Contextual Design Models
Models
Data in the Life Team Consolidated Model
Our interviews revealed a series of challenges that could potentially hinder reliable and continuous updating of data inventorying and cataloging. We determined that there is inconsistent reporting and storage of data, so it can be difficult for departments to know what data they have; how to efficiently collect what they know they have; what data might be considered candidates for open platforms; and ultimately, what open data to report in inventories.
Inconsistent methods of data generation and management, and the variety of ways data is used, also have the potential to obscure inventorying because of the lack of standards. Part of the goal of data inventorying is to create a city-wide scheme that can be applied to the departments’ disparate data sets that will support open data efforts, but the heterogenous quality of the datasets themselves can inhibit accurate and consistent reporting and identification of potential open data sets.
Other issues that could hinder consistent reporting of data inventories are conflicts with other reporting obligations. Our model shows the different outputs of data after they are collected on internal and external databases. Pittsburgh City departments are required by law to report various datasets in regular intervals to local, state, and federal governments. In addition, collaborative obligations with other city departments include sharing pertinent datasets. Our interviews have shown that these responsibilities take precedence over any open data efforts.
Persona Model by Robyne Ivory
A common pain point that was consistently noted in all the departments we interviewed was the dilemma in managing analog data also referred to as “paper entries.” Out of necessity, some data from the pre-digital world is still used in many of the city’s departments. This data is in the form of handwritten or typed documents, maps, charts, and other archival material. The persona model highlights Richard Q. RecordKeeper. Richard is a composite of the four different viewpoints concerning analog data observed within the organization. 1) “I rely heavily on analog archival maps and charts to do my work” This mindset is common in departments where knowledge of the city’s layout is necessary. Sometimes this information is only available in its original paper format. 2) “I do not work at a desk; I rely on printed work orders to complete my assigned tasks.” This member of the organization does not have access to equipment needed to access digital data. 3) “We are not experts in records management, and we have no records management plan, no digitization, and no open data policy.” This organization member usually carries a heavy workload and does not have time or resources to learn current data handling methods. 4) “It would be great to have all this data available at the touch of a button, but I just don’t have time to add the information into a data management system.” While this member uses and understands data management tools, they would benefit from a designated data manager who would manage the department’s data. The ineffective methods noted produce challenges in creating consistent data management policies.
Identity Model with Matrices by Lisa Over
The Identity Model with Matrices includes a model of interviewees’ roles and experience with data and of what they like or think about data or data management. It also includes three matrices that compare employee data experience measures. Refer to the matrices key for interpreting the matrices.
The Identity Model with Matrices reveals three primary challenges to overcome:
- Poor data management practices
- Limited or no skills or availability for open data practices
- Identification as a data user as opposed to a data manager
The “I Like, I Think” section reveals that interviewees like standardized, accessible, and consolidated systems. Although department O has significant problems because of poor data management and lack of related resources, the “My Role and Experience with Data” section reveals that most departments have some concerns about data accessibility, standardization, or consolidation. Any issues that hinder data acquisition and preparation will hinder compliance with the open data policy.
In addition to poor data management practices, many interviewees expressed concern that they did not have the ability or time necessary for open data publication. For example, interviewees V1 and M2 both have data savvy data analysts who are technologically capable. However, V1 believes that their team does not have the skills to manage data or create dashboards. M2 says that their team has the ability but not the time because their time is spent preparing data for government reports.
The final challenge to overcome is that some interviewees identify as being ‘data users’ and not ‘data managers.’ Interviewees O1 and V1 have both identified very specifically as data users and have expressly stated that they are not data managers.
Collaboration Model by Andrew Williams
The collaboration model analyzes the processes conducted between the Department of Innovation and Performance (I&P) and other city departments that were either discussed or overlooked during interview sessions. While many of I&P’s roles were covered, there were several variations and inconsistencies, especially relating to interviewees’ understanding of I&P’s role with open data.
I&P handles various communication processes for the city, including handling 311 calls from residents and forwarding them to appropriate city departments. Interviewees 01 and Z1, representatives from two departments that deal with 311 calls, mentioned I&P’s role in the service while other departments interviewed that also use that service did not. I&P also manages the social media content for all city departments, but only interviewees H1 and V1 acknowledged that role. Those same two interviewees mentioned that they held regular meetings with I&P. H1 talked about working with I&P to create an alert system, while V1 discussed how I&P assists with large data transfers.
The collaborative model also displays inconsistencies in awareness of I&P’s role in making departmental data openly accessible. While most departments acknowledged that I&P played some role in making some of their data publicly available, most did not know what that data was. Even the two departments who were aware of the type of data that was accessible could not specify details about that data. Another lacuna in the processes of this model is the lack of awareness of the Open Data Ordinance itself. In conclusion, the model suggests that there is limited awareness of the collaborative roles and responsibilities between I&P and city departments, and that current communication channels are minimal. A better understanding of interdepartmental roles and improved regular communication and training would be beneficial in promoting consistent data inventories.
The affinity diagram process revealed several insights for our project. First, the City of Pittsburgh has a very diverse workforce with a diverse set of skills and perspectives. Although all managers support the Open Data Ordinance, they do not all believe that they have the resources, either time or skills, to comply with it. For example, O1 says that they oversee maintaining records, but that no one in their department is skilled in records management. Their data management practices are inconsistent, with their data stored on paper and six different digital storage systems. Another example involves M2 who says they have data analysts who could do this, but this team will prioritize government reporting over other data management tasks.
Lack of resources is not the only concern. Strong identification as a data user is another. V1 identifies strongly as a data user even though they are data savvy. V1 is afraid they will have a stack of requests sitting on their desk and that they will be responsible for noncompliance because they will focus on their problem-solving work, which is critically important and which they prefer. Identification as data users and lack of time or skills are the biggest threats to the Open Data Health Status checks.
Another insight is that several of the managers stated that they know very little or nothing about the Open Data Ordinance. Of the seven interviewed, O1, Z1, V1, and M2 said that they were not familiar with the ordinance. The City of Pittsburgh has not yet conducted the training for data coordinators and stewards. It is possible that some managers’ concerns may be alleviated after they receive training. For example, V1 stated that they do not know how to create dashboards. However, the Data Services Team will provide support for these tasks if they know what data they have.
Finally, it is notable that every department shares data with other departments, with other government agencies, and/or with the public per the Freedom of Information Act legislation. These shared datasets are strong candidates for open data publication. When developing the Data Governance Health Status Checks, our group should consider both the concerns the managers have about resources and the work that they already do to prepare data for sharing.