Chapter of the Month

Keeping Found Things Found
The Study and Practice of Personal Information Management

By, William Jones



Chapter 7

Managing privacy and the flow of information

Information flows from us and to us. We manage flow to protect our
time, our attention, our money, . . . ourselves. Who sees what information
when? What information, from whom, gets our time and attention?
We're easily pushed and pulled by individual informational events. If the
television is on, we'll watch. If the phone rings, we'll answer. Managing flow
means focusing on channels of information and not individual events. We
manage information not just to protect, but to project. We manage not just
for privacy, but for power.

?he closing of a door can bring blessed privacy and comfort the opening, terror. Conversely, the closing of a door can be a sad and final thing the opening a wonderfully joyous moment.?/b>

US commentator, producer, author Andy Rooney (1919 - ____).

7.1. Starting out

Bob was late to his appointment last week to see a specialist about an apparent heart palpitation. Bob was also a week late in paying the monthly bill on one of his credit cards. And late the month before, too. Bob and his wife Carol sold their old house for $349,000 in April and purchased the house they currently live in that same month for $449,000. They are currently re-modeling the kitchen. Their three children Christopher, aged 10, Molly, aged 8, and Sidney, aged 6, all attend the elementary school a half mile from their house.

What do these private facts about Bob, Carol and their children have in common? They are all, to different degrees, a matter of public record. Bob's visit to the specialist, the outcome of the visit and even the fact that Bob was late (and seemed ?lightly agitated? are part of a report that the specialist sent back to Bob's primary care physician and that is now a part of Bob's medical record. Information about Bob's tardiness in paying his credit card bill is kept by the credit card company. Bob's pattern of being late in payments is a part of his credit history and possibly available to banks and lending institutions for inspection should Bob ever apply for a loan. Information about the sale of Bob and Carol's old house and their purchase of their current house is maintained by the government of the county in which they reside and is available on-line via the Web.

Bob's story is not unique. Technologies of storage and search enable others government agencies, private companies, authorized parties and unauthorized parties to keep large amounts of data about each of us in readily searchable, accessible, combinable forms. As noted in Chapter 6, technologies of keeping (e.g., supporting data capture, supporting ever cheaper, larger capacity, smaller-sized storage devices) and of finding (searching, indexing, faster access times, etc.) have outpaced technologies and procedures for properly maintaining the data stored and insuring its integrity. As a result, even a simple update of a person's email address, phone number or marital status becomes extremely problematic. Bob is receiving email, for example, on an old email account that he had hoped to decommission two years ago.

This chapter is about privacy and, more generally, about the flow of information. How can we control the flow of our personal information and also personally relevant information to our best advantage? For personal information as kept and maintained by others, we seek control over who sees what and when (and under which circumstances). We don't want our tax records ?lowing to telemarketers, for example. But often the best way to control information about us ?ut there is to keep it from getting there in the first place.

This chapter is also about the constant flow of information directed towards us. People stop by our office at work. Telemarketers call us during the dinner hour. The dryer beeps loudly to tell us the cycle has ended. Alerts on our computers tell us of new email, new updates, the restoration of a lost connection and other information that distracts us from and is not relevant to our current task. This too is a flow of information directed in towards us. The in-flow of information, in many forms, must also be controlled if we're to keep our concentration, our productivity, our precious time with our family, our solitude and our sanity.

Consider some more examples of flow involving Bob:

  • Bob receives a message sent by a colleague at work to a large distribution list. The message describes an upcoming re-organization. Bob replies to the colleague expressing concerns and asking that his comments not be shared with others. Unfortunately for Bob, the entire distribution list has been included on his response. Bob's ?rivate and confidential' message is now resting in the inboxes of 300 fellow employees.

  • At home, Bob downloads the update to a free software application clicking ?K and ? accept along the way without taking the time to really understand what he is agreeing to. Bob clicks ?K on one page before realizing he left checked a box requesting that he receive regular email updates concerning ?ews and announcements from the software maker. How else will Bob's email address be used or circulated? What else has Bob agreed to?

