A picture is worth a thousand words – actually, the brain processes images 60,000 times faster than it can read, so it’s worth turning your data into pictures. When you’re trying to make a point, impress leadership or win the hearts and minds of process participants, graphs and charts are the way to go. In this 1-hour intermediate webinar we’ll give you some step-by-step training on how to take a column of data and bring it to life on the big screen.
We’ll be using SigmaXL during the webinar. SigmaXL is a Microsoft Excel add-on which we highly recommend for its ease of use, low cost and compatibility with both Macs and PCs. Click here for even more instructions on how to use SigmaXL.
- How Charts and Data Tell the Story
- What Does a Histogram Tell You?
- What Does a Run Chart Tell You?
- What Does a Pareto Chart Tell You?
- What Does a Box Plot Tell You?
- What Does a Scatter Diagram Tell You?
- How Charting Tricks and Tips Can Help
Tools & Templates
Elisabeth Swan, Managing Partner & Executive Advisor
Elisabeth is a Managing Partner, Executive Advisor and Master Black Belt of GoLeanSixSigma.com. Elisabeth has over 25 years of success helping leading organizations like Amazon, Charles Schwab, and Starwood Hotels & Resorts build problem solving muscles and use Lean Six Sigma to achieve their goals.
Tracy: Hi, everyone. Welcome to GoLeanSixSigma.com’s webinar. Thanks for spending some quality time with us today. We have about 500 plus people registered for this exciting webinar. We are glad that you were able to join us today. Lean and Six Sigma are the go-to improvement methods used by lots of leading organizations all over the world. And our goal is to every month craft webinars just for you, our global learner community to help simplify concepts and tools of Lean Six Sigma so that our community can understand and apply them more easily and be more successful.
Today’s webinar is 5 Ways to Create Charts and Graphs to Highlight Your Work.
And I’m Tracy O’Rourke, a managing partner at GoLeanSixSigma.com. Today’s presenter is also a managing partner at GoLeanSixSigma.com, my colleague, the wonderfully innovative and consummately passionate about learning, Elisabeth Swan. How are you today, Elisabeth?
Elisabeth: I’m good, Tracy. How about you?
Our Expert: Elisabeth
Tracy: I’m good. So Elisabeth is an executive advisor, a Master Black Belt, consultant, coach, and trainer. And for over 25 years, she has helped leading organizations like Amazon.com, Charles Schwab, Target, Volvo, Alberta Health Services, Starwood Hotels, and many others successfully apply Lean Six Sigma to achieve their goals. Elisabeth lives in Cape Cod with her husband and geriatric cat.
How to Interact
So here are a few housekeeping notes before we begin. During this webinar, all attendees will be in listen-only mode. At the end of the presentation, we’ll have a question and answer session. However, please feel free to ask questions at any time by entering them into the question area.
We’re also going to have some really interesting polls for you to answer. And if we don’t answer all your questions during the webinar, we’ll be sure to post answers as well as share a recording of this webinar on our website at GoLeanSixSigma.com.
So now go ahead. Please join us for our first activity, sharing where are you from? We love to hear where people are registering in from. What part of the world you’re in and how late people are staying up and how much coffee you had to drink to attend this webinar.
So please click on Ask a Question and type in where you’re joining us from today. All right. Let’s see. So we have folks from Ontario, California, Red Deer, Alberta, Eugene, Oregon, Santa Monica, California, Rolling Meadows, Illinois, Delhi, India, Chicago, New York City, Atlanta, Mexicali, Pennsylvania, Brazil, Portugal, Lagos, Nigeria, Utah, Olympia, Washington, Tampa, Florida, Wisconsin, Montreal, Bosnia, and there are a lot of people all over the world coming to listen to your presentation today, Elisabeth.
Tracy: So, over to you now.
Who is GoLeanSixSigma.com?
Elisabeth: Thank you. Wow! Thanks for the warm introduction, Tracy. And I’m sorry I can’t see all of you but I’m happy you join us. As Tracy said earlier, we’ve both been with GoLeanSixSigma.com since its inception. And our mission is to make it easy for you to build your problem-solving muscles. So that means we simplify complex concepts, we make our training really practical and I think really enjoyable. We provide a running case study at the Bahama Bistro, our restaurant chain. Team applies all the tools. We’re going to do it again today with them.
And aside from this webinar series, we put out blogs, podcasts, book reviews, lots of other information to help you get where you need to go. We’ve used and taught Lean Six Sigma for decades because it supplies the best toolkit for problem-solving. And thankfully, there’s a growing list of companies who agrees with us. And here are some of the organizations that we’ve worked with.
We’ve Helped People From…
So as you can see there’s bricks and mortar, there’s online, there’s diverse industries such as healthcare, financial services, manufacturing, and government. And the reason is because Lean Six Sigma is about problem-solving. And as Tracy loves to say, once you have an organization, you’ve got problems. So like all of you, these companies want to be the best at problem-solving. So you’re in good company.
So more on benefits later, but first let’s look at today’s agenda.
Today, we’re going to use a mini case study back to our buddies at the Bahama Bistro and we’re going to show how charts and data tell the story of an improvement project. And we’re going to focus on these 5 basic charts. They are the most commonly seen in improvement projects so we’re going to stick with these although there are other charts. Then we’re going to give you some tips on how to create these charts and graphs on your own.
