The Information & Social Media Industry needs to step up to help us differentiate higher quality submissions from poor ones. Providing a rating for each post will help get us there.
Not all information has the same value. This is something that we all inherently know. However, when it comes to social media it is difficult to differentiate the quality information from the bad. An independent industry association is needed to provide some reference to help us determine the quality of the information we are subjected to every day.
To understand how we might address this chasm on social media, we need to explore how we validate information we receive in our physical lives. Let’s imagine that you are on an elevator and two women are talking. The daughter of the shorter woman in the elevator has an overbite and may need orthodontic treatment, as she is explaining to the taller, blondish woman. The shorter woman heads up a department in your company that you’re interested in and your Aunt is an excellent Orthodontist, so you are now very interested. You listen carefully. You decide to file this piece of information away for an introductory conversation that you might have with the department head in the next few weeks.
Of course, it’s not always that easy to determine the value of the information. In this case, you know who the shorter woman is, and you know that she is an expert on the primary subject of the conversation – her child! You also know that she could be important to your career. Each of these factors make this information valuable and believable, but how?
You brought some information of your own to the elevator: you recognized the shorter woman and you were knowledgeable about her role in the company. This makes any information she imparts important to you and, frankly, to many others at your company. You also have a high degree of confidence that the information that she is sharing is legitimate. After all, it would be very unusual for anyone to board an elevator and start lying to a co-worker about a problem her child is having. She also holds a position of great responsibility in the company and you believe that makes her a more responsible and truthful person. Lastly, you have concluded that the information that she is presenting is actionable. That is, you can use it as part of a future conversation. Depending upon her level of satisfaction with the Orthodontist advising her, you may even have an opportunity to throw a little business your Aunt’s way.
Whoa! You are a pretty quick thinker to have figured this all out in the elevator ride to your floor. You politely shoulder past the two women, smile engagingly (but not too much) and head out to work. But, what if, on the way out, you realize that the shorter woman isn’t who you thought she was. She is actually a different woman who works in the office next to your future Department head and the two women were talking about the Department head and her child. Does the information now have the same value as it did a scant few seconds earlier? Is it as reliable?
Let’s examine what we know about how you judged the information. For tracking purposes, we’ll award two points to each factor. In the first scenario, you recognized the department head – you were assured of her identity. That’s worth two points. In the second scenario, you now realize that the conversation was only about her (she’s probably been in her office for an hour, already). However, you still feel reasonably sure that she was the subject of the two women and that the speaker has direct access to the department head. Let’s award it a single point.
How about the quality of the information? In the scenario where the department head is speaking, you are aware of her sterling reputation. She has a position of great responsibility in the company and you have attended company meetings wherein you have judged for yourself her sincerity and ability. In other words, she hasn’t steered you wrong before and you have no reason to think she’s doing so now: 2 points for reputation. She’s also talking about her own child – a subject where she is undoubtedly, the absolute expert: two points! She is also relating a condition explained directly to her from the Orthodontist, so she is a direct witness to the conversation: two more points!
In the adjoining office scenario, however, you have no idea if the woman had even met the department head’s child and you can’t be sure where she got her information. Did she talk to the department head or her assistant? Did she simply overhear a telephone call? Is she simply making up a story for the entertainment of the taller woman? You have no idea. You suppose you may have to ask around to verify that the department head has a daughter with an Orthodontic issue or some other question to help verify the conversation before attempting to talk to her. This means the information is of less value, because there’s additional work required to validate the accuracy of the data.
How do we score this? It’s unlikely that the shorter woman is making up the story out of whole cloth and any exaggeration is also likely to be limited. Unfortunately, the information is second-hand, at best, and may even have been the result of an overheard telephone call wherein only one side of the conversation is heard. Let’s be generous – we’ll award the woman one point because she appears to be a trusted employee of the company (she works on the executive floor, doesn’t she?), a point for information about the child and a point for information about the orthodontic need.