  • Bob's day has been a blur of interruptions and interruptions of interruptions. He has moved from meeting to meeting handling phone calls, email messages and instant messages in-between meetings and also within meetings. Most meetings, phone calls, email messages and instant messages relate to an upcoming re-org within his company. Bob is working hard to be sure that he and his group are minimally affected. By the day's end Bob feels exhausted but he can't really point to anything that has been accomplished. As Bob tries to relax with his family over dinner, they receive several unsolicited phone calls.

  • Bob remembers the blissful feeling he had a few weeks ago of ?unning on all cylinders as he moved through the work of his day. Everything just seemed to fall into place. The information he needed to write his group's marketing plan was readily at hand. Bob was very happy with the report's organization and direction. At the end of that day, Bob felt energized and even decided to stop by the hardware store on the way home to complete a task he'd been putting off (the return of a drill that isn't working).

Each scenario involves flow. Information flows from us. Information flows towards us. And sometimes we're in a ?low with concentrated time to complete a task and all the needed information at hand.

The management of privacy and information flow is explored in this chapter through discussion that moves through the following sections:

  • Getting oriented. Issues of privacy and information flow arise for each sense of personal information described in Chapter 2.

  • Managing the outflow. This section deals with issues commonly associated with privacy. Focus is primarily on the second sense of personal information information about us. The best place to exercise control is at points where information about us (e.g., our credit card number or email address) is about to flow outward from regions of the PSI we control to regions under the control of others (such as companies we do business with on the Web).

  • Managing the inflow. This section deals primarily with the third sense of personal information --information directed towards us. We want to protect not only the privacy of our information but also the privacy of our time and our solitude. Even good information can become too much resulting in information overload. In managing this kind of information the challenge is less to protect the information itself than to protect ourselves and our precious time and attention from its incursions.

  • Staying in the flow. A lawyer once told me that he felt that he got 90% of his ?eal work done in 10% of his time. Most of us occasionally have periods like those the lawyer described, periods when everything seems to come together, our energy levels are high and a great deal is accomplished. Not only is more work accomplished during such periods, but the quality of this work is often better. There is no sure formula for getting into such periods of flow but this section will review some of the requisites for creating opportunities for such periods of flow to occur.

7.2 Getting oriented

The previous two chapters have talked about the meta-level activities of organization and maintenance as an extension to the keeping of information. This chapter is about the management of privacy and the flow of information. This too is a meta-level activity in both the ?bove/about and ?fter senses of ?eta?

    About/above. Questions about privacy and the flow of information cannot be dealt with piecemeal. A decision to give a company our credit card number raises questions concerning cost (risk of misuse) and benefit (the convenience of payment).

  • After. Dealing with issues of privacy and information flow is not a primary goal for most people. People are busy trying to get things done. Also, people are unclear how best to address these issues. The result is that the management of privacy and information flow becomes an afterthought to be dealt with in a ?ater that never comes.
Relation to ?e?/b> Privacy issues Flow direction and means of control
 1  Controlled by, owned by me Security against break-ins or theft. In/Out. Virus protection. Firewalls.
 2  About me Who sees what when (under which circumstances)? Does it ever go away? Out. Security controls in web browser (e.g. disabling certain uses of ?ookies?. Legal protections Support for P3P. Detection of ?hishing?
 3  Directed towards me What information gets our attention (energy, time) from whom (or what source) and when? What's the right balance between blocking everything and blocking nothing? In. Junk email filters. Pop-up blockers. Do-not-call lists.
 4  Sent (posted, provided) by me Who sees what when? Did the message get through? Out. Distribution lists. ?o not distribute requests.
 5  Experienced by me Do we want a web site to know that we've been there? Can people see what books we've checked out of a library? Out. Security controls in web browser (#5 is a special case of #2 information about ?e?.
 6  Relevant (useful) to me; about to be experienced by me How to filter out or otherwise avoid information we don't wish to see? (How to do likewise for our children?) In. Filters. Content blockers.
Table 7 1. Issues of privacy and flow arise for each kind of personal information.

Concerns of privacy bring us back to the several senses of ?ersonal described in Chapter 2 and summarized here in Table 7 1.

