Disruptions: Data Without Context Tells a Misleading Story

Description: Several years ago, Google, aware of how many of us were sneezing and coughing, created a fancy equation on its Web site to figure out just how many people had influenza. The math works like this: people’s location + flu-related search queries on Google + some really smart algorithms = the number of people with the flu in the United States.

Source: NYTimes.com

Date: Feb 24, 2013


In today’s digitally connected world, data is everywhere: in our phones, search queries, friendships, dating profiles, cars, food, reading habits. Almost everything we touch is part of a larger data set. But the people and companies that interpret the data may fail to apply background and outside conditions to the numbers they capture.  “Data inherently has all of the foibles of being human,” said Mark Hansen, director of the David and Helen Gurley Brown Institute for Media Innovation at Columbia University. “Data is not a magic force in society; it’s an extension of us.”  Read Rest of Story 


 Questions for discussion:

1. How can Data without context be misleading?

2.  Will Big Data eliminate error, uncertainty, and risk?


20 thoughts on “Disruptions: Data Without Context Tells a Misleading Story

  1. Navreet

    When one is making predictions like these, one needs to be absolutely certain that the data is reliable. Experiments can be misleading when the data is not collected long enough and it can be misleading. It is highly necessary to collect data from not one but many highly reliable sources and test it accurately before making bold public statements. Researcher also needs to see other side of the story and go in depth before making assumptions. Collecting a lot of data is not enough; it needs to be made useful by intelligence and wit. Determining the appropriate way to utilize it depends on the researcher and the given problem. There will always be some amount of margin of error but the prediction should not be very far away from the truth. This is where data mining and information systems come into play. Accuracy and precision of the big data plays a huge role in arriving at a conclusion. Data can be deceptive and it should be kept in mind that, “Little knowledge is dangerous.”

  2. Ryan Orr

    The article discusses how data without context can be misleading; without context data can put data analysts or anyone reading the data onto tangents that can be exaggerated or just false. Without proper knowledge of what is going on researchers can be mislead and can fault the entire project. Big data is an interesting subject as it is controversial; similar to the misleading context, big data must be approached reasonably. Big data will help to eliminate risks as smarter software and hardware systems are able to extract more and more relationships. However if all that is considered is the cold facts of Big Data we run the risk of not considering context of it all. Both context and Big Data need to be approached reasonably as they are both a big part in research and development.

  3. Brendan Schlossberger

    When data is collected it simply tells you what happened and where it happened. It leaves out the most important W; Why. Without the why this article shows how the data can lead in the opposite direction you thought. Data without human assistance and supervision and ultimately end up being useless. This is why data can only ever be a supplement to human reporting and analysis. Big data certainly cannot eliminate errors, uncertainty, and risk because nothing can. No matter how much technology we use or how much preparation we have there is always going to be some degree of risk and uncertainty. As the article has shown sometimes big data can actually increase the error and uncertainty of an experiment. Big data will have to be continually analyzed and we will have to continually try to better our understanding of what the big data is telling us and what it is not. One major problem of Google’s Flu tool is that when they developed it people began to trust it immediately and didn’t think it could possibly give out the wrong analysis. This over trust of technology and data is dangerous because of the uncertainty involved and the restrictions with data need be seen.

  4. Brandon

    Data without context can be entirely misleading. When ever someone starts an argument, statistics and data are usually the first thing to be skewed. However, where do we as a society draw the line for what we allow people to skew? Do we continue to lower our moralities to the point where we almost inadvertently allow people to use anything to prove any point? By continually saying nothing to people who make these ignorant assumptions we are almost encouraging them to do it. Therefore, how do we change what is allowed? Simple, we need to start using data and statistics to portray what they were created to portray. Otherwise, we may find ourselves making decisions based on half-truths that will create major problems for us in the future.
    Now, will big data eliminate error, uncertainty, and risk? I would answer with a big NO. Living in the digital age, I would have to repeat the quote, “To error is human, but to really screw things up takes a computer.” Big data is no different. Even though big data does offer more detailed information, and it will be a game changer, it does nothing more than what people extrapolate from it.

