One long-standing argument in the testing industry as to why manual scripted test cases can be useful in one sense is this:
"Manual scripted test cases are a great way for new testers to learn the product while they test"
That statement as useful as this:
"Using a GPS to find your way is a great way to learn to navigate the city"
It just isn't true.
The main problem is that while following a script, or while driving by instructions from a GPS, you are not actively engaged in the process.
Therefore after you have arrived at your destination you will be hard pressed to remember exactly how you got there and to find your way again.
The GPSResearchers have in the last years started studying as to what effect driving by GPS has on your brain. Interestingly enough it turns out that the GPS is disrupting something that our brain is naturally supposed to be good at, namely finding our way.
When you're driving "the good old way" what happens is that you are forced to be engaged in the process. It means that as we drive we make notes of the landmarks around us, intersections, surroundings and roads, and while doing that we create a cognitive mental map of the way.
A researcher at University of Nottingham, Gary Burnett, showed in 2005 that while putting volunteers into a driving simulator and letting half of the people have detailed instructions for four routes to follow, and letting the other half having a traditional map to follow, then afterwards when they were asked to sketch out a map of their route the ones given detailed instructions did significantly worse in all aspects.
They even failed to notice that they had been led past places from different angles.
A study in 2008 by Toru Ishikawa working at University of Tokyo generated results that supported Burnett, and later other studies followed as well.
One quite famous study done by British neuroscientist Eleanor Maguire in 2000 showed that taxi drivers in London that had navigated the city for years had more gray matter in the posterior hippocampus compared to people who were not taxi drivers.
The brain is plastic, and the parts we use grow larger, and the parts we don't use grow smaller.
Solving puzzles by instructions or by figuring it out yourselfMy friend Duncan Nisbet ([@DuncNisbet](http://twitter.com/DuncNisbet) is actually running an interesting experiment where he observes people’s ability to solve problems either by getting instructions, or by solving them by figuring it out for themselves. When we met at last DEWT we agreed to start doing this project together, hopefully something interesting could come out of this, or it could just be an excuse for me and Duncan to get together and hang out ;-)
So at this point you might think how all of this is connected to testing?
Well the things is that, as I pointed out in the beginning of this text, testing by running manual scripted test cases is not a way to learn the product.
Just like when driving by GPS we are not engaged into the process, learning doesn't take place, and we don't form spatial memories of the product we are testing.
Manual scripted test cases are not a great way to learn the product
Adding value as a testerIn his book [“The Lost Art of Finding Our Way”](http://www.amazon.com/The-Lost-Art-Finding-Our/dp/0674072820) John Edward Huth talks about the impact driving by GPS has had on us as a society and the abilities that we lose.
One of the things he mentions is how, without driving by GPS, we are actively turned into the physical world.
Testing by manually scripted test cases has the opposite effect, it disengages us from our surrounding environment.
Huth also says that:
“You’re losing this chance to have a greater awareness of your environment”
One of the things a lot of us testers knows is that being a good tester is so much more than technically testing the product, in other word only testing to find out what works and what doesn't work.
A good tester is also observant on how the product is being used, potential ways it can be improved, dangers and risks lurking around the corner and much more.
If I'm driving by manually scripted test cases, what do you think that does to my abilities as a tester to react and respond to these types of things?
If I don't "see" the application, how can I be expected to see problems, what can be improved and to report on those findings as well?
Learning from our mistakesOne aspect of the GPS that is a double-edged sword is that when driving by GPS we don't make mistakes. If we don't make mistakes then there isn’t any place for reflection and learning.
"Hmmm, I wonder how I ended up here?"
Manual scripted test cases guides us along "the right" path every time. Therefor there is little to no opportunities for exploring, trying out new things, and learning.
Now take exploratory testing, it is the very opposite. As a tester I have to be very engaged with the product and with my own mind in order to do it well.
I'm trying out new thing, learning from them, forming new test cases, constructing hypothesizes, rejecting them, and building mental models of the product.
Landmarks being folders, log-files, and config-files, surroundings being the OS, hardware, and roads are the ways we navigate through our tests in the terminal or the GUI.
We construct a mental map/model as we test through the product, if we are engaged.
Throw out all the manual scripted test cases
So what am I saying here? Running manual scripted test cases will reduce the size of areas in your brain?
Absolutely not, and there isn't any research on how doing complex tasks by lists of instructions affects the effect the brain, at least no research that I have been able to dig up.
Am I saying throw out all the manual scripted test cases?
Hardly, although I admit having a slight aversion towards them. In some situations and contexts some manual scripted test cases have a time and place.
All I'm saying is that the age-old statement that manual scripted test cases are a great way to learn the product is simply put not true.