Throughout this, we’re going to be using – everything I show you is going to be in SigmaXL. There are two different – there are two software packages that we use with people and that’s SigmaXL and Minitab. And even if you don’t have them, there are ways. I’ll tell you at the end how to create these charts on your own.
And we’re not going to go into screenshot by screenshot in case anyone was hoping for that although I’ll send you to a link at the end so you can see screenshot by screenshot.
Mini-Case Study Problem
So first up, let’s dive into our mini case study. All right. So read this carefully. It’s full of really critical information so I want you guys to study it. OK now stop.
Consider how many times you’ve been in this position. How often are you reading an email or trying to decipher a project charter, trying to determine what’s important of what you’re reading? What does the person who wrote it trying to tell you? And it takes a while as you can see here.
How often are you reading an email or trying to decipher a project charter, trying to determine what’s important of what you’re reading?
“ABC It for Me”
So we’re going to try a shorter version. All right. So the short version is 6 pieces of dishware break per day on average here at the Bistro. OK. Now, the title of this slide is ABC It for Me, which is something I learn from a team lead who was quoting his boss. She had no time for long explanations. If he started out, she’d say, “Just ABC it for me.” And I really love that. It boils it down to, “It’s your job to make this understandable to me. I don’t have the time.” So this team lead had to get to the habit of editing and tightening his work.
Another favorite quote of mine, Mark Twain once said, “Sorry for the long note. If I’d had the time, I’d have written a shorter one.” Well, that’s funny because it’s true. Editing your emails, reviewing what you’ve put into a project charter, that all takes time. And if you don’t take the time, then you put the burden on the reader.
“Sorry for the long note. If I’d had the time, I’d have written a shorter one.”
So I once had someone described getting emails back from their boss and they had this 3-letter acronym that’s all the female would say, TMW. TMW. And I said, “Well, what does it mean?” He said, “It meant too many words.” So it’s another way of saying, “ABC it for me.” So funny again but it’s true.
So let’s take the same issue and think about numbers. OK. I want you to study this carefully. This is important data and we’re going to work with this. OK. Stop. All right. If you’re already on to me, good for you. But again, I’m just trying to make a point. I see this kind of wall of data all the time.
In fact, I pulled this. I re-edited all the names to protect the innocent. But this is from a team lead giving me some information. Everybody has spreadsheets of data. But like texting, it takes a while to read and decipher. So we’re going to study the greatest hits of charts and graphs to see how they can help us get our point across faster. Tell our story.
5 Ways to Visualize Data
So we got 5 ways we’re going to visualize data. So research shows that the human brain processes images 60,000 times faster than text. Think about that, 60,000 times is a lot. When you’re working on a project, your job is to tell the story of improvement and your data gives you this great opportunity to tell that story visually. So we’re going to use these 5 charting methods to see how they can help.
But first, we want to hear from you. So we’re going to go to a poll. All right. So Tracy, I’m going to launch this one and that is, which is the bigger pet peeve? And here we go.
OK. How about you, Tracy?
Tracy: Well, I really like that you have an option for both because they both drive me nuts, lots of text. But also, because pictures as you have said, they tell us a thousand words in a picture and it’s so much easier to talk to stories. So I really love to see that. How about you?
Elisabeth: Yeah. No, I think you referenced that a picture is worth a thousand words. But they updated the research to say, “It’s 60,000 times faster with an image.” So that’s like left in the dust. That’s like old school now or that’s like a myth that has been busted or something.
But yeah, I feel like this – it’s work. It does take work but it’s also practice like trying to figure out what are you supposed to say? If you don’t have practice, if you don’t know about charts, if you haven’t had the exposure then it’s not necessarily an obvious thing to go to a visual. What would the visual be? That takes a little getting used to.
OK. Well, can we see the results, Tracy?
Tracy: So, let’s see. Actually, it’s funny. A lot of people did agree to this, 57% said both using a lot of text and using “walls of data” or both of their pet peeves. And then that was followed closely – well, not actually closely, a distant second was when people use just the “wall of data” followed by the text, 14% there and then “I don’t really thought about it,” 7% and neither 5%. I wonder what their pet peeve is.
Elisabeth: We’ll find out. OK, good. So this is an issue for most you listening in today. So let’s go back to our presentation. All right.
First up, we’re going to cover the histogram. So what is a histogram? They are charts that show the distribution of values produced by a process. They are visual displays of how much variation exists in the process. So that sounds OK. But again, words not so helpful. Images are more helpful.
What’s a Histogram?
So let’s look at a basic example. So this is an example. It’s not related to our broken dish issue we just heard about. We’ll use this to understand the parts of a histogram. This is often referred to as a snapshot of our data. In this case, it’s a snapshot of the time to seat customers in the month of June. The X or horizontal axis has the scale. In this case, it’s minutes. We can see that across. The Y axis shows frequency at each level. So if we take the shortest bar to the left, it shows that one time it took 14 minutes to seat the customer. So the bar only goes up to 1 on the Y axis. That’s how we read that.
