SAS Support Communities

Thank you Steve. I did similar to my merit increase model based on each employee performance ( high solid low) and their current salary based on market. We had  specific % allocated per each 3*3 cell ( i.e high performance and high salary will get 2% increase, low performance and high salary will get 0% merit etc.) and did goal seek function to stay at our budget. However on my current case that % is not allocated and need to do some kind of statistical modeling to have optimal %. I am not a statistician at any means but have solid understanding on the insides of model if somebody could  guide me what model/test to apply based on my business case. I will investigate linear programming. Do you have any  article or link that is relevant ? Please do share. Much appreciated   ... View more

Actually I am struggling in the approach. Any idea to start? I will do some research  if I get some direction ... View more

Thank you Miller I thought of 5% increase in all manager's current average sales. The only thing I was not sure is.... is it ok to have flat 5% to all of them because every sales manager has different in terms of # of product sales per day, the cost of product, etc. what if one manager is already exceeded the goal ,,, is it fair for him to again improve the sales by 5 %?  is there any way to adjust  that? I know its not straight forward.. I am here to seek some guidance as I am very lost here ... View more

The average is calculated just by sum of total sales / # of rows I need to have weighted target for each manager. Some manager have higher influence on average overall and their sales is skewed to right and some are in lower quartiles. I am not sure how we can do  but generally speaking if we want to target for each managers to meet the  5% increase goal how do we model it? ... View more

Hi Miller I apologize  for confusing title. To summarize here is my two cases: 1. I want to understand how much each group has to reduce in an average to meet the goal of 5% reduction in average LOS. Firstly I need to run a model how each group contribute  to have an average overall LOS. And then .. now how much each group had to reduce compared to its weighted factor( some group have higher influence in los because  some group consist of 50% of total patients and some have just 3 patients under that groups.. which means not all groups should be treated equal) 2. Now we need to find out which area should we focus among these top 3 contributer group? Theymay think we have higher LOS because we treat  high risk patient or we treat medicare patients. So we have to run a model where if insurance is same, age of patient is same, severity is same,  how each groups have a influence in LOS. In general we want to know where should we focus to be honest and we are making up all these hypothesis among groups to first target.   Does it make sense?   Thank you in advance ... View more