Today’s session (14 August 2017)

Hi everyone

Today I started by checking on students’ progress with Assignments 1 and 2. You should be underway with Assignment 1 by now, and this includes looking through the Resource Book to see the sort of articles that we are interested in looking at.

We then returned to where we began two weeks ago(!) when I showed you a picture of the planet Mars and asked you what sort of problems you thought there might be, when planning a trip to Mars. NASA is trying to solve these problems by building a Mars simulation on a red lava field in Hawaii and they expect this to better inform their decision making for living in space. Issues involved isolation, weightlessness, stimulation, etc.

I then introduced you to Katey Forth (http://www.radionz.co.nz/national/programmes/ninetonoon/audio/201810131/katey-forth-nasa-researcher,-entrepreneur-and-frisbee-whizz) who is an ex-NASA neurophysiologist who was involved in research around how the mind and body respond to weightless environments. Katey’s particular area of interest was around how the human body responds to zero-gravity and this led her into research around how we may be able to predict the chance of someone falling over.

This is a far more complex area of study than we might imagine, as walking (and running) always involves controlled falling, followed by recovery of balance. We also know that healthy people can fall unexpectedly (think back to the last time you might have lost your balance) and that people who look doddery or infirm can be very good at managing their balance (which can be why they do look doddery in their movements!)

Forth and her team have developed models which help them to make accurate predictions about who is actually in danger of falling. Her predictive algorithms allow us to differentiate between people who (on the one hand) look and act doddery but are in no real danger of falling over and (on the other) people who are at risk.

I can see exactly what she is talking about in the case of my own elderly parents: My mother looked incredibly unsteady on her feet in her latter years but she only ever fell over once that I know of. This was a combination of her ability to adapt to her limitations and a low neurophysiological risk of becoming unbalanced.

So, in order to solve the problem of how to prevent falls, Forth is suggesting that we should look less at whether someone ‘looks’ like they are in danger of falling over (or even whether they are in the ‘danger demographic’) and more at what measurable data tells us about sensory / motor-skills / impairments.

Being able to correctly predict falls is a good example of a problem which has BIG implications in other areas (e.g. for ACC, health, aged-care, retirement homes, etc.) In fact, Katey has now left NASA and she is working on her own business, building algorithms to predict falling.

This discussion then led us to think about what an algorithm is and why it is useful in problem solving and decision making. An algorithm is:

* A self-contained step-by-step set of operations to be performed to reach a solution.

* A set of rules for solving a problem in a finite number of steps

* A procedure for reaching a solution.

For more discussion on algorithms and their use as a problem solving tool, see http://www.comp.nus.edu.sg/~cs1101x/4_misc/jumpstart/chap3.pdf

Algorithms help us to clearly and accurately interpret information in order to reach sound decisions. As such, they are part of the scientific method and an important part of logical, sequential and critical thinking.

We then tried our hand at some simple algorithms. Our first attempt was to make a simple algorithm for opening a twist-top jar. We reduced this task to three simple steps:

* Pick up the jar in one hand

* Place your other hand on the lid of the jar and grasp firmly

* Twist the lid in an anti-clockwise direction until it separates from the jar.

That was a relatively simple algorithm but you should note that it also contains a lot of assumptions which might (in another context) require further explanation. For instance, what would you do if the jar had a reverse thread lid or if you only had one hand? Similarly, what is a ‘jar’ and what is ‘twisting’? The level of detail that we provide might be considerable different if we are writing our algorithm for a Martian!

I then asked you to write your own algorithm for a task which involved turning over playing cards and arranging them in numerical order. At the end of the exercise, I brought someone into the classroom to see if she could follow your instructions and accomplish the task (which she did).

We then moved on to looking at various problem solving models. The models that we looked at were reasonably simple and generic and many of them contained the same steps, framed in different way.

We started by talking about some of the things which could affect our ability to solve problems effectively. The class identified a number of factors including the people in the team, our perceptions, the quality of the information we work with and whether we understand the problem correctly.

I then showed you a slide that listed stages in the PS and DM process where errors could creep in (e.g. at planning stage or when we take action) and it gave some examples for each stage (e.g. Misinterpretation or not understanding; Poor technique; Poor assumptions at Problem definition stage). In reality, I cheated a little bit here because I used the Simplex model to give me each of the eight stages which I listed on my slide. More about Simplex in a moment …

We then moved on to look at several PSDM models.

There are literally thousands of models which can be applied to problem solving and decision making. Some of these models are intended to be very generic and applicable to many everyday situations. Others have been designed to address specific business-related situations (such as quality in manufactured goods or changing organisational behaviour). This means that you can choose any recognised PSDM model that is applicable for Assignment 2. You do NOT have to choose from the models which we discussed today.

I pointed out that many of the basic PSDM models share similar stages, often under slightly different names. Whether you ‘analyse’ your results or ‘study / check’ is partly a matter of preference and partly a shift in emphasis depending on what the originator of the model you are using wants you to think about.

