Irvine Minnesota Inventory (IMI) Materials

Background

Inter-rater reliability is an important element of environmental audit tools. This paper presents results of reliability tests of the Irvine–Minnesota Inventory, an extensive audit tool aimed at measuring a broad range of built environment features that may be linked to active living. Methods: Inter-rater reliability was measured by percentage agreement between observers. Reliability was tested on a broad range of sites in both California and Minnesota.

Results

For the variables that remained in the inventory, in tests conducted at the University of California–Irvine, 76.8% of the variables had 80% agreement among the three raters. In tests conducted at the University of Minnesota, 99.2% of the variables had 80% agreement among the two raters.

Conclusions

Reliability was high for most items. The inventory was modified to eliminate items with low reliability. Differences in the use of the inventory and the goals of the research led to generally higher reliability in Minnesota. Those differences, limitations, and directions for future research are discussed. (Am J Prev Med 2006;30(2):153–159) © 2006 American Journal of Preventive Medicine

Background

Researchers and policymakers increasingly identify active living—including walking and bicycling for travel and recreation—as a potential strategy to increase rates of physical activity in the United States. Understanding the impact of the built environment on physical activity levels requires reliable methods to measure potentially relevant built environment features. This paper presents an audit tool—the Irvine Minnesota Inventory—that was designed to measure a wide range of built environment features that are potentially linked to active living, especially walking.Methods: The inventory was created through a literature review, focus group interviews, a panel of experts, and field testing in 27 settings. The inventory was developed in 2003–2004.

Results

The Irvine Minnesota Inventory includes 162 items, organized into four domains: accessibility (62 items), pleasurability (56 items), perceived safety from traffic (31 items), and perceived safety from crime (15 items). (Some items are in multiple domains.) The inventory includes both a paper version and a version in Microsoft Access, to allow data to be input directlyinto the computer.

Conclusions

Limitations of methods used to develop the inventory are discussed. Strategies are offered for using the Irvine Minnesota Inventory to systematically and reliably measure characteristics of the built environment that are potentially linked to active living.

Abstract

The Irvine Minnesota Inventory (IMI) was designed to measure environmental features that may be associated with physical activity and particularly walking. This study assesses how well the IMI predicts physical activity and walking behavior and develops shortened validated audit tools. A version of the IMI was used in the Twin Cities Walking Study, a research project measuring how density, street pattern, mixed-use, pedestrian infrastructure, and a variety of social and economic factors affect walking. Both bivariate and multivariate analyses were used to assess the predictive value of the IMI. We find that while this inventory provides reliable measurement of urban design features, only some of these features present associations with increased or decreased walking. This article presents two versions of shortened scales—a prudent scale, requiring association with two separate measures of a physical activity or walking behavior, and a moderate scale, requiring association with one measure of physical activity or walking. The shortened scales provide built environment audit instruments that have been tested both for inter-rater reliability and for associations with physical activity and walking. The results are also useful in showing which built environment variables are more reliably associated with walking for travel—characteristics of the sidewalk infrastructure, street crossings and traffic speeds, and land use are more strongly associated with walking for travel, while factors that measure aesthetics are typically less strongly associated with walking for travel.