Many graduate-level research projects involve distributing surveys and analyzing the results that come in. The Likert scale is one of the more popular metrics for attitudinal research. If you're taking a Likert survey, you'll see a series of statements, and you'll be asked to indicate whether you "strongly disagree," "disagree," "slightly disagree," are "undecided," "slightly agree," "agree," or "strongly agree." Whichever answer you choose is assigned a point value, and the researchers conducting the survey interpret the results.

Assign each response a point value, from 1 to 5 or 1 to 7, depending on how many possible responses there are. Some survey designers do not include the "slightly" options on the agree or disagree side. Common values for the options start with "strongly disagree" at 1 point and "strongly agree" at 5 or 7 points.

Tabulate your results and find the "mode," or the most frequently occurring number, and the "mean," or the average response. If your sample is large enough, both of these metrics will be valuable. The mode will tell you the most common response to each statement. And while the numerical values for each response aren't as objective as counting numbers would be, the mean will give you the overall average response.

Create a graphic representation of the responses using a bar graph, giving one column to each of the response choices. Under the horizontal axis, label each of the response choices with the point value, and mark off lines crossing the vertical axis with different numbers -- 50, 100, 150, 200 and so on. These numbers will vary depending on the number of respondents. Choose a scale that will fit all of your response totals but will also show the differences among them meaningfully. If you only have 30 respondents, and your first number on the axis is 100, you won't be able to show meaningful differences among the various columns.

Disaggregate your data as needed for your research needs. You may want to separate data out by age groups, gender, ethnicity, religion or other variables. Create a bar graph for each separate group you want to analyze.

Use one of a variety of variance analysis tests to analyze your data. Many attitudinal surveys are done at two different points in time, to test attitudes over time. Others are just done once, to see how groups of people feel about statements at a particular point in time. Tests like the Kruskal-Wallis, Mann-Whitney, and chi-square analysis can all take attitudinal data from Likert surveys and provide different forms of analysis.

Determine whether your results show significant differences that either match or contradict your hypothesis. The definition of "significance" will vary depending on the test that you use. However, if your results do show significant differences, for example, in the way adherents to different religions feel about the way models dress on the covers of fashion magazines, then you can find applications of that research for fashion editors.