Techniques of Univariate Statistics Discussion- Response to another learner

Techniques of Univariate Statistics Discussion- Response to another learner

Choose one major technique of univariate analysis and provide a fairly complete description of its primary use or purpose. Be sure to include relevant terms, concepts, assumptions, limitations, and any other information that will provide a useful review for the reader.

YOU ARE TO READ THE FOLLOWING POST DISCUSSION FROM ANOTHER LEARNER AND RESPOND BY SUGGESTING ONE SITUATION FOR WHICH THE GIVEN TECHNIQUE WOULD BE PARTICULARLY APPLICABLE AS WELL AS ONE SITUATION FOR WHICH THE TECHNIQUE WOULD NOT BE A GOOD CHOICE.

The one-way between subjects ANOVA is a statistical test that is used to explore the means between two or more groups on a quantitative dependent variable (Warner, 2013).  For example, say an instructor would like to compare the average final exam test scores across three difference sections of an educational psychology course (an 8:00 am class, 12:00 pm class, and an evening class, 5:00 pm).  The dependent quantitative outcome variable would be final exam test score while the independent variable is the educational psychology course split into three groups.  The one-way ANOVA is considered an omnibus test which means the test checks for whether a difference does or does not exist overall.  It does not, however, tell the researcher where the difference exists (Field, 2013).  Post-hoc tests such as the Tukey HSD can be used to determine which groups’ means are statistically significantly different from one another (Warner, 2013).       

     The assumptions of ANOVA include one dependent variable measured on an interval or ratio scale, one independent variable consisting two or more groups or categories, independence of observations, homogeneity of variance, the absence of significant outliers, and normal distribution of the dependent variable (Laerd Statistics, 2013).  Field (2013) pointed out that some researchers refer to the ANOVA as “robust” (p. 444) meaning that even if some of the assumptions are not met, the test can still be correct.  However, Field cautions the reader to continue to take care to identify and report violations of assumptions. 

     In conclusion, the one-way between subjects ANOVA is preferred over multiple t-tests which compare the means between only two groups because the researcher can limit the chance of making a Type 1 error or rejecting the null hypothesis when the null is true (Field, 2013).  In other words, using ANOVA reduces the chance of the researcher indicating a difference exists when it in fact does not.

Thank you,

Robyn Zeltinger