3. Dataset Summaries
Cities often provide a short description of their 911 data along with the dataset itself. The Reimagine 911 review team refers to these as the “dataset summaries.”
While dataset summaries are brief and sometimes lack essential details needed to perform an analysis, they are a useful starting place for determining what type of 911 data is contained in each dataset. For example, if there is a significant change within a dataset, such as a systems upgrade, the summary might explain why the change was made and how it affects the dataset. Or, if you see a dataset summary mention a weekly override of data, you’ll know right away that this is the wrong dataset for a year-over-year analysis.
Our Open Data Review found that about four out of five data providers supplied a summary of the dataset, though the quality varied.
Helpful Summary Information
Here is a brief summary of valuable information that would ideally be provided in a summary. Many of these topics are covered in depth elsewhere in this article:
Data providers: Most of the publicly available 911 data is a compilation of different emergency service datasets (see Open Data Providers).
Historical availability and consistency: It is generally easy to tell how many years of data is available, but information about changes in data collection is less common. For example, did the data collection process change over time or did the columns and schema change at any point in the past? (For more information on this, see Historical Records.)
Data usage licensing: Depending on how you hope to use the information, the licensing agreements may have important implications (see Data Licensing).
Data access: Open data is available in many different ways across different cities. The ways that cities make their data available may dictate your processes (see Data Platforms and Exports).
The summary can also provide helpful information on what is not included in a dataset. In Reimagine911’s review of datasets, we recorded this in an “exclusions” column. Information excluded from the dataset summary can often be found in other columns in the review, which will be explained in this article.
Locating Summary Information
In the most straightforward situations, each dataset will be prefaced with a descriptive paragraph either on an introductory page or perhaps above the actual data table. Unfortunately, this is not always the case. In our review, we found that 27 out of 129 data providers (20%) supplied no summary or an insufficient summary of their dataset in an intuitive location and our team members had to do additional research to provide summaries for all datasets.
Even when the data is available it can be difficult to find. Here are common places descriptive information is kept on the major platforms:
Socrata’s collapsible fields: They can be easily missed; in some cases the only indication of summary text may be the small gray triangle itself. This is illustrated below in Mesa Arizona’s open data portal.
Fig. 3.2 Socrata’s collapsed summary information.
Fig. 3.3 The expanded summary field.
ArcGIS’s detail pages: They may be hard to find when collapsed unless you know where to look. Look for the “ⓘ” icon in a collapsed sidebar on the left side of the screen. You may have to navigate around several linked pages to get a full understanding of the dataset.
Fig. 3.4 ArcGIS’s “View Full Details” button is easily found once the info tab is expanded.
Fig 3.5 Valuable information such as data dictionaries may be maintained off-site.
The bottom of OpenGov’s CKAN pages: Look here to find more summary information. This platform tends to display useful information such as data dictionaries and licensing information at the bottom of the page, below a long table of data.
Fig. 3.6 On CKAN sites, scroll down below the data tables to look for valuable summary information.
Last updated