memory  of  extension  officer  which  is  native  who  understand  conditions  in  the site.  This  study  also  got  information  about  existence  of  the  rice  miller  from
society, owner of the rice miller visited, and so on. Then, this list was used as a guide to found rice miller in the field. Based on
field survey, there were 44 rice millers in Gekbrong. This number was insufficient to determine the efficiency of rice milling industry. In order to get more accurate
results,  this  study  was  expanded  to  neighboring  sub-district.  Purposively, Warungkondang selected as the study site by the main reason that Gekbrong was
part of this sub-district before regional expansion occurred. Thus, it was suspected that both sub-districts have similar characteristics of rice miller.
Similar to Gekbrong, this study also did not have a clear sample framework. So, sampling was done same way as in Gekbrong. Due to some limitations such as
materials,  time,  and  field  conditions,  this  study  only  used  50  rice  millers  in Warungkondang.
4.2 Data Processing and Analyzing
For  this  study,  primary  unit  of  observation  is  the  rice  milling  business. Expected outputs of each business are rice, broken rice, rice bran, and chaff. All
outputs used in form of weight by kilogram. While input variables are used in this study include grain, working hour, and fuel.
Primary  and  secondary  data  are  processed  and  analyzed  by  using quantitative  and  qualitative  methods.  Qualitative  analysis  is  used  to  describe
characteristic  of  rice  milling  industry  in  study  site.  Meanwhile  quantitative analysis  is  used  to  determine  relative  efficiency  of  rice  milling  industry  in  this
site. Model used in this study is DEA model. This model is used to analyze the
relative  efficiency  of  rice  milling  industry.  Efficient  score  value  of  efficiency equal to one shows that the rice milling business has been relatively efficient and
inefficient score value of efficiency less than one indicates rice milling business on the other conditions.
Referring to Koopmans definition about technical efficiency, the study used two requirements that both requirements must achieved by rice milling industry to
determine efficiency of rice milling, namely: 1.
Rice milling industry must have efficiency score equal to one. 2.
Rice  milling  industry  must  achieve  zero  slack  condition  for  all  variables used.
DEA  was  created  as  a  tool  to  evaluate  the  performance  of  an  activity  in  a unit entity organization Charnes, Cooper,  Rhodes, 1978. DEA calculates the
efficiency  of  a  DMU  in  one  group  of  observation.  The  working  principle  of  the DEA  model  is  to  compare  the  data  input  and  output  of  an  organizations  data
decision-making unit DMU with other input and output data on the same DMU. This comparison is performed to obtain an efficiency score.
Assumptions of DEA: 1.
Entities evaluated by using a set of the same input to produce likewise set of the same output.
2. The data is positive and the weight is limited on positive values.
3. Input and output are variable.
4. In  the  presentation  of  Hayes,  2005,  there  are  strengths  and  weaknesses  of
DEA.  DEA  model  used  as  a  device  to  measure  the  performance  has  several advantages over other models, including:
5. DEA model can measure many input and output variables.
6. It  does  not  require  an  assumption  of  a  functional  relationship  between  the
variables measured input and output. 7.
DMU is directly compared with one another. 8.
Input and output variable can have different measurement units. Meanwhile, limitation of DEA model, including:
1. It has sample specific.
2. It is an extreme point technique so that measurement error can be fatal.
3. It just measures the relative efficiency of DMU, not absolute efficiency.
4. Hypothesis  testing  is  statistically  difficult  to  be  done.  Using  linear
programming formulation separately for each DMU calculations is manually difficult, especially for large-scale problems.