The Return Of Dendrograms

What is Dendrogram-Based Testing? Well, what is a dendrogram to start with?

A dendrogram is a tree diagram that visualises hierarchical clustering. If that didn't help, a dendrogram basically groups objects in a tree view based on how similar they are. The closer the objects are drawn, the more similar they are.

Thanks for the maths lesson, but how is that useful in testing?

Good question. I'll come back with a final conclusion later in this post, but I can think of two uses for dendrograms:

Clustering defects: Visually show how similar the defects previously found are.

Clustering test charters (test cases): Visually show how similar planned test charters (or test cases) are.

In order to create dendrograms we need the objects, e.g. defects, to have such properties that we can measure distances between them. This is where it starts getting tricky - how do we measure the distance between two defects? The simplest thing to do is to think of properties we believe to be important and then assign them numeric values.

One example could be the property "User" (P1) and we could assign a defect a value between 0 and 5 for this property depending on how affected we think the user is by this bug. Another property could be "Performance" (P2) or "Business" (P3). Imagine we are testing a web shop and have two defects:

B1:The "This is a gift" checkbox is missing in the GUI.
B2:Memory issue that slows shopping down when you have more than 10 items in your cart.

Each of the two bugs have the properties P1, P2 and P3, and we might to assign values as follows:

B1: P1=5, P2=0, P3=2
(the user is affected, the performance is not but the business flow is also affected)
B2: P1=3, P2=5, P3=0
(some users will be affected, the performance is affected, the business flow is not affected)

Based on these properties we can now see how similar the defects are in a dendrogram. In my earlier post I explained how to create a defect dendrogram with simple example, and I'm not going to repeat that.

Similarily we can assign test charters or test cases properties and create dendrograms. Here properties could be which actors, functions or areas that are involved, and the dendrogram shows a kind of test coverage. If all test charters are grouped together, they test very similar things.

So how do we base our testing on dendrograms?

A defect dendrogram would of course be used to decide where to focus testing. I think isolated defects would be my priority. A single defect far away from all other defects seems too unlikely, maybe there are more hiding that need to be discovered. Then again, if a large number of defects are very similar there is reason to believe that area requires special attention.

A test charter dendrogram would of course be used to help decide which charters to add. A single isolated test charter might be ok for a low-risk area, but might also be a warning flag.

Is this useful?

I have some serious doubts. Firstly, we need to find useful properties and assign them subjective values. The dendrogram will be based on those values and nothing else, so there is a huge risk of bias. Secondly, I have yet to find a good tool to use to draw dendrograms. With more than three variables (defects/test charters) and two or more properties it cannot be done by hand. Of course, writing your own tool would not be too complicated.

Right now I don't think the value gained outweighs the effort needed. I'm very interesting in hearing arguments that I'm wrong though.

Schools of testing?

CAST 2011 hosted an interesting debate between Doug Hoffman and James Bach on the topic of schools of software testing. The question under discussion was not whether there are different schools of thought within the testing community or not, but rather whether naming the schools and associating people with them is a good - or really bad - idea.

The debate was energetic, and clearly provoked a strong reaction in a lot of the attendees, which was only expected. The core issue is of course if it is ok to categorize people without bothering with their opinion. Most people categorize others, but hate when they themselves are put in a category that they do not approve of, or think they should belong to. It is a very touchy subject.

Personally, I like it.

To me, the fact that someone is associated with a school of thought corresponds to me being provided with a table of contents of a book. Let me try to explain. If person A says to me "- Person B belongs to the XYZ school", it provides me with a limited amount of information about person B, just like browsing a table of contents tells me something, but not everything, about the book. Immediately - without having to read the whole book (i.e. without having to have a deep discussion with the person) - I get a rough idea of the contents (i.e. the person's views and ideas). The same way I do not mind being associated with a school, or associating myself with a school. I find it helpful because I do not have to explain my general views over and over again, I just need to state which school(s) I consider myself belonging to. Sometimes I will disagree when others associate me with a certain school, but  that on the other hand gives me valuable clues as to how I am perceived. And it might even make me change my behaviour.

However, I do assume that people are mature and intelligent enough to realize that a table of contents can be misleading, and in order to get the full story you actually have to read the book. You cannot know a person without having talked to them and having formed your own opinion.

