Final Report on RREF 2001 - Department of Health
Final Report on RREF 2001 - Department of Health
Final Report on RREF 2001 - Department of Health
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
Part B: Development <strong>of</strong> Opti<strong>on</strong>s for a Revised <strong>RREF</strong><br />
Both groups c<strong>on</strong>sidered that the range <strong>of</strong> variables that had been c<strong>on</strong>sidered had covered the<br />
main factors affecting need for HACC services and that inclusi<strong>on</strong> <strong>of</strong> six variables in the base<br />
populati<strong>on</strong>s and six need factors for weighting would give appropriate recogniti<strong>on</strong> to these<br />
factors in the <strong>RREF</strong>.<br />
With regard to grounds for weighting any <strong>on</strong>e <strong>of</strong> these variables more than another, the expert<br />
opini<strong>on</strong> was that opti<strong>on</strong>s should be modelled to illustrate the effects <strong>of</strong> weighting, with the<br />
unadjusted variables providing the basis for comparis<strong>on</strong>.<br />
4.4 EFFECTS OF WEIGHTING OF VARIABLES<br />
There are a variety <strong>of</strong> opti<strong>on</strong>s for weighting variables and it was c<strong>on</strong>sidered useful to model<br />
some opti<strong>on</strong>s for the <strong>RREF</strong> without adjusting the weighting <strong>of</strong> need variables and to provide<br />
examples with adjusted weightings to dem<strong>on</strong>strate the impact <strong>of</strong> such adjustments.<br />
With unadjusted weighting, the c<strong>on</strong>tributi<strong>on</strong> that each variable makes to the overall outcome<br />
reflects the distributi<strong>on</strong> <strong>of</strong> scores across LGAs. For all variables, the highest need LGA gets a<br />
score <strong>of</strong> 9, and the lowest get a score <strong>of</strong> 0, but variable A <strong>on</strong> which a large number <strong>of</strong> LGAs get<br />
a low score and a few get a high score will have less impact overall than variable B <strong>on</strong> which a<br />
large number <strong>of</strong> LGAs get a high score and a few get a low score.<br />
If it is c<strong>on</strong>sidered that variable A has a markedly str<strong>on</strong>ger impact <strong>on</strong> the need for HACC<br />
services than variable B, there may be grounds for adjusting the weighting for variable A to take<br />
account <strong>of</strong> this relati<strong>on</strong>ship. Reas<strong>on</strong>s that might be c<strong>on</strong>sidered for varying the weighting <strong>of</strong><br />
different variables were canvassed in the C<strong>on</strong>sultati<strong>on</strong>s and are noted below.<br />
Modelling <strong>of</strong> opti<strong>on</strong>s with and without adjustments to the weighting <strong>of</strong> the need variables is<br />
useful for illustrating the effects <strong>of</strong> varying the weighting between variables. It should however<br />
be noted that the effects <strong>of</strong> changing the weighting <strong>of</strong> any <strong>on</strong>e variable are c<strong>on</strong>strained as the<br />
weightings for all variables have to total 100%. Increasing the weighting <strong>of</strong> <strong>on</strong>e variable thus<br />
reduces the weighting <strong>of</strong> others; for example, in a five variable model in which each variable<br />
happened to c<strong>on</strong>tribute 20%, doubling <strong>of</strong> <strong>on</strong>e variable would increase that variable to 33% and<br />
reduce the other four to 16.6% each.<br />
4.5 OPTIONS FOR CONSIDERATION FOR WEIGHTING OF NEED<br />
VARIABLES<br />
The Opti<strong>on</strong>s Paper illustrated two sets <strong>of</strong> opti<strong>on</strong>s for adjusting the weightings <strong>of</strong> the need<br />
variables, by way <strong>of</strong> (1) reducing the number <strong>of</strong> variables to be included in the <strong>RREF</strong> from 6 to<br />
4, and (2) by doubling the weighting <strong>of</strong> some variables. The C<strong>on</strong>sultati<strong>on</strong>s sought views <strong>on</strong><br />
these opti<strong>on</strong>s.<br />
What are the<br />
opti<strong>on</strong>s for<br />
weighting the<br />
variables?<br />
• Should all variables evenly weighted?<br />
• Which if any <strong>of</strong> the variables should be weighted more than<br />
others, and for what reas<strong>on</strong>s?<br />
<str<strong>on</strong>g>Final</str<strong>on</strong>g> <str<strong>on</strong>g>Report</str<strong>on</strong>g> July <strong>2001</strong> 45