Source: Summary of previous studies.
Based on the typology of isomorphism, the
research hypotheses are as follows:
Coercive isomorphism is done with compulsion
to achieve compliance with the authorities.
Normative isomorphism is done with awareness
to achieve acceptability in society.
Mimetic isomorphism is done voluntarily to
achieve individual success.
3 METHODS
Based on the secondary data from BI and OJK, the
population of all BPDs in Indonesia until March 24,
2017 is 26 BPDs. Given the small number of
population, then this research uses all 26 BPDs in
Indonesia. The domain of this research is in
management strategy that emphasizes on observation
of top management behavior in managing
organization. Thus, the observation unit in this study
is the top management of all BPDs namely the
President Director.
Recognizing that this study has a small population
size and the potential consequences that the predictive
validity of the study is lower, the solutions undertaken
to strengthen the research findings are: (1) The
sampling method is censused to 26 BPDs and (2)
Interviews and FGDs are conducted intensively to
produce a thick description in the discussion of
research findings. So in this research, the mixed
research method becomes very relevant to be
implemented.
This research uses explanatory design a two-
phase mixed method, where the first phase involves
collecting and analyzing quantitative data and the
second phase includes qualitative data collection and
analysis done to strengthen the result of quantitative
research (Creswell, 2013: 203).
In QUAN process using PLS-SEM. PLS-SEM
can still be implemented for small data, if the
population surveyed is small and the consequence of
predictive validity is lower (Hair et al., 2014: 48).
Further, Hair et al. (2014: 16) provides a detailed
description of the key characteristics of PLS-SEM-
related data. It is said that the data size for PLS-SEM:
(1) Unidentified problem with small sample size, (2)
In general, although the sample is small, PLS-SEM is
still able to produce high-quality statistical analysis;
and (3) The bigger the data, the expected estimation
of the research the more precise (consistent).
The process of sending and collecting
questionnaires was conducted throughout July-
August 2016. The mechanism of distributing and
collecting questionnaires was through Asbanda. Of
the 26 questionnaires sent to all BPDs, all were
returned and answered in full by the respondents.
Details of respondents who answered the
questionnaire were: 12 BPD President Directors, 4
BPD Directors 8 BPD Division Leaders, 1 BPD Sub-
Division Leader and 1 BPD Section Head.
All indicators can also be stated reliable with
Composite Reliability (CR) value of 0.937 and
Average Variance Extracted (AVE) value of 0.663.
Both of these reliability indices have values greater
than the minimum limit of 0.70 for CR and 0.500 for
AVE. The CR value of 0.845 states that 84.5% of the
mimetic dimension variation can be explained by its
indicators and the AVE value of 0.646 states that
64.6% of all indicators’ variation can be well
explained by the mimetic dimension. This indicates
all reliable indicators for measuring the mimetic
dimension. The result is as presented in table 4 below:
Table 4: Measurement model of isomorphism variable.
Source: Quantitative result of the study.
The result of measurement model for
isomorphism variable shows that all dimensions have
factor loadings greater than 0.500, thus all dimensions
are concluded as valid in measuring the isomorphism
variable. In addition to valid, the three dimensions
can also be declared reliable with CR and AVE values
greater than 0.700 and 0.500 respectively. Based on
the calculation results, it can be seen that the
dimension of mimetic isomorphism is the dimension
with the largest factor loading, which means that this
dimension is most closely related to isomorphism
variable.
In the QUAL data collection process, the
confirmation of the problem formulation is done by
conducting FGDs with banking experts to confirm the
findings in the quantitative phase and to get input on
what, how and why this occurred as a result of the
quantitative findings.
ICIEBP 2017 - 1st International Conference on Islamic Economics, Business and Philanthropy
454