statistics.com

[ksipa]

I am looking for more clarification on how a census survey is usually analyzed and reported. While there is a lot of information on how to go about applying statistical tests and estimates for sample data to make inferences about the population, there is not a lot of information on how to interpret census data. The information sources I have come across just states that it is not practical or economical to survey everyone in the entire population therefore sampling should be used. However, the population I deal with is small and it is feasible to survey everyone.

I understand that statistical tests rely on the sampling distribution to determine how certain you are that you are within a range from the population parameter. However, in a census survey wouldn’t I have the population parameter and therefore would not need tests of significance? Would I just summarize the data as found on the frequency tables (depending on my data type of course)? What about measures of associations to test relationship with subgroups? Any guidance that can be provided would greatly be appreciated. Thanks!

[alethephant]

The term "census" doesn't necessarily mean 100% sampling. It tends instead to indicate a cross-sectional survey. E.g., the US Census next year will be based on sampling.

If you have canvassed the entire population at one point in time (a 100% sampled cross-sectional survey), then, as you indicate, you problem is to identify descriptive parameters of the population that would be of interest, and relationships of interest among response and covariate variables. Insofar as you wish to generalize from your population to a larger one, your population would become a sample and the parameters statistical estimates.

The purpose of population statistics is to reduce a mass of measured data to manageable proportions. Exploratory data analysis techniques are a good place to start, as are multidimensional contingency tables. Multivariate methods are also applicable to identify clusters.

It would help if you actually described your data instead keeping it hypothetical with no structure. Otherwise all you will get is hypothetical unstructured answers.

[ksipa]

Thanks alethephant,

Can you provide more clarification on what you mean by "canvassed the entire population"? Maybe some additional information on the population I deal with would help. When I say "census" survey, I am referring to all individuals within an organization or within a division. Literally, I would be surveying all individuals and of course there will be some non-response. I have received some conflicting information on whether I should treat my survey data as population data or sample data.

Also, when you say, "relationships of interest among response and covariate variables", are you referring to measures of association between independent and dependent variables?

[alethephant]

If you wish to draw conclusions ONLY about the particular group you are studying, then your group is a population, and all you can do is describe what's going on inside it.

If you wish to make inferences about other groups similar to the one studied, your group is a cluster sample. Your inferences will be then subject to error.

Sample vs. population is a matter of your intent to make inferences. Any sample can be a population, and any population a sample, depending on the way it's used to draw conclusions and about what.

[alteriunhun]

Hi alethephant,


Thanks for the suggestion that's a good help for me and to other too i can draw conclusions ONLY about the particular group you are studying, then your group is a population, and all you can do is describe what's going on inside it!!!thanks!!
internet marketing training courses
internet marketing training courses