0:02 hi there we are now on the third
0:05 graphical analysis tool that we will
0:07 cover inside your Lean Six Sigma yellow
0:10 belt curriculum this time we will have
0:14 Pareto chart Pareto chart is mainly used
0:17 for priorities Ishod but let's look into
0:19 the details for us to better understand
0:23 how to use this new Pareto chart is a
0:25 set of bar charts where the bars are
0:26 arranged in descending order of
0:29 magnitude the bars may represent defect
0:33 categories locations departments and
0:36 others the problem-solving tool that
0:38 involves ranking of potential problem
0:42 areas or sources of variation according
0:45 to their contribution to Aust or total
0:49 variation typically 80% of the effect
0:51 comes from the 20% of the possible causes
0:52 causes
0:57 meaning 80% of the problem has accounted
1:01 to 20% of the causes or the factors so
1:05 efforts are best spent on this vital few
1:08 causes temporarily ignoring the trivial
1:12 many causes as said earlier it is a
1:15 prioritization tool now we're looking
1:17 for 80% of the problem which is
1:21 accounted to 20% of this factors let's
1:24 say you have 10 possible root causes
1:27 probably two or three among those 10
1:30 root causes is contributing around 80%
1:32 of the problem so you might want to
1:35 focus your energy in your efforts on
1:37 solving those problems that we call
1:40 vital view this is most commonly used
1:43 for a prioritization of resources and
1:46 also for validation of root causes now
1:48 let's study the anatomy of a Pareto
1:51 chart as you can see here this is the
1:56 category on the x-axis and this contains
2:01 the bar chart the data which is arranged
2:03 in descending order containing the
2:07 frequency of the categories and this one
2:10 is the cumulative line cumulative
2:13 percentage line meaning we are adding up
2:15 the contribution of this one to get this
2:18 point and this one and this one to get
2:21 this point this one this one and this
2:22 one to get this point
2:25 so on and so forth until it reaches the
2:29 100% summation of contribution so we're
2:32 finding the 80% of cumulative
2:35 contribution of the factors that is
2:37 being presented here so let's practice
2:40 Pareto chart using it is study about
2:44 floss pin floss a quality engineer for
2:46 an automotive supply company wants to
2:48 decrease the number of car door panels
2:50 that are rejected because of paint flaws
2:52 the engineer wants to determine whether
2:54 a relationship exists between the type
2:57 of paint floss and the ship during which
2:59 the door panels are mean but in this
3:01 case we will use Pareto chart to
3:03 determine what particular type of paint
3:06 flow is affecting the operations now
3:09 let's go to the worksheet and then let's
3:13 click flow our data is represented by
3:17 log data meaning there is a lock for one
3:20 scratch one scratch pill deal not
3:23 summarized so you don't have to worry
3:29 just copy it and go to Minitab create
3:31 another new worksheet and then you can
3:36 close this paste your data so you have a
3:40 data that that's in-text okay now we
3:43 want to go to Pareto chart so click that
3:47 and then find quality tools then you
3:48 will see para to chart on the second level
3:49 level
3:52 click para to child in a Pareto chart
3:54 you will be asked to provide at least
3:57 the number of defects or attribute data
4:00 and the frequency now because we don't
4:02 have any frequency so you can see here
4:04 if we click this field there will be no
4:07 data that can be assigned to this but if
4:10 you can click this one there will be the
4:13 flaws so you only have to do is double
4:16 left-click floss okay now there is an
4:19 option here that will ask you if there
4:21 are too many factors that are available
4:24 let's say 50 but the other factors don't
4:27 have any value that is significant
4:30 can actually combine them into the 95
4:33 percent remaining cumulative percentage
4:35 or if you don't want to do that just
4:37 click don't combine visually we click
4:40 combine so that we can consolidate those
4:44 none significant factors so click OK
4:48 then there is your palette a chart by Flo
4:49 Flo
4:52 so let's study the output palette to
4:55 chart again this axis talks about the
4:58 frequency of problems this is the
5:02 cumulative percentage line and this axis
5:06 talks about the categories of the
5:10 problem we have ill scratch other and
5:12 smudge because we only have four
5:14 categories then we don't need the other
5:18 consolidated factor here ok so let's
5:22 focus on the numbers we have 15 for pill
5:26 13 for scratch 6 for other and smudge 6
5:28 so we have here the individual
5:31 percentage contribution and here is the
5:34 cumulative percentage contribution so 37
