Looking for clues. Picture from www.pixabay.com CC0
Some men love fast cars. Others love women. Me? I love spreadsheets and databases. But when it comes to teaching such things, how can you make them interesting? I think this applies particularly to databases. Flat file databases are, it seems to me, inherently boring. Relational databases are much more interesting. Perhaps because they are more useful, though I suspect the real reason is that they are so much more complicated.
Like most teachers I imagine, I introduced flat file databases before tackling relational ones, but after teaching the basics, and how to do simple and complex searches, I thought I'd try an experiment.
I had particular objectives in mind:
- The database had to be very large, because I didn't want the pupils to be able to find data simply by looking through the records or by sorting them.
- The searches had to have a real purpose, or at least real in the sense that they could be searches that someone would carry out "in real life".
- The activity had to be fun.
The solution
I created a very large database of over 1,000 records and 27 fields. Then I set them a murder mystery to be solved. This was in the form of a brief story and a number of statements to the police.
For example, one statement said something like "I saw someone with blonde hair who was eating a cheese sandwich. I was in the Chinese take-away and listening to some jazz music on the radio at the time".
The pupils had to work out whether witnesses were telling the truth or not. Thus in the above example, if they discovered that the person making the statement doesn't like jazz music or Chinese food, they would put them on a list of suspects. Eventually, by finding the person who had told the most lies, that person would be identified as the culprit.
Not exactly robust evidence, but still.
The whole exercise was very successful. Why?
What made this a success?
- It met all the objectives cited earlier.
- The pupils were doing Boolean searches but they first had to work out what the search actually was. Taking the example cited above, they would have had to search for people who had blonde hair and liked cheese sandwiches before they could go much further. But there wasn't an instruction to do that: they had to infer it for themselves.
- The lesson had an element of competition: pupils were trying as hard as they could to be the first to solve the mystery.
The really nice thing about compiling such a large database was that it was a good investment of time, as I was able to use it in other lessons too. I even gave it to the science department, so they could use it to teach pupils to test hypotheses. For example: only old people like blues music.
I was also able to use it show the maths department how to dig deeper into the data by using pivot tables.
So even though the data had been made up, the database proved to be a fantastically flexible resource.