So, it's come to this.

Tuesday 8 May 2012

Can programming be taught?

This blog post is specifically targeted at Teresa again as she rewrote much of the first year undergraduate Computer Science Programming Principles course a couple of years ago.

A friend pointed out Jeff Atwood's blog at codinghorror.com. If you're not aware of Jeff, he's a co-founder of Stack Overflow and has many thought provoking articles. The one that really caught my eye is from back in 2006 called Separating Programming Sheep from Non-Programming Goats. The paper mentioned (the blog links to a draft here, but look on your preferred academic repo for the final version) discusses the high rate of failure on these types of foundation programming courses and suggests that students who fail the course can be determined before they even start the course with a high degree of accuracy.

I find the following quote, like the author, quite disturbing and also a little hard to believe.

But it's still a little disturbing that the act of programming seems literally unteachable to a sizable subset of incoming computer science students

Could it really be possible that some people simply can't be taught how to program? If so, what are the root causes to this?

Also linked to this, is the UK government's intended changes to secondary IT education to make it more computer science relevant likely to change the results if the experiment was to be rerun in a few years?

Assessment - Required in all learning? - Revisited

So, it's been a while since my last blog post and I haven't been quite as active as I would have hoped, but I've got quite a few new posts all half written, begging to be finished, so what this space.

Since my last post on assessment, Teresa has offered a retort to my ideas here. I couldn't resist picking up on a few issues that I'm sure Teresa won't be surprised me bringing to her attention.

The first point Teresa has even admitted I'd be shaking me head at. She was spot on. I am and here's why:

A student can either do something, or they can’t – a miss is as good as a mile. The only skills or knowledge that we can assume a student has are the ones they can demonstrate.

My first issue with this is that whether someone can demonstrate a skill depends totally on how well that skill or competency is defined. Driving is a fairly easy situation to define what the competency is an how to assess it; there is a whole book dedicated to defining it's rules: the highway code. When you move away to anything more grey and difficult to define, this assertion falls apart.

My original example of where this is the case diversity training for Hampshire Constabulary. To even start to assess this type of training you need to set out what the competencies are. It's very tempting to qualify this in terms of what the tutor is trying to teach: "To demonstrate to students that different cultures and backgrounds can affect the way an Officer should deal with a situation" rather than what the student should learn. The problem arises when you try and define it for the student as the outcomes can be different for different people.

What you're trying to do it change beliefs and subtle behaviours in everyday encounters. How do you assess change in beliefs? Are there a set of rules that define exactly what a person's beliefs should be in this instance? I would suggest not, and that to attempt this would be very difficult. The teaching here is more about giving people a set of tools, giving people a brief set of instructions and (for want of a better word) hope that they use them properly. Here's an analogy: say you give someone a knife and show someone how to use it. You can never be sure what someone's beliefs on the knife are. Do they respect it? Do they think it's a utility rather than a weapon? Do they appreciate it's power? Will they ever forget to use it or misuse it? More importantly here, can you ever assess any of these things? People often are unsure of what their beliefs really are until they are tested in real life. Real life is often far removed from simulation or the classroom, I can't see how you can pass or fail someone on this.

Teresa makes the point that this is about the teaching, not the learning, but only a learner could even start to assess beliefs. Teresa also says that the "assessment" could be from the learners perspective and that they can be small, atomic and unconscious. I agree this is useful, but is it assessment? The though provoking image is a good example, people could challenge their initial perceptions and perhaps change them in future, but is this really assessment and do it really qualify the standard of teaching?

Saturday 7 April 2012

Assessment - Required in all learning?

Last Friday I was jogging with a friend of mine Teresa when an interesting conversation came up. Just to give a bit of context, Teresa is currently working an her PhD, the topic of which I always ask but never seem to remember. Anyway, it's very much related to e-learning, so she's pretty clued up about the academic side of education theory. I'll mention now, that my knowledge in this area isn't great and what I'll say here may be fundamentally flawed. If it is, I'd love to know in the comments, or you could just comment if you just disagree with me.

I was talking about some recent Police training I had been on, and my feelings on it. I had then mentioned that some of the training I do for the Police isn't assessed. Teresa's feelings on this were made very clear: if you don't assess learning, then there is no point taking it; I disagreed with this.

I'll be honest and say I don't think I really explained my reasons for this particularly well, but we moved on to other topics and/or we became out of breath and therefore decided that running in silence may be more sustainable.

So in particular Teresa style, she's been stewing on this for a week and decided to challenge my assertion via a few questions in an email:

a. Is it worth teaching or learning something if you are never going to use it?
i. Where “using it” includes incorporating or translating it into another, larger ability that can be used.
b. Which is better in situations that require correct, full knowledge about X:
i. Knowing nothing about X
ii. Having half-knowledge or a misconception about X, but a belief that you know it fully

I'll directly answer these questions in a moment, but the context in which I made my assertion was quite specific. I've done a few courses for the Police where the majority of the course isn't necessary to gain new knowledge per se, but to alter or reinforce (depending on existing position) particular behaviour. This focus on soft skills is rather different to normal teaching.

