Strategy Dynamics Forum
Latest Topics
 
 
 


Reply
  Author   Comment  
AnnMarieWhite

Junior Member
Registered:
Posts: 2
Reply with quote  #1 
Trigger alert: This topic contains a reference to a traumatic event

In a case of business system trying to improve covered lives as its focused objective (i.e., reducing # lives lost to suicide - a stock), what are relevant models to consider or adapt?  
0
KimWarren

Subject Expert
Registered:
Posts: 104
Reply with quote  #2 
hi Ann-Marie. First, a small but important correction - lives lost to suicide per year is a rate, not a Stock [the clue is in the 'per period' unit of measure. Next to note is that this very important issue goes well beyond [but not excluding] business considerations. Presumably, employers are concerned about this risk, hence its relevance to business cases - they are certainly concerned about much less serious 'triggered' events, such as resignation.

Given the seriousness of the issue, you should definitely contact the wider system dynamics community for any academic or practitioner expert who may have direct experience of the topic. Any such individuals who read this post are most welcome (please!) to add to or correct what follows by replying in this Forum.

The issue appears to be a specific example of a class of dynamic phenomena involving [a] state-of-mind factors and [b] thresholds. We look at these in class 9 of our modeling course.

In principle how this works is that a certain 'state-of-mind' for an individual is added to by events [nice events add to positive states of mind, e.g. nice meal-experiences build our favourable feeling about a restaurant ... unpleasant events add to 'negative' states of mind, e.g. poor service experiences add to our feeling of annoyance]. The more frequent and large are those experiences, the more 'full' the related state-of-mind becomes. I think it is true that such positive and negative states of mind are distinguishable, rather being a composite state-of-mind running from negative to positive. We mostly cannot hold on to any strong feeling [+ or -] for ever, so if events do not recur, the related state-of-mind 'fades away'. See the example from our 'frameworks' course concerning the impact of flight-delays on an individual's propensity to switch airlines.

Extending this to a group needs to recognise that a larger population [a] will have a distribution of states-of-mind - among 100 people exposed to the same flight delays, some will report higher or lower levels of annoyance and [b] will have different thresholds of annoyance at which they are willing to incur the hassle and risk of switching. The simplest way to model this is to have a Stock of 'average annoyance' for the group, then use a look-up relationship to get a value for the frequency with which people switch [at avg annoyance of 0.8, 0.85, 0.9 ... 1%, 2% 4% etc of customers will switch per month [see sdl.re/sygfull clip 32]. So - a persistent high level of work pressure, or frequent incidents of bullying by superiors across a whole team may fill a Stock of average stress for that team, and raise the frequency of resignations [or in your case, suicides]. 

It may be possible and valid to adapt the same principles to the suicide issue. You may find evidence that negative, work-related incidents raise average levels of stress among employees, and you may find that this higher average stress level is related to an observed rate of suicides [this is indeed a 'rate', not a Stock]. You may then be able to model the impact of changing employment policies on mitigating that stress level, and thereby the suicide rate. All of this, of course, being complicated by a host of non-work-related factors [debt, divorce ...] for which employers may have no knowledge or means to respond.   

Other knowledgeable readers please reply to this post ...
0
AnnMarieWhite

Junior Member
Registered:
Posts: 2
Reply with quote  #3 
Thank you for tackling this difficult but important topic.

In terms of thinking about stocks - couldn't one model a total number of suicide completions (e.g., in people's memories in the system) that have accumulated in a system at any given point in time (with the number per year being the inflow)? Objectives become to reduce this accumulation (i.e., reduce the total # completions in people's memories; outflow could be number "forgotten" per year, etc.).

The threshold reference is very helpful thank you! 
0
KimWarren

Subject Expert
Registered:
Posts: 104
Reply with quote  #4 
Yes you might want to track the accumulated Stock of suicides in a year .. and you can always report "the number in the latest 12 months" with a Variable and Formula "the Stock" - delay ("the Stock",12).

But the number you want to explain is the current rate [suicides/quarter or month], and this is the item that any revised policy will directly change. If the Stock for the latest year is lower than for the previous year, this will be because the flow-rate has been lower due to your actions.

Note too that you may want shorter time-units [maybe quarters] because management intervention can be changed faster than 1 time per year.  
0
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.