficacy Scale (NGSE), and readiness for change measured with Health Risk Instrument (HRI), while HPB measured with Health Promotion Lifestyle II (HPLPII). The author used a quantitative non-experimental cross-sectional survey research method and bivariate/simultaneous linear regression (Research design) to determine the relationship between the NVS, NGSE, and HRI to HPLPII. Below is the RQ, the null hypothesis (H₀), and the research hypothesis (H₁)

 RQ:     To what extent is there a relationship between health promotion predictor and health

            promotion behavior?

H₀:       There is no significant relationship between health promotion predictor and health

            promotion behavior

H₁:       There is a significant relationship between health promotion predictor and health

            promotion behavior (Walden University Academic Skills Center, 2021).

            The independent scale variable (NVS, NGSE, and HRI) is used to predict the dependent scale variable HPLPII, thus the author’s choice using the bivariate/simultaneous linear regression as it is the most appropriate tests statistic for the combination and type of variables.

            The author did not use table to displays the data but clearly highlighted with figures that there was a positive correlation between the independent variable constructs to the dependent variable as follows: NVS (r = .244, p < .002), NGSE (r = .312, P < .001), HRI (r = .444, p < .001). Where r is the sample Pearson’s correlation coefficient value or closeness of the observation in a scattered plot of line regression ranges from -1 to +1, the sign is the direction (negative or positive) while the value is the strength (weak, moderate, or strong); variable strongly correlated as r tends to 1 (Walden University Academic Skills Center, 2021).

            As shown from the above figures, even though the three constructs positively correlated to health promotion behavior (HPLPII), the readiness for change (HRI) was most highly correlated (r = .444), followed by self-efficacy (NGSE, R = .312) and the least was health literacy (NVS, r = .244). Health literacy with the perspective of level of education, r = .414, p = .001. The p-value indicated that the study is statistically significant when related to the threshold of p < .05 (Frankfort-Nachmias et al., 2020). Thus, the result is stand alone as one can easily interpret the study from the highlighted figures in which H₀ is rejected, confirming H₁ that there is a significant relationship between health promotion predictor and health promotion behavior.

            The author reported an effect size, R squared (R²) of .298, which means that 29.8 percent of health promotion behavior (HPLPII) is related or associated to health promotion predictors. The effect size and p-value indicate that the study is both meaningful and statistically significant, respectively. The remaining 70.2 percent enable the social change advocates and public health stakeholders to strategize community-based and cultural-oriented programs that include readiness for change, health literacy, and education, especially among black women, to reduce health disparities in the United States.

                                                            References:

Frankfort-Nachmias, C., Leon-Guerrero, A., & Davis, G. (2020). Social statistics for

            a diverse society (9^th ed.). Thousand Oaks, CA: Sage Publications. 

Hepburn, M. (2018). The Variables Associated With Health Promotion Behaviors Among

Urban Black Women. Journal of Nursing Scholarship, 50(4), 353-366. https://doi.10.1111/jnu.12387

Dendukuri, N. & Reinhold, C. (2005). Correlation and Regression. Fundamentals of Clinical

Research for Radiologists. American Journal of Roentgenology, 185(1), 3-18.

https://doi.org/10.2214/ajr.185.1.01850003

Walden University Academic Skills Center. (2021, October 10). Strengthening Your Statistical

Skills. Correlation Bivariate Linear Regression [Video]. YouTube. https://www.youtube.com/watch?v=PKdKqBFX-aU