Can you make use of these data? Is the information resulting from that data, when interpreted, useful to you? Let’s look at the score. When you were sure that the shorter woman was your future boss, the information became relevant. Additionally, both the authority and the quality of the information were rated very highly. These factors were good for eight points. In the adjoining office scenario, the best we can muster is four points. These results are summarized in the table below:
This example of evaluating the quality of information we encounter is something that each of us does every day. We generally don’t use math or a strict process to score the data and the useful information that we might glean from it, but we do make value judgments about that information and choose to ignore it, investigate it, or act upon it, based on that value judgment.
Over the past 25 years, our access to data and pre-digested information has exploded. The internet is awash with opinions, facts, data, propaganda, and, yes, fake news. We, as consumers of that information, have very few tools that enable us to effectively value the worth of that information. Just think how many times you’ve seen a post on Facebook or some other social media platform where a friend has reposted a news story or picture or saying that your other friends fervently discussed only to have some spoilsport post that they checked it out on Snopes.com (or some other truth-telling site) and found that it wasn’t true.
It becomes even more challenging when real facts are presented in a way that makes them seem like they are saying something else. I have two favorite examples here. In the first, a company has experienced 4% sales growth for 5 of the last 7 years. In the sixth and seventh year, it actually saw 6% sales growth. This is a success story, but not earth-shattering. How can it be made more dramatic and attract investors? In the two charts below, the exact same data is presented. Which looks better?
The one on the right, of course! On the left, Sales are graphed from zero dollars. On the right, the graph starts at $140,000. In the leftmost graph, it’s nearly impossible to see where sales increased from 4% to 6%. In the one on the right, it is much more obvious. On top of that, the right graph makes all the years look better! The only thing that changed is the way the data were presented.
My other favorite example hearkens back to my youth. In those days, the battle against tooth decay in the nation’s youth had become a great suburban initiative and one of the toothpaste manufacturers would routinely state that “2 out of 3 Dentists surveyed” recommended their brand. It struck the younger me that there were a lot of ways this statement could be true. The manufacturer could have surveyed 300 dentists or a thousand Dentists and received 200 recommendations or 667 recommendations, respectively. Or, they could have asked three Dentists and when they got two recommendations, stopped and said, “Hey, this looks good! Stop the survey – we’ve got two out of three.”
Now, to be clear, I have no idea how many Dentists they surveyed, and I have no doubt that they conducted a proper and statistically significant survey. There was probably an asterisk following the claim that referenced the particulars about the survey in small print at the bottom of the screen that I couldn’t read no matter how close I got to the TV (which I was going to get yelled at for, anyway).
My point is, I can’t tell anything about the quality of the information from the simple claim of “2 out of 3 Dentists surveyed.” It could have been that the three dentists were the Marketing Director’s Uncle Mickey, Cousin Sheldon, and that Dentist on the East side. I do understand the purpose, however. It is clear: 2 out of 3 Dentists recommend this product, so you should be assured that it is a high-quality product and you should buy it.
We’ve not talked a lot about purpose. Once you realized the shorter woman wasn’t the Department Head, what was her purpose? Could she have been telling a story about an important Department Head as a way to gain elevated status in the eyes of the taller woman? She may have wanted to be seen as a confidante of the Department Head. Similarly, what is the purpose of a social organization (or PAC or bot or political party or historical society or business or industry group or union or tourist destination, et cetera) posting a story on social media? Are they trying to educate, to find fellow believers, to get us to think a certain way, to get us to distrust another group or something else? Are they trying to divide us or unify us or make us so mad our stomachs will ache, and we’ll buy more antacids? What about news organizations? Are they trying to inform us or are they simply trying to get more clicks or more viewers, so they can sell more ads?
Product Marketers have spent decades trying to make us feel uncomfortable about body odor and teeth whiteness, jealous of others with more expensive automobiles, and worried about falling and not being able to get help. Some of these are legitimate concerns, but some of the methods used manipulate our most basic concerns are questionable. You are being manipulated every day. And, now, with the internet and social media, it’s not just to sell product – it’s to sell ideas.
