You may have noticed that if someone asks MAST a question, they can get their answer right away, at little cost. But if they try to get a custom tabulation in a more traditional manner, they have to pay a lot of money and wait a long time.
How can that be? Is someone scamming the public here?
No, there’s no scam going on. This is the reality of data-processing. There are certain questions that are really easy to answer, such as “How many females aged 60+ live in Michigan?”. You can get answers to that kind of question in pre-tabulated tables that are freely available to anyone who wants to find them. No problem. But suppose you want to know how many 5 year old children live in the same household with 7 year old children. Or perhaps something far, far, more complex than that. In these cases, special purpose software needs to be developed in order to answer the question.
To develop that software normally takes a large investment of time from highly paid and skilled data processing folks. A few thousand dollars really isn’t a bad price for that type of work, but keep in mind that it usually costs more than a few thousand dollars. I consulted for a telephone company, and data users in the company could go to a couple of different places within the company to get data. My client (within the company) absolutely loved to turn down data requests (he was too busy), and send people over to the IT department. The user would then go to the IT department and get an estimate for their special tabulation – usually the estimate would be around 3 months time and $30,000. They would then come back to my client and plead their case. He would then give them their answer the next day. For free.
How did he do that? The IT department had to put it’s programmers to work writing special-purpose software to create the special tabulation (when they could fit it into their schedule). That’s why it took 3 months and cost $30,000. My client just had to have one person spend a little time running a code-generator. Special-purpose software is being written in both cases. The difference is that in the typical IT shop, it is written by a human, whereas when a code generator is used, a computer writes the code.
The obvious question then is “Why didn’t the IT shop use the code generator?” Suppose you are a manager in a company, and your competition for raises and promotions are other managers in the same company, and you have a “magic genie” that answers pretty much any question you can think of, quickly and accurately. How hard would you work to insure that only your shop had access to it? Do you think you would publicize it and demonstrate it to the people you are competing with for raises (especially when they have as much legal right to use it as you do)?
If you’ve ever worked in data processing, think about this for a minute: you can spend 5 minutes explaining a need to the IT shop, then it takes them weeks or months to fulfill the need. Why not develop software that allows you to spend the 5 minutes explaining to the software what you want? Then the software writes the program. This is how MAST works. And this is why I can answer a question in a short time (for little or no cost) that would take much longer when using traditional methods.