Geoscience Reference
In-Depth Information
Contrast
Significance
Difference
+/- limits
Cho - Dot
*
0,336232
0,324017
Cho - Prop
-0,164918
0,326829
Dot - Prop
*
-0,501149
0,328589
Table 2:
LSD Multiple Range Test for Total Direct acquisition answers.
Figure 7:
Percentage accuracy for direct acquisition task
The cross-tabulation for results on this task also permitted the considera-
tion of the simultaneous occurrence of map used and preferred map against
the sum of correct answers (
Figure 8
). For this result, it can be observed
that users who matched map usage and map preference (DD, PP, CC) did
not always achieve answers of higher accurate when compared to non-
matching (CP, CD, DC, DP, PC, PD) users. Best results were achieved by
proportional maps (PP), but differences were only statistically significant
against DC, DD and CC classes. It could indicate evidence that matching
preference and use can lead to good results for specific visual variables.
However, once there were more than twice as many occurrences of
choropleth map preference, and sampling was not designed to support sub-
sample analysis, a higher and equally distributed number of respondents
would be necessary to confirm this trend. Another possible analysis would
be to consider each map technique separately. In this case, for proportional
and dot maps (PP and DD classes), it is true that matching cases had better
performance, but this was only statistically significant for the proportional
map technique.
