Decision algorithms

[Tolman's Rat]

Anyone create decision algorithms that you'd like to discuss? I'm thinking that I will create a list of variables, each with a set weight, and then score each program to which I am accepted on those variables. This would create a simple composite score that, in theory, would make selecting a program much easier. 

 

Has anyone done this, and if so, how did you weight each factor (e.g., research fit, research impact, funding, location, personality match)? I'm sure there are others I am forgetting.

[laminator]

Anyone create decision algorithms that you'd like to discuss? I'm thinking that I will create a list of variables, each with a set weight, and then score each program to which I am accepted on those variables. This would create a simple composite score that, in theory, would make selecting a program much easier. 

 

Has anyone done this, and if so, how did you weight each factor (e.g., research fit, research impact, funding, location, personality match)? I'm sure there are others I am forgetting.

 

Something like this...

 

Maximize F = (research*personality match+program prestige)*(isFullyFunded)

subject to: Location > mediocre

 

Or just lexicographic order on funded > research*personality match > prestige > location.

[TakeruK]

I made a spreadsheet to directly compare a few factors. Some of it was financial -- I listed the stipend, subtracted off necessary costs that change on location (rent, health insurance, copays) so that I would be able to compare the stipends directly (I figured small differences in e.g. cost of food is not going to be a game changer). I considered how much I liked the courseload/degree format (how many exams? orals?) and time to degree. I also considered research but at this point (down to 3 choices), the research everywhere was good enough that the real deciding factor was how happy I would be at each place. 

 

I don't like numerical scores because you can just fudge the weighting factors to get the result you want. On the other hand, if you find yourself fudging the weighting factors to get a certain result, maybe you have already made a choice!

[Kleene]

if (accepted)

    if (funded)

        attend;

else

    cry;

Edited January 30, 2015 by Kleene

[HIMJoel]

Double[] college;

 

college[]= new double[uCSC,Pitt,CMU,Drexel,Northwestern,Kent State]; 

i=0

While (true) {

     if (accepted){

       if (funded) {

         if (math.random>.05) {

            print ("are you happy  with outcome heads"+ college

         else print are you happy with outcome tails?+ college

        if no, 

         attend.

 

 

 

i++

[poweredbycoldfusion]

This is a really fun idea    I'm not sure how I'd quantify certain factors. There's a monetary aspect to it that's easier to judge, but program prestige vs lab prestige vs likability of individuals in the program vs oppertunities for students to grow as scientists is a bit harder to quantify.