Introduction:

Popularly referred to as the playground of the North, the Northwoods has not been studied as a cultural region. A previous study conducted on North American regions, indicated by names of metropolitan enterprises, included the North as a popular region, but not the Northwoods as a distinct, separate region of the Midwest (Zelinsky, 1980). This study on religion in the Northwoods was done in collaboration with other research that focused on climate, soils, agriculture, land cover, tourism, recreation, cultural cuisine, and mining, in order to delineate the Northwoods as a region of the United States, in the states of Minnesota, Wisconsin and Michigan. Previous studies have been conducted on the geography of religion in the United States (Meinig, 1965; Zelinsky, 1952). In this study, evaluation of the distribution of religion by county in the Northwoods was used to determine if a pattern of religion exists in the Northwoods. After a preliminary analysis of 18 denominations, 11 denominations were chosen for further analysis, consisting of several likely to be dominant in the region and major Christian denominations. Data published by Glenmary Research Center, Nashville, Tennessee, in Religious Congregations & Membership in the United States: 2000, reported by region, state and countywere used. The sponsoring organization for the Glenmary study was the Association of Statisticians of American Religious Bodies (ASARB). This study produced results indicating that no pattern of religion exists that would assist in delineating the Northwoods, with the exception of the distribution of total adherents. A statistical analysis showed that total adherents were greater in the Northwoods in comparison to total adherents in the southern portions of the three states included in the study: Minnesota, Wisconsin and Michigan.

Methods:
                 
Obtained Glenmary data and CD-ROM from the University of Wisconsin-Eau Claire McIntyre Library:

• Downloaded data from Glenmary Research Center database (CD-ROM)
• Created a Microsoft™ Excel file selecting data for Minnesota, Wisconsin and Michigan
• Sorted data from 1952, 1980 and 2000, using eleven selected denominations
• Created .dbf files for each of the 11 denominations selected, for 2000 data.  Microsoft™ Excel was used to organize data from Glenmary’s database into .dbf files for each of the preliminary denominations.  Created .dbf files for 7 additional denominations and finally for the selected denominations. Each .dbf file was joined to a map of Minnesota, Wisconsin and Michigan, with North American Lambert Conformal Conic projection, using ArcGIS™.  Choropleth maps were produced, each displaying the distribution for the respective denomination in the three states.    
• Based upon a visual analysis of 36 choropleth maps of denominational adherents by county, eleven new categories were chosen:  nine individual denominations, combined Lutheran denominations, and total number of adherents. In order to analyze the correlation between the number of adherents by county to the north, a list of counties which could be considered Northwoods was selected.  A Pearson’s r statistical analysis for correlation was performed on each of the 11 categories using SPSS software (statistical software from IBM Company). 
• Loaded mapping data (.shp file) of Minnesota, Wisconsin and Michigan (North American Lambert Conformal Conic projection) into ArcGIS™. Loaded religion data into ArcGIS™, joined it with the corresponding mapping data, and produced choropleth maps.
  

Chart 1.  This flow chart briefly depicts the methods used to research religion in the Northwoods. Rectangles and ovals are used alternately for visual clarity.

Results:

In order to analyze the correlation between the number of adherents by county to the north, a list of counties was selected which could be considered Northwoods. Northwoods Counties Used for Correlation Analysis (Figure 1) does not serve to delineate the Northwoods, but displays counties arbitrarily selected as being in the Northwoods for the purpose of this analysis, in order to determine if a significant statistical difference exists. 

Figure 1.  Northwoods Counties Used for Correlation Analysis. Glenmary Research Center data for 2000 was used.

The results of the statistical analysis for correlation revealed a positive correlation between the Northwoods and religion (Table 1). The Pearson’s r value of .308 for the ‘total adherents’ shows a positive correlation, the only statistic in this analysis with a possible correlation between Northwoods and religion, because in the area of human geography a pattern is evident at a Pearson’s r value of .25 to .3, at the lower limit (www.sdobson.net, 2009; www.uiowa.edu, 2009).

Table 1. Northwoods Denomination Correlation Adherents as Percent of County Population: 2000 Pearson's r Significance Total Adherents 0.308 0 Evangelical Lutheran Church in America 0.14 0.03 Lutheran Church Missouri Synod 0.037 0.565 Free Lutheran 0.014 0.826 Wisconsin Evangelical Lutheran Synod 0.193 0.003 Combined Lutheran 0.187 0.004 Southern Baptist -0.097 0.134 Evangelical Free Church of America 0.029 0.65 Episcopal -0.181 0.005 Catholic -0.138 0.032 American Baptist -0.005 0.942

Table 1.  The standard statistical analysis for correlation commonly called Pearson’s r was used. Pearson’s r has a range of -1 to +1. Negative one represents absolute negative correlation while positive one represents absolute positive correlation with zero showing no correlation. Table 1 displays the significance value for each correlation. The commonly accepted significance or confidence level of 95% is shown by a value of .05 or less on this table (McGrew and Monroe, 1993).                      

