ASQ Six Sigma Green Belt – Course Wrap Up and Final Practice Exam

1. Top Ten Notes CBSA Exam

In DMEC, we started the project with D. There we defined the project, defined the problem, we took the management approval, we made a plan, we made a team which was working on this improvement project. So that was the defined phase. Then from the defined phase we moved to the measure phase. There we measured the current performance of the process which we wanted want to improve. Where do we stand today? That was the measure phase. And then after we took measurement of the current performance, we analyzed that, analyze that data in analyzed phase. And if you remember, in analyzed phase we had lots and lots of statistics. So having completed that was a major achievement for you, because now you have analyzed the problem, now you are ready to improve. When it comes to improvement, there are a number of things which you could improve in your process.

So you would have come out with number of ideas, number of suggestions to improve the process from the current performance to the new level of performance. But then out of so many options, you would select with some of these options and start implementing that. So this is something which we do in improved phase when it comes to the ASQ body of knowledge. For green belt exam, we have three main topics which we need to cover in the improved phase and these are design of experiments, root cause analysis and lean tools.

So let’s briefly talk about these topics. In design of experiments, we will be experimenting, we will be changing inputs and we will be seeing the effect of that on output so that slowly we can change the inputs, keep on changing the input, keep on doing experiments so that we can take our process to the most optimum level. That’s what we will be doing in design of experiments. In root cause analysis, you go to the root of the problem, you don’t attend to the symptoms to resolve a problem, what you want to do is you want to go to the root of that and solve that problem so that your problem is gone forever. In lean tools we’ll be talking about things like waste. How do we reduce waste, how do we reduce cycle time so that we can make things faster? How do we implement Kaizen? Kaizen will be to improve process step by step, slow improvement.

So with this background to the improved phase, let’s move on to the topic of design of experiment. What are we going to learn in this topic of design of experiments, which is the first topic here? First we will understand what the design of experiments is and then we will learn about some basic terms related to that. And then we will look at some design of experiments, graphs and plots. The design of experiments is a big topic in itself, so we are not required to master this topic. In the green belt, a lot of this will be discussed in black belt because this is an advanced topic. But as a green belt you need to have some basic understanding of the topic.

So that’s what we will be doing here. Look at the basic terms and look at some graphs and plots and try to interpret those graphs and plots as green belt. And when it comes to the basic terms, which is the first part of the design of experiments here we will be looking at these definitions. So if you look at this slide, there are a lot of terms here which are listed on this slide. Independent and dependent variables, factors and levels, treatment and response errors, replication block and randomization, repetition and effects. So these are the topics which we will be covering in basic terms for the design of experiment.

2. Blockchain Terminology Review

Do these experiments on our day to day life as well. So when you think of design of experiments, don’t think of this as a rocket science. This is something which you do in your life on a day to day basis without even knowing about this. So let’s take the example of a car. Now, what we are interested is in having more mileage from the car. So now we have numbers, number of things which affect mileage. So when I drive car and I am thinking to improve the mileage, the question comes to my mind does AC, the air conditioner affects the car mileage, yes or no, does number of passengers affect the car mileage? So if I have five people sitting in the car or one person driving the car, how much is the effect of that on car mileage? Same thing about the tire pressure, the speed and so many other factors which could affect the car mileage. So studying this, studying the effect of these factors on the car mileage is the design of experiments. You might have done this without even knowing. When you bake a cake, how much amount of each of these ingredients go into that, how much is the baking time, those are the input variable and output of that is the quality of the cake which you produce. So you keep on changing the ingredients so that you want to make a good cake.

Same thing in case of popcorn, how much time you want to put the popcorn in the microwave oven? Less time, more time, less moisture, more moisture. You keep on changing those inputs and you want to study what is the output of that, how you could make good quality popcorns there. Also you are doing design of experiments. When I’m selling these courses, the courses which I make, I’m studying the effect of number of factors on the course sale. Whether having a good video as the introduction, whether that makes a big difference in the sale or not, whether the course length affects my course sale, having quizzes, having closed captions. So there are a number of inputs which I can change to improve the course sale there. Also I’m doing design of experiments. So I think by now you will have a good understanding what a design of experiments is. So, to put in simple terms, design of experiments or Doe is a method to find out the relationship between factors affecting a process and the output of that process. Let’s take the same example of car mileage.

