I keep reading news articles about the inferior nature of email, instant messaging, web forums, and other forms of non-verbal communications. I have finally found the source of one of these articles and have found some very interesting details out.
The news article I’m referencing is “Your Emails Aren’t As Funny As You Think”:http://www.digitaljournal.com/news/?articleID=4787, which is based on the research in “Egocentrism over E-Mail”:http://www.gsb.uchicago.edu/fac/nicholas.epley/Krugeretal05.pdf. The research study is mentioned in various places throughout the internet, but I need to stay focused if I want to tear both the research and the article apart.
The general belief I’ve gotten from the research is that when you’re communicating verbally, you can read body language and hear tone and voice inflection to get a better idea of the true meaning of the message – sarcasm, annoyance, flirtiness, humor, etc. In email, for instance, there is no body language and no tone to be read.
h2. The flaws in the research study
I have found that the study has made some decent points, but seems to very specifically avoid certain aspects of email communication that would likely have helped show email in a better light.
First off, I’m not talking about study 5 here. This is a study where they took preselected Jack Handey quotes from Saturday Night Live to see how often people’s humor meshed with somebody else’s, when the message was shown in video vs. email. I’m just not sure the point of the study when the humor is not from the actual person. I don’t think anybody will question that humor is better spoken by a practiced comedian than sent in email by an anonymous stranger, so if that’s all they meant to prove, I think they wasted their time.
h3. Most of the studies relied on pre-written text
Studies 1, 2, and 4 had people deliver a certain sentence word for word. These were to be delivered in a certain tone: sarcastic, serious, angry joking, maybe a couple others. These sentences were not divulged (I think that without these sentences, it’s very difficult to determine the validity of the tests), so I can’t comment on how useful they may have been… but think about this: if the sentence is “Your mother is such a bitch for making you pay for your own car”, it’s totally ambiguous whether the speaker/emailer is being sarcastic if all you do is read the text. If I were trying to say that sarcastically, I would probably start off with “Oh yeah, your mother is just such a bitch…” and end with “God forbid….”. Maybe even laugh.
With vocal inflections, I can certainly make my meaning clearer no matter the full text I read. But in email, we rely on things like context to make a message’s intent more clear. Even in normal vocal conversation, we change our message to indicate a different tone, though it’s not nearly as necessary.
The problem I have with these three studies is that the email group of the studies wasn’t allowed to modify the message in any way. No bolding, italicizing, capitalization, smilies, or laughter (LOL, ROFL, hehe, etc) could be added. This, in my opinion, makes for very fallible results. Look at the next section for more details….
h3. Email and IM communications most certainly can have a tone.
As I just said: there are things like bolding text, italicising text, CAPITALIZING text, and using smilies (:D, :), =), ;), :-P, etc.) to get the point across about your intent. Read the following two sentences:
I think the direction our company is headed is absolutely correct, and I’m glad to be a part of it. I won’t be looking for a new job anytime soon.
I think the direction our company is headed is absolutely correct, and I’m GLAD to be a part of it. I won’t be looking for a new job anytime soon ;).
It’s not crystal clear, but the second sentence has a different tone than the first, and will give people a better chance of “getting” the real message (“I hate the direction we’re going, and I’ve already posted my resume to monster.com”).
Now consider that some email programs (and most forums and IM clients) even have graphic smilies for showing even more specific emotions. Add an eye-rolling smilie (
) to that prior message, and I doubt many people will mistake the tone.
The interesting thing to note here is that the study admits that smilies (referred to as “emoticons”) might help send the right tone, but they claim that won’t make much difference.
The way they “prove” this conclusion:
* Some smilies are ambiguous, such as “;-)”. Is that a happy response? Flirty? A “just kidding” response?
** This is true, but the same is true of real life! If somebody says “I like that shirt” and winks, I won’t know if they’re being friendly-but-wierd, flirty, or just kidding.
* A “follow-up” study was done that allowed emoticons, and found that overconfidence wasn’t affected between the emoticon-users and non-emoticon-users.
** Um… what emoticons were used? What tones were available? What size was the group? In other words, without showing specifics about that follow-up study, how can you use it to dispute emoticons?
** Along the same lines, what was found in that study? Overconfidence may not have changed between the groups, but did accuracy change? If accuracy went up, the level of overconfidence may well not have changed, but that would very nicely prove my point about email tone!
h3. Emailing strangers will lead to more misinterpreted messages than emailing friends or even coworkers.
Study 3 “proves” my above statement incorrect. But you see, here’s where I get into context again. Familiarity is all fine and dandy, but if users can’t bold, italicize, use smilies, or otherwise convey context, then you’re not testing their ability to communicate!
In email, if I’m sarcastic, I’ll add a smilie or “p’shaw, whatev” or something. In fact, to different people my sarcasm will be different. To a good friend, I can say “Oh dude that is totally so like awesome man! I’m so stoked about it, sign me up, brotha!” My friend will know I’m being sarcastic because I don’t normally IM/Email like that. My father, on the other hand, whom I speak to more formally, won’t know how to interpret that message.
Surrounding context is even more important in my opinion. Read on…
h3. The overall conversation’s context isn’t even evaluated!
