Minitab 17 presents the most frequently used options and tools in Minitab.
Most statistical analysis requires you to follow a series of steps that are often determined by prior knowledge or the subject area you are researching.
Example: Explore data with graphs, Perform statistical analysis, Evaluate quality, Design an experiment.
An online book store has three regional shipping centers that distribute orders to customers. Each of the shipping centers uses a different computer system to enter and process order information. The store wants to identify which computer system is the most efficient to use at the three shipping centers.
Throughout Getting Started with Minitab 17, you will analyze data from shipping centers as you learn to use Minitab. You will make graphs and perform statistical analysis to identify which of the shipping centers has the most efficient computer system. You will then focus on the data from that center. First, you'll create control charts to test whether the hub's shipping process is in control. You will then perform a capability analysis to see if the process is within specification limits. Finally, you will conduct an experiment to determine how those processes can be improved.
You will also learn session commands and how to generate reports, prepare a worksheet, and customize Minitab.
Before starting the analysis, open Minitab and examine the Minitab user interface. On the Windows taskbar, select:
Start> All Programs> Minitab> Minitab 17 Statistical Software.
By default, Minitab opens with two visible windows and one minimized.
Session window
The session window displays the test results in text format. You can also enter session commands in this window instead of using Minitab's menus.
Worksheet
The worksheet, which is similar to a spreadsheet, is where you enter and organize your data. You can open multiple worksheets.
Project Manager
The third window, Project Manager, is minimized under the worksheet.
In a project, you can manipulate data, perform analysis, and generate graphs. Projects contain one or more worksheets.
Project files (.MPJ) store the following items:
- WorksheetsChartsSession window resultsSession command historyDialog box settingsWindow layoutOptions
Step 2
In the task bar select "Stat", then "Basic Statistics," and then "Display Descriptive Statistics."
Step 3
Select in the variables window the one required to find the mean, it can be by means of a double click or by means of the "Select" button to move the name of the variable to the window on the right.
Step 4
Step 5
Select the “Median” box and press “OK” 2 times. The median will be displayed in a new window.
The median for this particular exercise will be: 256
In this study, it can be observed that there is a lot of dispersion between the data, since the standard deviation is very large and the range confirms it, 334.
CHART OF SCATTERING
A scatter plot gives you a visual idea of what is happening with your data. Scatter charts are similar to line charts. The only difference is that the line chart has a solid line, while the scatter chart has a series of dots.
Scatter charts in Statistics create the basis for simple linear regression, where we take the scatter charts and try to create a usable model using functions. What all regression actually does is try to draw a line through all those points.
EXAMPLE:
A pizzeria wants to know how the baking time affects its product, since they have observed defective portions depending on said times, for this a sample of 7 pieces with different baking times (variable X) was taken and thus review the results of the defective portions (variable Y).
Step 1
Enter the information in 2 columns. One column must be variable X (the independent variable) and the second column must be variable Y (the dependent variable). Be sure to put a heading to your data in the first row of each column - this will create the scatter plot.
Data of Interest
A scatterplot or scatterplot or scatterplot is a type of mathematical diagram that uses Cartesian coordinates to display the values of two variables for a data set. The data is displayed as a set of points, each with the value of one variable that determines the position on the horizontal (x) axis and the value of the other variable determined by the position on the vertical axis
(and).
Step 2
Select "Graph" from the toolbar and then select "Scatter plot."
Step 3
Select the "Simple" option. In most cases, this is the option to use for scatter charts in basic statistics. You can choose any of the others (such as a scatter plot with lines), but they will rarely be used.
Step 4
Select the name assigned to the variable "y" in the left window, then select "Select" to move that variable "y" in the box of said variable or with double "click".
Step 5
Select the name of the variable x in the window on the left, then select "Select" to move that variable into the variable "x" box or double click.
Step 6
Select "OK" to create the scatter plot in Minitab. The graph will appear in a separate window.
As can be seen in the graph, when the baking time increases, the defective portions also increase, this indicates a strong positive correlation between variable X and variable Y, what the pizzeria must do is find the optimal time, where the pizza will be properly cooked in the shortest baking time.
After determining that a process is under statistical control, you want to know if that process is capable. A process is capable if it meets specifications and produces acceptable parts or results. To assess process capability, compare the spread of the process variation to the width of the specification limits.
IMPORTANT NOTE
Do not assess the capability of a process that is not under control, because estimates of the process capability might be incorrect.
PPK
Indices, or statistics, of capacity are a simple way to evaluate the capacity of the process. Since capability indices reduce process information to single digit numbers, it is easy to compare one process with another.
It is recommended before evaluating the capacity of the data processing, to make a normality analysis, the test results indicate whether or not the null hypothesis that the data comes from a normally distributed population should be rejected or not. You can perform a normality test and produce a normal probability plot in the same analysis. The normality test and the probability plot are often the best tools for judging normality, especially when it comes to small samples.