The set of desired behaviours aren't really taught, instead the instructors tend to challenge the perceived "undesired" view and this is generally done within group discussions. Sometimes, these are pseudo assessed via role-plays, but how do you really assess whether someone has the desired behaviour when they are out on duty? I don't see how you can. Often behaviour won't be changed in the 2 hour training session either. Perhaps the trainer gives an interesting anecdote which challenges your existing beliefs, it may take may days or weeks of mulling over to accept it's merits and change your beliefs and therefore future behaviour. Again, how could you possibly assess this?

So to answer the questions put to me, no, I don't think it's worth teaching or learning something you never have a possibility of using it, however, where you are changing beliefs and behaviour, is this really an "ability"?

I agree that having misconceptions or not fully understanding certain topics can be damaging, especially in the Police, and I agree that when we're taught legislation and we have concrete information to learn that assessment is vital. That said, how can you possibly assess whether someone is more accommodating of other races or whether we understand what it's like to be mentally ill so we can treat them the best possible way? Answers on a postcard please.

Friday 30 March 2012

Playing the System

I mentioned in my previous post that after failing an evidential breath test, you will get a “statutory option” if you’re reading was under 50µg. This means that you can opt to replace that evidential breath test with either a urine or blood test, but that test will be the new result regardless if it’s higher or lower. The other thing to note is that you don’t get to choose whether its urine or blood, the police officer taking the test makes this decision.

Most people of good character would say that they’ve failed the test and there is no point opting for another, especially if the reading is closer to 50µg than the 35µg limit. However, unfortunately for all of us, the people that do this regularly will always choose the second test because as it stands they will already be charged and a second test gives them precious time to process more alcohol before we get a doctor out to take bloods for example. The worst that will happen is that they will just get another fail, at which point they are in no worse position than what they started with.

It’s often the case that people won’t know the implications of choices such as these, and however much we feel for someone, we can’t advise them one way or another. The people that do know the implications are often the ones that we don’t want to take the option, but we can't stop them and sometimes they do get off prosecution because of it.

Wednesday 28 March 2012

Drink Driving

Following on from my post about road safety, I mentioned that I attended a RTC recently that was carrying a group of four girls. As part of any RTC, we breathalyse both drivers as a matter of routine; drunken behaviour can often be masked as shock sets in, so getting a definitive reading is useful to rule out any other offences.

The 18 year old girl that I spoke to was shaken but otherwise ok bar a few bruises and was happy to blow into out breathalyser; unfortunately for her, she blew a roadside reading of 53µg per 100ml of breath. The legal limit is 35µg, so the rather shaken 18 year old had her first arrest. Perhaps I’m being a bit naïve here but I honestly believe that this girl didn’t intend to drink and drive, nor did she think that she was unfit to drive. She said that she had a bit of a heavy night and stopped drinking a bit after midnight. She was most certainly over the limit in the morning and this had come to a bit of a shock to her, so on top of having the guilt of seeing one of her friends being stretchered into an ambulance and another in pain with some broken bones, she now had the ordeal of being arrested by me.

I spoke with one of the ambulance staff to see what their intention was for the young driver. Usually, we would take her to the police station and require her to use the evidential breathalyser for the readings that would be used in the charging decision. We try and get them to custody as soon as possible because every minute they are on the roadside their body is metabolising the alcohol and we are essentially losing evidence. However, the ECP (Emergency Care Practitioner) said that they recommend she be taken to hospital just to check there were no unobvious injuries. I knew at the point that this job would keep me tied up for the rest of the shift.

We have a procedure for people being taken to the hospital; it involves a rather long form and a kit for taking blood. One of my colleagues remarked that the form is like an adventure game book, you read a paragraph and then get sent on a quest to page 78 for the next part of the story. It’s not quite so confusing when you get used to it, but it takes a minute or so to figure out. Once the patient/detainee has been assessed by a doctor we need their consent to take a sample. We then have to call out our own doctor to the hospital to draw the blood for impartiality reasons, all whilst the person is losing alcohol from their system. It took about 2 hours between being arrested and finally getting consent from the A&E doctor. After having the consent form signed, we were told that she was being discharged from hospital, so although we had started the hospital procedure, we decided that going to a police station would be quicker than waiting for our doctor to arrive, as well as realising A&E wanted the bed back.

So around 3 hours after being arrested my prisoner is finally on an evidential breathalyser. My colleague and I were convinced that she would not only blow under the limit but quite substantially below the limit. She hadn’t had a drink for around 12 hours and her body wasn’t absorbing any more alcohol in her system and at a rough guess we estimate that most people metabolise 10µg per hour of alcohol, so we expected a reading of maybe 20-25µg, well under the legal limit.

The result surprised us both: the lowest reading of the two we take was 45µg, remember the legal limit is 35µg. As she had blown over the limit but was under 50µg she had what is called “the statutory option” of having a urine or blood sample instead. We can’t offer advice on what to take but she was given the options and declined. I’ll follow up on the implications of this in another blog.

As she was still over the legal limit, we couldn’t interview her until she had sobered up a bit more which means she probably wouldn’t have been legal to drive until around 1530 in the afternoon at the earliest. Many of the general public don’t realise quite how long after the night before it can take to properly sober up and this really prove that.

So after just wanting to get a Mc Donalds, this otherwise law abiding 18 year old had written off her car and another, injured three of her friends and one other motorist and will likely get a drink driving conviction for simply not realising that she needs longer to sober up the next day. The funny thing is that it could have been so much worse and they are actually quite lucky.