Number of passengers is one factor which could affect the car mileage. Having AC on or off is another factor, tire pressure is another factor, speed is another factor. I’ve just listed four factors here. There could be ten other factors which could affect the car mileage. So studying the relationship between these factors and their effect on the output and output is the car mileage. That’s what design of experiment is. So when we talk of design of experiments. The conventional approach to deal with this type of question was one thing at a time. You could call this as one variable at a time. Ovat one variable at a time or o fat or one factor at a time. That’s what probably you would have done earlier. You make change to one factor and see the effect of that.

Let’s say I want to see the car mileage. I drive my car for let’s say 100 kilometres or 500 kilometres with AC on and drive again 500 km with AC off. And I see how much gas did I use when the AC was on, when the AC was off? That will give me the effect of AC on the car mileage. Similarly, I would have used two different car tire pressures and studied the effect of that on the car mileage. This is one factor at a time. But what you do in design of experiments is you change multiple factors at a time. And I will come to the advantage of that because the conventional wisdom says you change one thing and keep everything constant and you study the effect of that factor. But design of experiments, we change almost all these things in a systematic way and study the effect of that. And based on that we could find out that what is the effect of each of these, which of these are major factors, which of these really do not affect the output and why we do number of changes at a time rather than going one factor at a time? Because when you change number of things you can conduct the same experiment with minimum effort and expenditure. Because if you change one factor at a time, then you will have to do lots and lots of experiments compared to using the design of experiments approach. There you could find out the same thing with less number of experiments. So less number of experiments means less effort and less experimental.

3. Blockchain Roles

We have a term called y is equal to FX and y is the output and x is the inputs. So what does this mean? Is that output or the y is the function of inputs. And that makes sense. When you want to change the output of a process, you need to change inputs because output is affected by inputs. So if we go back to the same example of car mileage, car mileage is the output. Output are also called as dependent variable. And then we have number of inputs. Inputs are number of passengers, AC on or off, tire pressure, speed. These are inputs. These inputs are something which we can control, these are something which are independent, we can independently control these inputs, we can change AC to on or AC to off and the effect of that will be on the car mileage.

So car mileage is a dependent variable depending on these inputs. When it comes to design of experiments, the technical term for inputs is the factors. So when we change inputs, we are talking about changing factors and the effect of that or the output is called as the response or the outcome. So when you look at any process, look at that process in terms of inputs and the output, output is something which you are interested in. Input is something which you can change. Input is called as factors in case of design of experiments and output is called as the response. So here we have the definition of response and factor. So, as we earlier said, response is the output of a process. Sometimes this is also called as dependent variable. So response here in our example was the car mileage. And when it comes to factor, a factor of an experiment is a controlled independent variable, a variable whose level is set by the experimenter.

These can be numeric or these could be categorical. So factor is something which we change. Factor was having AC on or off. Factor was having different level of tire pressure. Factor was having different number of passengers in the car. And when I say that factors could be numeric or categorical, numeric is in terms of tire pressure, tire pressure could be 30 PSI, 31 PSI, 32 PSI, 32. 3 PSI, whatever level we want to set. So that is a numeric factor. Categorical factor is where you are having categories. And the example of that was having AC on or AC off, having passengers as 1234 or five. Those were five categories. So you could have factors which are numeric or you could have factors which are categorical. So when we talked about factors in that we said that the experimenter could set the factors at different levels.

Let’s say number of passengers, the experimenter could set the number of passengers at 1234 or five, whatever number the experimenter is interested in. Experimenter will put the AC factor as on or off tire pressure as 25 PSI, 30 PSI or 35 PSI. And same way speed also at different levels. So here we are talking about factors and levels. So we have four factors here. Factors are number of passengers, AC, tire pressure and speed. Levels are levels for each of these factors which are listed here. So you need to be clear about the factor and you need to be clear that each factor has different levels. So this was one set of definitions.

4. Taking the actual CBSA Exam

These were factors and levels. Now, let’s look at few more definitions here, and in this video, we will be talking about two main terms which are treatment and response. Let’s talk about treatment first. So, treatment is a specific combination of factor levels, whose effect is to be compared with other treatments. So, when we said that, number of passengers, air conditioning, tire pressure and speed, these are factors, and each factors had different levels. So, if we take one particular combination of these. So, let’s say my first experiment is that I will use one passenger, I will keep my AC on, and I will use my tire pressure as 25 PSI. And I will use, let’s say speed as 120 km. So what I’m doing is I am picking different levels of factors. So combined together, this becomes one treatment.