  5. Logan Reiter

    Data without context can be misleading for a number of reasons. If there is no context for the data then an individual can put it in whatever context they want and can turn it into something that is completely false. This type of data can look better or worse than it is depending on the context. I dont believe that big data will eliminate error, uncertainty and risk. How can it possibly do this? No amount of data can do this. There will always be some error and uncertainty. And with both of these things comes risk. It can help to decrease the error and uncertainty but it can never be completely eliminated.

  6. Dallon Martin

    Data without context can be very misleading, because you don’t really have any other knowledge of the data. It can lead people to thinking that the situation is either better or worse then it actually is. Especially with the vast amounts of data out there this constantly increasing. Will big data eliminate error, uncertainty and risk? I do not think so. I feel that computing systems and research can provide great amounts of basic statistics but in term of what Google did, i think that it is taking it too far and that when doing this type of research they have to look at the context behind the searches. The main thing i think is that we need to have an understanding of the context of data, because like it says in the article data is an extension of us.

  7. Colin Stevenson

    Data has become a key part of our society. Everything we do practically becomes part of a data set that can be used by someone to target us for marketing etc. However, data only reflects a certain aspect of the overall picture of life. Without recognizing all the factors and variables that contribute to the data it can easily be misinterpreted. In the case of Google using algorithms to determine how much of the population has influenza this fact is evident. With the proliferation of media and social media the coverage of the flu this year was extensive. Many people who did a Google search for the flu may not have personally had influenza but were looking for information. Associating a search engine with actual flu numbers could be somewhat of a hasty generalization. As Google refines their algorithms and adds other factors to their calculation it is possible they will come closer to real life results.
    I do not believe that big data will eventually eliminate all error, uncertainty and risk. Data is just a tool, a set of information. It is up to a human or in this case a human created program to interpret the results. Data and humans all possess the same fallibility, so while this use of data may reduce to some extent these issues; it is impossible to completely eliminate them.

  8. charles chow sau pui

    Nowadays, the society collect load of data. If data are without context, it would cost miscalculation in data collection. If corp use those data to build up their products ,the products could satisfy market needs. Therefore, i feel data without context must be misleading. If the
    Big data cannot eliminate error,uncertainty and risk unless it is with context. Data without accuracy could not reflect the real and realistic situation. Therefore, the data useful or useless depends on the quality but not the quantity. If we want to reduce the risk and uncertainty in data, we have to get the accurate data.

  9. Abraham Sarthak

    Data nowdays has become a useful part of our society , in business people use it to predict what would be the reaction of the people towards a particular product .Government use it to estimates problems in the society and sometimes to compete in the election.But is the data we collected relevant , this is a big question .How do we get to know that the data we have is of any use or not , or is it significant or not .Sometimes companies do research using the data to enter the market , but still they fail in that .Thats because there data is not relevant and accurate .In accurate data can lead too huge losses and misconceptions .Nowdays data is everywhere ,but how do we use that data is the problem , we need to figure out the way to identify which is the best data for us .So the data without any context could be really misleading as in the examples of the New York university .The professor came to a result without investigating the validity of the data. That tells us how unpredictable can be data and before coming to any conclusions we need to identify the significance of data.

  10. Ollie Sherwood

    Google were wrong to have even conducted this study because at the end of the day, who are they to research how many people are affected by the flu or have the virus. This type of research should be done by professional within the medical field and not an internet company. How can data without context be misleading? Well, as this article shows it can be very misleading. Google ” calculation” of the number of people with the influenza turn out to be twice as high as the national research and it also turned out to be wrong. Without the true background behind the information the research could mislead many people into taking preventative action that would be unnecessary. The context and background behind the data can dramatically alter the outcome and there is no way that a computer or computer system could truly understand the context behind the information. Will big data eliminate error, uncertainty and risk? personally, i do not think so. I feel that computing systems and research can provide great amount of basic statistic and that it but in term of what Google did, i think that it is taking it too far and that when doing this type of research they have to look at the context behind the searches etc.

