Community Key Informant Final

Forty-two water decision makers in cities in Utah were identified representing elected official positions as well as staff (e.g., public utilities, public works, etc.). Three valleys in the rapidly growing Northern Utah Wasatch Range Metropolitan Area (WRMA) are represented. In smaller cities where staff play multiple roles, those who performed some operations in water management were selected. Those selected for interviews were identified through city websites and, in a few cases, phone calls to city hall. Participants were contacted by email first and followed up telephone as needed.

All of the interviews were conducted in-person between November 2015 and July 2016. During this time, city elections complicated contact and identifying key informants. When able, we interviewed the incumbents. Only one potential respondent who had initially agreed to an interview canceled without follow-up, for a response rate of 97.6%. Interviews were audio-recorded and tended to last between 20 and 90 minutes each. Each interview was transcribed with the help of two transcribers and deductively coded for themes by a team of three using NVIVO 11 Pro. The team started with an a priori coding matrix based on the interview guide and allowed for additional themes to emerge through the revision of categories and the coding agenda, reaching inter-coder reliability (<80% kappa coefficient). The database in NVIVO titled CKI_project_TEAM contains 40 transcribed interviews. One interview was not coded due to irrelevance and the pilot interview was not coded. Interview 013 does not exist because the respondent canceled. Overall, coders maintained a range of kappa coefficients with % minimum agreement. The final agreement measurements were calculated on Interview 38 which was coded by all three coders. High dual-coder agreement was also attained on the following interviews: 001, 003, 004, and 011. Coders met weekly to retain alignment in nodes and definitions (qualitative agreement). Coders were instructed to code every respondent sentence to the period (quantitative agreement). If the respondent's answer was short (e.g., Yes/No), the coder coded the interview question along with the answer to retain context. Respondents were asked the following: 1) the one key water issue facing their city today; 2) if their city had an adequate water supply to meet their city’s needs today, and 3) did they think their city had an adequate water supply to meet their city’s needs in the future.

在犹他州的城市中，确定了四十二位水资源决策者，他们代表的是选举产生的官员职位以及工作人员（例如，公共事业、公共工程等）。这三个山谷位于快速增长的北部犹他州沃萨奇山脉都会区（WRMA）。在规模较小的城市中，由于工作人员承担多重角色，那些在水资源管理中执行某些操作的人员被选中。通过城市网站以及少数情况下通过拨打市政厅电话，确定了接受访谈的人员。首先通过电子邮件与参与者取得联系，并在必要时通过电话跟进。所有访谈均于2015年11月至2016年7月期间以面对面形式进行。在此期间，城市选举使得联系和确定关键信息提供者变得复杂。在可能的情况下，我们采访了现任官员。只有一位最初同意接受访谈的潜在受访者取消了访谈，未进行后续跟进，访谈回应率为97.6%。访谈均以音频形式记录，每次访谈通常持续20至90分钟。每位访谈都由两位转录员协助转录，并由三名编码员使用NVIVO 11 Pro对主题进行归纳编码。团队从基于访谈指南的先验编码矩阵开始，通过修订类别和编码议程，允许出现额外的主题，达到编码者间可靠性（小于80%的Kappa系数）。NVIVO中的数据库“CKI_project_TEAM”包含40份转录的访谈。有一份访谈因不相关而没有编码，试点访谈也未编码。访谈编号013不存在，因为受访者取消了访谈。总体而言，编码员保持了各种Kappa系数的范围，达到最低的百分比一致率。最终的一致性测量是在编号为38的访谈上进行的，该访谈由所有三名编码员进行编码。在编号为001、003、004和011的以下访谈中也达到了高双编码员一致率。编码员每周开会，以保持节点和定义上的对齐（定性一致性）。编码员被指示将每位受访者的句子编码到句号（定量一致性）。如果受访者的回答较短（例如，是/否），编码员将编码访谈问题以及答案以保留上下文。受访者被问及以下问题：1）他们城市今天面临的一个关键水资源问题；2）他们的城市是否拥有充足的水资源来满足当前的城市需求；3）他们认为他们的城市是否拥有充足的水资源来满足未来的城市需求。