Page 201 - Dark Shades of Istria
P. 201
9.4 Istrian Dark Commemorative Events and Conflicts of the 20th Century
variables: variables with f = 0 were removed, some related options were
sensibly merged. The final selection is shown in Figure 9.4.
9.3 Istrian Dark Commemorative Events and Conflicts
of the 20th Century
Exactly 88.9 of analysed regional electronic media reports at least men-
tioned historical facts highlighted at the events (event’s background),
which means that reports and events themselves (journalists usually re-
fer to speakers or other information obtained at the event) are history-
centric. Nearly 66 of events occur at memorials or internment sites
and 28.6 at sites of individual or mass deaths. On the other hand,
the mythologisation of historical facts was observed in 19.8 of reports.
Hence, 102 unrepeated electronic media reports were identified (see Table
9.2). This also means as many different regional public dark commemora-
tive events, of which memorial services prevail with 75.2 – see Table 9.3.
In relation to the conflicts of the 20th century, 52.5 of events are related
to w w i i. From the Crosstab, a special type of table for demonstrating
the relationship between two categorical variables, it can be concluded
that dark commemorative services related to wwi i are the most frequent
dark commemorative events in Istria. Dark commemorative events re-
lated to the conflict of the 1990s are much less frequent and thus less often
subjects of electronic media reporting. Interestingly, almost 9 represent
wwi-related events, which are definitely associated with the anniversary
of the war. All these are answers to rq4.
9.4 Clusters of Dark Commemorative Events
On the basis of expectations arising from rq5, the TwoStep cluster anal-
ysis was employed. Both possibilities, a i c and b i c, were implemented
first (default settings were used) for the automatic clustering algorithm.
aic, chosen as the automatic clustering algorithm, creates seven differ-
ent clusters with 4 ≤ f ≤ 13.The identified low f of some clusters was not
satisfactory, which means that the proposed solution cannot be accepted.
The clustering procedure was repeated where automatic clustering was
limited to a maximum of five clusters.⁷ This resulted in four clusters (Fig-
ure 9.4), with f ≥ 10 and the quality of the model being preserved as ‘fair.’
This means that the new solution can be accepted (see Figure 9.2).
⁷ Interestingly, the determination of max. four clusters offers the same solution, while other
options associated with fewer clusters offer poor cluster quality.
201
variables: variables with f = 0 were removed, some related options were
sensibly merged. The final selection is shown in Figure 9.4.
9.3 Istrian Dark Commemorative Events and Conflicts
of the 20th Century
Exactly 88.9 of analysed regional electronic media reports at least men-
tioned historical facts highlighted at the events (event’s background),
which means that reports and events themselves (journalists usually re-
fer to speakers or other information obtained at the event) are history-
centric. Nearly 66 of events occur at memorials or internment sites
and 28.6 at sites of individual or mass deaths. On the other hand,
the mythologisation of historical facts was observed in 19.8 of reports.
Hence, 102 unrepeated electronic media reports were identified (see Table
9.2). This also means as many different regional public dark commemora-
tive events, of which memorial services prevail with 75.2 – see Table 9.3.
In relation to the conflicts of the 20th century, 52.5 of events are related
to w w i i. From the Crosstab, a special type of table for demonstrating
the relationship between two categorical variables, it can be concluded
that dark commemorative services related to wwi i are the most frequent
dark commemorative events in Istria. Dark commemorative events re-
lated to the conflict of the 1990s are much less frequent and thus less often
subjects of electronic media reporting. Interestingly, almost 9 represent
wwi-related events, which are definitely associated with the anniversary
of the war. All these are answers to rq4.
9.4 Clusters of Dark Commemorative Events
On the basis of expectations arising from rq5, the TwoStep cluster anal-
ysis was employed. Both possibilities, a i c and b i c, were implemented
first (default settings were used) for the automatic clustering algorithm.
aic, chosen as the automatic clustering algorithm, creates seven differ-
ent clusters with 4 ≤ f ≤ 13.The identified low f of some clusters was not
satisfactory, which means that the proposed solution cannot be accepted.
The clustering procedure was repeated where automatic clustering was
limited to a maximum of five clusters.⁷ This resulted in four clusters (Fig-
ure 9.4), with f ≥ 10 and the quality of the model being preserved as ‘fair.’
This means that the new solution can be accepted (see Figure 9.2).
⁷ Interestingly, the determination of max. four clusters offers the same solution, while other
options associated with fewer clusters offer poor cluster quality.
201
