Analysis: Work in Progress
We have started a comparative analysis of Canadian, Greenland, Chuktoka and Alaska data, using the
five international themes. Greenland and Canada are taking the lead on Themes 2.
Chukotka (Health) and Canada (Regional Autonomy) are taking the lead on Theme 5. Alaska
is taking the lead on Themes 1 and 3. Theme 4 (Education and missionaries)
requires data not yet available from Greenland and/or Chukotka).
To keep the initial Alaska analysis separate from the comparative analyses,
you will find a link to Comparative analyses associated with each theme. We have
left the preliminary Alaska analyses "as is". We have also left "as
is" an initial analysis of combined Alaska/Canada data.
Note: There are specific rules for disclosure for the Canadian APS
data set. Unweighted cell sizes have to be at least 10. Percentages have to be
calculated by generating weighted counts, exporting these counts to Excel, using
a StatCan procedure for rounding the weighted counts, and then calculating
rounded, weighted percentages in Excel. Download
view an example of such a spreadsheet. The sheet containing unrounded, weighted
counts has been deleted by StatCan.
Note: We also learned that asking for vetted output from Stat Can now may preclude
us from asking for vetted output that differs marginally from that previously
released (e.g. using a different grouping of the same variable). This is because
the two results conceivably might be used to calculate smaller groupings that would be below
the minimum cell size. To avoid this problem, we decided to record the strength
of each observed relationship qualitatively - that is, "very strong", "strong",
"significant but not strong", and "not significant". These research results
appear as color-coded cells in Excel spreadsheets by theme.
SLICA Results for Review