You can calculate the range. That’s the maximum which is 22.8 minutes minus the minimum which is 14 minutes, so almost a 9-minute range here. That’s the variation in the process.
And the other thing we can see is the center of the data. The tallest bar is the mode. The stats offer for these charts also calculates the median, the middle most point, the mean, and other basic stats. You can also see the shape of the data. So it’s a normal bell curve or it’s skewed. This one looks a little normal. And it’s a great chart since it’s packed with information. You can see it at a glance.
What Does the Histogram Tell Us?
But let’s get back to our broken dishware. What can a histogram tell us about broken dishware? The original issue was that there’s an average of 6 pieces of dishware that break per day. And we see in projects, and it’s helpful. Averages are helpful. We deal in averages. And they’re great but your process is more going on than the average, some classic misuses of averages.
To tell a funny story, there is a – 1984, those of you familiar with the US colleges, the graduates of the University of North Carolina, starting salary was over $100,000. And that’s 1984. It’s a long time ago. And that sounds very promising. It might make you decide or might have made a lot of people decide to go down there in 1985.
But it also happened to be the year that Michael Jordan graduated from the University of North Carolina and signed with the Chicago Bulls, a professional basketball team. He was a superstar. So that’s skewed the average. So it made everyone – everything looked great. So that average wasn’t really helpful.
Another one, the average of diaper wear is 40 years old. OK? This one seems a bit silly but in this case, the population slipped down the middle. So very young, very old so average doesn’t work.
So let’s take a histogram of the – take a look at the broken dish histogram. So this is – the average was 6 days but the mode, the highest bar is 5 and that’s because it got skewed because one day, there were 20 dishes broken. So you can see that out on the side there. It’s that one little bar showing us that that happened once.
So, I don’t know if Michael Jordan was using dishes to shoot basketballs but the data skewed. They seem to break at least 4 dishes a day. But that seems like a lot. And we don’t know when the 20 dishes were broken. Was that early in the month? Was it late? Of course, there’s snapshot but it doesn’t tell us when anything happened.
What Else Can a Histogram Do?
So what else can a histogram do? Another helpful thing about histogram is when you add customer requirements. You can see here, we’ve got a target of 5 minutes. It would be great if we can seat customers in 5 minutes. But USL, which stands for Upper Specification Limit, that’s customers say, “We don’t want to stand any longer than 10 minutes.” So that’s your upper spec limit. And that’s unacceptable.
So this chart can include those customer requirements so then you can compare how you’re your process perform compared to what customers want? You can also change the color of the bars. On this case, we highlight everything over 10 minutes to show basically customers will be seeing red. They would be really peeved if they had – if one of these times was their experience.
So it’s a great way to highlight the beginning of your story, the baseline. Where are we starting? How bad is this problem? And you want people to see how bad it is so they’ll support you. So part of telling your story is look at this. This really needs fixing. Then you get people, leadership, stakeholders, team members saying, “Yeah, yeah, let’s do this.” It looks bad literally. The image is not good.
Then you get people, leadership, stakeholders, team members saying, “Yeah, yeah, let’s do this.”
The Run Chart (Time Plot)
OK. So that was really helpful. Now, let’s see what else we can learn by showcasing our data with a Run Chart, also called a Time Plot. So this is a time series. It’s a graph that shows data in sequence over time. And again, let’s start with a basic example.
What’s a Run Chart (Time Plot)?
So the keys that run chart shows you your data and shows you what’s happening over time. That could be a day, a month, a year, just a shift. The X or horizontal axis is always over time, as a time period.
Things like time to seat the customer, defects per day can be on the Y or the vertical axis, so whatever you’re tracking. It’s a very good all-purpose chart.
The classic center of a Run Chart is the median and half the data points are above and half the data points are below. So in this case, there are 20 data points, so the highest 10 values are above the median and the lowest 10 values are below the median. And that’s also helpful to see.
We’re used to seeing time plots in newspapers and you should always have one of your project data. Your project should always include a time chart so people can see what are the trends and shifts of your data. Is it random fluctuation? Is it going up? Is it going down? The Run Chart tells you what is happening over time.
Your project should always include a time chart so people can see what are the trends and shifts of your data.
What Does the Run Chart Tell Us?
Back to our broken dishes project. What can the Run Chart tell us? What is happening to the dishware breakage over time? Aha! It’s getting worse. And the one day when someone broke 20 dishes happened on the last day in November. As soon as you have 7 points in a row above or below the median, the process has shifted. So it’s time to do more digging since this is costing the company money, probably alarming the customers.
What Else Can a Run Chart Do?
Let’s look at what else a Run Chart can do. Run Charts are time plots, share what happened before and after a problem was addressed. So you can insert right here on this date, if you had dates going across your horizontal axis, you could see exactly when it happened. Say, that’s when we implemented our solutions. Now, you see things got better. And that’s what you want to see.
So in the case of time to seat customers, the team would want to highlight the fact that the time decrease once they fixed the issue. There’s a shift here but it’s a good one. That’s what you want to showcase at the end of a project. Again, that’s a great thing to show stakeholders, leadership, and other team members.
The Pareto Chart
So let’s move on to another chart that we can use. And this one is the Pareto Chart. So a Pareto is a quality chart. It helps identify the most significant sources of defects. And again, it sounds good. Let’s take a look at an example.
