Feb 14

It Must Be Love! (or Non-Specific Arousal)

Posted on by

Don Williams’ hit single “It Must Be Love” (also recorded by Alan Jackson) features the lyrics

First I get cold then hot
Think I’m on fire but I’m not
Oh, what a pain I’ve got
It must be love [...]
It must be love, it must be love
I fall like a sparrow, fly like a dove
You must be the dream, I’ve been dreamin’ of
Oh, what a feeling, it must be love

Around Valentine’s Day, I advise those of my students who looking to get a special someone to fall in love to try to get this individual to go along on a roller coaster ride.  It turns out that the brain often has some difficulty in determining the apparent cause of arousal–is it because of fear or because of love?  The brain has to make an interpretation that may or may not be correct.  Many individuals have found themselves to be particularly vulnerable to falling in love during periods of stress. Students experience this at finals’ week; a colleague of mine reported that this would happen to him when he was at academic conferences to present a paper in front of his colleagues.   It turns out that, in trying to make sense of the arousal experienced, the mind may conclude one is in love–why else would one be experiencing this feeling?

In 1962, the psychologists Stanley Schachter and Jerome E. Singer conducted an experiment that today would be considered patently unethical.  Male volunteers were asked to participate in a study ostensibly on the impact of vitamins on vision.  Thus, each participant received an injection.  It turned out, however, that in half the cases, the injection actually consisted of epinephrine (adrenaline), triggering arousal and “fight or flight” responses.  The other half of the participants, the control group, received a saline solution as a placebo.  As an ostensible part of the experiment, the participants then waited for twenty minutes in room in which a confederate acted in one of two different ways:  “euphoric” or “angry.”  There was actually an additional manipulation by which some subjects were made aware of a potential “side effect” of arousal–thus knowing what to expect–while others were either given no information or were actively misled with expectations of alternative, non-existent side effects.  Those subjects who received the epinephrine injection and were either given no information or were actually misled were significantly more likely experience the ostensible feelings of the confederate.

 

Jan 19

Why is it so difficult to get viewers to Oprah’s OWN Channel?

Posted on by

The Wall Street Journal reports that Oprah Winfrey’s much publicized interview with Lance Armstrong drew a live audience of only 3.2 million.  Although this is the second highest audience reach on Oprah’s OWN network so far, it is significantly below what high profile events get on the major networks–some one fifth, for example, of recent American Idol Viewership and a very far cry from the 48.5 million viewers who watched Barbara Walters’ 1990s interview with Monica Lewinsky.  OK, for all the sensation that goes with Armstrong’s belated admissions, is not, of course, in league with the Lewinsky scandal in terms of its Juiciness Quotient (JQ), but Oprah’s interview was still a major event.

As one would expect, a large number of viewers later caught the interview on the Internet, but this audience is much less likely to become long term viewers.

This difficulty in gaining critical mass could result in a vicious cycle.  With disappointing viewership, celebrities will be more reluctant to choose this outlet for major appearances, in turn making it more difficult to attract and sustain a sufficient viewership.

Apparently, one reason for low ratings is that many cable and satellite subscribers find it difficult to locate Oprah’s channel on their boxes.  With the large number of channels now available, ambiguously named channels such as this one often get lost among what can amount to over one hundred choices (in addition to music and other limited function channels).  In retrospect when one thinks about it, it makes perfect sense that OWN stands for the “Oprah Winfrey Network,” but that may be less obvious to many viewers when they attempt to make their way through a long list of obscure three letter codes in their cable listings.  Those who have access to a Tivo or another DVR system can, of course, search for key phrases such as “Oprah” or “Lance Armstrong” if they plan to record in advance, but that requires advance planning.  Further, such ad hoc recording may not be conducive to viewing of the network’s subsequent programming.

