Page 27 - World Heritage and Tourism Innovation

Hiyab Gebretsadik Weldearegay Critics on Heritage Laws and Tourism Politics in Ethiopia

and monitoring attrition rate in each round; and, de- ence was on how diffused the opinions were among
claring whether or not a particular round is closed, panellists or how fast the panellists reached the pla-
based on stop criteria, and, saying that consensus pla- teau. The study timeframe was determined to be five
teau is reached. and a half months, which is consistent with the rec-
ommendation of Fish and Busby (2005), which states
Reliability and Validity that four months suffice for three rounds of Delphi.
Fish & Busby (2005) indicate that we can estimate The continued tasks of this forming Phase were dis-
the reliability between the first and second rounds by patching a general introduction of the problem state-
exploring the consensus rates of the respondents. In ment, an invitation to participate in the Delphi study,
other words, if a reasonable level of consensus is on developing consent acquisition forms, and completing
many items on the second questionnaire, it is likely ethical clearance from the responsible body. Forming
reliable. The issue of validity is directly related to the a Delphi Panel with an anonymous respondent code
selection of the panel of experts. Against the field’s throughout the inquiry was part of this task. The last
appropriation, reviewed expert selection criteria is an part was general (brainstorming about ways forward.
insurance of the validity. In addition, the experts have
expertise in the study area. Accordingly, reliability First Delphi Phase; Identifying Salient Issues Phase
and validity concerns have been addressed by strictly This Phase included brainstorming about the first task
following Fish and Busby’s (2005) recommendations: of identifying salient issues, identifying salient issues
in each research objective, composing content, and
The Five-Phased Delphi Flow Chart producing a Delphi 1 summary. First-round questions
Although the classic Delphi technique recommends with epistemic assumptions were circulated to 91 ex-
at least four rounds, this study used a five-round Del- perts enquiring about identifying salient issues under
phi, except for the forming Phase, as seen in the dia- each research objective. We subjected the responses
gram below. to defuzzification and produced an intermediate sum-
mary of this round. We monitored the attrition rate,
Pre-Delphi Phase: Forming resulting in 82 panellists at the end of the first round
The forming Phase included preparation, introduc- of Delphi with a 90% response rate. We did not close
tion, and general (brain) storming. The practice in- the process of identifying salient issues under each
cluded: objective at this Phase and did not reach the conclu-
sion plateau. We structured the next rounds based on
• Designing the overall Delphi methodology, in- the responses to the previous stages.
cluding selecting relevant fields of panellists.
• Designing panel selection criteria. Second Delphi Phase; Consensus Measuring Phase
• Determining panel composition (size/expert on Lists
breadth). In the second Delphi Phase, we sent a 5-Point Lik-
ert-based questionnaire back for validation to rate the
• Choosing the Delphi period.
identified salient issues under each specific objective;
The following action was to decide the number of this aimed to investigate the level of consensus on the
rounds and iterations. Delphi rounds were predeter- consolidated summary of lists of salient problems
mined to be five according to the typology of informa- presented by the panel members in the first round of
tion required from the experts and to be careful not Delphi. We indirectly and methodologically cultivat-
to compromise panellists’ response rates and enthu- ed divergence. We developed striking epistemic theses
siasm. However, we did not have predetermined iter- and called for antitheses through iterations until we
ations; instead, we were supposed to undertake them reached the consensus plateau. The iteration process
until we reached a consensus plateau. So, the depend- allowed participants to comment on the responses of

Proceedings of the 7th UNESCO UNITWIN Conference | 23