What’s a Pareto Chart?
So this chart, we’re moving on to just an example. We’re going to come back to our broken dish issue. This one is showing different types of food orders. The Pareto Chart is good for categorizing data. In this case, the team was categorizing what people ordered out of 100 orders. So we did the 100 orders. And the X axis shows us what categories, in this case, categories are salads, combinations, solids, and so on. The Y axis shows us the frequency or the number of each category.
In the case of these orders, 22 out of 100 were for salads. And 20 out of 100 were for combinations. So if you add the salad orders to the combination orders, you get 42% of the total. This helps the team zero in on the category of highest impact, highest importance. It could be defects. In this case, it’s just what’s the most common thing they order. That might help them with their ordering process.
So in the case of broken dishes, let’s come back and see what can a Pareto tells us about the dishes. There was that percentage on the second Y axis and there is the accumulative lines on your Pareto showing you as you add those bars together, the percentage goes up.
What Does the Pareto Tell Us?
So coming back to broken dishes. In the case of broken dishes, the question is what are we measuring? What are we breaking the most of? And it turns out, it’s bowls. Out of 188 broken dishes in the month of November, 150 or thankfully for our second Y axis, 80% were bowls. You can see it right there on the accumulative line. It shows us. So that’s very helpful. It’s going to help the team narrow the focus of their investigation. This chart caused them to change the project scope so they narrowed it down to just deal with broken bowls.
Paretos are a great way to highlight data. They help others see at a glance the biggest source of problems. They can also be used to dig deeper. So let’s go back to those 100 food orders and how do – now, there’s 150 out of 188.
Paretos are a great way to highlight data. They help others see at a glance the biggest source of problems.
What Else Can a Pareto Chart Do?
So what else can a Pareto do? In this food order chart, the salads were the most common order. So it’s even more helpful to the team ordering supplies to know which salads are most popular. So let’s take a look. Let’s say the salads. They broke it out and they said, “Well, of the salads, the most popular one is the flying fish Waldorf salad.” It sounds delicious.
There’s a great feature of these charts. One Pareto can lead to another and they are called nested or cascading Paretos. It requires that you collect more data but it’s digging to root cause. It’s trying to understand what’s underneath. And that can often be very helpful. So results and understanding what’s inside the biggest bar. What’s part of why this particular issue is the biggest issue?
The Box Plot
All right. Let’s take a look at another one of our charts. Next up is the Box Plot. I think it’s a very misunderstood chart but it’s my favorite. And what is it? It is known as a Box Plot or a Box and Whisker Plot. It’s a graph dividing the data set into core tiles. There’s a new word. And with the center box containing 50% of the data and two whiskers that each represents 25% of the data, which might sound like some kind of cartoon cat. So let’s get a closer look.
What Is a Box Plot?
So this has this new term, not necessarily intuitive. It deals in quartiles. But that just means fourths. You divide the data up into fourth. So it shows the distribution, sort of like a Histogram. The highest quartile, the middle 50% of the data and the lowest 25% of the data. So it gives you that distribution. It shows you whether it’s skewed. The center is the median.
No Really – What Is a Box Plot?
OK. But it’s not an intuitive chart as I said. So really, what is a Box Plot? Now, you can see the box plot divides data into four. So let’s take a look at this and see how does a Box Plot relate to a Time Plot. So the Time Plot, it’s already cut in half. There is the median. So half my data, I’ve got 20 data points, so 10 points above the median, 10 points below, lowest 10 values, highest 10 values.
Now, if I’m going to divide it into fours, then I’ve got to divide it again and look at the lowest 5 points. So we’re going to look at the bottom 5 values and let’s draw another line there so we can see where that would land.
And then the top quartile represents the highest 5 values. All right. So top values once again, I’m going to put a line and those are the highest 5 values. So you can see the Box Plot is just a representation of that distribution. So that’s hopefully making it a little clearer again, using pictures better than explaining what a Box Plot is.
So experiment with these. Get a feel for what they do. This is just two methods of displaying data, how they’re related. We’ve been showing you a Single Box Plot. But this chart is best for comparisons. So let’s come back to the broken dishes.
What Does the Box Plot Tell Us?
OK. What is a Box Plot tell us? So we can use the Box Plot to stratify the data by shift, breakfast and lunch and dinner. And breakfast and lunch look basically – breakfast and dinner look lower than lunch. And the Box Plots can do that comparison better than putting histograms next to each other. Histograms should be harder to get that visual comparison but Box Plots are thinner so you can see it.
You can also see there’s that outlier again. So once again, it tells you this is the normal set of data. Here’s the median. Here’s the spread of your data. And lunch has more variations, a higher median. It also shows you that outlier. It will do a calculation or most offer packages to do that calculation and figure out what doesn’t fit, what are the outliers here.
So it’s just letting you know. And some packages will put an asterisk there. It could be a dot. It could be an asterisk. But it’s – and it might have more than one. It would be more than one outlier. It’s just letting you know that data point was different from the rest of the data set.
What Else Can Box Plot Do?