This difficulty for viewers in getting to lesser known channels makes it difficult to understand why Al Jazeera was willing to pay a reported $500 million for the Currents channel founded by Al Gore in 2005.  This channel, too has had limited viewership. Yes, the New York Times reports that Gore lobbied cable service providers, suggesting that balking at carrying Al Jazeera might come across as buying into stereotypes of the Arab world.  Although there is relatively abundant space available with today’s cable carriers, had Al Jazeera approached the networks on its own with a “new channel,” resistance might have been greater.  Still, what good is a network going to do with a limited viewership?  It is not as though limited number of Americans who want to watch an English language edition of Al Jazeera can’t get this on the Internet as things were.  Perhaps Al Jazeera–which was historically not operated for profit–hopes to be perceived as a more mainstream alternative if it is available on U.S. cable networks, but the price still seems rather high.

By now, a number of specialized cable channels do receive a fair viewership.  One thing that many of these networks have going for them is that they have common, or at least overlapping, ownership, making it more cost effective to promote upcoming programs on “sister” channels.  Although the OWN network could, in principle, exchange such program promotions with other networks, it is unlikely that these competitors would be eager to promote events on a network that could, if it gained more habitual viewers, gradually siphon off more and more of their audience.

Dec 19

FTC augments rules on collecting online data on children

Posted on by

The New York Times reports that the Federal Trade Commission (FTC) has broadened its rules on online data collection on children to better accommodate new portable technologies such as cell phones and tablets.  Generally, online sites must obtain permission from parents before collecting information about their children.  Existing regulations already require parental consent before information that can be used to “identify, locate or contact” a child–e.g., a photo or video–can be collected.  A significant addition in the new regulation, however, is that “persistent identification systems”–tools such a cookies that do not reveal the specific identity of a visitor but tracks visits by the same person across time–are now covered as identifying information.  Since online advertising is often based on prior activity–e.g., an individualized advertisement in an online article for a product for which a person has searched on a site such as Staples.com–is a rapidly growing area, this is a major limitation.

The protection and privacy of children are clearly important and the potential loss of some online content for children that it may no longer be cost effective to provide may be warranted.  An interesting question, however, is the extent to which children under thirteen who sign up for online accounts actually answer truthfully when asked about age.  There are also questions about children on both sides of this age limit.  Are safeguards that might be adequate for children aged 11-13, for example, sufficient for those under 10, an increasing number of whom by now have smartphones?  Further, although children over thirteen may be better able to make judgments about their own more theoretical privacy protection, does this age cutoff send the message about the types of content and offers that might be sent those over thirteen who are subject to extensive protection and close supervision in other parts of their lives such as school and extra-curricular programs?

Is this type of regulation merely a “feel good” measure or does it offer significant practical protection for children?  The answer is not clear.

Nov 28

Scroogled? Microsoft Attacks Google on “Shopping” Search Results

Posted on by

An article in the Washington Post reports that Microsoft is launching an attack on Google, gleefully warning searchers that this previously trusted information source has changed the algorithm for the “Shopping” search feature results such that only those firms that pay to be listed show up.  Although paid search and “sponsored” search results show up on the right side and in a shaded region at the top of ordinary search results, Google prides itself on the objectivity with which its “organic” (merit based) search listings are determined.

For many years, Microsoft has attempted to take share away from Google in the search engine market.  Despite the tremendous resources that Microsoft has available, its Bing search engine has very limited traffic relative to Google.  Some people may end up with Bing as their default search engine when they buy a new computer where this setting is made in Internet Explorer, but otherwise, few searchers seem to make the move to Bing despite its touted advantages.  Getting people to try out Bing out of fear that Google search results in the “Shopping” section may be “manipulated” could be a powerful way gain trial among consumers who may be vaguely aware of Bing but never took the time to try it out.  Realistically speaking, most searchers probably will not feel much difference between the two search engines.  Some people may turn out to prefer Bing for idiosyncratic reasons and may stay.  Others may find Bing interesting as way of getting variety.  Some may stay with Bing in protest.

In its newly created web site Scroogled, Microsoft chastises Google:  “In the beginning, Google preached, “Don’t be evil“—but that changed on May 31, 2012. That’s when Google Shopping announced a new initiative. Simply put, all of their shopping results are now paid ads.”  Google founders Sergey Brin and Larry Page are quoted saying that “Furthermore, advertising income often provides an incentive to provide poor quality search results” to underscore the ominicity of Google’s policy.