Data on Table 1 shows a weak negative correlation (-0.138) for the Catholic Church entry, with the 0.032 level of significance showing the computation of the data to be mathematically valid, because it falls within the accepted confidence level: below 0.05. But a visual analysis of the Catholic Church Percentage by County: 2000 choropleth map (Figure 2), seems to indicate a possible pattern of a greater presence of the Catholic Church in the north. This is contradiction is caused by faults inherent in choropleth maps, which are useful for getting a first glance at distributions (Robinson, et. al., 1995). The large dark areas in the North representing large counties, interspersed with smaller counties of lighter colors, gives the visual impression of a larger number of Catholic Church adherents in the north, but further evaluation through statistical analysis indicates otherwise.

Figure 2.  Catholic Church Percentage by County: 2000. Glenmary Research Center data for 2000 was used.

Similarly, results of the statistical analysis for the Episcopal Church show a weak negative correlation (-0.181), indicating that it is slightly more present in the Southern part of the three states than in the Northern part. Again, Episcopal Church Percentage by County: 2000, seems to show a slight presence of the Episcopal Church in the North (Figure 3).

Figure 3.  Episcopal Church Percentage by County: 2000. Glenmary Research Center data for 2000 was used.

A Pearson’s r value of 0.14 for the Evangelical Lutheran Church indicates that it is slightly more present in the North than in the South. Evangelical Lutheran Church Percentage by County: 2000 gives a visual representation of the data in a choroplethic map (Figure 4).

Figure 4.  Evangelical Lutheran Church Percentage by County: 2000. Glenmary Research Center data for 2000 was used.

A Pearson’s r value of 0.193 for the Wisconsin Evangelical Lutheran Synod indicates that it also is slightly more present in the North than in the South. Wisconsin Evangelical Lutheran Synod Percentage by County: 2000 displays the data in a choroplethic map (Figure 5).

Figure 5.  Wisconsin Evangelical Lutheran Synod Percentage by County: 2000. Glenmary Research Center data for 2000 was used.

A Pearson’s r value of 0.187 indicates that the category of ‘combined Lutheran’ is slightly more present in the North than in the South. The choroplethic map All Lutheran as Percentage of Total Adherents: 2000 gives a visual representation of the data (Figure 6).

Figure 6.  All Lutherans as Percentage of Total Adherents: 2000. Glenmary Research Center data for 2000 was used.

The ‘total adherents’ category, with a Pearson’s r value of 0.308 is the only category showing a statistically significant correlation between the Northwoods and religion, because it meets the commonly accepted criteria mentioned above. Total Adherents as Percentage of Population by County: 2000 displays a visual representation of the data (Figure 7). 

Figure 7.  Total Adherents as Percentage of Population by County: 2000. Glenmary Research Center data for 2000 was used.

Summary:

The initial assumption that Lutheran denominations, because of their cultural ties to Scandinavian ethnicity, would delineate the Northwoods as a distinct region of the Midwest proved to be false. Analysis of 18 other denominations or combinations of denominations through choropleth maps, failed to show a significant difference between the North and the South. Further analysis, through Pearson’s r statistical analysis for correlation, revealed that the category ‘total adherents’ was the only area showing a possible correlation between the Northwoods and religion. This weak to moderate correlation shows that there are more adherents in the North (of the denominations evaluated) than in the South. But further research would be advised before these findings could be used to delineate the Northwoods region. This could include analyzing data using different combinations of denominations or other normalizations or statistical approaches. 

References:

Jones, Dale E., et. al. 2002.  Religious Congregations & Membership in the United States 2000: An Enumeration by Region, State and County Based on Data Reported for 149 Religious Bodies

McGrew, J. Chapman, Jr. and Monroe, Charles B. 1993.  Correlation. In An Introduction to Statistical Problem Solving in Geography (Chapter 12--pp 244-264) Dubuque, Iowa. Wm. C. Brown Publishers.

Meinig, D. W. 1965. The Mormon culture region: strategies and patterns in the geography of the American West, 1847-1964. Annals of the Association of American Geographers 55(2): 191-220.

Robinson, Arthur H., Morrison, Joel L., Muehrcke, Phillip C., Kimerling, A. Jon, and Guptill, Stephen C.. 1995. Symbolization: Feature Attribute Volumes (Ch. 26) in Elements of Cartography. (6th Ed.)  New York: John Wiley and Sons, Inc.

www.sdobson.net (accessed November 29, 2009)

www.uiowa.edu (accessed November 29, 2009)

Zelinsky, Wilbur. 1962. An approach to the religious geography of the United States:
Patterns of church membership in 1952. Annals of the Association of American
Geographers 51: 139-193.

Zelinsky, Wilbur. 1980. North America’s Vernacular Regions. Annals of the Association of American Geographers 70(1): 1-16.

