Community CharacteristicsCommunity characteristics are mainly considered resources - as in the availability of medical care.
We think that community population size (total and Native) is correlated with many differences in the resources available in communities and in the structure of social and economic relationships within the community. Community population data is not obtained in personal interviews; rather, we obtain these data from the most recent government census or government administrative records. We can attach community size data to personal interview records by matching community identifiers.
Privacy Act requirements in Canada may require Statistics Canada to remove community identifiers from the public use micro-data set (i.e. the file of individual records released for public research use). If so, it will not be possible to attach community population size data to the personal interview data.
There are two alternative strategies to address this problem:
1. attach a community-size variable to each personal record in the public use data set (e.g. an ordinal 4-point scale).
2. group smaller communities in the same region.
The disadvantage of alternative #1 is that there are likely to be other community attributes important to the analysis (e.g. economic structure of the community). Each of these attributes would need to be added as a community-level variable to the personal interview data. The combination of these attributes might allow a user to identify the community. Thus care would have to be taken in the number and level of detail in the community-level attributes attached to the file.
The disadvantage of alternative #2 is that grouped communities are certain to differ in some important characteristic.
We recommend that the final decision on how best to deal with community-level attributes be made on the basis of preliminary analyses.
In the Alaska dataset, we are using the following community-level analytic variables:
PLACETYP and VILLAGE: both variables are functionally identical and differentiate regional centers (Barrow, Nome, Kotzebue) from the smaller villages.
PLACE and PLACENUM: both variables are identical and identify sampled places.
PLACEREG: identifies sampled places as PLACE and PLACENUM but include a regional identifier in the text value so that places are grouped by region.
There is one region-level variable: REGION, which identifies the three Inupiat regions (N.Slope, Bering Straits, Northwest Arctic).
WHALECOM Whaling communities in sample: Barrow, Kaktovik, Nuiqsut, Wainwright, Kivalina, Point Hope Savoonga (250,1560,1710,2436,2660,2910,3490)
D4 Availability of doctor or other medical professional in community.
D12 Ability to get medicine respondent needs.
D19 Availability of traditional healing in community.
ENVIRON Count of 10 types of environmental problems in region or community.