  11. yuxuan.hou

    It is an information society. Data and database information system help to provide information to many kinds of organization and business trade. Data with context can help information to pass to people more quickly and convenient. Data with context is good for improve the characters of data such as veracity, timeliness, and transitivity. However, sometimes, data without context is useless because data without context can be misleading. Data needs context especially for the context background that has the ability to link the whole information. If we can just get a part of data without context, we are very easy to misunderstanding the real meaning of the data provides. Accounting to the article, I agree with the words of Mr. Mark Hansen said. That is “data is an extension of us”, we gather many kinds of information and create the datasets and databases. If we lack of necessary context, the data wound become useless and do the opposite effects. I think the big data is very useful because it involves much information. However, the huge scale, the worse error, uncertainty, and risk are exist. After all, each coin has its two said. People should maintain data and database system in order to make sure they are accurate.

  12. Mohadese

    Data and Big Data help to provide the information for different field and various career and organization. Data would help decision makers to have more effective and reasonable decisions. Data would reduce the uncertainty, risks and errors. For example researchers via technology and big data would access to some sources that are critical for their researches and help them to achieve more appropriate and useful results. In organizations also data would help decision makers to access the information, which reduce the risk of bankruptcy in the high competitive market environment.
    Data and information could reduce the risk and uncertainty only if they have critical, relevant and beneficial content and context. Gathering many irrelevant data would be just loosing time and would result irrelevant and high risks decisions. Also, it is very important who interpret the data and how wisely they do that.
    Overall, the goals of gathering data, tools to access the data, a sample or population size for collecting data need to be clear. Also members in the process of gathering data and information need to know what is the context and content of the data and how much data and information are related to what they need. Moreover, people who use the information and data for making decision need to understand the context of them and use and interpret them wisely.

  13. Antonia

    Data without content can be very misleading you may get one result but with all of the content you may get an alternate result. You can’t base statistics on the amount of Google searches. Everyday I search for things that I am curious about. Anyone curious about the flu and how to prevent it may have done a Google search, that doesn’t mean that they had the flu. Without having all the data around the information you get errors and may mislead people. I think that Big data can be very useful but only to a point, it will help clear things up and reduce risk but we still need that human element to review all the possibilities that technology could have missed to make sure that the data is accurate.

  14. Kristian Rochon

    Data without context can be misleading and potentially dangerous. With such enormous amounts of data being inputted every day, it is increasingly getting easier to access data in almost any field. This is no doubt in benefit to a vast amount of individuals and businesses, however it also poses a large risk. As described in the article about elevator use, the class relied to heavily of technology, and didn’t realize that one of the elevators was broken. This is a simple and honest mistake, but in fields such as medicine, engineering, construction, ect. where perfect specifications and test results are absolutely crucial, a simple and honest mistake due to low context could have severe consequences. With careful implementation and practices big data will in fact help eliminate error and risk. There will always be error caused by machine, but also there will also be error caused by human. Cases when technological error most likely outweighs those of human error. With smarter software and hardware being seen everyday, it would be naive not to think that big data could reduce or perhaps eliminate risk. If humans practice careful habits and work together in harmony with technology, technological advances and big data will continue to increase uncertainty.

  15. Crystal Smallface

    Context with data is organized data, so data that is not organized is useless which in turn can be misleading. Depending on who is using the data or what it is intended for would determine whether or not the data has contextual data that is useful. When datasets become so large the data within that dataset can be seen as useless. Why such large numbers? Would this not increase error? Uncertainty? And risk? I believe it does and will increase these obstacles instead of eliminating them. We have studied research methods which indicate that as datasets become larger and larger the data can be construed which becomes insignificant over time. As these datasets or big data become even bigger the context is no longer there as when they are smaller datasets with contextual data which is useful for studies. Therefore, I believe that Mr. Mark Hansen was right in saying that “data is an extension of us”, because we create the datasets and when we let it get out of hand to become so large that we can longer use it anymore and no longer find it useful, then we as humans are in error.

