Page 11 - HIMES_2007_

Performance Indicator Importance in MPA Management 611

the community obtains” from the presence of the MPA. All this may be due to the relative
emphasis that researchers, namely biologists like those interviewed, seemed to place on
Downloaded By: [University of Leeds] At: 13:49 18 January 2008
environmental protection and conservation of marine organisms.
It was expected that fishers would rank “increasing their catch” as the highest priority;
however, this ranks considerably lower than “increasing income made from local tourism”
and relatively even with “increasing community benefits from the MPA” and “increasing
biological resource biomass” in the area. Although it did not score highest, the priority that
fishers give to increasing the number of fish caught is significantly higher than the other
groups (ANOVA, F-value = 2.658, p-value = .047). Interestingly, none of the priorities
that fishers placed on seven of the eight indicators was especially high or distinct from
the other priorities given, with the exception of “increasing tourism income.” In fact, the
priority fishers have given to increasing income derived from tourism is significantly higher
than all other stakeholder groups (ANOVA, F-value = 8.187, p-value <.001).
Both management indicators and increasing community involvement in management
were given the lowest priority of fishers. This is surprising because it was hypothesized
that fishers, being the only stakeholder group whose livelihoods are directly affected by the
EIMR, would take a larger interest in management procedures and interventions.

Consistency and Transitivity
Due to the complex nature and variability of the needs and interests of local stakeholder
groups, a negative effect on the degree of consistency achieved by respondents was expected.
Standard AHP practice is to accept the responses of individuals where their inconsistency
is less than or equal to 10%. It is unclear in many studies if this is adhered to; however,
in fisheries studies up to 20% has been accepted (Mardle & Pascoe, 1999). However,
the consistency ratio traditionally measured in AHP studies resulted in a high level of
inconsistency among responses in the present study. There are two explanations for this.
First, in this study, lower level objective comparisons have been made. In the general
case, only elements belonging to the same parent criterion are compared. As a result, in the
present study comparisons at the lowest level were made between indicators under different
branches of the hierarchy (Figure 2). Higher than average inconsistencies could then be
expected as respondents may have difficultly trying to make direct comparisons between
concepts that are not easily comparable (e.g., decreasing pollution vs. more income from
tourism).
Second, upon analysis of individual preference matrices, it was determined that the
majority of responses with more than 20% inconsistency were problems of scale as opposed
to inconsistency. Inconsistency is directly related to the respondent’s comprehension of the
AHP scale. Respondents wanted to compare the chosen indicator with everything in general
(e.g., is increasing available information important in general) instead of just with the pair at
hand. Furthermore, it was noted that respondents occasionally treated the scale differently.
Many wanted to make strong statements of their opinions and proceeded to rank the
indicators they felt the strongest about with a 9 and those that they did not feel strongly
about with a 1, regardless of the indicator with which it was being compared.
Each respondent’s answers were examined in detail and the number and type of
circular triads were counted for each respondent. To maintain a logical order of preference
for objects A, B, and C follows if A is preferred to B and B is preferred to C and consequently
A is preferred to C (Arrow, 1951). In a complex situation, it is recognized that this logic
may be broken. Kendall (1962) describes this concept as circular triads, and it is a trait in
human decision-making to which the AHP takes account (Saaty, 1977). Using the concept