What else can a Box Plot do? So one common method of comparison is to reflect on the process after it has been improved. The Box Plots are great for showing before and after. And it’s much easier to compare. Like I said, Box Plots are easier to compare than Histograms. Histograms are wider whereas this gives you that nice simple construct of the distribution so you can see it.
And this is also a great thing to put in your project and show the leadership. Look where we started. Look what we got to. It’s great to show that disparity. It tells a great story. It highlights your good work. And it’s very simple chart and people get it. They can see that difference between the two. You just have to label it and let them know what are they looking at. So that one again, tells your story really quickly.
Mystery of the Broken Bowls
So let’s come back, now that we’ve learned a little few things about our mystery of the broken bowls. So breakages increasing over time, 80% of the breakage is bowls. Most of the breakage is happening at lunch time. When asked about this, the staff said soup and salad special was driving the sales of clam chowder, hence the use of bowls. They said customers keep asking for the chowder to be hotter. They are saying it is November. They like it nice and hot when they come in here.
They suspect the staff is dropping the bowls because they’re too hot to handle. So if we had a chart, we had some other tool that could tell us whether or not they heat of the chowder affected the breaking of the bowls. So that’s where we’re going now. We want to try to understand the mystery, the root cause. We have our suspicious. What chart would be helpful to prove or disprove this theory?
The Scatter Plot
And that brings us to a Scatter Plot. So the last chart helped us understand if there’s a correlation between two variables. It’s a great chart to use when trying to prove disprove hypotheses. A Scatter Plot is a chart that shows whether there’s a relationship between two variables. So let’s figure out what do we mean by that.
What’s a Scatter Plot?
Here’s an example. It’s comparing outside daily temperature, this is just a simple example, to daily sales, trying to see if there’s a relationship between the daily temperature and the sales. And these are paired variables. We have to know both the temperature and the sales for the day. So we pair those together.
And that’s two columns of data. The Y or the dependent variable is on a vertical axis. And then the X or the independent variable is on the horizontal. So we’re trying to see if changes in X result in changes in Y.
So this chart is telling us well, yeah, as the daily temperature increases, that’s the independent variable, the daily sales increase, that’s the dependent variable. So yes, as the exchanges, you will see changes in Y. And then you can probably caution that correlation does not guarantee causation.
And one of the classic examples is that as ice cream sales go up in New York City, crime rates also go up. So can you say that ice cream causes crime? No. But you can say there’s a lurking variable there that the heat is leading to the ice cream sales which is leading to more access to buildings, hence the crime rate is going up.
So this graph shows that correlation and whether there is one there or not and whether it’s positive, negative. There are different types of correlation.
So this graph shows that correlation and whether there is one there or not and whether it’s positive, negative.
What Does the Scatter Tell Us?
So let’s get back to the hot chowder. Does the breakage go up as the temperature of the chowder goes up? Let’s take a look. Yes. We can watch the waiters and waitresses to confirm. But the data shows correlation, a positive correlation. So this is helpful. The hotter the chowder, the more bowls they dropped. We’ve got some data here. We’d like to do some observation. Always back up your data with some process observations especially when you’re looking at correlation. But this is pretty helpful.
Always back up your data with some process observations especially when you’re looking at correlation.
What Else Can a Scatter Plot Do?
Then let’s ask the question. What else can Scatter Plots do? OK. Of course, there may not be any correlation. I worked with a team in a hotel. They were trying to understand why their staffing model wasn’t working. They didn’t necessarily have the right number of people on stuff at any given time of day. And we looked at their model and they were always basing their staffing in their on-site restaurant on the number of guests reservations for that day.
But when they compare the reservations to the number of orders, they found nothing. A chart like this that just showed scattered dots. No real correlation. And it’s only when they remove lunch and dinner orders that they found correlation. So if you think about that, the number of reservations did correlate to the number of breakfast orders.
And that makes sense if you’ve ever been to a hotel eating breakfast there, kind of a no-brainer. You might meet your colleagues there. You might spend some time briefly there in the morning and then you’re off to a meeting. You’re off into town. It makes sense that it would correlate to breakfast.
So this Scatter Plot helped them change staffing model for the better. It made a big difference and they started applying that to all their hotels. So it helped not just one property but they were able to translate that improvement across the board.
Let’s come back to our clam chowder. So, helping you with our hypotheses statements. These charts are really helpful. So problem solved with the clam chowder. It’s official. Clam chowder has the biggest selling lunch items along with that sandwich, that combo, the soup and salad were still popular but didn’t have enough trace to go around. And since they were making it hotter to respond to customers, they didn’t have enough trays. The bowls were too hot to handle. And since the soup and salad special is not going away, the Bistro is investing in more trays.
Let’s get another poll from you guys and think about what chart or graph are you most comfortable using? And if you’re not really using them now, you don’t have to engage in this one. But let’s launch – let’s go back.
OK, Tracy. I got it. So Tracy, what’s your favorite?
Elisabeth: The one we didn’t include was Controlled Charts. I do see a lot of those. But that’s just a form of a run chart, a time plot, in a more sophisticated one. So that one I think makes sense. I think those are – the ones you mentioned are helpful. I’m always trying to sell the Box Plot. That’s where I am.
Tracy: Me too.