I think the concept of schools of testing is helpful, and in all honesty - even if it was rejected people would still categorize each other 'secretly'. I would rather have it done openly so you at least can have a discussion.

CAST 2011

I'm back from CAST 2011, and I've had some time to digest the experience and think about what thoughts I want to share. There have been many excellent write-ups that give detailed account of what transpired at the conference, and there is no need for another (worse) one. Instead, I'll make some short remarks on my impressions.

I had very high expectations, but I'm still amazed.

What drives me in life in general and as a tester in particular is a continuous strive forwards and a desire to learn and progress. I have no sympathy whatsoever for people who seem to consider testing to be nothing more than a way to pass time and earn your paycheck. What made CAST 2011 such a fantastic experience was that it was a gathering of enthusiastic, engaged, creative and ambitious testers. All willing both to learn and to teach. Everyone was friendly and approachable and willing to share. I was in awe of all the experience and knowledge that was surrounding me.

Even though there were of course different opinions on various topics on a smaller scale, it was fantastic to see such a large body of people all strive in the same general direction, sharing the same goal. And I find it very comforting to see people (testers) take such pride in their work.

I learned a lot and got a bunch of new ideas to try out, but mainly I was just soaking up the joy and energy. Thank you everyone who attended and thereby contributed to making CAST 2011 one of the best conference I've ever been to.

I'm proud to be a tester.

Dendogram-Based Testing

Friday afternoon I was looking through the latest tweets when my eye was caught by the phrase Dendogram-Based Testing. I like all words that have a Greek origin and sound like science, so I had a closer look, and of course it was James Bach introducing a new concept. One that - as far as I understand - is still pretty much missing a definition. No reason to let a small detail like that stop you, right?

After reading up on dendograms I realized that I have actually used them before, but didn't know they were called dendograms. A dendogram is basically a way to take data points and cluster them based on their properties. They are commonly used in computational biology, and that's were I encountered them. About a year ago I spent a week of my vacation making dendograms from genome data.

In testing, one way to use dendograms would be to cluster defects. In order to do this you would need to define a set of properties for each defect, and based on these properties it would be possible to calculate distances between the defects and cluster them in a dendogram. I will save the discussion on whether this is useful or not for later.

Example: The android game SuperTester

To the best of my knowledge this game does not exist, but if it did it would be a game in which the player has to find critical bugs in imaginary applications under time pressure. Just for the record, I haven't thought too much about this so I'm just making it up as I go along.

Let's say five bugs have been found when testing the actual SuperTester game:

1. Can't save game (D1)
2. Can't change sound volume (D2)
3. It is possible to register the same bug twice (D3)
4. Application crahses if you find exactly 13 bugs (D4)
5. Application crashes if you play for more than 59 minutes and 59 seconds (D5)

Now we need to assign these defects properties in order to cluster them. This is the tricky part and requires some careful thinking, Which properties you pick decide what information you will get out of the dendogram. For now I'm just going to pick two properties for the sake of creating an example,

1. Frequency of occurrence on a scale 1-10, where 1 is rarely and 10 often (P1)
2. Severity of defect on a scale 1-10, where 1 is not severe and 10 is very severe (P2)

Time to make a table:

Defects with assigned properties P1 and P2.

Ok, the values might not make so much sense but let's ignore that for now. We have everything we need to calculate the distances. Note that all this assumes that the properties are numeric, if you have other properties such as "red" or "green" you need to decide how to calculate the distance between "red" and "green". For numbers we use the Euclidean distance. I'm not going to go through all the boring details, but the distance table will look like:

Distance table.

D2 and D4 are closer to each other - that is, more similar - than any other defects. Hence we cluster them in cluster A. And so it goes on. What we end up with is the following dendogram:

Dendogram. Defects are clustered based properties.

That's it! Two important points here are i) you need a tool because you definitely do NOT want to do this by hand and ii) how about non-numeric properties, how do you measure distances? I definitely think dendograms can be useful, but a tool needs to be found and then there must be some thought on which properties to use - what kind of information do we want from the dendogram?

This post is just a collection of initial thoughts and is focused on what a dendogram is. I will now crawl back under the rock I came from and think more about how to actually use it for testing.