    Information controlled by/owned by us. Included in this category is the information we keep on our computers, in email accounts, on network shares, etc. We need security protections not unlike those we seek for protection of our physical belongings our houses, cars and the valuables within. We need protection against break-ins, vandalism (e.g., in the form of computer viruses that may corrupt our data), theft, snooping and outright take over (e.g., the use of our computer to store and distribute pornography). Issues of computer and data security are beyond the scope of this book.1

  1. Information about us (outflow). Information about us often starts out under our control but then flows into the hands of others and out of our control. The best place to exercise control, then, is often at the point where information passes from our immediate control to an outside person or organization. In practice, such control is often difficult or impossible to exercise. When reading through a document filled with legalese whether on-line or on paper our inclination is often simply to mark ? accept or sign to get on with the task we're trying to complete. And we're generally not given alternatives short of ?ake it or leave it? Information concerning our finances and our health is especially sensitive. Yet, forms relating to both kinds of information often give us little choice other than to sign to release (if we want advice or treatment).

  2. Information directed towards us (inflow). The whole world, it sometimes seems, is trying to get our attention (and our time and money). Phones ring, TVs blare (especially during commercials), billboards loom. Web pages include advertisements that move to grab our attention. Even trusted applications and the operating system of our personal computer may distract. The Microsoft Office Assistant, dubbed ?lippy (actually officially ?lippit?, distracted people even though it was designed to help. If Clippy is gone, other distractions still abound ranging from the balloon notifications that pop up (and disappear) without warning to the red-squiggly lines under words in a document that invite us to break our train of thought to fix a spelling error (that could just as easily be corrected in a later pass through the entire document).

  3. Information sent (or posted) by us (outflow). Many of us have had the experience of pressing send only to realize that a response intended for only one or two people is going out to an entire distribution list. Even as we send information outward, we desire to keep some control over who sees this information. We might at least like to slow the spread of information through requests we include in the message (?his information about the company re-organization should not be shared with your direct reports for the time being.?.

  4. Information we experienced (outflow). Our access to information is personal even if the information itself is not about us or owned by us. We may prefer to keep private the books we check out of a library or purchase at a bookstore (?rick-and-mortar or on-line). Similarly, we may prefer to disable Web ?ookies that might provide the web sites we visit with personal information concerning who we are and how we can be reached.

  5. Information we will (will want to, will not want to) experience (inflow). We try to create a private space for ourselves and our children in which our values are preserved. A space that is free of offensive or gratuitously disturbing encounters. In our physical world, we may accomplish this by avoiding the proverbial ?ed light district and maybe also by avoiding certain people. It is harder to do this on-line as we read our email or surf the web. We seek filters that can screen out material that is offensive or ?unk?

?'ve never looked through a keyhole without finding someone was looking back.?/b>

US actress, singer Judy Garland (1922 - 1969).

Note that #6 information we will (will want to, will not want to) experience -- is in opposition to #5 information we have already experienced.2 After we've experienced the information, it reverts to sense #5 of the personal. We can't ?n-experience it. The issues change. If the information is useful, we might want later to return to the information again. Search support both for filtering out or in new information and for re-finding information already experienced is explored in more detail in Chapter 11. Issues of privacy that relate to #5 have to do with the record of what we've experienced. This record, separate from the information itself, is a kind of private information and we may want to control who has access to this information. One important aspect of #4 -- information sent (or posted) by us is the email we send and this is discussed in greater detail in Chapter 10.

Remaining sections in this chapter focus primarily on control over two other senses of personal information information about us (sense #2) and information directed towards us (sense #3). Discussion is cast in terms of managing flow managing outflow and managing inflow. Again, aside from legislation, we can't stop other people (companies, advertisers, the faceless senders of junk email) from directing information our way. Our main control is over what gets in to grab our time and attention. Similarly, leaving aside legislation, the best way to control the information that others have about us is to keep them from getting this information in the first place.

7.2. Managing the outflow

An ancient belief has it that a ringing in our ears means someone is talking about us.3 If the belief is true then perhaps the ringing we hear in our ears is not a result of tinnitus or some inflammation of our ears. At any point in time, we could be the subject of discussion as various people and organizations collect and exchange information about us.

