• Hi All

    Please note that at the Chandoo.org Forums there is Zero Tolerance to Spam

    Post Spam and you Will Be Deleted as a User

    Hui...

  • When starting a new post, to receive a quicker and more targeted answer, Please include a sample file in the initial post.

Regression Analysis - Biological Age Calculator...

A

andrewpearce

New Member
Working with a couple of guys from my gym on building a Bio Age calculator. I am pretty comfortable in Excel but don't have a math/stats back ground so need some direction.


For an individual we capture responses to approx. 17 multiple choice questions which will give us 17 'bio markers'(BM), with either a positive or negative influence on life expectancy, the aim is to take the average life expectancy of an Australian male/female & calculate the individuals biological age based on their BM and chronological age (CA).


I've read some research papers which suggest multiple linear regression is the way to go as our BM have no correlation to CA. Currently we are researching the impact of each BM on average life expectancy for use in weighting each BM.


I want to get started on the stats part and this is where i need 'HELP!'. I've read Hui's Are you Trendy? posts and have got a broad overview of the subject. What I'd like to ask the forum is advice on which method I should use & why? I need an outline of what numbers I need to calculate & formulas required in layman's terms. Due to my lack of knowledge on this subject I can see myself using a flawed methodology & don't want to produce rubbish!


Many thanks in advance! AP
 
Andrew


Just because you can do linear regression analysis, doesn't mean you should.


Excel and the Data can be used to derive a formula for anything, but it doesn't necessarily mean that it is correct.


To achieve whatyou want you wopuld need a lot opf data of the various measures of a wide range of people of differing Age, Sex, Ethnicity, Body Shape, Height, Weight, General Fitness etc etc and even then I'd be very careful.


Investigate what others in the same industry are doing to work out similar stats.

Read technical papers on Population/Public Health or Sports Medicine etc where they have bound to hae studied these types of relationships


I'd also suggest using a mathmetican to assist as they can do lots more smart things like determining wether an individual measure has any correlation or not and hence can be eliminated from the stats.
 
@Hui: Many thanks, I've found myself a mathematician so I'll see how we go! Will endeavor to share the finished product when we get there. Regards AP
 
You must log in or register to reply here.
Back
Top