Example System Overview - brianhigh/data-workshop GitHub Wiki

This research project will use a data system to support a study involving the collection and analysis of bacteria samples. The data system will store sample information and laboratory test results. The results will be used for statistical analysis.

The sampling technician will enter sample information into the data system (or a separate data collection tool) as samples are collected in the field. If a human subject is being sampled, either the human subject or the sampler will enter some basic personal information about the subject into the system or tool. For a human subject to be involved in the study, they must have first returned a signed consent form to the sampler. The sampler will deliver the samples to the lab for testing by a laboratory technician.

Once the samples are in the lab, they will be placed in refrigerated storage cooler. A lab technician will run lab tests on the samples and store the results in the data system. The samples will be kept in the cooler in case further testing is needed. After the samples are no longer needed, they will be disposed.

These interactions with the data system will be through manual data entry into a user interface connected to the data system or through the import of data from external files. For field sampling, it would be nice to be able to use a tablet for data entry into some sort of “app”. In the lab, it would be nice to use either a computer with a database “form” or a web interface for simple data entry, but using a spreadsheet would be okay, though less desirable. A tablet with an “app” would also be okay for simple lab data entry. Automated lab tests with lots of numerical results coming from a lab instrument would be best exported from the instrument as a text file, like a CSV. And that CSV would be imported into the database. Or the instrument data might be imported into a spreadsheet, processed there, and then exported to the database.

Additionally, the data system will be used for statistical analysis. The statistician will either connect to the database directly (through the campus network) from the statistical software to run queries, or will run queries using the query tool supplied with the database product and then export the results to an external file for later statistical analysis. It is preferable to minimize the importing and exporting of files where possible.

This draft overview is a good starting description of the study needs regarding a data system. It is the sort of description that might come out of a casual meeting between the researcher and an IT person. The The researcher is fairly flexible about the user interface requirements, yet still stated preferences, which is nice. The mention of specific tools for data management, like “database”, might seem a bit premature, as the system is yet to be designed. However, it is good to capture the researcher’s ideas of how the system might be implemented. Such details make excellent starting points for further discussion. There are still many details left unmentioned (like security issues such as how personal identifiers will be protected, or who will be allowed to access the data), and so the other requirements documents will be used to elucidate more specific needs. The next step will be to create a Example Goals List for the data system. Later on, once the other requirements documents are prepared, we will be able to provide a more specific Example System Overview to be included with the compiled Example Requirements Document sent to a system designer, or equivalent.