(Part 4 in a 4-part series)
In the last post, I put forth that it's sometimes useful, both in practical planning and academic research, to distinguish between types of Christians. In this post, I give some caveats in doing so.
When typologies are created from data, they reflect decisions made by the researcher.
1) Content of groups. Foremost among these decisions are which variables to analyze in creating groups. For example, suppose we had a large data set with lots of religion-related variables. One researcher might create a typology using measures of religious affiliation, and the resulting typology would emphasize denominations and other group measures. Another researcher, however, might look at variables in which people describe their relationship to God, and that researcher might end up with a typology similar to that used by the Reveal Study. A third researcher might emphasize behaviors and come up with a typology of religiosity.
2) Number of groups. Another issue regards how many groups to identify. Researchers specify various thresholds in their statistical software that guide the formation of groups. Different thresholds would give different number of groups. For example, the study described earlier in this series by Christianity Today identified four groups. Changing around the default settings in their statistical package could decrease or increase that number.
3) Naming groups. Once groups are identified, they are usually described by their characteristics on various characteristics. At this point, the researcher usually assigns some sort of snappy name for each group, to make it more understandable. For example, Barna labeled a group "revolutionaries." Another researcher, however, might look at the same data and give that group a different name. For example, perhaps "revolutionaries" could also be named "cutting-edge" Christians or maybe "highly-committed" Christians. While various labels would fit any given type, each label has its own connotations, giving subtly different meaning to the taxonomy.
What's the main point here? Typologies are inherently arbitrary. Two very competent researchers could look at the same data and create two very different groupings of Christians.
As such, presentations of typologies miss the mark when they present the types as the "true" types or the "real" types found in the data. In reality, there are hundreds of possible different groupings of Christians, and so any presentation of typologies should emphasize the usefulness of those types but not claim any unique reality of it.
Properly used, typologies certainly have a place in analysis of Christianity.
Thanks David & Stacey!
In the last post, I put forth that it's sometimes useful, both in practical planning and academic research, to distinguish between types of Christians. In this post, I give some caveats in doing so.
When typologies are created from data, they reflect decisions made by the researcher.
1) Content of groups. Foremost among these decisions are which variables to analyze in creating groups. For example, suppose we had a large data set with lots of religion-related variables. One researcher might create a typology using measures of religious affiliation, and the resulting typology would emphasize denominations and other group measures. Another researcher, however, might look at variables in which people describe their relationship to God, and that researcher might end up with a typology similar to that used by the Reveal Study. A third researcher might emphasize behaviors and come up with a typology of religiosity.
2) Number of groups. Another issue regards how many groups to identify. Researchers specify various thresholds in their statistical software that guide the formation of groups. Different thresholds would give different number of groups. For example, the study described earlier in this series by Christianity Today identified four groups. Changing around the default settings in their statistical package could decrease or increase that number.
3) Naming groups. Once groups are identified, they are usually described by their characteristics on various characteristics. At this point, the researcher usually assigns some sort of snappy name for each group, to make it more understandable. For example, Barna labeled a group "revolutionaries." Another researcher, however, might look at the same data and give that group a different name. For example, perhaps "revolutionaries" could also be named "cutting-edge" Christians or maybe "highly-committed" Christians. While various labels would fit any given type, each label has its own connotations, giving subtly different meaning to the taxonomy.
What's the main point here? Typologies are inherently arbitrary. Two very competent researchers could look at the same data and create two very different groupings of Christians.
As such, presentations of typologies miss the mark when they present the types as the "true" types or the "real" types found in the data. In reality, there are hundreds of possible different groupings of Christians, and so any presentation of typologies should emphasize the usefulness of those types but not claim any unique reality of it.
Properly used, typologies certainly have a place in analysis of Christianity.
Thanks David & Stacey!
2 comments:
You cited some general studies, but can you give some specific examples of poorly typology vs. helpful typology?
Hey J.R.,
In a sense, what is helpful for a typology is up to the user. E.g., in our church, we have lots of UConn students, so we constantly have to think about university students vs. regular old community members. Other churches wouldn't need to.
Post a Comment