When many of us get sick, we try self-diagnosing the problem. Maybe we look up information online. We may even go to a Web MD to find out what’s wrong with this or that. Through this process we might find all sorts of symptoms we think we exhibit and conditions we think we have. The problem is that we are not experts in the medical field. We don’t understand the nuanced differences between the conditions. We may think that there’s more – or less – wrong with us than there actually is. Also, we may not factor into consideration the outside conditions that may cause these symptoms to happen.

Why is this? Sometimes it’s because we think we know best. If we can reason it out ourselves, why do we need someone else to tell it to us? Other times we may just not want to take the effort to talk to an expert about the issue. Either way, we’re shortchanging ourselves because we are not getting all the information we need or we’re misunderstanding the information that we do get.

The same is true when individuals request data. They may think that they want one thing when they really need another. They may think that there are worse problems than there actually are. Or maybe they think that the issues they find are either more serious or not as serious as they actually are. In any case, it’s best if they rely on the person that writes a report to gather that information for them.

Self-Diagnosing the Need: I know what I need

Has this ever occurred to you? Someone requests a report. (We’ll refer to this person as the data consumer.) They tell you exactly what they need. Then, when you deliver the report, they examine it and tell you that it is not what they need, even though it meets all of their preset qualifications.

Let’s use an example here. Say the English department asks you to generate a report of all the times that students dropped one of their 100-level courses. You give them exactly what they asked for. They were expecting to see maybe a hundred records. Instead, they see thousands. When they ask you why there are so many, you explain that there are a number of drops that occur anytime from the point that registration opens until the add/drop period ends. Some of these are students switching sections of the same course. Other times, they are indecisive on what they want and may drop and then re-add the same course and section. Through an exploration of the data, you discover that only a small percentage of the drops occur once the semester has begun.

What they meant

Once you explain all of this to the data consumer, they tell you that what they really needed was a list of the students that withdrew from their courses. Why was there such disconnect between the request and the final product? The data consumer thought that they knew what they needed. Instead of explaining the symptoms, they tried to self-prescribe what they needed. In this case, they did not even use the correct terminology; they asked for drops instead of withdraws.

What they should have done is to explain the symptoms. This may have played out as such: “I have a number of students withdrawing from my classes, and I need to monitor it.” With this statement, the data consumer indicates what the problem is, or at least what they think the problem is. They are leaving the prescribing of the necessary report to the report writer. Now, the expert can analyze the situation and offer the best report or set of reports to monitor the situation.

Self-Diagnosing the Treatment: I can write the report myself

We spoke earlier about self-diagnosing. But what about self-treatment? After self-diagnosing a medical condition, some individuals will self-treat as well. If you do not need a doctor for the diagnosis, why would you need a doctor to prescribe and administer the treatment? Many doctors may caution you against self-diagnosis. This does not mean that all stomachaches require a trip to the doctor, but there are definitely some situations in which you should consult a physician for something severe that could be appendicitis or a variety of other conditions. When it comes to conditions like that, you should definitely consult a physician on the treatment plan.

Let’s take this back to our data request example. Say the data consumer had the access and tried to prepare the report themselves. Let’s also say that the data consumer is not well versed with the data content or architecture, regardless of how well they know the front-end system. They may search through the data until they find something that they think is what they need. (You can sense the cringing here.) Is what they found what they think it is? Do they know what to do with it now that they have it? For this scenario, let’s say that they do not.

What could go wrong?

First, they could find the wrong information. As mentioned above, the person from the English department asked for course drops. If that is what they found in the data, it is not what they need, because the real scenario was withdraws after the term began. If there is no one else reviewing the data to determine that this is not what they need, then they could make bad business decisions derived from the bad data.

Second, the data may not be tied correctly together. We have all seen where the lack of a proper connection causes a report to populate many more rows than it should. The wrong fields could also be tied together, inadvertently removing rows of information. In either case, the data is not accurately gathered, and as mentioned above, bad decisions could follow.

I have witnessed both situations. Maybe you have too. The outcomes can be severe depending on the importance of the data, the level of the error, and how quickly it was caught. Once the errors are caught, usually some action takes place to make sure that the data consumer does not try to write the report for themselves again.