The reality is that only a minority of shoppers actually use the “Shopping” feature of Google.  Most consumers, even when they are looking to buy, tend to use the ordinary Google search.  The “Shopping” feature does not even show up on the first level menu on Google; you have to click on “More” to get that as a search option.  This will, of course, bring up paid search results along with the merit based listings, but that is nothing new and is not different from a Bing search.

Although screen layout is subject to constant change, at the moment, the Google “Shopping” page does not appear to acknowledge the paid nature of search results.  However, in sharp contrast to the “clean” look of the general Google search page, the search screen does look a lot more “commercial” and cluttered.  Thus, although consumers may or may not infer that the search results are paid, the consumer may nevertheless get the impression that this may somehow be a less “pure” search.

Nov 21

An early Thanksgiving means an early Black Friday

Posted on by

Historically, the Friday after Thanksgiving marked the “official” start of the holiday shopping season.  The day was so named because this was the day on which the balance sheets of many retailers would turn from “red” into black as the stores had sold enough to cover their fixed costs.  Any margins after that were profit.  (Today, with greater competition and lower margins in the retail industry, the actual turning point may come a bit later–especially when consumers are still weary of the economy.)

Even when Thanksgiving falls on November 28–the latest date on which the fourth Thursday in November can fall–one would hardly think that people would be truly strapped for shopping days.  Yet, the reality is that holiday shopping is, for many consumers, not a particularly disciplined and well planned process.  Many people procrastinate.  Although few people would entirely neglect to buy presents for their close family members, the quantity of overall gifts bought will depend heavily on the number of days available to shop.

Up until 1941, Thanksgiving was generally held on the last Thursday of November.  In 1939, in response to concerns expressed by retailers that Thanksgiving would have fallen on November 30 that year, President Franklin D. Roosevelt moved up Thanksgiving one week.  Retailers at the time feared that consumers would find it unseemly if holiday gift merchandise were displayed prior to Thanksgiving.  This rescheduling was apparently a rather unpopular move with the public, and the date became known derisively as “Franksgiving.”  However, soon afterward, Congress passed legislation fixing the date of Thanksgiving at the fourth Thursday in November.

Historically, Black Friday was not so much a day of great sales as it was a genuine start on holiday shopping.  The real sales happenedafter Christmas as the retailers had to liquidate leftover merchandise that now longer held the same relevance.  Today, however, Black Friday has become a much more significant event in and of itself due to dramatic sales offered by many retailers which often open at what might traditionally have been seen as “unseemly early” hours.

It is certainly a positive factor for overall holiday sales that Black Friday falls early this year.  However, as consumers may see more days ahead before Christmas and as more and more shoppers become tired of the long lines and early lines on the day after Thanksgiving, it is very possible that the early timing of Black Friday itself will put a bit of a damper on the actual sales that day.  That, of course, would leave more bargains for those who are prepared to dare the crowds and darkness in the early morning hours.

Nov 16

Twinkie Hoarding

Posted on by

CNN reports that with the impending shutdown of operations by Hostess, hoarding of remaining inventory is likely.  Although Twinkies will likely be manufactured within months as the brand is sold off in the course of bankruptcy proceedings, for the moment, supplies will likely dwindle fast.

Twinkies apparently have a loyal following despite the rather mixed reputation of the brand.  In 1979, attorneys for a man charged with several highly publicized San Francisco murders invoked the now infamous “Twinkie defense,” alleging that their client’s depression had caused him both to commit the murders and go on a diet of Twinkies.  Counsel for the defense did not actually argue–as it is commonly believed–that Twinkies caused the murders–but rather that both the shocking diet and the murders were said to be caused by depression.  This association helped make the brand the butt of jokes.  A product consisting of highly refined wheat and sugar, Twinkies come across as the epitome modern of food gone bad.  Eating these rolls comes across at best as a guilty pleasure and, in the view of many, as utter and complete lack of culinary sophistication and responsibility.

