[comp.software-eng] Prediction from Unit Testing

schaefer@chance.mitre.org (Carl Schaefer) (12/05/90)

I am looking at the predictive value of unit testing for large
software development projects.  In particular, I am looking for
real-world experiences that provide empirical answers to questions
like: Was the number of problems encountered in unit testing a good
predictor of the number of problems encountered in later testing? Did
a less-than-thorough job of unit testing predict trouble in later
testing? Did the results of unit testing say anything about how
maintainable the code proved to be?  Etc., etc.  Can anyone point me
to such information?

		Carl Schaefer, Lead Scientist
                Mail Stop Z646
		MITRE-Washington Software Center,
			C3I Division
		7525 Colshire Drive
		McLean, Virginia  22102-3481

		e-mail: schaefer@.mitre.org
                tel   : 703-883-6150
                fax   : 703-883-5519


other projects in which MITRE has been asked to look at (the
implications of) unit testing?  Pointers to literature would also be
appreciated.  Thanks.

  -- Carl

vladimir@prosper.EBB.Eng.Sun.COM (Vladimir G. Ivanovic) (12/06/90)

In article <SCHAEFER.90Dec5101829@chance.mitre.org> schaefer@chance.mitre.org (Carl Schaefer) writes:

   I am looking at the predictive value of unit testing for large
   software development projects.  
   [...]

Perhaps someone can help me and Carl at the same time.  I remember a short
article in {Computer, IEEE Software} in the last few years which gave a
formula for calculating the expected failures in the field given the number of
hours of testing so far and the number of failures found.  The article was
perhaps in Computer magazine's "The Open Channel" section.  I seem the
remember that the author was not an academic.  I've checked back copies of
Computer for the last several years to no avail, but I'm missing some issues.

Help would be greatly appreciated...

-- Vladimir
--
==============================================================================
   Vladimir G. Ivanovic				Sun Microsystems, Inc
     (415) 336-2315				2550 Garcia Ave., MTV12-33
    vladimir@Sun.COM				Mountain View, CA 94043-1100

      Disclaimer: I speak only for myself.  Your mileage will vary.
==============================================================================

edschulz@cbnewsj.att.com (edward.d.schulz) (12/06/90)

In article <VLADIMIR.90Dec5201616@prosper.EBB.Eng.Sun.COM>, vladimir@prosper.EBB.Eng.Sun.COM (Vladimir G. Ivanovic) writes:
> I remember a short
> article in {Computer, IEEE Software} in the last few years which gave a
> formula for calculating the expected failures in the field given the number of
> hours of testing so far and the number of failures found.

Check out "Quantifying Software Validation: When to Stop Testing?," by John
D. Musa and A. Frank Ackerman, IEEE Software, May 1989, p. 19-27.
-- 
Ed Schulz, AT&T, Room 2P276 200 Laurel Ave., Middletown, NJ 07748
+1 201 957 3899      Ed_Schulz@att.com    or    eds@mtdcb.att.com

bwb@sei.cmu.edu (Bruce Benson) (12/07/90)

In article <1990Dec6.140439.24925@cbnewsj.att.com> edschulz@cbnewsj.att.com (edward.d.schulz) writes:
>In article <VLADIMIR.90Dec5201616@prosper.EBB.Eng.Sun.COM>, vladimir@prosper.EBB.Eng.Sun.COM (Vladimir G. Ivanovic) writes:
>> I remember a short
>> article in {Computer, IEEE Software} in the last few years which gave a
>> formula for calculating the expected failures in the field given the number of
>> hours of testing so far and the number of failures found.
>
>Check out "Quantifying Software Validation: When to Stop Testing?," by John
>D. Musa and A. Frank Ackerman, IEEE Software, May 1989, p. 19-27.

Or his book: Musa, Iannino, Okumoto, Software_Reliability_Measurement_
Prediction_Application, McGraw-Hill,1987

* Bruce Benson                   + Internet  - bwb@sei.cmu.edu +       +
* Software Engineering Institute + Compuserv - 76226,3407      +    >--|>
* Carnegie Mellon University     + Voice     - 412 268 8469    +       +
* Pittsburgh PA 15213-3890       +                             +  US Air Force

vladimir@prosper.EBB.Eng.Sun.COM (Vladimir G. Ivanovic) (12/07/90)

Sigh.  

I almost included in my original message that I was aware of the Musa article
and of the book - both of which I have.  What I remember was more like a war
story than regular article or book.

Well, thanks for trying anyway.

-- Vladimir (still hoping)
--
==============================================================================
   Vladimir G. Ivanovic				Sun Microsystems, Inc
     (415) 336-2315				2550 Garcia Ave., MTV12-33
    vladimir@Sun.COM				Mountain View, CA 94043-1100

      Disclaimer: I speak only for myself.  Your mileage will vary.
==============================================================================