Conference rusure::math

1014.0. "Population Analysis Help Requested" by IAMOK::ROMANO (Consistently wrong = right??) Wed Jan 18 1989 13:59

    I'm not sure if this is the right conference but seeing that my
    problem is math-related I'll try.  
    
    I work is personnel and perform demographic studies on populations
    within DEC.  I follow such things as hiring, terminations, retirements,
    growth, and internal movement.  I have a few questions/problems
    on how things are currently calculated and wanted to get some opinions
    to either back my ideas or prove me wrong.  There are two main areas
    that I need feedback on:
    
    1)
    
    For hiring, termination, or retirements: the rates are currently
    calculated as:
    
    New Hires or New Terms or New Retire
    ------------------------------------ = Hiring, Term, or Retire Rate
    Average Population (Prev+Present/2)
    
    
    This doesn't make sense to me for some reason.  I think that the
    denominator should be the base population only:
    
    New Hires or New Terms or New Retire
    ------------------------------------ = Hiring, Term, or Retire Rate
    Previous (base) population
    
    
    2)
    
    Measuring internal movement within a population (termed Churnover)
    is a highly debated topic.  Not many people agree on a method. 
    I can't even describe the present method.  Right now Churnover is
    calculated by:
    
    (position change + group change - overlap figure - new hires)
    -------------------------------------------------------------
                            Average Population    
    
    I, myself, have no great idea on how to measure 'internal movement'.
    I hope that some people may have some ideas.  My closest guess
    (simplistic) would be:
    
    Int. transfers into Pop. + Hires + Int. trans. out + Term + Ret
    ---------------------------------------------------------------
          Previous (Base) Population
    
    
    If anyone has any ideas, or other notes conferences to turn to please
    let me know.
    
    
    Thank you for your help,
    
    Don
    

>    For hiring, termination, or retirements: the rates are currently
>    calculated as:
>    
>    New Hires or New Terms or New Retire
>    ------------------------------------ = Hiring, Term, or Retire Rate
>    Average Population (Prev+Present/2)
>    
>    This doesn't make sense to me for some reason.  I think that the
>    denominator should be the base population only:
>    
>    New Hires or New Terms or New Retire
>    ------------------------------------ = Hiring, Term, or Retire Rate
>    Previous (base) population

For a very large company it probably doesn't matter which formula you use, 
but for a startup the latter version blows up (division by zero), so it 
makes sense to stick with the first version.
    
>    ... Right now Churnover is
>    calculated by:
>    
>    (position change + group change - overlap figure - new hires)
>    -------------------------------------------------------------
>                            Average Population    
>    
>    I, myself, have no great idea on how to measure 'internal movement'.
>    I hope that some people may have some ideas.  My closest guess
>    (simplistic) would be:
>    
>    Int. transfers into Pop. + Hires + Int. trans. out + Term + Ret
>    ---------------------------------------------------------------
>          Previous (Base) Population

By definition the internal transfers in and out should be identical. I 
assume what is important is the processing cost of each change in job, and 
an internal transfer therefore costs roughly twice as much as a new hire.
However, I must admit that I don't know what you are trying to measure and
for what purpose, and I don't know what all the terms mean, so I don't have
much to add. 

Lynn Yarbrough