From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.4 (2020-01-24) on polar.synack.me X-Spam-Level: ** X-Spam-Status: No, score=2.1 required=5.0 tests=BAYES_40,INVALID_MSGID, REPLYTO_WITHOUT_TO_CC autolearn=no autolearn_force=no version=3.4.4 X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: f8499,80ec92800f85d9e7 X-Google-Attributes: gidf8499,public X-Google-Thread: 109fba,80ec92800f85d9e7 X-Google-Attributes: gid109fba,public X-Google-Thread: 103376,80ec92800f85d9e7 X-Google-Attributes: gid103376,public X-Google-Thread: f43e6,80ec92800f85d9e7 X-Google-Attributes: gidf43e6,public From: "Scott P. Duncan" Subject: Re: Do college AP tests require C++ knowledge? Date: 1997/12/03 Message-ID: <3485A3B2.4BCA@mindspring.com>#1/1 X-Deja-AN: 294915420 References: <65vqjn$9tk$1@flood.weeg.uiowa.edu> <3484657B.2F3F@gsg.eds.com> <3484CD3B.CC7FA342@oma.com> X-Server-Date: 3 Dec 1997 18:23:15 GMT Reply-To: softqual@mindspring.com Organization: SoftQual (Quality That Lasts) Newsgroups: comp.lang.ada,comp.lang.c++,misc.education.science,comp.software-eng Date: 1997-12-03T18:23:15+00:00 List-Id: Tim Ottinger wrote: > > If the test has the right questions, and you learn everything that's > needed to pass the test, then is that a good thing or a bad thing? My college Economics 101 professor gave out a list of questions at the beginning of the semester (about 100 as I recall) and said that the final for the course would come from that list. He would not say exactly what the questions would be, only that about half of them would be a part of the test. He said anyone who felt they could get the answers to the questions without having to come to class was free to do so -- hardly anyone ended up doing anything close to that. > The concern is if you didn't learn everything you need to know, but > just memorized a set of answers. It also places the onus on the > designers of the test to be very comprehensive. The latter is the key issue in testing, doing customer surveys, etc. > A class that tailors > to a shallow or narrowly-defined test could be a disaster, but one > that caters to a really good test might be fine. If designers of test/survey "instruments" don't do a good job, even well-run training/education will not show up effectively through the badly designed test. It would be fascinating to have an employer send out a list of questions to potential candidates and say, in effect, "Here's what we expect you to be conversant with if you want a job here." Want ads for programmers, test specialists, etc. often mention platforms, languages, tools, etc. that the applicant should know. But an actual set of concepts, ideas, etc. would be fascinating. [It used to be that employment "tests" were problematic because they had to be certified to show the knowledge on the test had job skill relevance and a lot of companies just stayed away from that possible legal pit. I was given verbal tests during a few interviews: impromptu "How would you code this" or "What's wrong with this code." But, as I recall, only AT&T had an actual sit-down test.]