I have begun to go through the process of getting back to work after some time off. I have passed all but the last of the qualifying exams in my program. The last test has given me some trouble. I have taken it twice and failed it both times. I have petitioned for the right to take the test again and have been granted that chance. Based on my medical tests and other factors at the time of the test I qualify for such a chance. That chance comes at the price of going through the DePaul PLuS program. I was active in the program a number of years ago. In my opinion I know more about my condition than they do. Even in that light I have agreed to work through the PLuS to get another chance at the test. More on my thoughts about the test later. On Thursday of last week I responded to the PLuS program director about possible times for us to meet. I told him this Monday, Tuesday or Wednesday. I have been trying to contact both my adviser and this director for the entire summer and were not successful. Not until last Thursday had I finally heard from both of them. Now I wait to get word on what day I will meet with the director. I have begun to get ready for the test ahead of anything that the PLuS director or my adviser would have me do. One of the questions on the test will most assuredly be on Zed, often called the Z notation. This language is for formally specifying software. It has been on every software engineering qualifying exam. This is a frustration for me since I have had a great deal of classes and every class that is supposed to teach Zed has one assignment or less and less than one full lecture on the subject. One more tidbit is adviser's dissertation involved Zed. Last night I loaded tetex and latex. I am an emacs user so my editor works well with these type setting technologies. On top of latex you put Zed. The major accomplishment last night was getting fuzz to work on my PowerBook. Fuzz is a collection of tools that help with formating and printing of Zed specifications. It can check them for compliance with the Zed scope and type rules. I got fuzz to work last night. This will make it possible for me to use the fuzz tools to gain proficiency in Zed. It seems that this requirement is not one possible to waver on and pass the exam. I must be cable to get expert proficiency from a text book. Zed is not taught to the level needed to answer the questions on the exam. So I must use the technology at hand to validate my learning. It is interesting to me that every time I have gone to review any of my candidacy exams the professor doing the review with me prefaces the experience with the statement "Some of these questions are not in my area of expertise." This is baffling to me. Other professors have said "These test are designed to show other universities that you have a level of knowledge worthy of a PhD. A PhD is a credential that is accepted all over the world. We want you to measure up to that standard." I perfectly understand academic testing terminology such as "Establishing Test Reliability" and "Establishing Test Validity" These actions don't seem to fit the experience. Four types of reliability are used to establish whether a test produces scores that do not include much random variance. 1) Internal reliability - Is each test taker consistent across different items within a single test? 2) Test-retest reliability - Is the performance for each test consistent across two administrations of the same test? 3) Inter-rater reliability - Is performance for each test taker consistent if two different people score the test? 4) Parallel forms reliability - Is performance for each test taker consistent across different forms of the same test? I can see at least a few things wrong with the university strategy for these exams. I would say that the test is not designed to show proficiency in multiple ways. Five entirely different questions on different areas and choosing four does not give the opportunity to demonstrate knowledge in different ways. Test-retest reliability is in jeopardy because there is much subjective analysis needed to score the test. If no two professors can cover all the test because "Some of these questions are not in my area of expertise" how can we say that this is a general test? They will all grade it differently. I don't believe that there are multiple forms of this test. If for some reason you cannot do this kind of test but know the material you fail. All test I have taken for software engineering have been nearly identical. Test anxiety, reading skills, flat organization of three dimensional facts without a skill for organization in a flat space, historical clues to answers with no historical significance, all these and more play a part in making a test that is not about the technology and all about passing a test. At the current time I would judge this test unreliable. I plan to work hard and pass any kind of test regardless of whether I think it is fair or not.