  16. Katie Guccione

    I think that the amount of data and the rate at which it can analyze the data can be done much quicker on google, but as seen it is still not as accurate as having an actual person do the tests or research. I think that Big Data and computers have their time and place, I think that they are very useful but as seen with the influenza example, they need to coupled with human experts to make logical sense. With the influenza example it shows that the data was slightly off because it did not take into consideration all of the data that needed to considered such as the season, the actual number of people etc. When this data was released people worried and immediately took this data to heart and acted in a way that was likely not beneficial to their health. I think that Big Data, when used correctly, can help to aid a problem or possible facts that lie within our society, such as flu season, but I think that it needs to coupled with the data that is collected and tested by people who are experts in the field and who have the ability to test the data properly. I think that Big Data and computers are very helpful, but they cannot be completely reliable in eliminating errors, so there will always need to be human input to make sure that data is as correct as possible.

  17. ibrahim.akinola

    1)Data without context can be misleading because without background or additional information, you cannot exactly tell the accurate result of an experiment, also the kind of data acquired can tell you a story about a specific finding but when you gather more information and add context to that finding the result will be different. To back up this point ; based on the sensor example we can see that the researcher was feeling confident with her result until she was informed by the security guard on whether she consider the fact that one of the elevators had broken down few days ago before her research. Based on this additional information by the security guard, we can completely see that this bit of context had changed the research result and has placed doubt on the researchers mind on her first result; thereby making her to carry out another experiment. It is important to understand the value of context and make sure to gather background information when carrying out a research so as to arrive at the accurate result of an experiment.

  18. Marc

    1. The same as it always has been… traffic jams in LA and annual snowfall in Lethbridge have had interesting correlations in the past, so the data looked useful, but any human would struggle to find out (1) why that’s true and (2) why that would be useful.

    2. I love the Mark Hansen quote, “data… it’s an extension of us,” as are the algorithms and the human actions that create the data. People get wrapped up in data like people get wrapped up in money. Money doesn’t actually matter, except as a means to get what we need or want. Data doesn’t actually matter, except when it’s used to close a gap between human action and human understanding.

    The scientific method is used to reduce error and uncertainty, and part of that is collecting data. We can’t forget in the world if Big Data, that it’s only part of the scientific method.

  19. Brogan Mueller

    Data without context can be very misleading because the data can be portrayed in a certain way which leads people to believe the situation is better or worse. As seen in this article, the number of influenza infected people was double what is should have been, based on data without context. This data made the flu outbreak seem much worse that it was in fact. Big Data and computers do not have the ability to eliminate all error and uncertainty, because they can not always have all the information. The Big Data collected from college students using the stairs vs. elevator seemed very accurate, until the security guard said that the elevator was broken for a few evenings. This context alters the results of the data collected. There is no way that a computer could know this information before the test in order to give results that are completely reliable. I believe such technologies as Google and computers with light sensors have the potential to collect data and give useful results. However, in order to have context that makes this data reliable there will always be a need for human interference.

    1. Josh

      I liked how the article pointed out that this is not a new issue or phenomenon. Misused, deceptive, and even outright dishonest “data” is used every day to indoctrinate and control populations around the world. History textbooks and government propaganda (often times one and the same) are filled with statistics and data points that prove theories or philosophies that other data sets attest to be false. Incarnations of Big Data have been used for decades, to even justify opposing sides of the same argument. This problem is larger than most people realize. Our perceptions of other cultures, our understanding of historical events, and even the fables we accept as truths are generated and molded by massive collections of often hand-picked or distorted data. How we view the modern era is literally shaped by the findings, analysis, and presentation platforms of research data. A simple, specific, and hopefully less-than threatening example, the entire US war on drugs was founded on studies and data sets completely misrepresented and stripped of contextual evidence and essential facts. But, the scary part is that that the same could be said about the next news story to come across your screen. The use of Big Data has great potential in reducing redundancy and discovering inconsistencies, but it cannot eliminate errors. If anything, it will only perpetuate them further and faster.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s