With the amount of choice that consumers have today, one might wonder why there is such as attachment to a brand for which very similar substitutes are in abundance.  Many people–if given blindfolded taste tests–might not be able to taste the difference from similar–and often significantly lower priced–products, but old habits die hard.  Today, the taste store and lower tier cola brands is becoming increasingly difficult to distinguish from that of Coke and Pepsi; yet, these lower priced competitors have made only modest inroads.  For some people, far from just being a matter of getting “the real thing,” Twinkies might be a sorely missed opportunity to express a not so subtle rebellion.

We may have another interesting phenomenon coming on.  Scarcity is a strong motivator and a driver of great interest.  There may be a number of people who haven’t eaten Twinkies for years who–now that these may not be available for a while–will develop a sudden interest in getting their hands on any available supply.

Nov 06

Would the Democrats be sadder than the Republicans would be happy?

Posted on by

The economists Daniel Kahneman and Amos Tversky pioneered a line of research that has consistently shown that people tend to react more negatively to “losses” than they react positively to “gains.”  Currently, there is a Democratic President in office.  Thus, if Barack Obama loses this election, this would be a loss for the Democrats.  A win by Mitt Romney, however, would be a gain for the Republicans.  One might hypothesize, then, that the Republicans would be less happy than the Democrats would be sad.

Conversely, if Romney loses the election, this would not be a loss (relative to the status quo) for the Republicans, thus limiting (to some extent) their sadness.  According to theory, neither would the Democrats be as happy to have the President re-elected as they would be sad not to have him re-elected.

Still, when the results of the election are known, there will probably be plenty of happy and sad people.

Nov 01

Are consumers really spending LESS time online than last year?

Posted on by

A recent report by Forrester Research indicates that consumers report spending less time online than those surveyed last year.  Can this actually be true?  Granted, a few individuals may kick the Facebook habit, and others may find that they have to focus more on actually getting productive work done.  Yet, with more and more content available, one would likely suspect, at the balance, that consumers are actually spending more time online this year.  Maybe the rate of growth is less now that social media have already been adopted by so many, but negative growth sounds rather unlikely.

The analysts and Forrester have concluded that what is going on is not an actual decline in use, but rather a result of consumers’ “mental accounting” standards in deciding what constitutes “being online.”  With the proliferation of iPads and other tablets, and as more and more people are becoming comfortable with the smaller screens and keyboards of smart phones, it appears that some people are simply not thinking of certain mobile activities as being online.  “Being online” reflects what is known as graded structure–the idea that some activities are better “exemplars” than others.  Using a desktop to surf the Internet is clearly going online.  Most people might also agree that reading one’s e-mail–even if this has been downloaded by one’s e-mail program ahead of time–constitutes being online.  People surfing the web or using a Facebook app on a cell phone would probably also think about this as going online, but may think of checking e-mail as less so because the sending and receiving of messages may be less obvious.  If  you sit at home at your desktop and watch a streamed movie, that is clearly going online.  Strictly speaking, if you download a movie to view on a flight that may not feature high speed wi-fi, watching it on the plane is not being “online”–and might not be counted as such–but it is an activity that is contingent on having gone online ahead of time.

One of my students insightfully observed that with more experience, people may have become better skilled at online activities and might be able to accomplish these more quickly.  However, most people have been using desktop and notebook computers to surf the web for a long time, so their learning curve is probably not that steep anymore.  Ironically, although people may have become more efficient at using portable devices, we notice that some of the activities performed on these devices less likely to be considered being online and being counted in the total amount of time spent.  Even if people did become significantly more efficient in their online activities, there is a question as to whether the time saved would go toward productive work (or even some unproductive off-line task) rather than being spent on additional online activities.

The lesson here is that we have to be careful in how we interpret survey data.

Oct 26

Presidential polls in tight races

Posted on by

In recent history, there have been some very tight Presidential races where it has been impossible to predict the outcome with any degree of certainty ahead of time.  In principle, it only takes having one more vote in a particular state to win it.  Why, then, is it so difficult to predict who will win?