And then in the next experiment, I will make another combination of these factor levels and make another treatment. So that’s the treatment. Treatment is a combination of factor levels, the effect of which we want to compare. So here we were talking about four factors. To make things simpler, let’s just use two factors. Let’s say we are doing this experiment with just two factors, and these factors are number of passengers and air conditioner. So here on the table, if you see on the left, there are ten possible treatments which are possible in this particular experiment. So the treatment number one is having one passenger and having AC on. The second treatment would be having one passenger again, but this time, AC is off. Same thing we do with two passengers, AC on and off.

Three passengers, four passengers, five passengers. So in total, we end up with doing ten experiments. Or here we have ten treatments. Now, what we do is let’s add one more factor here, and the next factor is, let’s say tire pressure. So if I add a tire pressure, then I will have five multiplied by two multiplied by three experiments to be done. So, in the earlier case, the number of treatments which I needed was five for the number of passengers and two for the air conditioner. So, five multiplied by two gave me ten experiments or ten treatments. Here, if I add tire pressure also as the factor here, then I need to do five multiplied by two multiplied by three experiments, which will be 30 experiments. And each of these treatment is the combination of these factor levels.

Let’s say the first treatment or the first experiment will be one passenger AC on, and the tire pressure as 25 PSI first time, 30 PSI second time, and 35 PSI third time. Then I again go for one passenger, but this time AC off. And I do three experiments with three tire pressure levels. So, if you see here in this particular case, I will end up with 30 experiments, or I will give 30 treatments and I will study the effect of those treatments. Since this was becoming very complicated because I had five levels for number of passengers, I had two levels for AC and I had three levels for tire pressure and I had four levels for speed. So here, to do this design of experiment, the full factorial or the full experiment, I needed to do five multiplied by two, multiplied by three, multiplied by four experiments. So which will be 120 experiments. So many times you have limitations on how many experiments you can do.

So, to make things simpler than what I decided was, instead of having various number of levels, let me stick to two number of levels. So in case of passengers, I will have just two levels, one passenger or five passengers minimum and the maximum. And in case of AC, it’s on and off. In case of tire pressure, I decided to go for 25 PSI and 35 PSI. In case of speed, I decided let me go for 60 and 100 km/hour. So here I have two levels for each of these four factors in your exam. Six sigma green belt, six sigma black belt exam there is a good chance that you will be asked a question where you will be given levels and factors and you will be asked how many experiments are needed. So, if you have equal number of levels for all the factors, here I have four factors, and for all the four factors I have just two levels only. In that particular case, the number of experiments will be level to the power factor. Here, if I am having two levels for each of these four factors, then what I need to do is I need to do two to the power four or 16 experiments. And here I am talking about full factorial. And the full factorial experiment is where we are looking at all the possible treatments. So here we are looking at all the possible combination. That is called as full factorial experiment.

So in full factorial experiments, you will be needing 16 experiment. We have not talked about half factorial, but let’s say if you are asked how many experiments you will need to do, in this particular case, if you are doing half factorial experiment, half factorial means half of these. Instead of 16, you would just need to do 16 divided by two or eight experiments in case of half factorial, quarter factorial, even half of that. Another example is if you have five factors, let’s say if I add one more factor into this, let’s say city drive versus the highway drive, and that also has two levels, that in that case I will be needing two to the power of five, which is 32 experiments in case of full factorial experiment. So these are number of experiments which I need to do.

So, we talked about treatment. Treatment was a combination of factor levels and we talked about experiment. In experiment, we are looking at the result of these treatments. Next thing which we need to understand here is the response. And the response we have already talked. Response is the output of the process. So if we go back to the simpler example where we just took two factors, number of passengers and AC, on and off. So there we needed ten treatments for these two factor levels. In case of passengers, I had five levels. In case of AC, I had two levels.

So I needed ten treatments. What I do is I record the output of that, and the output of that is how many liters of gas I needed for driving 100 km. This is the result of the experiment. And that’s what I put here. This is called as response. So when I experimented with with one passenger and AC on, I ended up in spending 13. 8 liters for driving 100 km. So that’s the response. Next response was 13. 6. So I keep on recording these because that’s what I need.