[Berk21]

I have been working on a spread sheet on quality control for telecommunican technicians.
I am stumped on one or two things. I am trying to get a % of failure according to the work that has successfully completed according to what was QC'd and what failed. Is there such a way? Or formula? And would it be fair to evaluate them on such a equation?

Example:
Joe completes 25 jobs successfully. Quality control evaluates 32% of that work. Joe fails 5 of that 32%.. What can I base his failure % on?


I am having trouble trying to attach the excel file to show what I have. Any ideas?

[JBeaucaire]

Hello Berk, welcome to AMHD. I will be interested to see other's input. For me, there are only two choices based on that data:

1) Joe failed 5 out of the 8 QC'd jobs, meaning a 62.5% failure (so far)... the number will go up/down as QC completes more checks.

2) Joe failed 5 out of 25 jobs completed, meaning a 20% failure (so far)... the number will go only go up as QC completes more checks.

[JBeaucaire]

TO attach a workbook, EDIT your post, GO ADVANCED and click on the paperclip.

[mdosh01]

I would agree with JB, but with the following caveats:

1) 62.5% failure rate overall is calculated or "assumed." That is, statistically, if we were to evaluate all 25 jobs, statistics would say he most likely failed QC for 15 or so of those jobs. The complexity of the jobs need to be considered. Did QC, by chance or design, select the more complex jobs or was it a truly random sample?

2) It is "fact" that he failed 20% of jobs completed. But remember that not all jobs were run by QC.

My approach would be first to determine the audience of the information or who asked you to compile this data. Ask them which they would prefer to see. Second, if you are sharing this information with the workers, be sure they understand how their performance is being measured.

Not really an Excel answer! Sorry. If you were just looking for formulas, let me know.

[Berk21]

1st I would like to say hi to JB since he seems to be in every forum known to man... and also, that was the most information you gave me based on all I talked to you about.
I am here to see others input as far as being mathamaticly possible. I found out that there seems to be too much Probability in my assumtion that this will work accurately if at all.

So far I like the failed jobs divided by the QC'd work... which is a hair better then Pass divided by fail. Since I trying to get a round about answer on failure to QC.

As far as the attachment goes, it told me that it was too large... I tried it when I posted the first time.


This is being evaluated on a week by week basis. My boss is the one that is looking for this information to do a report (every Fri) on the work completed and who failed.
The complexity of the jobs I believe can't be a factor because, The jobs taken to quality control are the most complex ones... The easy ones are overlooked more due to the High standards we try to hold in this company. Some what like if you were a cook. No one will keep judging you on boiling water. Instead they will constantly monitor your 7 course meals to finalize how good of a cook you are... Thank you for all your input and I will check the thread to see if anyone else has something to say.

[JBeaucaire]

Just a thought, it seems the one factor you WANT to take into account and so far don't seem to be happy with is the concept of the "7-course-meal" vs. "Boiling Water" jobs.

Perhaps it is intimated by the fact someone does 25 jobs, but only 10 are taken to QC. But it's not implicit.

So, maybe your "tally" sheet of jobs might need to expand to indicate clearly to the bosses how many of each "type" of job the tech is completing. You might spot someone who seems only interested in/capable of completing "boiling water" job each week. That might be good to know.

Since you appear to have a "complexity" concept at play already, could you break out the jobs into a "complexity matrix" so it's easier to see what kinds of jobs are being done by each tech:

• Tier 1 Jobs - "Boiling Water" - anyone can do this, not much thought, crossing T's and dotting I's
• Tier 2 Jobs - "Baking Bread" - requires following simple but sequential steps to complete properly, being able to call it "complete" is still enough of a test, no QC required yet
• Tier 3 Jobs - "Casseroles" - Several unrelated sets of steps required, finished product not immediately indicative of success of job, QC recommended
• Tier 4 Jobs - "Multi-Course Meal" - Most complex jobs we do, many things can go wrong or get skipped and not be evident by results, Full QC required on all of these jobs

If your chart indicated all of these types of jobs, you could spot some "water boilers" and "bread bakers" much faster...