One issue is the sample size needed to make precise inferences.  National opinion polls typically sample a little over one thousand individuals.  This usually allows for predictions that, if assumptions are satisfied, come with a margin of error of plus or minus three percent.  That is, if 51% of individuals polled in the sample indicate that they will vote for candidate A, in 95% of cases, the candidate could expect to get between 48 (51-3) percent and 54 (51+3) percent of the vote.  Note that the 50% dividing line is squarely within this interval–thus, we cannot confidently predict the outcome.

Getting additional precision comes at a heavy cost.  Note that when a greater level of precision is desired, the needed sample size increases exponentially.

Another problem involves determining who should be surveyed.  Some polls, for example, sample registered voters in general.  However, not all voters are equally likely to actually show up and vote.  Therefore, surveys that sample “likely voters”–voters who are more likely to actually show up to vote–tend to give more accurate predictions (assuming that the polling organization has identified appropriate criteria to gauge likelihood of voting).  In general, Republican voters tend to be slightly more likely to turn out than Democrats.  Therefore, surveys of registered voters in general are likely to over-estimate the actual support for the Democratic candidate.  Given the very close race between Barack Obama and Mitt Romney, it is not surprising that different types of polls seem to make different types of predictions.

The problem with needed sample sizes is compounded by the fact that U.S. Presidential elections are determined by the Electoral College rather than by the “popular” nation-wide vote.  In most states, the candidate that receives the greatest number of popular votes gets all the electoral votes of the state.  Although fluctuations among states to some extent cancel out each other, an accurate prediction would, in principle, require separate polls–each with large sample sizes–to estimate the Electoral College votes from states where the outcome is in question.  (Relatively speaking, California leans so heavily to the Democratic side that no Democratic candidate could hope to win the national election without winning California.  Similarly, no Republican candidate who failed to carry Utah could possibly hope to win.)  “Swing” states such as Florida, Ohio, and Missouri take on special importance here, but in a tight rate, the votes of less populous “swing” states such as Nevada also become important.

Another difficulty with predicting actual votes ahead of time, of course, is that some voters remain undecided and others change their mind before the election. Although an increasing proportion of voters tend to make up their minds as we get closer to the election, each has the option to change his or her choice up until the ballot has been cast.

Oct 12

Launching Windows 8 with a bang

Posted on by

The Wall Street Journal reports that Microsoft is expected to unleash a major advertising campaign this weekend to introduce the new Windows 8 operating system.  Part of the reason for this elaborate campaign appears to reflect the complexity of what Windows 8 entails.  This is not just a new operating system for traditional computers–there are parallel and tightly integrated versions for cell phones, tablets, and non-Intel based computers.  Apple has had its own operating systems for both computers and cell phones, but although there has been some integration between the two, these systems are nevertheless separate.  Apple periodically issues upgrades and improvements, but these are incremental with no real leaps.  Windows 8, in contrast, will be much more consistent in its appearance and functionality across versions.

Most people have at least heard of Windows 8 and have at least some vague idea of what is involved in the PC version.  Only the geekier element of society has a broader understanding of what is about to be unleashed.  Now comes the hard work for Microsoft in introducing the bigger picture of just what this next step entails. The question is how to get this idea across without creating information overload or otherwise confusing consumers who may have only limited motivation to struggle to understand the details of this new innovation.  This limits the amount of information that can be crammed into one advertisement.  Thus, different ads may need to address different issues.  Since most people are not going to spend the bulk of their weekend watching TV, Microsoft cannot count on people having seen any previous advertisements.  This means that each new ad will need to be understandable in its own right; in this short time span, it is not possible to count on people understanding the background from previous ads.

Back in the 1995 when Microsoft released Windows 95, they also bought extensive advertising time during the launch weekend.  In addition, they reportedly paid the Rolling Stones several million dollars in royalties to use the tune “Start Me Up,” emphasizing the “start” icon on the new operating system.  Although there was a need at the time to develop interest in Windows 95, the need to inform and communicate complex information was much less since consumers were already familiar with the basic idea of a graphical interface from the Macintosh.  This time, the leap is much greater and potential customers are likely to be much more confused.