With each email we send, with each item we buy, with each web page we view, possibly with every step we take, we potentially provide information about ourselves that can be used by others. Living, even living as a reclusive shut-in, provides information to others. This fact is not new, as anyone in a small town can attest. What is new is the extent to which information about us about the ?mpression we make on the world around us is recorded in digital forms that can be stored, transmitted and then later retrieved by nearly anyone, anywhere at any time. Our digital information age has brought several profound changes that impact our privacy:

    Digital storage continues to increase in capacity and decrease in cost. Michael Shamos (Shamos in press) notes that, by the year 2010 according to current projections, ?1000 will be able to buy 50 terabytes of disk, enough to store 200K bytes of information on every person in the United States. Keeping information about people is cheap. Ever-increasing amount of information are maintained including our credit history, medical information, tax records, purchase decisions, academic records, employment histories and so on.

  1. Devices to record information digitally are becoming ubiquitous. Certainly, our computer-mediated interactions with people and information sources are easily recorded including our exchanges via email and instant messaging, our web site visits and our web-based purchases. Digital records are also made of purchases done off-line at the grocery store or a gas station, for example. Moreover, cameras may record us as we stand in line to make the purchase.

  2. As supported by computer-based tools of search and data mining, large quantities of digital information can be rapidly searched and combined in ways that would have been far too costly and time-consuming to complete for paper-based information. Even as we struggle to bring our scattered information together in order to get things done, others can readily do so. Items of information combined can tell a story concerning who we are, where we live, the names of our children, the assessed value of our house, etc., even though each item by itself might seem to say very little (Sweeney 2002; Malin and Sweeney 2004).

  3. In our digital age,the advantage of anonymity is not with us, who wish to protect our privacy, but with the people who wish to violate this privacy. A Peeping Tom in physical space risks discovery. Not so his cyber analog who may be able to sift through large amounts of information about you or me in complete anonymity and from the privacy (his, not ours) of a home office or a coffee shop. There is now a verb reflecting the enormous popularity of the Google search service: We talk about ?oogling someone. ?oogling limited to the publicly available information on the Web seems harmless enough. But maybe not. How much of our life is available somewhere on-line for someone to piece together if they take the trouble? Are your ears ringing?

  4. As a corollary to observations #1 and #2, information kept in digital form can readily stay in digital form forever. The pictures and viewpoints a person posts to the web in an act of youthful indiscretion can lie dormant in a Web archive only to come back decades later when the person is trying to build a life.

One comment I sometimes hear in discussions of PIM and privacy is that young people seem not to care much about privacy. The conclusion is sometimes reached that a generational shift has taken us to a point where issues of privacy matter less than they used to. Perhaps the young are merely the first to accept Scott McNealy's dictate that ?ou have zero privacy anyway?a name="4p">Get over it.?sup>4 But there are reasons to believe youthful nonchalance concerning matters of privacy will fade when age brings careers, families and reputations that might suffer or even be lost through the revelations of youthful acts. Even if we understand our acts to be those of a distant self we barely recognize, others may still attribute these actions to us.

Michael Shamos, an attorney specializing in matters of personal privacy provides an example from his own life that illustrates the points above and the ways in which large quantities of information kept in digital form profoundly alter the privacy landscape. A Web-accessible database is kept by the government of the county in which he residesresides5 that contains information about houses and other buildings in the county:

?ou'll learn my wife's name, how much we paid for the house, its assessed value, how many bathrooms it has, that we have central heating and air conditioning, how much we pay in real estate taxes, whether we were ever delinquent in paying, how much we were assessed in penalties, and a lot more data you didn't imagine the county even knew. You will also be treated to a photo of my house and its floor plan. (Shamos in press)

The county provides the information as a public service with many legitimate uses. In particular, people can view the database to judge the fairness of their property assessment. But this public disclosure of personal information can be used in bad ways, too. Thieves planning a break-in would find it very useful to know the layout of the house and of neighboring houses as well. Knowing that a particular house doesn't yet have central air conditioning could certainly provide a useful cover should a concerned neighbor discover the would-be thieves and ask what their intentions are (?e're here to install central air conditioning ?).

7.2.1. Don't let it out in the first place?

Once personal information is out, it's out ?he genie out of the bottle? There is very little we can do to control or even correct the information that others have about us. To the extent that we have any control at all it is usually in deciding what information gets out in the first place.

One policy is to let out as little information as possible. But this is not a practical solution for most of us. Some distribution of personal information can be very useful. Alan Westin noted that ?ach individual is continually engaged in a personal adjustment process in which he balances the desire for privacy with the desire for disclosure and communication of himself to others, (Westin 1967, p7).