Context is incredibly important. If I’m asked to convey anger in a single sentence, I don’t know how I’d do it in a reliable way, other than “I’m really angry” (and note that this too can be interpreted many ways depending on context). Measuring the results of effective communication based on a single sentence is simply measuring the wrong thing. They’re seeing how well people can convey an emotion in a single, context-free instance. They then use those results to claim that email is inferior to verbal communication, even though in normal communication, a huge amount of interpretation is based on the context of the conversation.
If my wife and I have been joking around and she suddenly says, “You’re such a jerk!” I’ll know she’s joking, even if her tone would suggest otherwise. The same sentence, spoken similarly, could mean anger, hurt, frustration, or nothing at all. All depending on the surrounding conversation.
h3. The study is flawed by the nature that the participants knew what was being studied!
This may be a controversial statement, but I believe it’s true. Let me explain. If I yell out “You ASSHOLE! I’ll kill you!” in the meanest voice I can muster, and your options for my tone are: angry, sad, sarcastic, or joking, you’ll probably pick angry. If you know me well enough, though, you’ll know that in real conversation, if I yell that out, I’m joking. So if you and I know we’re being tested for tone, I’ll speak the way I expect the average person will understand me, and not the way I would speak in a normal conversation. In fact, I’ll likely exaggerate my speech (sarcasm: “OOOOHHHHH I’M SOOOOOOOOOOOO EXCITED”) to “get the test right”.
In a normal conversation, my angry tone is barely different than my serious tone. I don’t yell; I rarely even swear (out of anger, at least). If you had to interpret a real tone from a real conversation, you would not have nearly as easy a time, and friends and family would have a huge advantage over strangers.
My point is that the experiment should have measured tone in a different way. The speakers/emailers could have been told to create a message as if it were to various family members or friends, for a specific scenario. After typing it up and speaking/emailing, they would have been asked to rate each the overall tone of the message. It could have been presented as some study in effective communication. The recipient of the message would be asked various questions, some related to the study, most not. How well did they get their point across? Was it too wordy? Too brief? Etcetera.
h3. The Implications section of the study is flawed
So we’ve got, in my opinion, some flawed results. The idea that we communicate in a way that is egocentric makes plenty of sense, but most of the other conclusions I’ve seen are, at the very least, misguided. But the final section blows me away.
The claim is that other forms of nonverbal communication are going to be as bad as, or worse than, email. They explicitly include instant messaging.
I’m convinced this study was done by people who view email as a necessary evil, and not by people who “get” it. Their conclusions would suggest this much, but this section really convinces me. Instant message programs like AIM, Yahoo, and MSN all have some very animated smilies for conveying tone. As shown above, the eyerolling animated icon can do wonders for a message. Now imagine dozens of these, all available in one or two clicks. For those of us who like to type, these animated emoticons are even easier to put in a message.
Look at the variety of emoticons in most IM programs and tell me you can’t effectively convey sad vs. angry vs. sarcastic vs. serious. Hell, I could run a study using Yahoo Messenger where people only get to use one icon to convey those four emotions, and guarantee better results than this study….
h2. Gripes about the news
The news article that references the study is flawed as well. The one I’ve referenced above draws conclusions that aren’t in the study – they go from the SNL Jack Handey jokes losing funniness in email to the conclusion that emails you don’t find funny are inherently flawed.
This sentence is just the beginning: “According to a recent study by a trio of business scholars, people think their emails are twice as funny as they really are.” The study is talking about going from a comedian reading of a very specific joke to a FLAT EMAIL. Jokes forwarded around the internet may not be funny to a lot of people, but they tend to circulate well because they’re the kind of humor that doesn’t need to be heard! Jack Handey quotes most definitely gain a lot from their reader.
What’s more, people rated a flat reading of certain jokes at a higher level than the recipients (for instance, I rate a certain quote at 7, but the person I send it to rates that one lower because they like a different one better), showing us that flat reading -> flat reading loses something! Study 5 (the one about humor and Jack Handey) taught us that people have vastly different tastes in humor. Any given person choosing the 5 funniest Jack Handey quotes, even in text-only form, will find that, on average, other people don’t find those 5 to be the funniest! WTF does that have to do with this news article’s conclusion?
Then this article specifically mentions “Photoshopped celeb pics” and “a hilarious clip of a napping cat” as problematic emails addressed by this study. But pictures don’t fall into the boundary of email communication problems! Again, WTF?!? When (and why) did the author think to jump from email communication problems to pictures that he doesn’t find funny? How the hell do those even relate?
Then he goes as far as to say the study is “a much-needed slap in the face to the forward-frenzied emailers out there”. Whereas the study drew a lot of incorrect conclusions, and tested the wrong data, it was at least paying attention to something, and did have a lot of research behind it. The author of this article, David Silverberg, apparently didn’t even READ the damn study! He probably heard about it, wanted to make himself look clever, and chose to revel in his complete ignorance rather than actually research the facts.
As much as I disliked the lack of proper scientific method in the study, this article (and others like it) makes me sick! How can we trust any journalists anymore, when so many of them just recycle other people’s data? And fuck if they can’t even do that right!
h2. Conclusions
Email is almost certainly inferior to verbal communication. But c’mon people, let’s measure the right data next time! And Mr. Silverberg, please try doing the tiniest iota of research before you write again. Might save you from coming across as an ignorant, lazy twit. Oh wait, too late for that….