5:39 bring down 37 here 37 plus 32 0.5 will
5:42 give you 70 70 plus 15 will give you 85
5:47 and 85 plus 15 will give you 100 now
5:51 this is the cumulative percentage using
5:54 Pareto chart we are asked to determine
5:58 the 80 percent of contribution that is
6:02 accounted to 20 percent of the factors
6:05 in Pareto chart analysis we can use
6:08 three possible algorithms or heuristics
6:11 as to decide which are the vital few
6:14 factors first we can check on the
6:17 cumulative percentage and look for the
6:19 cumulative percentage that is near 80
6:23 percent so for this case we have 85
6:27 maybe we have 85 therefore we can say
6:30 that the vital few factor includes till
6:33 scratch and other okay but as you can
6:35 see if we will use this for
6:38 prioritization it's a little bit weird
6:40 because 3 out of 4 has been
6:43 now if you will look into the cumulative
6:45 line if you will look closely to this
6:48 there is what we call the breaking point
6:50 here they the breaking point
6:53 it breaks the tone of your cumulative
6:55 line just like what happened here
6:58 meaning even the closest to 80 is 85
7:01 here you have to consider the breaking
7:04 point here it tells you that inside out
7:08 including this one the other you have to
7:11 stop here on pill and scratch because it
7:15 gives you the idea of the majority of
7:19 the problem surrounded to minority so
7:21 for you to be able to satisfy the rule
7:24 of thumb or the arbitrary points that is
7:27 being given to us when you use Pareto
7:29 chart analysis so far now we will focus
7:33 on this pill and scratch as are vital
7:36 few causes it so happened that we have
7:38 the breaking point that is why we
7:41 haven't chosen 85 pass the start of our
7:44 vital view so when we have this you will
7:48 refer to this and then get whichever is
7:50 in close to the left so that is scratch
7:54 and then pill okay so that is how you
7:56 prioritize and how you use Pareto chart
7:59 for prioritization purposes
8:03 now let's look at another example we
8:06 have a clothing manufacturer traveler
8:08 number and type of defects in a line of
8:11 clothing so our source data could be
8:13 found including defect so let's go to
8:17 excel this one then you have to copy
8:21 this all of these columns and then we
8:22 have to create another worksheet and
8:28 then close this one okay okay we have to
8:32 copy paste now we will be doing another
8:35 set of Pareto charts so we have to go to
8:39 stat let then quality tools we have
8:41 Pareto chart here you have to erase
8:46 floss okay so we are asked to where is
8:49 the data of defects or attribute data in
8:54 the defect categories is in c1
8:57 so you have to double-click see one now
8:59 we are asked where are the frequencies
9:01 so we have the count of each defects on
9:05 c2 so we have to click see two and down
9:10 then choose okay now we have our title
9:15 chart scroll down little okay now let's
9:19 focus on the categories we have each
9:22 count here the data is here the
9:24 percentage and the cumulative percentage
9:27 we want to understand how we can
9:29 prioritize the problem-solving
9:31 initiative therefore you have to figure
9:34 out the vital few causes we have to find
9:36 on the cumulative percentage a
9:39 percentage that is near 80 so we found
9:42 out that it's eighty two point five
9:45 eighty two point five so that's third
9:47 one two three
9:50 meaning the vital few causes include
9:53 missing button stitching errors in loose
9:55 thread okay
9:58 we don't have any clear breaking point
10:02 here so we have to use another method so
10:05 eighty two point five now we can use the
10:07 method that is by tracing on the
10:10 cumulative percentage line so on the
10:11 cumulative percentage line you have to
10:15 pick 80 and then chase it to the left
10:18 until it reaches or it touches the
10:21 cumulative percentage line when it
10:24 intersects with document the percentage
10:27 line trapped it down to what particular
10:30 causes will it fall it fell on loose
10:33 trail so our right of yukos's now is
10:36 missing but on stitching errors in loose
10:38 thread meaning eighty percent of the
10:40 problems on to almost 20% of the causes
10:45 or lesser so this is how you use Pareto
10:47 chart for your prioritization activities
10:51 there are cases that frequency is not
10:54 enough for prioritization let's see what
10:58 if those who has higher frequency only
11:00 reflects less cost to the organization
11:03 and those who have smaller frequency has
11:05 higher causing back into the
11:08 organization so this is where
11:11 the value of this particular people here
11:13 will come into the picture as you can
11:16 see the defect categories provided as
11:17 well as the heart are this is the usual
