Using PROC CORR, looks at the correlation between the independent variables. 1)

Question

Using PROC CORR, looks at the correlation between the independent variables.
1) which variables are likely to be redundant ( i.e. They are likely to be similarly associated with the dependent variables)
2) how does this influence the starting model for regression?
3) identify possible sets of sets of variables for use of regression. Page 9 of s output Question Association of communicable and non-communicable disease with economic and health resources. Partia -Correlation The CORR Procedure 10 variables: gov pcs all pcs income pc phys pt drink95 sanit95 drink00 sanit00 drink05 sanit05 Simple Statistics Mean Std Dev i sum Minimum Maximum 1036 166240 14.60000 gov pcs 237 701 43291 all pcs 237 484.49536 778.41901 114825 1.70000 income po 225 9594 12037 2158710 200,00000 69040 126 106478 10313 130 te204 drink95 227 go 67B41 12.16289 20584 14.00000 100 00000 sanit95 227 70.00881 28.68119 15892 9.000000 100.00000 drink 231 91.24675 10.70005 121078 36.00000 100 00000 ,oooooo 10000000 B 51514 sanitio5 230 70 53013 30.84453 16.224 11.00000 100.00000

Explanation / Answer

1) Following are highly and significantly correlated variables:

gov_pcs -> all_pcs, income_pc
all_pcs-> income_pc
drink95->drink00
drink95->drink05
sanit95->sanit00
drink00->drink05
sanit05->sanit00

so, when we are cosnidering left side variable, right side variables are redundent.

For instance, if we consider:

gov_pcs
drink95
sanit95

following variables should be dropped:

all_pcs,
income_pc,
drink00
drink05
sanit00

2) The presence of these redundant variables in the regression model causes multicollinearity issue. This causes the change in sign of another variable and leads to misinterpretations of coefficients and their significance.

3) Potential sets of variables should be used:

gov_pcs
drink95
sanit95

phys_pt

Related Questions

Navigate

• Browse (All)
• Browse U
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.