The Hypochondriac: I think that there are errors throughout the data

What do you envision when you hear the term hypochondriac? Do you picture someone that is certain that they have ailment after ailment? This term refers to someone with an abnormal anxiety related to their health. You may not know, but many medical students experience something similar. It is called medical student’s disease. As they study diseases and conditions, they may start to think that they have contracted some disease that they are studying. This also can happen when they treat someone with a specific disease. The fear of contracting it themselves can cause them to think that they have it.

Why is this? I think it relates to having enough knowledge to be overly concerned about what they are seeing without the experience to know that it may not be as bad as they think it is. Bringing this back to data requests, the data consumer may be worried that there are a number of errors in the data. They could try to find it themselves, as with the self-treating data consumer listed above, or consult a report writer to gather the data for them. In either case, their overly concerned mentality can cause a number of negative outcomes.

The Works Report: Give me one with everything on it

The data consumer may ask for everything. Anything they can think of, they will ask to have it included in the report. They want to monitor it all so that they can pinpoint all of the problems. There are a number of inherent problems here.

First, with that much data, there is bound to be a lot of noise. It may be hard or impossible to pinpoint what the true data anomalies are between all of the ebbs and flows. If anomalies are found, their significance can also be hard to determine among all the noise.

Second, if you are looking hard enough to find a problem or anomaly in the data, you will likely find it. How important is that anomaly? Is it significant enough to cause problems? Is it worth the effort to fix it? The data hypochondriac may think so, but an experienced report writer can demonstrate the significance of that anomaly amongst the rest of the data. Their experience of how the data operates, is stored, and is modified can give the necessary perspective.

Mixing Prescriptions: A little of this and a little of that

When you go to the doctor’s office, they ask you to bring all your prescription medications with you, or at least a list of them. Why do they do this? They want to get a full view of what your symptoms are, but also any medications that you are already taking to determine if there are any potential interactions between the prescriptions you are currently using and what they may be prescribing. There may be some unexpected outcomes. The same is true with data.

Let’s say the data consumer already has some report they are using for some specific information. This report could be perfectly fine and give them exactly what they need for that purpose. Now they need a second report, and this information is slightly different from the other report. The information however is not completely different. This means there could be some overlap in the data.

Using a similar scenario to the one above, let’s say that report A looks at drops of 100-level English courses for all on-campus students. For report B, they are asking for detailed information about all of the drops of on-campus 100-level English courses. At a quick glance, these might seem to involve the exact same base data. The data consumer does not think to mention report A when requesting report B because it is just a deeper dive of report A.

You write report B for the data consumer. When they review the data, they think there must be some error because the information from report A and report B are not matching. You examine reports A and B and discover the reason for the discrepancy. Report A is looking at for drops of on-campus students, while report B is looking at drops of on-campus courses. Depending on your university, an on-campus student could take off-campus or online courses, so report A may have a different total than report B.

How can this be fixed?

Like when you go to the doctor and are asked to bring a list of medications, it is good to include a list of similar reports when requesting data, especially when they will be used together. An experienced report writer could see that the reports are not much different, but will likely produce different results. The report writer could then consult the data consumer on how the reports might conflict or determine if the reports are indeed supposed to be different. Maybe the data consumer misspoke or misunderstood how the data was connected.

Conclusion

Whether it’s illness or requesting data, it can be tempting to self-diagnose or self-treat. In either case, it is better to go to the expert. Let that expert determine what the problems are and what significance they carry. Let them determine if there are any overlaps or gaps between reports and how best to handle those situations. This is a time saver because time is not spent looking into errors that are not there or that are not significant. Misunderstandings due to a lack of knowledge in the underlying data can be avoided, as can bad decisions that can result from that error.

Bryan Fortriede

Information Specialist for Enrollment Systems Technology and Reporting at Ball State University
Bryan Fortriede is the Information Specialist for Enrollment Systems Technology and Reporting at Ball State University in Muncie, IN (home of Dave Letterman and Garfield). He has worked as an Argos Datablock designer for over 6 years to generate dashboards, reports, and hundreds of schedules.

Bryan lives in rural Indiana with his wife, an Academic Advisor, and his two wonderful daughters, as well as a few dogs, cats, chickens, rabbits, and eventually goats. Bryan's hobbies include writing, gardening, and all things Star Wars, Marvel, and LOTR.
Bryan Fortriede
Share This
X