11:20 data table that we have for a Pareto
11:22 chart but in case that we want to check
11:25 the cost contribution then you have to
11:27 put cost per accordance on the third
11:29 goal loop and then get the product of
11:32 count times cost that is 217 times point
11:35 17 which will give you 36 point 89 and
11:39 then you reach 180 1.89 for hemming
11:41 errors okay so this is now total cost
11:44 rather than frequency now let's create
11:47 another control chart so if you want to
11:50 rerun a function you just have to click
11:54 we just have to hold ctrl and press E
11:58 and then the last dialog box will appear
12:02 so let's replace count this time we want
12:04 to check the Pareto effect using total
12:07 cost that is c4 so you have to replace
12:10 count by c4 and then you just have to
12:15 click OK now we have this now what do we
12:18 see we have a different type of vital
12:20 view now we have a different set of
12:23 vital fuel now now if we want to have a
12:25 side-by-side comparison we just have to
12:29 click this chart and then click this
12:32 drop down button and then click layout
12:35 tool let's reduce the number of columns
12:45 to 1 and let's get this chart okay so we
12:48 now have two charts here and then finish
12:51 now we have two charts this is the first
12:53 chart that we did and this is the second
12:56 chart that we need comparing them we
12:59 have different sets now of vital view we
13:01 will based on frequencies the vital few
13:04 are missing but on stitching errors in
13:07 whose thread but if we will use total
13:12 cost it is different now we have fabric
13:15 floss hemming error and stitching errors
13:17 now using that particular example we
13:19 have seen the effect of combining
13:22 frequency and cost
13:24 now in actual application you have to be
13:26 very careful in doing Pareto chart of
13:29 course is a critical component to what
13:31 you are doing so this is how you use for
13:34 at a chart in analyzing and prioritizing
13:37 your problem categories or your sources
13:40 of problem later on when you do your
13:42 case study you can also use this
13:44 whenever you want to prioritize your
13:47 list of potential root causes now when
13:50 using Pareto chart there are things to
13:52 consider data collected during a short
13:54 period of time specially from an
13:57 unstable process may lead to incorrect
13:59 conclusions because the data may not be
14:02 reliable you may get a misleading idea
14:04 of the distribution of defects and
14:07 causes when the process is not in
14:10 control the causes may be unstable at
14:12 the vital few problems may change from
14:15 week to week short periods of time may
14:17 not be representative of your process as
14:20 a whole so before you get the data we
14:23 have to check first the process
14:26 stability so for you to have a more
14:29 clear picture of the problem now if you
14:31 are dealing with data collected during
14:34 long periods of time there can be
14:37 changes examine the data for
14:39 certification are changes in the problem
14:42 distribution over time if there has been
14:44 a change then you move that data and
14:47 start a new choose categories carefully
14:49 if your initial predator analysis does
14:52 that yield useful results that is you
14:55 cannot see any potato child effect that
14:57 would help you then rerun it and ensure
14:59 that your categories are meaningful and
15:01 that your other category is not too
15:03 large this is the combined category
15:06 because if that's the case maybe there
15:09 is something that is wrong with the data
15:12 collection or the analysis itself choose
15:15 weighing criteria carefully for example
15:19 we use cost per gear cost me be a more
15:21 useful measure for prioritization the
15:23 number of occurrences
15:25 especially when the cost of various
15:28 defects differ concentrating on the
15:29 problems with the highest frequency
15:31 should decrease the total number of
15:34 items needing rework concentrating on
15:36 the problems with the highest score
15:38 should increase the financial benefits
15:40 of the improvement so you have the
15:42 balance between the two there are always
15:45 work around and just have to figure out
15:47 what are those for you to find the
15:50 balance last the goal of a Pareto
15:52 analysis is to obtain maximum reward
15:54 from the quality efforts but that
15:57 doesn't mean that small this whole
15:59 problem should be ignored until the
16:02 larger problems are so so this is by
16:05 applying quick Queens so that during the
16:07 analysis period until we are reading the
16:10 major improvements to be done we are
16:12 doing something for us to improve little
16:15 by little so that our tips that you have
16:18 to consider when you're using Pareto chart