Step 1
Write the data in a Minitab worksheet from bottom to top (For this EXAMPLE we will use data taken from a resistance measurement of a fuse).
Data of Interest
As in Cpk, Ppk is a capability index that indicates whether a process is capable of satisfying two-sided specification limits. However, Ppk uses the current standard deviation for the process variance calculation, while Cpk uses an estimated standard deviation. The target value is taken into account with Ppk, so the system does not have to focus on the target value to be useful. A higher Ppk than 1 indicates that the process can meet the specifications.
Step 2
We select "Stat" in the toolbar, then we select "Basic Statistics" and finally "Normality Test"
And the following box will appear:
Step 3
We double “click” on variable C1 Resistance in the left window to move it to the right box in variable.
Step 4
We select “OK” and it shows us the Normal Probability graph in a separate window.
To determine if the data is normal, the P value must be> 0.05, in this case, the data is normal, since the P value is greater than 0.05.
Step 5
We give a "click" in the toolbar, in the window of "Show worksheets folder" to go back to the data taken
Step 6
In the toolbar, we select "Stat", "Quality Tools", Capability Analysis and Normal as follows:
And the following window will appear:
Step 7
Automatically the cursor is positioned in the first box on the right "Single Column", double "click" in the Left window on C1 Resistance and it will automatically move to the box on the right, then place 1 or 5, in the box of "Subgroup size" depending on how the readings were taken, if it was a single batch of 30 pieces, or it was 6 lots of 5 pieces, and the limits of the specification are set, minimum in "lower spec" and maximum in "Upper Spec ”And the OK icon is pressed.
Step 8
Once the graph is obtained, we can review the values of the process capacity, in this case we are interested in the Ppk.
A minimum value of Ppk of 1.67 is recommended, to consider that the process is capable, in this case, the result is 0.19, it means that our process is not capable, however, our Cp (Potential Capacity) is 3.40, and our standard deviation is very small, this indicates that perhaps the limits were set wrongly and only a limit adjustment would be needed, if the pattern remained the same in the following runs.
Step 2
In the toolbar select: "Stat", "Control Charts", "Variables Chart for Individuals" and "Individuals".
And the following window will appear:
Step 3
In the window on the left, double click on the variable X1 x1 and it is automatically transferred to the window on the right of variables.
Step 4
Press OK and the graph of individual values is generated.
In which it can be observed that all values are within control, with the limits provided by Minitab equivalent to 3 sigmas.
If you want to include in the graph all the tests to determine if the process is stable, in the toolbar select the option "Edit Last Dialog"
And in "I-Chart Options" in "Tests" choose "Perform all tests for special causes" and press OK
In this particular case, the graph remains the same, since the data does not show instability in the processes.
GRAPH C
In general, C charts are used to plot the total number of defects in a sample when the sample size is constant. You can inspect a type of defect, for example dead pixels.
You can also inspect for multiple defects together, such as dead pixels, stuck pixels, streaks, and fuzzy dots. An LCD screen can have 2 or 3 dead pixels and still be acceptable.
Example: An LCD screen manufacturer wants to monitor 17-inch LCD screens for defects. Technicians record the number of dead pixels for each subset of 10 screens per hour. They use a C chart to monitor the number of dead pixels.
Step 1
Load the data into each of the Minitab runs, using the first column.
Data of Interest
A defect is called the lack or lack of the proper and natural qualities of a thing or any natural or moral imperfection.
It is not the sole lack of a quality or perfection or the limited degree of possessing it, but rather these same negations when they affect a subject, whose concrete nature required said quality or perfection or a higher degree in it.
The U chart also graphs the defects. However, the U chart shows the number of defects per unit. The U chart is useful when the size of the subgroup is not constant.
Step 2
In the toolbar choose "Stat", "Control Charts", "Attribute Charts", "C"
Step 3
Double “click” in the left window on c1 number and the information will automatically move to the right window of “variables”
And choose OK.
On average, technicians find 7.56 dead pixels in each sample. Sample 5 is out of control. Technicians should try to identify any special causes that may have contributed to the unusually high number of dead pixels.
BIBLIOGRAPHY
- https://www.minitab.com/uploadedFiles/Documents/getting started / Minitab17_GettingStarted – es – pdf https://es.wikipedia.org/wiki/Mediana_(estad%C3%ADstica) https: //es.wikipedia. org / wiki / Dispersion_Diagram% C3% B3n http://www.manufacturingterms.com/Spanish/Ppk.html http://support.minitab.com/es–mx/minitab/17/topic–library/quality–tools/ control charts / understanding – attributes – control – charts / what – is – a – c – chart / http://support.minitab.com/es–mx/minitab/17/topic–library/quality–tools/capability analyzes / capability – metrics / ppk – ppu – ppl – and – cpm – measures – of – overall – process – capability /
