Multivariate analysis deals more with the differences in variables or combinations of variables as a result of a set of data. Multivariate analysis will more specifically examine how the mean value, standard deviation, and other statistically significant factors influence the inter-relationships of the data. Students will have a greater understanding of what relationships exist when multiple regression is taken into account. Students can even apply their knowledge of linear regression by taking a multivariate test.

Taking both regression and multivariate analysis multiple times on the LSAT will help students prepare better for the final test. Differentiated instruction and practice tests are available so that students can try each of these techniques multiple times before taking the LSAT. The most important thing for students taking the LSAT in order to do well on the LSAT is to get the most out of every section.

A question that might come to mind is why students should even consider regression and multivariate analysis. There are a number of reasons why students should consider these techniques on the LSAT. One reason is that regression and multivariate analysis is very easy to interpret. Students will often find themselves unable to make a distinction between the two unless they have strong pre-existing math skills. This means that a student taking the LSAT with weak pre-algebra skills will not be able to understand the implications of a regression.

Another reason why it is good to think about regression and multivariate analysis is that the results of the analysis can often be predictive of future scores. That is to say, if a student discovers that his or her scores are predicted to do poorly based solely on the raw scores, then doing some regression and using the data analysis can show that these predictions are indeed true. Once a student knows this, it becomes much easier to improve the scores in preparation for the LSAT. Also, students should know that because the study of regression is so easy to understand, they will likely be tempted to do it themselves before even taking the LSAT.

How does regression and multivariate data analysis work? Let’s imagine a situation where we are looking at a person’s raw scores. We want to see how well these scores relate to the characteristics that are listed in the person’s profile. We already know that the person has performed well in school. So we would like to see if we can predict how well the raw scores will relate to things that are important in a professional context (i.e., what kind of teachers the person is).

We begin by regressing the raw score onto the scores from the major subjects that we have just measured. In this case, we want to see how well our predicted raw score will fit with the person’s own performance on the LSAT. The next step is to take the adjusted raw score and compare it to the predicted score. If the difference is greater than zero, then we know that we are right about the nature of the students’ performances on the LSAT. If the difference is less than zero, then we know that we are wrong about the nature of the students’ performances on the LSAT.

This is a very simple example of regression that can be used in many other applications. Note that I use “regression” to mean the process of adjusting the values of the variables so that they result in changes in the corresponding variable values as the sample is being analyzed. In these examples, I am assuming that the analysis procedure has been done already and that there is some sort of reasonable normality to the data set. The multivariate techniques I have described here assume just that. It is a very important tool, and you really should practice using it in your research papers and your consulting assignments.