Related Works:

Annals of the Association of American Geographers, 96 (1). 2006. Forum on
“Theorizing and Studying Religion,”

James Proctor, Introduction to the Forum,” pp. 165-168.

Adrian Ivakiv, “Toward a geography of “religion,” mapping the distribution of an
unstable signifier,” pp. 169-175.

Michael P. Ferbver,”Critical realism and religion: objectivity and the insider/outsider
problem,” pp. 176-181.

Julian Holloway, “Enchanted spaces: the science, affect, and geographies of religion,”
pp. 182-187.

James Proctor,” Religion as trust in authority: theocracy and ecology in the United
States,” pp. 188-196.

Anne Buttimer, “Afterward: Reflections on geography, religion, and belief systems,” pp.
197-202.

Barcus, Holly R. and Stanley D. Brunn. 2004. Mapping Changes in Denominational
Membership in the Great Plains, 1952-2000. Great Plains Research 14 (1): 19-
48.

Brunn, S. D. and Holly R. Barcus. 2004. New Perspectives on the Changing
Religious Diversity in the Great Plains. Great Plains Research 14 (1): 49-76.

Fickeler, Paul. 1962. Fundamental questions in the geography of religions. In Philip L.
Wagner and Marvin W. Mikesell, eds. Readings in Cultural Geography. Chicago:
University of Chicago Press, pp. 94-117.

Heatwole, Charles. 1977. Exploring the geography of America’s religious
denominations: A Presbyterian example. Journal of Geography 76: 99-104.

Hill, Samuel S. 1985.  Religion and Region in America. The Annals of the American Academy of Political and Social Science 480 (1): 132-141.

Issac, Eric. 1965. Religious geography and the geography of religion. Man and the
Earth. University of Colorado Studies, Series in Earth Sciences No. 3. Boulder:
University of Colorado Press.

Kong, Lily. 1990. "Geography and religion: trends and prospects".  Progress in Human Geography 14 (3): 355-371. http://phg.sagepub.com/cgi/pdf_extract/14/3/355.

Kong, Lily. 1991. Mapping “new geographies” of religion: Politics and poetics in
modernity. Progress in Human Geography 25: 211-33.

Maffly-Kipp, Laurie F. 2007.  Putting Religion on the Map. The Journal of American History 94(2).

Mayhew, Robert T. 2000. “Geography is Twined with Divinity.” The Laudian
Geography of Peter Heylyn. Geographical Review 90 (1): 18-34.

Leonard, James M.2006. Local geography and church attendance: Wayne County, West
Virginia. Geographies of Religions and Belief Systems 1 (1): 28-50.

Levine, G. 1986. On the Geography of Religion. Transactions of the Institute of British
Geographers 11: 428-440.

Licate, J. 1967. The geographic study of religion: A review of the literature. Chicago:
University of Chicago. Unpublished MA Thesis.

Religion and Public Life in the Midwest: America's Common Denominator? Ed. by Philip Barlow and Mark Silk. Walnut Creek: AltaMira, 2004.

Robinson, Arthur H., Morrison, Joel L., Muehrcke, Phillip C., Kimerling, A. Jon, and Guptill, Stephen C.. 1995. Census and Sampling (Ch. 9—pp 159-166) in Elements of Cartography. (6th Ed.)  New York: John Wiley and Sons, Inc.

Robinson, Arthur H., Morrison, Joel L., Muehrcke, Phillip C., Kimerling, A. Jon, and Guptill, Stephen C.. 1995. Digital Databases (Ch. 13—pp 222-244) in Elements of Cartography. (6th Ed.)  New York: John Wiley and Sons, Inc.

Robinson, Arthur H., Morrison, Joel L., Muehrcke, Phillip C., Kimerling, A. Jon, and Guptill, Stephen C.. 1995. Geographic and Cartographic Database Concepts (Ch. 14—pp 246-254) in Elements of Cartography. (6th Ed.)  New York: John Wiley and Sons, Inc.

Robinson, Arthur H., Morrison, Joel L., Muehrcke, Phillip C., Kimerling, A. Jon, and Guptill, Stephen C.. 1995. Cartographic Design (Ch. 18—pp 315-338) in Elements of Cartography. (6th Ed.)  New York: John Wiley and Sons, Inc.

Shortridge, J. R. 1976. Patterns of religion in the United States. Geographical Review
66: 420-34.
Sopher, David E. 1967. Geography of Religions. Englewood Cliffs, NJ: Prentice-Hall.

Star, J., and J. Estes, Geographic Information Systems: An Introduction, Englewood Cliffs, NJ: Prentice-Hall, 1990.

Stump, Roger. 1986. The geography of religion – introduction. Journal of Cultural
Geography 7: 1-3.

Warf, Barney and Mort Winsberg. 2008. The Geography of Religious Diversity in the United States. The Professional Geographer 60 (3): 413-424.

Yeates, M. 1974.  An Introduction to Quantitative Analysis in Human Geography. New York: McGraw-Hill.