Craving conferences

On beautiful summer days like these, I have to admit that the main thing on my mind is being out in the sunshine, preferably eating ice-cream! However, I am also starting to feel keen ongoing to another test conference. Luckily, the late summer and fall look very promising.

For starters there is CAST 2011 which is held in Seattle by the Association for Software Testing in early August. The theme is "Context-Driven Testing", and I am of course going together with some colleagues - how could I resist the opportunity to hear about, and discuss, context-driven testing with like-minded people?  It will be a blast!

Then I am proud to announce that I am speaking together with Michael Albrecht at the STARWEST conference, October 2-7, in Anaheim, California. I am especially happy to be able to provide North America with some Swedish test thinking. Why don't you join us? Register using special promo code SKWS and save up to $600 if you register by Super Early Bird August 5th! Click here to register online.

Like most people around the office, I cannot wait for my vacation to start, but I am really looking forward to coming back to an exciting fall!

Sometimes what tastes like mold actually is mold

Last night I enjoyed a test-oriented after work with Simon Morley, Oscar Cosmo and Daniel Berggren. Quite a bit of time was actually spent on group dynamics and the value of working in a team where the members have mixed backgrounds and experiences. We also talked about the importance of having opportunities to discuss testing outside of your team in order to get fresh ideas as well as feedback on your own ideas.

Then came the mold discussion. We got on to the subject of gut feeling – sometimes you just know something is not quite right but you do not have any hard evidence. Or you have a weird incident that only happened once, and cannot be reproduced, but you know that it is important and should be investigated, and still no one can be bothered.

Recently I was in Portugal on vacation, eating good food and drinking great wines. One evening I ordered a piece of blueberry pie for dessert. The pie arrived, beautifully covered in blueberries and nicely presented on the plate, and I dug in. It tasted a bit funny though, sort of like…mold. But I was in a nice restaurant, recommended by a local whom I was having dinner with, the main course had been fantastic and the slice of pie looked delicious so of course I kept eating it even though I could not get rid of that nagging feeling that something was not quite right.

Finally, when I only had a piece of crust left, my eye was caught by something bluish and fuzzy. Of course the crust was moldy! Probably the entire pie bottom had been moldy, and I had eaten it all up. Not once had I stopped to question if that funny flavour really should be present, nor had I stopped to examine the pie more carefully. I was fooled by the fact that I had been told that it was a good restaurant, and that the pie looked good.

This happens in testing too. You might be testing a third party product that ‘is known to be stable’, or asked to ‘just have a quick look because we know nothing has changed’. And it looks so good! But still, deep down, you know something is wrong. Trust your gut feeling, be courageous and be persistent. Sometimes what tastes like mold actually is mold.

…oh, I survived the pie just fine, no unpleasant after effects. Still, I learned my lesson.

The return of the context-driven physicist

I signed up for CAST 2011 as soon as registration opened since Henrik Andersson had told me it was The Conference To Attend. I didn't think too much about the topic - Context-Driven Testing - until I realised "everyone" was discussing whether they were context-driven testers or not. It's even the current poll on the AST homepage! It was time to do some fact finding followed be serious thinking - am I a context-driven tester?

The physicist in me (still going strong three years down the road) is nonplussed. It has never really occurred to me that it is possible to not be context-driven. Physicists are trained to be context-driven. In physics there is no such thing as one theory or formula that applies under all circumstances - on the contrary everything is highly context dependent. Like speed - when objects move fast enough relativistic effects have to be taken into account.

The (somewhat more quiet) statistician that also lurks at my inner core agrees. In order to interpret your data you have to know the context. Without context you can't know whether the data is best described by the standard distribution, or maybe the chi-square distribution.

As a scientist my approach is to first evaluate the context and then try to find the most suitable technique for solving the problem.  You won't get far if you have a favourite formula that you insist on always using. Nature has no intention to adapt to you.

I did not give TBTM a try because I thought it was cool (it is) and it would make a good blog post (it did), but because I thought it would suit my context. My focus is always on solving the task at hand, not on the methodology or technique. I do try to learn as many techniques and methodologies as possible, but not to have a nice CV but rather to really be able to be context-driven. If my tool-box only contains one tool it’s darn difficult to adapt to circumstances. 

I am a context-driven tester and proud of it.