2 What Can We Do With Data?

Asking Good Questions

Just as we can ask good question to guide our lives, organizations must ask good questions in order to get the information they need to make good decisions. Rather than just making decisions based on intuition or observations alone, organizations should use data that has been gathered to make decisions. This is referred to as data-driven decision making. By using data as the basis for decision-making, organizations can avoid bias and false assumptions that lead to poor decision-making.

Let’s apply what we know about good questions to making data-driven decisions.

Know Your Mission

President Henry B. Erying said that it is hard to see which questions matter unless you have some vision of your own life and of the world. The same is true of organizations. It is hard to see which questions matter to the organization unless there is some vision of the company’s goals, focus, or its mission. Once you have a good understanding of these, you can determine what questions must be answered to achieve organizational goals.

The questions should be important and relevant to the organizational mission. A foundational knowledge of the industry and what problems your organization and others face in the industry will also help determine these questions. A clear purpose and a defined problem to solve will also bring focus on what questions are important. Remember we don’t want to focus on unimportant questions to the exclusion of good ones.

For example, a company may need to decrease expenses in order to remain profitable. The company must make a decision. If the question of how to decrease the expenses in order to remain profitable is important and relevant to the company and their goals, then getting as much data about how to solve this problem is the first step.

Identify Data Sources

Once we know what questions to ask, we need to find data sources that will answer the questions. Where does the data come from? What data does our organization create or gather that will contribute to this data store? If you know what questions need to be answered, you can work backwards and figure out what data is needed to answer those questions. Also, keep in mind any future questions or problems that might require that we store additional data for the future.

Data-driven decision making requires that the data is good, in other words, that it is high quality or credible data. Many times data can be inaccurate, outdated or unreliable. We must be good stewards of our data to make sure the decisions we make are based on good quality data. It is critical to follow a quality assurance process as we gather data. For example, a company would not want to downsize staff or cut employee benefits if there is not good data to support this decision. With good data, the company might see other ways to cut costs or, indeed, decide that downsizing is the right decision. As long as the questions are good and the gathered data is high quality, a company can make good decisions. But how do we use good data to help us make decisions?

Organize the Data

In this course, you will learn how data is stored and retrieved. As you learn about database design you will learn how to organize data into an efficient design so that you can retrieve the information you need from it. You will prepare the raw data and organize it in such a way that it becomes tables of useful information that you can ask questions of (or query) for answers to your questions. As you organize the data, you will ensure data integrity (making sure our data is accurate and reliable) and data consistency (making sure our data is useable) and eliminate data redundancy (storing the same data in multiple locations). As we are querying the data, we will take tables of data and sort the data in a specific ways to make the reported data more useable. We will also filter data or hide unwanted data, displaying only the records we want to see.

Analysis and Conclusions

Once we retrieve the information we need to answer our questions. We can analyze the information and draw conclusions from the information in order to answer our questions and make decisions that will help our organization succeed.

Statement of Work

Because data systems or other projects are built to benefit organizations, a document called a Statement of Work (SOW) can be used to plan a successful execution of the project. This works with data-driven decisions. Users of the data system need to make decisions based on the output from these systems or projects. This SOW document is the starting point of creating a data-driven system or project that will give us information to answer our organization’s questions.

It's important to remember that as the designer of a database system, we don’t get to decide what will and will not go into the system. The client (the ultimate users of your database) will decide what they need. The statement of work will also help keep the project on schedule and within budget. It helps everyone know what is expected. For example, the client might be expecting a complete working system once the database is built, but it is your understanding that only the backend database is being developed. It is situations like this that can be cleared up through the Statement of Work so everyone is in agreement.

Parts of a Statement of Work

Let's look at 4 basic elements of the Statement of Work.

History

The history section is where you will find why the organization needs the database system. What is the current system like or perhaps they may not even have a current digital system, and they need one. How did they come up with their current system? What problems is that current system causing? What are its limitations? At this point, it is helpful to realize you need to be prepared with questions you can ask to get the users to focus on the important aspects they need. Make sure to do a lot of listening so you can understand exactly what is needed. Keep your preconceived notions out of the discussion. You will take a look at those who will use the database system and what output they need from the system to do their jobs well. These users may not know anything about database design, but they do know exactly what they need to do their job.

Scope

Scope is a description of what needs to be done. It is a general statement of the requirements and expectations of the project. It does not go into details about how exactly this will be done but provides the overall description. This may also include general constraints such as time and budget constraints. What software and hardware will be needed and who will do the work?

Objective

The objective element of the statement of work involves the purpose of the project. What is it intended to achieve? This does not include specifics about each of the elements of the database, but what the database is supposed to achieve. What outputs will be produced by the system? Remember, users know what they need to do a good job. You will listen to them, look at their current outputs, screen views, reports, forms etc. and get an idea of what information your database will need to produce. In this course, we will be focusing on data-driven systems like databases and, from these databases, how decisions-making information is retrieved.

Timeline

In the timeline section,each task involved in creating the project is broken up into discrete tasks. Each task will have an estimated time frame of completion and what is to be done during that time frame.

Later in the course, as we begin looking at database design, we will revisit statement of work documents that were created as we begin the design process. The statement of work is the beginning point to see what will be going into the design and is completed before a database system is created. We will not get into too many details of this process for this introductory course. Many of the databases we will use, come with assumptions that we have already been through this SOW process. However, it is not something you want to skip in a real world design process. It is a mistake to think you know how to begin the design process before you have studied it out thoroughly with those who will be using the system.

A good database begins with good quality data and is designed with careful planning and with the proper people involved.