Elisabeth: OK. So what do we see?
Tracy: OK. So Histogram squeaked out as number one with 31% followed very closely by a Run Chart at 30% and then the Pareto at 28%. The Scattered Diagram is 7% and unfortunately, last and least is the Box Plot at 3%. And we’re hoping to change that, right Elisabeth? We want people to feel just as comfortable with the Box Plot as a Histogram.
Elisabeth: It’s elusive today, Tracy. But if we have our way, we’ll turn them into that. So with Scatter plots, the use of those is really related to whether you actually have paired data and you suspect there’s correlation. That’s not terribly common in projects. I think histogram really common. You can use that for tons of projects. One chart you can use for everything. Paretos, if you’re looking at defects, it’s certainly helpful to get that. So totally makes sense to me why those top there are the top three vote getters.
How Charting Tricks and Tips Can Help
All right. So let’s hide that and come back. And let’s move on. So let’s wrap up with some tips and tricks. The Histogram is known for showing you the shape, the center, and the spread of your data. I didn’t include it there but it often will also list for you what’s the median, what’s the maximum, what’s the minimum so you can work the range out. Really good all-around data. We call them sort of basic stats. Really helpful.
Next one up, the Run Chart. Looking at your data over time. Every project should be looking at their project why. What’s the main thing you’re trying to move up or down? And you should be looking at what’s happening to that over time which we’re trying to reduce the variation and if you’re trying to push it up or push it down then you should see it shift in your process.
….you should be looking at what’s happening to that over time which we’re trying to reduce the variation and if you’re trying to push it up or push it down then you should see it shift in your process.
Pareto Chart, this follows the 80-20 rule. You’re looking for where are most of my issues coming from? 80% of the breakage was the bowls. That made total sense for this team to narrow their scope and it helped everyone focusing on what’s happening with bowls that made them drive their hypotheses and it made them drive the next chart to understand why all the breakage. So great for understanding the sources of defects.
Box Plot, the main word there is comparison. Do you want to compare two strata? What’s happening between two different operators, who finishes faster, the most experienced person or the new person, if you want to understand the difference between two locations, if you want to understand before and after, some kind of comparison.
In the case of the broken bowls, it was which shift, what’s the biggest breakage by shift? And it was lunch time which again, helped them with their hypothesis. So you can use it to forward the analysis of your project and do it in a great visual way.
Last up, Scatter Plot. That’s looking for correlation. The chart shows whether or not two variables are correlated. Again, there could be lurking variable. But at least you can see are they connected in some way?
So those are the five charts.
What Data Goes With Each Chart?
So this is the behind the scenes. What does it look like? And a Histogram is just pulling a column of data. So in this case, breakage per day, they had a column. They looked at the month of November. They just wanted to know how many bowls broke in each of those days.
The next one is the Run Chart. It’s that same column of data but now, they have added the time, the dates in the left-hand column. And most stat packages let you include that column of data so you can show when things happened. So if you want to see things over time, it helps to be able to pinpoint when those data points happened. So that’s really helpful.
Paretos are different from everybody else. They like a table and on the left column, you got the categories. This is all the different types of breakage there was. And then in the second column, you have the totals. So you may have other spreadsheets that you’re pulling this data from but to run a Pareto, you need to put it into this table form so that it can do the counts and display them for you. So that’s a little bit different.
The Box Plot, again, this depends on what are you comparing. You can do a Single Box Plot just to look at distribution in a different way, similar information to a Histogram, diagram with a median, and looking at your data in quartile. So that’s helpful in terms of what is the top 25% of my data, where the middle of my data, and where is the lowest 25% of my data.
And for that comparison, you just need to know what are the stratifications. We looked at it by shifts. And again, you just need a column for each of those stratifications, if it’s different locations, if it’s different people, if it’s operators. You need a column for each. So that’s our Box Plot.
And lastly is our Scatter Plot. And this one as I mentioned, needs paired data. So in this case, we paired the clam chowder temperature, average temperature, to the number of bowls broken on that day. And that paired data was plotted on the Scatter Plot. So those two data points, where do they meet and then do they form some kind of a correlation? In this case, they did.
So that gives you a little background on what does each chart great for, why do we use, what’s it going to tell us in terms of our story, and then how do you set that data up? What does it look like in a spreadsheet? Kind of a basic question but it’s important. And for a lot of people, this is not obvious. If you don’t have practice, it’s not necessarily easy to know how to set your data. So this is an important piece of getting these charts and data to tell a story of your project.
If you don’t have practice, it’s not necessarily easy to know how to set your data.
So I want to know now in this next poll is, what’s most difficult for you when you’re creating charts and graphs?
So this one, Tracy, let’s see if I get any better at launching this. There we go. All right, Tracy. What do you see with folks you’ve been teaching in terms of what’s the most difficult for them?
Tracy: So I actually find that it’s selecting the right chart, which I guess would be A. And so sometimes, people call me and they will say, “Oh, I’m trying to get my presentation ready for the Green Belt project and I can’t get this data that run on Scatter Plot.” And I’ll say, “OK, send it to me and I’ll look at it.” And guess what? They only have one column of data. We could not run a Scatter Plot that way. So not only selecting it but setting it up right.