At least with products, we have all grown up with the idea that we need to buy things to have a successful life. Based on our personal preferences for status, appearance, security, and other concerns, we have learned to make decisions that meet the compromise of sufficiency and budget. However, most of us have not had to deal with the constant, highly inconsistent, deluge of information that is pushed at us daily, especially on social media.
A lot of things get said, or written, or posted on social media that just don’t provide us with enough information to validate the quality and truthfulness of the information. Sometimes, the information is accurate, but it is incomplete or presented so that it leads us to an inaccurate conclusion. Looking at the sales growth for XYZ Corporation, the data are absolutely accurate – it’s just that the rightmost graph is designed to make you think that XYZ is going gangbusters when their growth is much more modest. This isn’t a lie, but it’s not something that lets you make a visual comparison with a chart that doesn’t fool around with the starting amount.
You don’t even need a chart to play this sort of game. Words will do just fine. How often do we find that comments by people in leadership positions are taken out of context and twisted to serve the commentator’s narrow purpose? Especially egregious are headlines in newsfeeds. I can no longer count the times that a friend posted a headline from one of the minor news agencies (e.g. not NBC, CBS, ABC, FOX, CNN, or MSNBC) and I followed the link only to find that the headline was misleading, false, or over-stated. How many of us simply read the headline and react? Even with the majors, I find that people repost with their own interpretation that doesn’t quite fit the story. We should all go to the link and read the story and make our own decision, especially before reacting. And, if we don’t have time to read the story, we should simply not react.
But can we be that disciplined? Is it even reasonable to ask of people in our ever-more-complex and time-constrained lives? Maybe we could use a little help from the social media giants. What if we asked Facebook, Twitter, Google, and the like to share a measure of the potential quality of each post – a believability rating of sorts? This isn’t a crazy idea. Many industries have turned to self-policing on content.
Information quality has many facets, as we have seen. The identity of the person or organization providing the story, their reputation for truthfulness and fairness, the way in which the information is presented, the authority they bring to the data, and their purpose in sharing the information. It’s not unreasonable to think that these factors can be used to rate the quality of a submission.
Social media companies are in dire need of content rating. Social media companies have access to a great deal of information, about their users, their posters, and their data partners. They could be asking users to tag posts that they believe are false or misleading. They can set standards for grading the verification of the identity of individuals and organizations who post messages. They can compare the messages to other messages that have been proven false on truth-telling sites like snopes.com. “Big Data” and its manipulation into useful information is what makes social media companies so successful, so why not use it to assist us in making informed decisions about who and what to believe?
Social media is not the first media industry to be challenged to provide some sort of rating system for its content. In reaction to calls for decency in the motion picture industry, the Motion Picture Association of America instituted a rating system in 1930, while still a youthful technology. Similarly, the ESRB (Entertainment Software Rating Board) was formed in 1994 as a non-profit organization to rate gaming software. Both bodies came into being after some thrashing about following introduction of the new technology in question. However, a need was identified, and, in these cases, industry self-regulation came to the rescue and has been a reasonably successful way to protect people and help them to make informed judgments about how to spend their money and time.
The time has come to provide a believability rating on each post, especially those from organizations or persons with unverifiable identity. Each such post should be evaluated using a process similar to what we did for the elevator conversation. Factors should include (1) To what extent can we confirm the poster’s identity? (2) Have they posted quality information in the past? (3) Are they knowledgeable / authoritative about the subject matter and / or do they quote authoritative sources? (4) How is the information presented? (5) What is the speaker’s purpose for posting the information? The answers to these questions should be tabulated and presented as a believability rating along with the content.
Armed with this rating, we would at least have a chance to decide which posts are worth our time to review further and which ones are better ignored. And, maybe, at the same time, ratings would expose those bots or persons who are deliberately dividing our nation by posting false and misleading “information.”