Introduction to Stats Homework

 

So, why would I write another statistical software package?

Every feature of Stats Homework was designed with one goal in mind: to facilitate students’ engagement and learning in their first statistics course(s). To be specific, Stats Homework was designed to help students verify the results of the hand-written problems that they complete in their statistics class (i.e., to help with their statistics homework).  I believe that using professional statistical packages such as SAS®, SPSS®, or R in a student’s first statistics course has three undesirable features:

  • Professional packages tend to be expensive and limited in distribution on one’s campus. Even with academic discounts, the professional packages are still relatively expensive for students to purchase. If they cannot purchase their own copy, then students have to do their work in computer labs that often have limited availability.
  • Professional packages tend to be difficult for novices to learn. Requiring students to use one of the professional packages in their first statistics course(s) tends to increase the students’ learning load at a time when they are struggling to master the basic statistical concepts being covered in class.
  • The outputs from professional packages are typically not very helpful to students who are learning the basic concepts and doing problems by hand. Professional packages will give students the answer, but will do little to help them find their errors if their hand-written answer does not match the computer’s. In other cases, the output from professional packages will tend to confuse the student because it is organized differently than the output of their hand-written work.

Stats Homework was specifically designed for students who are learning statistics, so it has these features:

  • Stats Homework is free. Any student or instructor who wishes to use this program can download the software free from its web site.  Stats Homework has been written in Java — it will run on any modern PC operating system (i.e., Windows, Mac, and Linux/Unix).  Thus, every student has access to a powerful and useful statistical software package whenever they need it.
  • Stats Homework is extremely easy to use. I suspect that very few students will ever consult the documentation system for this program. Its approach to data management and analysis makes it intuitive and simple to use.
  • Stats Homework provides extensive preliminary statistics that students can use to track down their mistakes in case their hand-written answer does not match the one produced by the computer.

Of these features, it is the last one that most distinguishes Stats Homework from the professional statistics packages. Let me give you a quick example.  Let’s suppose that you are conducting a correlational analysis with a very simple data set. If you use SPSS® to compute the Pearson correlation, its default output will look something like this:This table includes three statistics: the correlation, the sample size, and the significance level. Imagine that you are a student trying to learn how to compute the correlation coefficient by hand, and that your hand-written work did not result in the same value of r. Where did go wrong?  In which step of the computation of r did you make a mistake?

This output just gives you the final value of r — it does not give you any of the preliminary statistics that are used to compute r. In order for you to get more information from SPSS, you would have to use some other procedures or options in your analysis. This would not be easy for you if you are just learning this statistic and this software package for the first time.

Submit the same data to a correlational analysis in Stats Homework  with all the options selected, and you will receive this output:corr8

..and this output:

corr9

…and this:

corr10

…and this:

corr11

…and this:

The full output from Stats Homework includes 29 different sample statistics and a scatter plot. In addition to the correlation coefficient and significance levels, the student is provided with all of the preliminary statistics that were used in the computation of the correlation coefficient. Thus, if their hand-written work does not match the computer output, they have all the information that they need to track down their errors. In addition, the output helps the student to verify the critical value of r that they had to look up in a statistical table, and reinforces important concepts being learned in class such as one-tailed vs. two-tailed testing, confidence intervals, and the importance of graphically plotting one’s data.

Shouldn’t students be expected to learn professional statistical packages?

Certainly — students should learn how to use a professional statistics package such as SPSS® as they move from their early methods and statistics classes to conducting supervised and independent research with their faculty advisers. That is why I designed Stats Homework to be very similar to SPSS in its data manager, and in many of its menus and user dialogs. Using Stats Homework in their early statistics classes should help students learn to use a professional statistics package later. Where features and function in Stats Homework differ from those of a professional packages, these differences result in a highly consistent and simple program that students can easily master in their first statistics class. They can then move on to a big professional statistics package later without much difficulty.

The goal is statistical education.

My intent in writing Stats Homework was not to create a program just to do statistics, but to create a program that will help students learn statistics. I sincerely believe that if students will use this program in their first statistics course(s), it will help to reduce their anxiety about statistics and statistical software, and facilitate a deeper engagement of the concepts being covered in their classes. This will inherently contribute to their long-term success.

So, whether you are an educator or a student, I hope that you will try using Stats Homework. And, please feel free to send me feedback about your experience. I would welcome any comments, suggestions, or questions that you might have. I’m always looking for ways to make this program more helpful to you as you learn or teach basic statistics.

If you have any questions or feedback about the use of Stats Homework, please contact Victor Bissonnette at: vbissonnette@berry.edu.


Return to Main Page

Return to Stats Homework Page

Return to Documentation Table of Contents

 

Leave a Reply

Your email address will not be published. Required fields are marked *