Elisabeth: Yeah. Yeah. I often have people, so they’ve got their data and they got some charting software and they send me a graph and say, “OK, what is this telling me?”
Elisabeth: Kind of a Hail Mary with, “Did I do the right thing?” And I’ll usually come back and go, “Well, it might be better both as a Pareto as opposed to that Histogram.” But let’s take a look at what these guys said.
Tracy: OK. So our highest one is figuring out what chart to use in the given situation was 34% followed by some combination of all of these things and then probably setting up a data in a spreadsheet, 21%. I actually thought that would be a little higher. Collecting the right data at 15% and using the charting software at 0%. That is really interesting, Elisabeth.
Elisabeth: It is.
Tracy: So that tells me that you focused on the right things for this webinar.
Elisabeth: I did.
Elisabeth: Well, that’s great. And that is helpful to see. We’re going to do another webinar about collecting data to build these charts and tell your story, sort of the background to this webinar. But this is really helpful information for us. Thank you.
All right. So I’m going to hide that and come back.
And this is just letting you know that on our website, there is a how-to on every single one of the charts we just ran today. And each one of these comes with a data set. There’s actually one data set for one Excel worksheet with a tab for each of these. And we’ve got two for each because we give it to you for SigmaXL users and we also give to you for Minitab users so you can have your choice there.
I want to point out that if you do not have a package and you want to try one, SigmaXL is the least expensive. It’s also can be used on both Mac and PC. And you can also get a 30-day free trial. So if you’re wondering whether you want to invest or not, that’s a great way to download that free trial and try out all these charts with the data set if you want to. But since you guys had no trouble with using charting software, I don’t even know if you need this. But it’s there for you if you do.
Tracy: And I’ll just – can I just add, Elisabeth, too that Minitab historically has been very expensive. But now, they have really good deals for students. And you can even rent the software for a few months for real inexpensive if you wanted to try it out and learn it too.
Elisabeth: Thank you. Thanks a lot, Tracy. I did not know. That was helpful.
Today We Covered
Elisabeth: So let’s review. We had a little visit to the Bahama Bistro and see how charts and data told their story of their broken bowls. And these 5 charts are basic but they have great versatility in what visuals they provide. So visit our website. Try them up for yourself. And just adding one chart in place of text or data will make a huge difference in how others view your work. Remember, people process images 60,000 times faster than text. And that is a huge, huge difference.
Tracy: They’re going to have to update that little disclaimer, a picture is worth 60,000 words, right?
Elisabeth: Or something like that, Tracy. I have to work out the math on that. I think there’s some calculation to do.
OK. We have question and answer period for you. While you take your time to enter your questions into the question area, we’re going to give you a little bit of a preview of what’s to come.
One thing is just remember, you can get all kinds of training and certification. Yellow Belt is still always free. This webinar is just a start. It might be a start of your journey. It might be in the middle of it. But the Lean and Six Sigma tools and concepts are there for you with more training.
And let’s look at the next webinar. Tracy, want to give a little primer on this one?
Upcoming Webinar – 8/21 11am PDT
Tracy: Absolutely. So go ahead and type your questions in again into the question window so that Elisabeth can answer any questions that you might have. And I’ll just tell you a little bit about our webinar coming up. So this is 5 Ways to Ensure Green Belts Apply Their New Skills After Certification.
So this is actually something we’ve seen a lot of organizations ignore is you’ve got these newly minted Green Belts. They completed a project that’s part of their certification and they’re just sitting there now, sitting pretty waiting to use those skills and no place to use them.
So, we’re going to talk about ways that you can put your Green Belts talent to work. And that will be coming up on August 21st at 11AM PDT.
Elisabeth: That’s a great topic, Tracy. I’m really excited to share this with clients just because this is a gap I see again and again and again. So I am excited to see it.
The other thing we’ve got going is our latest Just-In-Time Podcast. The podcast if you haven’t checked in is the latest apps to help with problem-solving, the latest problem-solving in the news, latest books about problem-solving, Q&As from our learners.
And at the end, there’s always an interview with problem-solving expert. This month’s expert is Mitch Ditkoff. He’s the President of Idea Champions, calls himself an Innovation Provocateur. He is very funny. He has a new book out on Storytelling at Work and it’s kind of related to what we’re talking about here, how do you tell your story? It’s visual but also being good at telling stories really helps. So he talks about how that has helped in the process improvement world. So please tune into our podcast.
Tracy, did you get any questions from our lovely listeners?
Tracy: At the moment, I have three. So if you want a question answered by Elisabeth, please type it in the chat window and I’m going to go ahead and ask her a few of the questions that have come up already.
So from Chris, he asked, “How do you know what the Y and the X values are on a chart?”
Tracy: Well, that’s a great question. So the Y is the thing that you are tracking. So in this case, we were trying to understand bowl breakage. So that is our Y. Because our project is focused on decreasing the breakage of bowls, that means that’s our focus and that’s our Y.
When I looked at – when the team looked at the different shifts, those were Xs. So breakage of bowls by shift. Now, you’ve got an X. Now, I’m looking at breakage of bowls by breakfast, breakage of bowls by lunch, breakage of bowls by dinner. They are still my Y but now, I’m looking at different stratification factors, also known as the X.
