At least some portion of the depression in IT students has to do with the cost of an IT degree measured in the amount of Student Loans vs. the number of years to Pay off said loans. Do you have a chart that compares the cost of an IT degree vs. Starting IT Salaries over the same period?

Since H1B visas will fill much of the demand, there are fewer entry-level jobs and those jobs pay lower Starting Salaries then they would without H1B visas filling jobs.

So, you have fewer IT students. Or to stretch Wayne Holder’s test to make my point, when you have to put a fist full of dollars in another bowl in order to hang out and work with the computers, it becomes much less attractive to even the most talented and interested among us.

You might enjoy reading the article “The H1B Visa Sham” By John Miano at FrontPageMagazine.com for a good statistical look at H1B visas.

Perhaps you would also take some time to read my (long) comments there. Although, my comments are based on my experience and not statistics.

Thanks for the observations and even more for the kind words. And your observations, as usual[*], are dead-on.

One of the most important things I’ve learned from your writings — starting with my original copy of The Psychology of Computer Programming, which I bought in the late 1970s as a recently minted CS graduate — is the profoundly human nature of software engineering and IT project management. Organizations keep trying to treat it like a manufacturing or chemical engineering process; I think it’s more like a sports team or an orchestra.

Oh, and I hope you don’t mind the “Weinberg’s Law of Metrics” that I coined for my column and re-used in the posting above. ..bruce..

[*] Except, of course, when I choose to disagree or pick a fight with you.

I would add one question to your questioner, up front. It’s implicit in your answer, but I like to ask it explicitly:

Suppose you had this measurement, what do you intend to do with it?

Frequently, I get answers that amount to:

- If I had this measurement, I’d know which testers to punish. (or reward, which amounts to the same thing, since the unrewarded testers will feel punished.)

I don’t bother to answer these people.

Most of the other answers amount to:

- If I had this measurement, I’d know how to improve testing.

To these people, I can usually look at what they’re doing now (eyeball measurement) and give them at least half-a-dozen suggestions that will improve their testing noticeably before they start looking for 1% improvements.

Gerald M. Weinberg
http://www.geraldmweinberg.com

Suppose we have a perfectly spherical elephant, whose mass may
be neglected … er, no suppose we have an organisation with
four levels of management. The lowest level is the one dealing
with the hardware, or the environment, and is therefore the level
with the most accurate and complete knowledge of reality. Let’s
say (for ease of calculation) that things are coasting along,
there’s equal amounts of good news and bad news at the coalface.

Now generally, people like to hear good news, and dislike to hear
bad news; in extreme cases, a “shoot the messenger” policy can be
in force. So it’s only human nature that reports of good news get
a bit exaggerated, and reports of bad news get trimmed back a touch.
Not by much; to make the maths easier, say good news gets increased
by 10 percent on each reporting, and bad news is reduced by the same
amount.

So after three levels of reporting, with equal amounts of good news
and bad news coming into the organisation, the man at the top hears
1 x 1.1 x 1.1 x 1.1 units of good news, and 1 x 0.9 x 0.9 x 0.9 units
of bad news; 1.33 to 0.73, or nearly twice as much good news as bad.
“Hey”, says the boss, “things are going really, really well.”

This, of course, assumes that information travels only through regular
channels, and doesn’t explain the thermocline effect, but does explain
most of the actual examples I’ve seen in managing large systems. My
(partial) cure was to be very pleased if anyone told me any bad news,
another approach would be to develop accurate reporting systems that
bypass multiple levels, but that would greatly upset the managers thus
bypassed. The real answer is to hire competent managers, but that’s
not always an option.

C W Rose