CS 305: Statistical Analysis

Highlights of this lab:

  • Statistical Tests
  • SPSS in Brief
    • In-Lab Exercise (Please remember to bring the test materials from last lab!)
  • Assignment

:: Statistical Tests ::

Last week in our lab we conducted a pilot usability evaluation. From the evaluation we received user data with respect to how real people use real systems. The analysis gave us "some" useful information, such as:

As well, in your first course assignment one of the tasks you were required to do was to summarize, in a csv file (or other), a user questionnaire that was issued to the Lyx community. The questionnaire attempted to acquire information pertaining to the general user satisfaction of Lyx in comparison to other word processing tools. The questionnaire asked such questions as:

All of this information is quite useful. Problem: How do we use it? Enter Statistics! (YAY...had to add this in...sorry ;-) In the study of human computer interaction, statistics plays a crucial role in our analyses. Conducting usability analyses are simply not enough. We also have to find what worked, what didn't work, and what comments do users offer that could open the door to an improved design? This is what statistics will tell us

Now, before we proceed, I'm not going to lie to you...Statistics can be a tricky, time consuming...and yes, sometimes tedious phase in our development plan. Nonetheless, it is one of the most crucial aspects to the usability analysis.

There exists a variety of statistical tools available for use. The statistical package R is one of these, and one I have just been introduced to recently. It is a very powerful application package. Problem is, it can be very confusing to learn. Lucky for us we have SPSS installed on the UDML computers. Although, arguably, SPSS is not as good as R (my opinion), it is wayyyyy better than excel and is used for statistical analyses by many researchers and research groups. Please note: SPSS is only installed on the PCs in the UDML

:: SPSS in Brief ::

Let us open SPSS and explore its interface. Before we do this, please download the following file:

This is the file that we are going to be using for the in-lab portion of todays seminar. Here is a list of things that we are going to be looking at in the lab:

There are two statistical values of interest we would like to see here:

Before we begin, let's populate this empty document with some data obatined from last weeks lab. I will ask some of you to give me your results and we will all enter in the data in the sheet (oh ya!...this will be fun! ;-) The only way to learn data entry is to do it so, what are we waiting for!

After the data is entered, let us try some statistics out:

:: Lab Assignment ::

Using the data that you collected for your first course assignment (the analysis of the lyxquest.mbox - the brief questionnaire that you had to convert into tabular format, or csv format), I would like you to run some analyses using SPSS. I would like you to:

Play around with SPSS and experiment. If you find other statistical tests, great! However, please justify using them. Try to analyze the data and report how you perceive the results. I would like you to submit:

This assignment is due Oct. 12 at 11:00am