ABSTRACT
A social work researcher might use a ratio as a base from which to determine the amount of deviation. The researcher would proceed to investigate what variables might account for the variance in scores and try to quantify how much each of the identified variables contributed to the total variance. Power is not a constant and varies in each application of a statistical test to the research data. Changing the significance level is very different for a researcher than increasing sample size or reducing variance which are design issues in terms of data gathering. Ordinal level data are the same as categorical level data except that the categories have ranking with a hierarchy of the values on those categories. For statistical analysis, interval and ratio level data remain ungrouped; however, sometimes for ease of presentation and understanding, the data is grouped.
