Google Translate is an amazing thing. You can take a chunk of text in just about any language, paste it into Google Translate, and it is instantaneously (if imperfectly) translated.
Since I can’t speak anything other than English, I’m not in a great position to say how good or bad the translations are, but my multi-lingual friends generally turn their noses up at Google Translate, saying it doesn’t do that great a job.
My response is that compared to any other alternative I know (like trying to track down someone who speaks Croatian, or going word by word through a Croatian-English dictionary), it seems like a miracle. I love it.
But even Google Translate has its limits. Read More »
Season 4, Episode 2
When Harvard professor Latanya Sweeney Googled her name one day, she noticed something strange: an ad for a background check website came up in the results, with the heading: “Latanya Sweeney, Arrested?” But she had never been arrested, and neither had the only other Latanya Sweeney in the U.S. So why did the ad suggest so? Thousands of Google searches later, Sweeney discovered that Googling traditionally black names is more likely to produce an ad suggestive of a criminal background. Why? In this episode of Freakonomics Radio, Stephen Dubner investigates the latest research on names. Steve Levitt talks about his groundbreaking research on names, economic status, and race. And University of Chicago economist Eric Oliver explains why a baby named “Cody” is more likely to belong to conservative parents, and why another named “Esme” was probably born to a pair of liberals.
Microsoft has now responded, with a blog post and a letter, to my post about an experimental study that I coauthored with Yale Law School students Emad Atiq, Sheng Li, Michelle Lu, Christine Tsang, and Tom Maher. Our paper calls into question the validity of claims that people prefer Bing nearly two to one.
In response to several commenters: I do not work for and do not have any consulting relationship with Google.
Microsoft claims that our study is flawed because it relied on their own blind comparison website. They now say that “Bing It On” is meant to be a “lightweight way to challenge people’s assumptions about which search engine actually provides the best results.” To be sure, companies often use fantastical or humorous scenarios for free advertising. However, Microsoft’s television commercials present the site as a credible way that people can learn whether they prefer Google or Bing. These commercials show people who discover that they really prefer Bing to Google. The challenge site that they created is either sufficient to provide insights into consumer preferences or it isn’t. The advertisements give the impression that the challenge site is a useful tool. Microsoft can’t have it both ways. If it is a sufficient tool to “challenge people’s assumptions,” then it is sufficient to provide some evidence about whether the assumed preference for Google is accurate. Read More »
Did you find this blog post through Bing? Probably not—67% of worldwide searches go through Google, 18% through Bing. But Microsoft has advertised in a substantial TV campaign that — in the cyber analog to blind taste testing — people prefer Bing “nearly 2:1.” A year ago, when I first saw these ads, the 2-1 claim seemed implausible. I would have thought the search results of these competitors would be largely identical, and that it would be hard for people to distinguish between the two sets of results, much less prefer one kind 2:1.
When I looked into the claim a bit more, I was slightly annoyed to learn that the “nearly 2:1” claim is based on a study of just 1,000 participants. To be sure, I’ve often published studies with similarly small datasets, but it’s a little cheeky for Microsoft to base what might be a multi-million dollar advertising campaign on what I’m guessing is a low six-figure study.
To make matters worse, Microsoft has refused to release the results of its comparison website, Bingiton.com. More than 5 million people have taken the Bing-It-On challenge – which is the cyber analog to a blind taste test. You enter in a search term and the Bing It On site return two panels with de-identified Bing and Google results (randomly placed on the right or left side of the screen). You tell the site which side’s results you prefer and after 5 searches the site reveals whether you prefer Bing or Google. (See Below)
Microsoft’s soft ads encourage users to join the millions of people who have taken the challenge, but it will not reveal whether the results of the millions are consistent with the results of the 1,000. Read More »
Last year, Google realized that its employees were eating too much free candy — M&Ms, specifically. So the company conducted a little experiment, and carefully tracked the results. Cecilia Kang, writing in the Washington Post, summaries:
What if the company kept the chocolates hidden in opaque containers but prominently displayed dried figs, pistachios and other healthful snacks in glass jars? The results: In the New York office alone, employees consumed 3.1 million fewer calories from M&Ms over seven weeks. That’s a decrease of nine vending machine-size packages of M&Ms for each of the office’s 2,000 employees.
The company has conducted similar experiments in an effort to reduce consumption of sugary drinks and encourage employees to consume less calories in the company’s cafeterias. “With a company as big as Google, you have to start small to make a difference. We apply the same level of rigor, analysis and experimentation on people as we do the tech side,” says Jennifer Kurkoski, a member of Google’s HR team.
A reader named Desmond Lawrence writes from London with further commentary on our “How Much Does Your Name Matter” podcast — specifically, about Harvard computer scientist Latanya Sweeney‘s research which found that online searches for people with distinctively black names was 25% more likely to produce an ad suggesting the person had an arrest record – regardless of whether that person had actually been arrested:
Read More »
So when I was listening to your podcast on “How Much Does Your Name Matter?” I was surprised to hear about Latanya and her story about these Google Ads that were being served.
Now as much as the company Instant Checkmate would like to say that they are not at fault here, I can guarantee that I know what has happened with their AdWords campaign.
When you set up an AdWords campaign you tend to do a fair bit of research. From there you will build a campaign around Broad match, phrase match or even exact match.
You can also do a thing called Dynamic keyword insertion. Now this is where I would suggest that Instant Checkmate went wrong. If you place the Dynamic keyword call code into an ad, it will place the keyword that has called the ad into the ad, thus increasing the effectiveness of the ad.
Our latest Freakonomics Radio podcast is called “How Much Does Your Name Matter?” (You can download/subscribe at iTunes, get the RSS feed, or listen via the media player above. You can also read the transcript; it includes credits for the music you’ll hear in the episode.)
The gist: a kid’s name can tell us something about his parents — their race, social standing, even their politics. But is your name really your destiny?
The episode draws from a Freakonomics chapter called “A Roshanda By Any Other Name” and includes a good bit of new research on the power of names. It opens with a conversation with NYU sociologist Dalton Conley and his two children, E and Yo. Their names are a bit of an experiment:
CONLEY: Of course it’s hard to separate out cause and effect here until Kim Jong-Un allows me to randomly assign all the names of the North Korean kids…but my gut tells me that it does affect who you are and how you behave and probably makes you more creative to have an unusual name.
Indeed, there is some evidence that a name can influence how a child performs in school and even her career opportunities. There’s also the fact that different groups of parents — blacks and whites, for instance — have different naming preferences. Stephen Dubner talks to Harvard professor Latanya Sweeney about a mysterious discrepancy in Google ads for Instant Checkmate, a company that sells public records. Sweeney found that searching for people with distinctively black names was 25% more likely to produce an ad suggesting the person had an arrest record – regardless of whether that person had ever been arrested. Read More »
The New York Times published an interesting article last week about an ongoing dispute in Europe between Google and European newspapers (and their supporters in government). The issue is whether Google must pay for the privilege of linking to those sites, or should be able to link for free. Of course, at stake is who gains the revenue that comes from aggregating and compiling links.
As the Times notes:
Read More »
Google got rich by selling a simple proposition: The links it provides to other Web sites are worth a lot of money, so much that millions of advertisers are willing to pay the company billions of dollars for them.
Now some European newspaper and magazine publishers, frustrated by their inability to make more of their own money from the Web, want to reverse the equation. Google, they say, should pay them for links, because they provide the material on which the Web giant is generating all that revenue.