Week7 D1 Research Design for One-Way ANOVA
The logic of ANOVA is the same as a t test, which is to test the null hypothesis (Warner, 2012). However, a t-test is limited to comparing 1-2 groups, while ANOVA allows for testing 2+ samples (Warner, 2012). When ANOVA procedures are applied to data with one dependent and one independent variable, such as race and mental health literacy (MHL), it is called a one-way ANOVA (Warner, 2012).
For example, Rafal, Gatto, & DeBate (2018), using a one-way ANOVA, seek to answer the research questions, Will race of college male undergraduate students (IV) cause differences in MHL (DV)? The authors create 3 levels of the nominal variable, race. They are: White, Asian, & Other. The null hypothesis states, There is no difference in MHL (DV) caused by race of college male undergraduate students (IV). The authors state, among undergraduate participants, statistically significant differences were observed by race for overall MHL; F(2,790)=18.953, p= <.001 (Rafal, et al., 2018).
A quick view of a table suggests 3 comparisons of MHL were made based on race, White Vs Asian, White Vs Other, Asian Vs. Other (Rafal, et al., 2018). The former two (White Vs. Asian & White Vs. Other) were statistically significant (p=<.001*), while the latter comparison group (Asian Vs. Other) was not (Rafal, et al., 2018). This means scores on MHL significantly differed when White was compared to Asian, and when White was compared to Other. Specifically, White males had higher total MHL than Asian male undergraduates (p <.001) and Other undergraduates (p<.001) (Rafal, et al., 2018).
Overall, authors can reject the null hypothesis because their data suggests race causes differences in MHL, and the differences are statistically significant.
References
Rafal, G. Gatto, A., & DeBate, R. (2018). Mental health literacy, stigma, and help-seeking behaviors among male college students. Journal of American College Health, 66(4), 284291. https://doi.org/10.1080/07448481.2018.1434780
Warner, R. M. (2012). Applied statistics from bivariate through multivariate techniques (2nd ed.). Thousand Oaks, CA: Sage Publications.