Grudin notes, for example, that ?espite well-circulated accounts of the extensive collection, aggregation, and interpretation of credit and debit card transactions to identify purchasing patterns, people would rather use them than make the effort to carry cash (p 279).

In general, people accept some risk in disclosure for compensating benefits. It is now common for people in an organizational setting to provide at least a busy/free level of access to their calendar information. There are risks of even this rough level of disclosure a workplace enemy might use this information to plan an office ?oup? But people consider these risks to be more than offset by the advantages of disclosure including much greater ease in scheduling meetings (Palen 1999).

A Harris Poll conducted in 2003 (Taylor 2003) divided the 1010 adults who responded into three categories based upon their privacy concerns: 1. ?rivacy fundamentalists (24%) are strongly resistant to any further erosion in their privacy. 2. ?rivacy unconcerned (10%) have no strong concerns about privacy. 3. The large middle ground (64%) was occupied by ?rivacy pragmatists? Privacy pragmatists feel strongly about the protection of their privacy against abuses or unauthorized uses of their personal information. But they are also willing to share their information with others when the apparent benefits outweigh the risks.

7.2.2. Walking the talk

Regardless of what we mean to do in order to protect our privacy and security, we're often drawn into situations where we must (quickly) make decisions impacting our privacy and security without clearly understanding of the implications of the decisions being made.

How many of us click ?o to the questions posed in either dialog of Figure 7 1? And yet, how many of us really understand the implications of clicking ?es? Or, similarly, how many of us carefully read the terms of an on-line agreement (for a download, for example) before checking ? Accept and clicking ?ontinue? The problem now is that we often have little choice between ?es and ?o and lack the time to decide even between these two alternatives.

Figure 7 1. How many of us click "No" to either dialog?

Certainly we need to avoid phishers (see sidebar) who try to con us into providing sensitive information such as credit card number or account number and password under the guise of being a provider with a legitimate need for such information.

But, all too easily, we can be drawn into dialogs with legitimate providers where we wind up providing information we would prefer to keep private. Do we faithfully enter our home, work and mobile phone numbers into a form when none of these is logically needed for completion of the transaction? Or, where forms insist on some non-null entry, do we think to satisfy the form by filling in a single number such as a main reception number for our office for all three questions?

7.2.3. Privacy controls people can understand

We may need to decide, case by case, what information to provide to whom and when. But that doesn't mean we should have to decide alone or without assistance. We can benefit from help to:

    Articulate a privacy policy we can use consistently and that we can use to ?enter ourselves against our inclinations to be too chatty, too trusting or too hurried in actual interactions.

  1. Match our privacy policy against the stated policy of the organization with which we mean to interact.

  2. Verify that the organization is actually following the privacy policy they espouse.

7.2.4. Sidebar: P3P: Platform for Privacy Preferences Project

P3P (standing for Platform for Privacy Preferences Project) is a World Wide Web Consortium (W3C)-sponsored initiative aimed at support of a common language for representation of privacy preferences and policy. P3P provides for the expression of personal privacy preferences and web site privacy policies in a comparable XML format, as well as mechanisms for locating and transporting these expressions.

7.3. Managing the inflow

Once upon a time, a group of students working on their PhDs at Carnegie-Mellon University would gather together every Saturday morning for breakfast at a little hole-in-the-wall diner in Pittsburgh called Ernie's. I was one of them. We ate pancakes and eggs and had great conversations. Breakfast at Ernie's was active entertainment. Everyone played. Each of us, in our turn, said something funny or silly or though-provoking. Then one day a television appeared, mounted high up near the ceiling on the wall. It was always on. Our breakfast at Ernie's changed forevermore. We gaped helplessly at the television even as we observed ourselves doing so. The harder we tried not to look, the more we looked. Conversation happened only in fits and starts, engaging only some of us whiles others attended, openly or surreptitiously, to the television. We stopped going to Ernie's.