We started with the basic IDEA model (Identify, Define, Explore and Action), which we looked at two weeks ago, in session 1. This is a really simple model in which we examine a problem, assess options and then do something to fix the problem. Put another way, we take a problem, we do something to it and after a while we get something else (remember that comment, in a moment …)

We then added a bit of complexity to the IDEA model by including a fifth stage (‘looking back’) and by looping the whole process back to the start so that we had a model which could continually loop until we solved our problem. These two additions allow us to review what we have done and to make further modifications if the problem isn’t solved the first time (or if the solution creates another problem). This is the IDEAL model.

The Problem Solving Loop is also a reiterative model (that means that it loops back to the beginning and restarts). In this case the five stages are:

1 Identify the problem

2 Explore information and create ideas

3 Select the best idea

4 Build and test the idea

5 Evaluate the results

(Repeat from the top, if required).

We also had a look at the Problem Solving Process slide and I drew your attention to the define-test-analyse-compare feature in this model (once again, more about that feature in a moment …)

We then changed our focus to creativity and arts, and had a look at a model used by Jonathan Milne at the Learning Connexion, in the Hutt Valley. Milne thinks of his process as a spiral which starts with an idea and then progresses to action and finally feedback. Each turn of the spiral is another stage in the process and we discussed whether we thought this represented ‘growth’ as the idea became more concrete and turned into something real.

Milne also refers to an idea expressed by American artist Jasper Johns. Johns described the artistic process as: ‘You take something, you do something to it and pretty soon you end up with something.’ In its own way, an artist creating something is another use of the PSDM approach to make a transformation and to achieve an outcome.

From there, we looked at the Simplex Problem Solving model. This is a model developed by consultant Min Basadur, who studied PS and DM in various organisational settings. He developed a process model consisting of eight stages, as follows:

1 Problem finding

2 Fact finding

3 Problem definition

4 Idea finding

5 Selection

6 Planning

7 Selling the idea

9 Action

The interesting thing about Basadur’s model is that he fits his eight stages around three simpler steps:

Problem formulation (steps 1 – 3)

Solution formulation (steps 4 and 5

Solution implementation (steps 6 – 8)

We then had a quick look at a very specific process model developed by John Kotter. Kotter is interested in change management (which we can think of as the problem of how to achieve change within organisations). Kotter’s model also consists of eight stages, as follows:

1 Create a sense of urgency

2 Build a guiding coalition

3 Form strategic vision and initiatives

4 Enlist volunteer army

5 Enable action by removing barriers

6 Generate short term wins

7 Sustain acceleration

8 Institute change

When we start talking about specific topics such as change management, then we are drilling down quite deep into one specific subject area. You should be able to see that a simple IDEA model is of limited value if your problem is about achieving change in the workplace. So, start thinking about which models are appropriate for the job (in this case, Assignment 2) and if you can’t find a model which is suitable then go looking for one that is!

Next up was a very simple quality management model, which underlies many other models. This is the Deming Cycle (or Shewhart Cycle), otherwise known as plan-do-study-act (or plan-do-check-act). Deming was an American engineer who spent time in Japan after World War II and played a big part in developing Japanese management systems. His interest was in statistical analysis and sampling defects in products.

(Remember my comment about the Problem Solving Process and how it uses define-test-analyse-compare (above)? How do you think this compares to Deming’s plan-do-check-act?)

Deming’s system is based around a four-stage process in which each stage informs the following stage before looping back to the start. So, we start by planning what we will do. Then, we do it to see what will happen. Next, we evaluate what happened. Finally, we make changes arising from our knowledge of what we did. Deming’s ‘problems’ were based around quality management in manufacturing environments but you might like to try this model for simple everyday problems and see what happens.

Then I showed you a simple representation of the Scientific Method, which is used to solve knowledge-based problems through the use of hypothesis testing. The interesting thing here, is that we can see plan-do-study-act steps from the Deming Cycle in the Scientific Method:

1 Make observations

2 Think of interesting questions

3 Formulate hypotheses

4 Develop testable predictions

5 Gather data to test predictions

6 Develop general theories

I also showed you the Ford 8D model and a simple engineering design process. All of these models help us to address problems in a systematic and logical manner. They also provide structure to our process and a framework that we can use to reach a conclusion (end point), sometimes through repeating the process over-and-over again.

We then did a practical exercise using the DECIDE model as our guide for solving problems and making decisions about available hospital beds in an emergency department at a hospital. The DECIDE model consists of the following stages:

1 Defining the problem

2 Establishing the criteria

3 Considering alternatives

4 Identifying the best option

5 Developing an action plan

6 Evaluating the solution

Remember to make a start on Assignment 1 and also have a close look at the jellybean jar in the Faculty’s admin office, sometime during the week. I need you to work your magic and come up with the correct number of jellybeans in the jar. At the mid-semester break whoever is closest to the actual number of jellybeans in the jar will win the whole jar.

Cheers

Phil