Question

A. For each of the scenarios listed below, specify which data analysis technique would be most appropriate and why?

I. A university lecturer would like to test whether students who attend tutorials (1 = less than 5, 2 = more than 5) achieve a high or low grade (1 = 4 or less, 2 = 5 or more).

II. A fast food restaurant wants to know if there is a significant difference in job satisfaction (1-10 scale) between employees who work the night shift and those who work the day shift.

III. A small business needs to evaluate whether, over the past 3 years, their monthly advertising spend (in dollars), employee training (in hours), and customer word of mouth (1-7 scale) affects the next month’s sales (in dollars).

Answer 1

A

I.

In this case we have only two key factors. That is

Less than 5 tutorials

More than 5 tutorials

The result is also divided in two

4 or less grade

5 or more grade

This means the most suitable data analysis for this is factor analysis. Something simple like a matrix or a scatter plot can be easily used to analyze the given set of data in this situation.

II.

Here the job satisfactions are quantitative in nature. They are marked from 1 to 10. However, the set of data sets are two. This means we should analyze various attributes of the day shift and night shift data. The best approach to do is to use ANOVA. It is the method of analyzing the variances from the two data sets, their means and basically compare similar type of two sets of data and compare if they are different or similar.

III.

Here the month’s sales is the dependent result we are attempting to measure or predict based on other variables. This means that the best approach to analyze such data is to use multiple linear regression.

Multiple linear regression is the best method to analyze a data set when the dependent variable (sales) can be attributed to more than one independent variables (advertising spend, training, and word of mouth).