Over twenty years ago, Ries and Trout in their book ?ositioning: The battle for your mind described an ?ssault on the mind brought about by a revolution in modern communication that ?as so jammed our channels that only a fraction of all messages get through. And not the most important ones either. (Ries and Trout 1986, p11). The battle for our attention has only increased in intensity since these observations were made. There is now an active area of discussion the ?ttention economy and ?ttention economics.?sup>6

As human beings, we are wired to attend to the ring of a phone, the appearance of a new email alert or the blare of a television set. We can't easily change our nature. What we can do is to adjust the flows of information in our environment in order to create spaces and times in which we are relatively protected from these informational intrusions. Control over the inflow as well as the outflow of information is essential if we're to protect our privacy. This section explores additional reasons for the importance of the management of inflow.

7.3.1. Attention capture

We often get distracted by sights and sounds that have nothing to do with the task at hand. Psychologists call this attention capture (Folk and Gibson 2001). We're especially wired to attend to movement and ?ooming on the periphery of our vision. Certainly sensitivity to change had survival value for our ancestors who were threatened by the warriors of a hostile tribe or by predatory animals (for whom our ancestors were targeted as actual meat, not just the ?oney, energy, attention, time of an information assault). Sensitivity to movement and looming has value even today as a way to alert us to the danger of an oncoming car or a mugger.

People working in the media, in general, and advertisers, in particular, have been very adept at developing devices for exploiting these sensitivities in order to grab our attention and keep it. Consider television. There has been a marked increase in camera cuts where a camera cut is defined to be ?ransition to a different camera perspective that results in the depiction of a new visual environment or entirely new visual information (Southwell and Lee 2004, p 655). In an analysis of Dutch episodes of Seasame Street between 1977 and 2003, for example, Koolstra, van Zanten, Lucassen, & Ishaak (Koolstra, van Zanten et al. 2004) found that the mean number of camera cuts per minute doubled between 1977 and 2003 from four to eight.

MacLachlan and Logan (MacLachlan and Logan 1993) found the number of camera cuts per minute to be especially high on programs and channels targeting a teen and twenty-something market, such as on MTV. Their analysis of camera shot lengths the time between camera cuts showed a steady decline in shot length between 1978 and 1991. For example, shot length on a 30-second commercial declined from 3.8 seconds in 1978 to 2.3 seconds in 1991. On average shot lengths for commercials were 50% or less than the shot lengths of the programs in which these commercials occurred.

The high number of camera cuts and the corresponding shortness of camera shots have the affect of keeping us glued to the television set. We may find it difficult not to attend even when the volume is muted. Some studies suggest that we're more likely to attend to an old object that moves than a new object that doesn't move (Franconeri, Hollingworth et al. 2005). This raises the possibility that the attention we pay may not always align with our expectations concerning information value. We may, for example, be more likely to attend to the blinking advertisement of a web page even the same advertisement that was there the last time we visited the web site than a new hyperlink possibly pointing to useful information.

Data also suggest that we may be nearing a breakdown in our ability to apprehend the message in a video segment that is laced with camera cuts. MacLachlan and Logan present data suggesting that as the length of a camera shot goes below a certain length (2.5 seconds in their study) the memorability and impact of the commercial also declines . Also Lang, Zhou, Shwartz, Bolls, & Potter (2000) found less memorability for short shot length (and more camera cuts). In a worst of both worlds, then, we may sit in rapt stupefied attention and then recall nothing later.

7.3.2. The availability heuristic

Research in human cognition provides a second effect, called the availability heuristic, which is also of direct relevance to a discussion of inflows and the management of personal information. The term was coined by Tversky and Kahneman (Tversky and Kahneman 1973; Tversky and Kahneman 1974) to describe a human tendency to estimate ?requency or probability by the ease with which instances or associations could be brought to mind. (Tversky and Kahneman 1973, p. 171)

To illustrate the effect, let's imagine a person named Tom who is running for office in a small local election a position, say, on the city council. Tom wants to know the likely percentage of people that will vote for him but he lacks the money to commission a poll. He decides, instead, to trust his own ?ut estimate that he is doing really well so well in fact that he decides to take a vacation rather than campaign during the final week before the election.

Tom loses by a landslide. What happened? He was so sure he would win! Tom's problem is that he based his estimates on the people he met personally who said ?e're going to vote for you or some similar expressions of support. Tom is much more likely to meet supporters than not. The people who didn't plan to vote for him weren't at his campaign rallies.

