Predictive Analytics is essentially a way of data mining that allows for the creation of predictive hypotheses’ of group behavior based on trends in the data stream. Simply making the prediction is hands off. You as the data miner have a guess about how a group may perform in certain situations and look for ways to apply it. This is the most difficult part, and the place where the most ethical dilemma comes into play. Once you have a prediction, how do you apply is where you decide who's best interests are at stake. I want to think about how this method of data mining can be both influential to change and production, as well as unethical if mishandled.
Let's think about the possibilities that could take place if people started to use trends in order to make organizational issues more apparent. Could the data serve as a tool for understanding how to more effectively coordinate procedure within large organizations?
What if the The department of transportation starts invoicing all of their customer interactions throughout the day. How many people are waiting in line at the busiest times of the day? What is the reason for their visit? Why are they coming in at this time? The DMV could then compile this information, and begin to make predictions of when there will be the most visitors and on what days. This could allow the DMV to make a variety of adjustments within their organization to restructure the flows of visits into a more manageable manner.
But how will this be used? Will the DMV make adjustments in the way they assign licence renewals? Will there there be a shift in times that you can get certain permits? We can't assume that the redesign of such an archaic system can be fixed with data alone. Here lies another big problem with big data. There is the possibility for new and fresh idea on how to create better mediated flows of communication on a daily basis. But are we putting too much trust in the numbers? We must consider how correct these predictions are and how much value should we place on their reliability.
By putting too much trust in an algorithm of prediction companies may encounter ethical dilemmas that could possibly compromise their integrity. Companies might start by taking models of predicted behavior and applying it the issues that address a problem or goal within the company. But the overwhelming potential of predictive analytics may lead the way for more automated predictions about customer and group behavior. Though automated decision making and algorithmic sorting makes things more efficient, there is somewhat of a negligence that come from this automated processing as well.
For example Insurance companies started to take the approach of data mining in exchange for rebates in the form of Snapshot devices to be installed in customers cars. Though in the short term the customer is getting a deal on their insurance in exchange for better driving, think about the long term analysis of these devices, and how they will play into the future of insurance all together. If all of the data from these “snapshots” of drivers, these companies can then start to make deliberate attempts to negotiate how people are insured based on behavioral predictions. But how universally accurate are these predictions? If a trend is found in the data that allows for better insurance distribution are the organizations willing to compromise the rights of the individual for the sake of the whole. (in my very drastic example) If there is a prediction that women drivers 5’4-5’8 with red jeeps had bad driving tendencies. Is the organization willing to raise the rates of this demographic in order to account for the behavioral analysis. Though this may be more efficient and reliable for the insurance company that is making adjustments where there is potential for reliability, what could this say down the line about the ethical side of the argument. Say that you are a very good driver, but you're 5’6 and drive a red jeep, your insurance rate is then raises because of recent finding; can the insurance company justify your raise rates based off of a prediction model. How could this new concept of prediction change the way we purchase and exchange effect the way people interact within the market?
Though I have addressed two potential extremes for predictive analytics, there is a wide range of possibilities for the process. Predictive analytics can supplement models of research within the medical field, enhance customer service, create small group marketing tactics, there are some many beneficial methods of applying the data to real life. But no matter the application there is still one side that is having fewer and fewer controls over the amount of data that is pulled from their assets. Unless you are in a very high position, you have no choice but to conform to the data pulling structure that has been built into our current communication systems. So there are two important concepts that organizations need to incorporate into the way they communicate with their publics. First is the concept of transparency should be front and center when It customer service, meaning companies should be upfront about how they are using the data they are cultivating. Medical practices are now using predictive analytics to improve current health care costs and practices, for more effective medical attention. As medical organizations move into a more data driven structure, they must keep their patients informed about how they are making their predictions and how it being applied to them.
Second organizations should be ethical in how they are getting the data in the first place. Apps like Facebook messenger and flappy bird, are among several that have been known to pull data from the user even when the app is not in progress. One of the main components of ethical business, is the practice of allowing for free will. Though the terms and agreements of these applications reinforce their intentions, there is a lack of freedom to choose how your information is being extracted, and therefor this is unlawful. Organization that choose to partake in this new strategy, should be ever cognizant of the rights of its subjects. The predictions and their applications will be more effectively integrated when the organization creates an open dialogue with its publics about the goals and results of their data mining.
Lastly there is a certain responsibility that belongs to us, the users. Though these concepts dealing with big data are new and somewhat complex, it is incredibly important to stay educated on the processes of big data. The Internet is regulated in a completely different way than most traditional media platforms, we as users must also seek to inform ourselves about data laws and power the struggle this creates between user and administrator. Big data and the online world are two very new and different fields in the modern world; but they are both so entwined within our day to day lives. That makes it unacceptable to be uneducated in the way they operate. Everyone that uses the Internet should have to be educated on the concepts of big data. These are the concepts that are taking over some control of our lives and behaviors. Though we hope that organizations will be open and transparent with their users, nothing is promised so users must stay educated.