Stuck on ASQ CQE Questions

Hello! My husband is going to take the ASQ CQE in a few weeks. There are a few questions from the two official ASQ CQE exams that he does not understand. Can someone help us figure out the logic behind these tests? Also, I recently discovered five sample CQE tests online from a place called IE&MS Research Center, which as far as I can tell, is a Korean company. Here’s the first one: http://www.iems.co.kr/CQE/cqedb/cqeproblem/CQEproblem1.pdf (There are five total, just change the “1” in the URL.) I’ve noticed with ASQ exams, a lot of the study guides don’t correlate to the actual test at all. Curious if anyone is familiar with these exams and if they are valuable to the actual test?


1. In a complex electronic system containing 1,000 devices in series, an increase in component reliability from 99.90% to 99.99% would improve the system reliability by a factor of approximately… the correct answer is 2.5, but we don’t know how that was solved.

2. A data set for a measured characteristic of a part has a standard deviation of 11.2. If the standard deviation of the measurement error is 2.1, what is the estimated true standard deviation for the characteristic? The correct answer is 11.0, but again, not sure how to solve it.

3. A process with a binomial distribution has a 5% probability of success. If 10 samples are drawn from the process, what is the approximate probability of having zero success? The correct answer is 0.6.

4. An internet service provider logged four primary types of complaints over a one-month period. A team used the data to create the following incomplete table. What test statistic should be used in a goodness-of-fit analysis?

Table:

Type | F0 Fe X2

Operating System | 98 45 62.4

Proxy | 90 40 62.5

Firewall | 99 40

Land Line | 93 45 51.2

Total | 380 170


Or, to see the chart, just go to question 74, PDF page 15, numbered page 13, here: http://asq.org/cert/resource/pdf/sample-exam/cqe-sample-exam.pdf


Thank you so much for any advice or help to prepare him to ace the CQE!!!

 

Hello,

I am getting ready for the test, it might be a little late to help you with your questions but here are my inputs:

1. In a complex electronic system containing 1,000 devices in series, an increase in component reliability from 99.90% to 99.99% would improve the system reliability by a factor of approximately.

The key thing here is that these devices are in series, so the reliability for the whole system is basically: R^n, where "R" is the reliability of the component and "n" the number of components.

So the original system is: 0.999^1000 = 0.36769.

The improved system is: 0.9999^1000 = 0.90483.

Whenever you are looking for a factor you are going for a correlation, therefore: (R improved) / (R original) -> 0.90483/ 0.36769 = 2.46085 which is approximated to 2.5.

--------------------------------------------------------------------------------------------------------------------------------------------------

2. A data set for a measured characteristic of a part has a standard deviation of 11.2. If the standard deviation of the measurement error is 2.1, what is the estimated true standard deviation for the characteristic?

How I solved this question was the inverse of how you add variability of two components which is sqrt((Std. Dev. Comp. A)^2 + (Std. Dev. Comp.B)^2). But here you want to remove the measurement error to find the estimated true deviation for the characteristic (without the error), so:

sqrt((Std. Dev. Comp. A)^2 - (Std. Dev. Comp.B)^2) -> sqrt((11.2)^2 - (2.1)^2) -> sqrt (121 - 4.41) -> sqrt (116.59) -> 11

--------------------------------------------------------------------------------------------------------------------------------------------------

3. A process with a binomial distribution has a 5% probability of success. If 10 samples are drawn from the process, what is the approximate probability of having zero success?

Here you use the Binomial Probability Distribution.

Success -> p = 0.05
Samples -> n = 10
Number of successes of interest -> r = 0

Cnr will be 1 as 10!/((0!(10-0)!)),

P(0) = p^r x (1-p)^(n-r)
P(0) =0.05^0 x (0.95)^10
P(0) = 0.59876 -> approximately 0.6.

The other option (how I plan to resolve on the exam is with the Poisson distribution table:

np = 10 * 0.05 = 0.5
r = 0

Looking for the table on CQE Handbook:
X -> 0.5
n -> 0

You have -> 0.607 which is an estimated answer, but it is way faster to obtain it.

--------------------------------------------------------------------------------------------------------------------------------------------------

4. An internet service provider logged four primary types of complaints over a one-month period. A team used the data to create the following incomplete table. What test statistic should be used in a goodness-of-fit analysis?

Table:

Type | F0 Fe X2

I believe this question was poorly worded, it took me longer to realize that Fo -> Frequency Observerd and Fe -> Frequency expected.

Then X2 is Chi-Squared.

When you apply the ((Observed - Expected)^2) / (Expected) you obtain the only missing value:

((99-40)^2)/ 40 -> 87.025.

Then you compute the test statistic with the sum of Chi-Squares -> 62.4 + 62.5 + (missing value) + 51.2 -> 62.4 + 62.5 + 87.025 + 51.2 -> 263.125 (option B).

--------------------------------------------------------------------------------------------------------------------------------------------------

Sorry if it was late, I hope you husband passed the CQE exam and maybe this post might be useful for someone in the future.

Please let me know if I applied any concept wrongly and I hope to pass my CQE exam next month!

Edit 1: Sorry to apply the formulas like this but it was the best way to write them here on the forum, hope you guys can understand it.

 

I am taking my CQE in December here in the US. I have studied for the last year off and on till 4 months ago when I really started digging in. One thin g I have found helpful is the practice exam CD that came with my primer. I have taken practice exams many time to get a feel for what I might run into during testing. Even though I might get duplicate questions taking so many exams, I have to treat it as though I have never seen them before, especially things such as Sum of Squares and exponential distributions. P, nP , C and U charts are just now coming around without much decision making.

Open book helps with all the formulas cause your not a walking index.

One thing I did find is that ISO 9001:2015 is not in the 2015 Primer, it reflects ISO 9001:2008. Has anyone seen anything on a test that might reflect that issue?

 

While you are studying try to focus on the calculation questions: Sum of squares, linear regression, reliability calculations and distributions.

Be sure to have notes, be familiar with them and know where to look for them, these calculations can be a little bit time consuming if you are not familiar with them.

About the material date that should not be an issue.

I recently passed both CQE and CSSBB , both on my first attempt, just focus on the primer cd, understand why you are picking the answer and not the other ones and you will have no problem.

For CQE I used the third edition of the handbook and a primer from 1997.

CSSBB I used the second edition of the handbook and a primer from 2007.

Best of luck!