The knowledge of statistical tests is now becoming an area of interest to nursing research, to come up with evidence-based interventions whose primary objective is to provide first-level health-related care to patients (Clifford & Gough, 2014). Therefore, as elaborated by Main & Ogaz (2016), there are statistical tests that measure the relationship in groups and on the other hand, others compare the difference between groups depending on the research questions of the researcher. One way ANOVA is one of the common statistical tests that are of interest to nursing research. The latter is used when a researcher is interested to know whether there is a difference between groups (if means of the groups differ) of an independent variable in a case where you have a continuous response variable (Gravetter & Wallnau, 2012).
One way ANOVA is best considered for research designs that have either a nominal or an ordinal categorical variable, as the independent variable, and a continuous dependent variable. However, a test for homogeneity of variance has to be done to confirm if the populations of each group have a normal distribution, by testing if they have a common variance (Assaad, Zhou, Carroll & Wu, 2014). If the assumption of homogeneity of variance is violated, then the researcher has to consider non-parametric tests for his/her study. One proceeds with the way ANOVA test if the assumption is tenable. Moreover, once the ANOVA test is conducted and there is a statistically significant difference in the groups, then a post hoc analysis is performed as a follow-up analysis to determine which groups differ and how they differ (Assad et al. 2014).
A study on the participants’ hemoglobin scores based on their diet group can be well analyzed using one way ANOVA. In this case, the diet is an independent variable with three-factor levels; ‘iron-rich foods & sulfate tablets,’ ‘sulfate tablets only’ and ‘normal diet.’ Hemoglobin scores is our dependent variable, and it is a continuous variable. The research question here is whether there is a difference in mean hemoglobin scores across the three diet groups. In this example, the test is statistically significant (p-value (sig) = 0.00 < .05). There is a significant difference in the participant’s hemoglobin mean scores based on their diet group. The post hoc analysis shows that every diet group differs from the other. Upon testing the effect size, 0.05 was the effect size, and we define it as a medium effect in reference to a study by Sullivan & Feinn (2012).
Assaad, H. I., Zhou, L., Carroll, R. J., & Wu, G. (2014). Rapid publication-ready MS-Word tables for one-way ANOVA. SpringerPlus, 3(1), 474.
Clifford, C., & Gough, S. (2014). Nursing and health care research. Routledge.
FREDERICK, J., GRAVETTER, W., & LARRY, B. (2012). Essentials of Statistics for the Behavioral Sciences. CENGAGE LEARNING EMEA.
Main, M. E., & Ogaz, V. L. (2016). Common Statistical Tests and Interpretation in Nursing Research. International Journal of Faith Community Nursing, 2(3), 5.
Sullivan, G. M., & Feinn, R. (2012). Using effect size—or why the P value is not enough. Journal of graduate medical education, 4(3), 279-282.