Want to protect your cyber security and still get fast solutions? Ask a secure question today.Go Premium

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 642
  • Last Modified:

How to apply Six Sigma DMAIC to software defects?

I am using the DMAIC model for software defects in a product that I am working on.

How would you best suggest I approach the 'measurement' phase.

Below is an excerpt from what I wrote for the 'definition' phase:

Multiple users have reported that they are unable to view changes made to a newly created process.

The application should allow a user to create a new process, and allow the user to proceed to make changes to the process.

While in edit/design mode, the application should allow the user to preview any changes that they make.  Multiple users have reported that when attempting to preview their changes, the preview screen remains blank.

Upon exiting the edit/design mode, if changes to the process were saved, the application should allow the user to select the newly created process from the process tree hierarchy and display a preview of the process map to the user in the preview pane.  Multiple users have reported that the application displays a blank screen in the preview pane when a newly created process with changes is selected from the process tree hierarchy.

Multiple users have reported that if they log out of the application, after making changes to a newly created process, then proceed to log back into the application, and then select the process from the process tree hierarchy, that the application does correctly present the process map with the respective changes to the user in the preview screen.
0
cnotley1982
Asked:
cnotley1982
1 Solution
 
redrumkevCommented:
 cnotley1982,
 
 How are you doing?
 
 For the measure stage, you need to find a measurement (or measurements) that are applicable to your project. This can be basically any quantifiable (measurable by numbers) means, Ex: Estimating Process/Project Capability.
 
 For any measurement, you must determine the key variables, key processes and what the process/project deems as key outputs. Examples could be costs, time, quality, number produced per shift, etc. What is most important that we can put into the process and what are we trying to get out? Once we know this, we can focus on measuring the results of changes made to these variables.
 
 For example, does more time help or by automating a sub-process can be see a gain in the number of outputs per shift. But also, is another variable changing, such as more outputs (a good thing) but quality goes down (of course this is a bad thing).
 
 In your definition, you have noted that multiple users are unable to view changes. How many is multiple? 3 out of 100 users or 8 out of 10, etc?  
   
The application should allow a user to create a new process, and allow the user to proceed to make changes to the process. Thus here you are looking for 100% accessibility to the changes by all (again 100%) of the users. Any user (even one) that cant access the new changes means a measured failure of the overall process change.  
   
While in edit/design mode, the application should allow the user to preview any changes that they make.  Multiple users have reported that when attempting to preview their changes, the preview screen remains blank.  
-> How many report this out of a total number (1 out of 3, 33%. 3 out of 20, 15%)  
-> Less than 100% able to see the preview =failure.  
-> What is the failure percentage? 20 users, 5 processes (thus 100 user processes) say only 3 cant be seen, 2 by one person and 1 by another, this means that we only have an overall success rate of 97% (failure, as it is <100%), however this is better than my other examples above of 33% and 15% respectively, thus there is an improvement. This is helpful when not solving the problem but making progress through changes (iterations).  
   
The upon exiting& I would follow a similar approach to the above, making changes that are specific (in both detail and nature) to the particular application/definition of this process.  
   
The log out, log in solution, does this result in 100% success? If so, perhaps (this is my own idea, which should go in the improvement phase, later tested in the control phase) the application should automatically restart, after a change, i.e., automatically exit the user and re-log them in, thus getting you back to 100%.

A great example of this use is in Windows, where Vista/Windows 7 can often continue to operate (no restart required) following the installation of software or changes to programs, where as XP and 98, almost always had to be restarted (possibly resulting in many restarts) following a change. Since you may not know what the O/S is, an auto restart might remove this possibility from the mix. But you may also want to add another process (such as -> check O/S, if <Windows XP & then restart, if => Windows Vista, do not restart).  

Hope this helps,  
Kevin  
   
   
0

Featured Post

Concerto's Cloud Advisory Services

Want to avoid the missteps to gaining all the benefits of the cloud? Learn more about the different assessment options from our Cloud Advisory team.

Tackle projects and never again get stuck behind a technical roadblock.
Join Now