iUTAH researchers contacted municipal water provider organizations in the 12 cities represented in the 2014 household survey that maintain billing records or other water use records. Of the 12, 11 cities released data under a strict confidentiality agreement outlined in a memorandum of understanding to link water bills from months in 2014 to parcels or buildings where individual survey respondents were located. The water bills were matched to results from a 2014 household survey. Researchers at Utah State University and the University of Utah implemented the ‘2014 iUTAH Household Survey’ with over 2,300 randomly selected households in 2014 in 23 neighborhoods in 12 Utah communities. The survey included detailed individual- and household-level information about water management behaviors, perceptions of water resource conditions, and attitudes toward a range of water policies and programs.

The survey research team leaders agreed to:
• Treat any water use or billing records with care and discretion and to respect the privacy rights of individual water system customers.
• Aggregate the results of our analysis so that the historic water use levels and water bills of any individual customer, building or parcel are not released in any publicly accessible document, presentation, or report.
• Never share the detailed water use records with any other individual or group without the expressed written permission of the municipal water provider organization.
• Ensure that any person who has access to the raw individual-level survey and water use datasets have completed institutional review board human subjects research training, are currently certified and authorized to work with the data, and agree to the stringent confidentiality protocols listed above.
• Not reveal the specific location or identity of individual respondents to the 2014 iUTAH Household Survey to any other individual or organization, including the partner municipal water provider organization.

The municipal water provider organization representatives agreed to:
• Provide an electronic dataset of billing or water use records that permit a reliable estimate of actual rates of water consumption at the parcel or building scale.
• Address, tax parcel, or other information that allows these records to be linked to the individual parcels, buildings, or customer addresses.
• Not require the research team to reveal to the municipal provider the identity of which specific parcels or households were either sampled into or responded to our survey.

The data cleaning process included the following steps:
a. Calculate monthly estimates
b. Calculate per capita based on household size
c. Calculate per acreage
d. Calculate tiered cost
e. Match household survey responses with water bill data