FW: SQL Proficiency test

From: Powell, Mark <mark.powell2_at_hp.com>
Date: Wed, 9 Apr 2014 20:26:09 +0000
Message-ID: <1E24812FBE5611419EFAFC488D7CCDD1188153C7_at_G6W2491.americas.hpqcorp.net>

The problem mentioned below appears to just be a problem with the screens not clearing the previous solution as there is a button to see the solution though I never bothered to use it.

From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Powell, Mark Sent: Wednesday, April 09, 2014 3:52 PM
Subject: RE: SQL Proficiency test

The official solution for question two appears wrong. The actual question does not specify to filter the results using the same criteria as question one but the solution posted includes the filter conditions which potentially changes the number of single word stations.

From: oracle-l-bounce_at_freelists.org<mailto:oracle-l-bounce_at_freelists.org> [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Stéphane Faroult Sent: Wednesday, April 09, 2014 6:18 AM
Subject: SQL Proficiency test


   I'm more and more involved in database education, and one of the things I feel much concerned by is the correct assessment of what people learn and of their ability with SQL. It's easy to build a MOOC with a video watched by 5,000 people, much less easy to grade 5,000 papers (going to have 150 to grade before one month ...). I have a very limited confidence in multiple choice questions as an assessment tool, and I have tried to do something better. So far, what I have found best on the web is what Gordon Russell did at http://sqlzoo.net; I have read his papers with Andrew Cummings, and I think one can improve on what they have done. If I expect a query to be select a, b from t where cond, I don't want to count select b, a from t where cond order by c as wrong to take just a simple example. I have been therefore standing on the shoulders of Russell and Cummings. I am checking a lot of things (even whether people are cheating), but not that the text of the submitted query exactly matches my solution. I also want to put people in realistic conditions, where you don't have everything wrong because of a tpyo, where you can try queries, check results against data, and only submit when you think it's correct. I don't want people to be stressed by time because I teach my students to build their queries bit by bit, layer by layer.

  I have a working prototype and I found nothing better to validate my algorithms than to publish a kind of game, an "SQL Challenge" at http://edu.konagora.com/SQLchallenge.php. It's not something as complicated as the SQL competitions organized by the NoCOUG :-). There are three levels, the queries to write in the two first levels are pretty straightforward (but for a few banana skins), the last level is a bit tougher. If some of you want to play with it, I'd greatly appreciate feedback (off-list), especially about grading as it's the main purpose.

Some of you might also find this useful for technical interviews, and checking that an applicant's claims are matching his or her abilities.

Sample database and questions could be easily changed on a regular basis, the most difficult is finding suitable ideas ...

Now, integrating all this into Blackboard/Moodle/... will be another fun part.

Hope that some of you will enjoy it ...

Stéphane Faroult
RoughSea Ltd<http://www.roughsea.com>
RoughSea Channel on Youtube<http://www.youtube.com/user/roughsealtd> Author, SQL Success<http://www.amazon.com/SQL-Success-Database-Programming-Proficiency/dp/1909765007/>, The Art of SQL<http://www.amazon.com/Art-SQL-Stephane-Faroult/dp/0596008945/>, Refactoring SQL Applications<http://www.amazon.com/Refactoring-SQL-Applications-Stephane-Faroult/dp/0596514972/>


http://www.freelists.org/webpage/oracle-l Received on Wed Apr 09 2014 - 22:26:09 CEST

Original text of this message