Page 156 - Contributed Paper Session (CPS) - Volume 3

CPS1979 Francisco N. de los R.
In the analysis, three values of M were used: twenty thousand, thirty thousand
and forty thousand all with 50% burn-in. The posterior samples are then used
to construct samples () = ( 1 () , … , () ), using the inverse logit transform

() () () () ()
= ( + ) / [1 + ( + )].
0 0

The jth sample from the posterior distribution of D is constructed as

() ()
−
|

() = ∑ 2 () (1− () | ) , = 1, … , where () = (∑ () )/(∑ ).
=1
=1

=1

Finally, D can be estimated by the median of { (1) , … , () }, while a 95%
credible interval is obtained from the 2.5 and 97.5 quantiles of
th
th
{ (1) , … , () }.

Voter Participation in the Philippine National and Local Elections
(NLE) of 2016, official data from the Commission on Elections was used in the
research. Since voter turnout is viewed here in a spatial data analysis paradigm,
and therefore contiguity-sensitive, turnout from overseas voting was not
included. Provincial level information on number of registered voters and
actual voter turnout for 86 areal units comprise the entirety of the dataset. This
includes special readings for the cities of Isabela and Cotabato, which are
labeled special provinces, and the four districts of Manila. Proximity due to
common-border cannot be used since there are 15 provinces which islands.
These are Batanes, Biliran, Bohol, Camiguin, Catanduanes, Cebu, Dinagat,
Guimaras, Marinduque, Masbate, Palawan, Romblon, Siquijor, Sulu and Tawi-
Tawi. Here, connectivity was based on a nominated critical distance. Special
consideration arose for the island of Palawan since a large critical distance was
needed for it to have just one neighbor. Thus for this specific province,
indication of geographic integration like presence of boat routes and trade
with a nearby province was used. This led to Iloilo being set as Palawan’s
neighbor. The proximity matrix was then revised to force a neighbor for
Palawan. Areal centroids were identified via the Universal Transverse Mercator
(UTM) coordinate system. Inter-unit distance was computed via these
coordinates. Created a proximity matrix W based on distance of at most a
1
nominated .
1, ( , ) ≤

= { } has = { 1 ; = 0

0,
( , ) ≤
1

Several values were tried for but only in = 500 UTM units did all areal
units, except Palawan, had at least one neighbor. Iloilo was forced to become
Palawan’s neighbor by virtue of transportation and trade relations. A distance
145 | I S I W S C 2 0 1 9