If you’re dealing with a Scatter Plot, again, that Y is the dependent variable. It’s the breakage. What is causing the breakage? It’s dependent on some X or some Xs. And in the case of the Scatter Plot, we looked at, did the temperature of the chowder impact the breakage. So the temperature of the chowder was the X because that would cause the breakage. So in your mind, think about what’s my main measure? What is my metric? What am I trying to understand? What am I trying to reduce or increase? That’s my Y.
And then how am I stratifying? Those are Xs. Where am I looking for correlation? That’s an X. So I hope that helps in terms of X and Y.
Tracy: Wonderful. Thank you. Very helpful. OK. This one is from Olga and she said, “Can you recommend a training or a webinar to get more familiarized with Minitab?” And I’m going to assume that she means besides the GoLeanSixSigma.com website that has the tutorials on the website for free.
Elisabeth: Yeah, I’m trying to think. I know we did some research on Minitab. I know most of these vendors provide you with videos and introductions. And Tracy, do you have more recent experience with the Minitab website?
Tracy: I will say this. I used YouTube a lot and I know a lot of people – when I first started using Minitab, there was even back then decades ago, there were a lot of videos on how to use them. And there still are. So I find that the go-to place is you can YouTube it and find out how they are running it. So I don’t have a particular webinar to recommend but I find that that can be really helpful because it’s very specific example that you can usually run like how to run a t-test or those kinds of things. So I’d recommend that.
And then they also, if your purchase Minitab or if you’ve got it, they have tutorials within Minitab to learn it. So it’s already built into the software.
Elisabeth: I will say that having learned Minitab years and years ago, that was the only or the premier stat package when we working with clients 25 years ago. So it has been around a while. But even back then, they had the largest manual back when they printed it. And when they stopped printing it, they still had the largest online treasure trove of explanations and instructions throughout the training. Throughout the software, you could always go to it. So my remembrance is that they were quite strong with it but I am not sure, Tracy, to your point whether they provide it before you’ve had a chance – before you’ve had to buy it.
But I do back your experience with YouTube that you pretty much find anything.
Elisabeth: OK. That was a great question, Olga. Thank you.
Tracy: OK, great. So then we have a question, “How do you check capability through a Histogram?” So I’m guessing process capabilities.
Elisabeth: Yeah, great question. When you go to create your histogram, you’re going to have some options. Most software and the one I just used SigmaXL, gives you an option. You can use a simple one or you can get a little more fancy. They always have a few layers you can go up in terms of sophistication of the chart.
So if you go up one layer, then you got some options. And thing it allows you to do is put in that target, that idea of what’s ideal. In that case for example of sitting, the ideal was 5 minutes before a customer is seated. The upper spec limit was 10. So they give you that opportunity to put in both a target and an upper spec limit. And that comes early in your project when you are in the define phase and you’re focusing on voice of the customer. What do they want? What are they looking for? What are you trying to achieve in your projects?
So, those are options when you go to build your histogram and before you complete it, you just enter those and it will put those two values and it will put them right on the histogram. Also, a great question.
Tracy: OK. Thank you. So we have on other question and I don’t know if you’ll be able to answer this without more information but we’ll give it a shot. What chart would I use to show volume of patients scheduled per user?
Elisabeth: Volume of patients scheduled per user.
Tracy: I’m not really sure what the user part is.
Elisabeth: Yeah. The user is tricky. But if you were say, looking at every day, say a month of volumes. So you were tracking each day and you had a spreadsheet that the day was the row and the columns were the different users. Then each day you would end enter how many patients did that user have for each of those users, and then complete that for the month and take that three columns and that 30 or 20-data points and turn that into a Box Plot so you can compare the volumes per day for the different users.
That would be one way if I’m understanding your question correctly. And I’m not still clear what a user is. But that’s one way to look at your – Tracy, may I know what the question was again? Was it capability?
Tracy: I was going the same way. What chart would you use that show volume of patients scheduled per user? So it looks like continuous. It’s potentially continuous or account and then you could have users, 5 users. User 1, let’s say there are 5 users and you’re looking at volume of patients per user, I would go with the Box Plot too.
Elisabeth: Yeah, because what the underlying – I just want to make sure we got comparison right and that capability. I was trying to remember what the word the question used. But comparison sounds like was key and that’s Box Plot.
Tracy: Yup, or it could be Pareto, right? It could be Pareto. So volume of patients by user, user 1, user 2, user 3 and then the count. It could be another one that you could use as well.
Elisabeth: If you just did total counts as opposed to looking at it by day.
Elisabeth: Yup. OK.
Tracy: Cool. Yeah. So there you got two potential charts depending on how you put the data into your spreadsheet.
Tracy: Alright. I think that brings us to the end of our webinar. And I want to say thank you all of you for joining us. Please join us again next month for the next webinar. And come visit us anytime for the blogs, podcasts, and other excerpts. Take care, everybody.
Elisabeth: Thank you. Have a great day.
View our upcoming webinars and join live so you can ask questions and let us know what you’d like to us to cover next. We’re busy building new webinars all the time. And we’re happy to know you’re busy too – building your problem-solving muscles – keep it up!