We are frequently in situations like Tom's. We must estimate likelihoods. ?f I go this way to work, what are the chances I'll be late because of a traffic jam? ?hat are the chances that so and so will finish the project on time? ?o I need to take my umbrella? (What are the chances for rain?) For some decisions, like the chances of rain, we can consult the newspaper or the Web. For many other decisions, we base estimates of likelihood on a sampling of information readily at hand information we can recall from memory or information we can readily apprehend from the hot regions of our PSI the headlines of today's newspaper, the papers on our desktop, or the email messages still listed in view on our screen.

What else, after all, are we to do? Decisions requiring some estimates of likelihood come up many times in a typical day. We certainly can't hire a pollster or statistician each time we need to decide something! But what we can do is to be aware of the ways in which our use of the availability heuristic biases our estimates (even our perceptions) of reality and of the decisions based upon these estimates.

Which is more frequent in the English language: 1. Words that begin with the letter ? or 2: words that have a ? as in the third position? If you answered with the first alternative you are in good company with a large and statistically significant majority of subjects in a Tversky and Kahneman experiment (Tversky and Kahneman 1973) In fact, a typical English-language text contains roughly twice as many words with a ? in the third position as in the lead position. The explanation for our mistakes in estimation is that we're able to recall many more words that begin with a ? than words that contain a ? in the third position. Our memories for words are indexed by lead letter, not by embedded letters.

Mistakes in the estimation of word frequencies is not likely to cause most of us trouble. But the use of the availability heuristic, with its reliance on the memorability of events, can lead to systematic biases in our decision-making with real-world impact. Tversky and Kahneman notes that ?ontinued preoccupation with an outcome may increase its availability, and hence its perceived likelihood. People are preoccupied with highly desirable outcomes, such as winning the sweepstakes, or with highly undesirable outcomes, such as an airplane crash. (p. 230).

The media tendency to focus on sensational events has been linked to a tendency for people to overestimate the likelihood of these events (Combs and Slovic 1979). Sensational events such as terrorist attacks, incidents of road rage or molestations by pedophile priests actually get a double boost in availability: 1. We are more likely to hear of them through media coverage (than, for example, the many more frequent deaths that result from disease). 2. The lurid nature of the events covered means we are more likely to remember these events later on.

Efforts to counter the biases of the availability heuristic are multi-pronged.

7.3.3. Information overload

Figure 7 4. Lucy eats the unwrapped candy in the "Candy Factory episode" of "I love Lucy".

Some of us have seen the classic ? love Lucy episode called ?ob Switching (but often simply referred to as the ?andy Factory episode?. Lucy is trying to manage in her new job as a worker on a candy factory assembly line. The candy comes down the line. Lucy carefully wraps each individual piece. So far so good. But then the flow increases. More pieces of candy go down the line. At first Lucy tries to go faster too. The individual pieces are not wrapped as carefully but they're still wrapped. But when the flow of candies goes still faster the situation turns chaotic. Lucy stuffs candies in her mouth and in her blouse when she can not keep up. Even so, unwrapped candies flow by and out the other end.

If we move beyond the slapstick hilarity of the episode, we can easily relate Lucy's situation to many we face in our own efforts at personal information management. Information comes in. And then more information And then still more information. The email messages that arrive on the first day of our job may be answered with care. As more email messages arrive, our responses become briefer. And as the flow of email increases still further, we may begin to ignore or overlook email messages altogether.

E-mail for many of us certainly seems to be a prime example of information overload. The term is itself overloaded with meaning. Information overload is sometimes equated with the exponential growth in the amount of available information. But an explosion in the world's supply of information needn't cause us personal stress unless we're a ?enaissance scholar hoping to keep pace with it all. Scholars dropped any pretense of doing this well before the onset of the Renaissance.

Kirsh (Kirsh 2000) suggests another sense of overload related to the notion that, while the amount of available information is rising exponentially, the supply of ?uality information is growing only linearly. This Malthusian relationship is leading to a steady decline in the density of quality information. But again, this needn't be cause of personal distress, either because we're blissfully ignorant of the information we're missing or because we've found effective ways of locating the needle in the haystack (e.g. through use of search tools).

Information overload becomes personally distressing when our personal systems for managing information begin to breakdown. As discussed in Chapter 5, signs of breakdown are evident in ever-increasing mounds of paper documents waiting to be filed according to a system that seemed great when we thought it up but now takes too much time in practice to maintain. Signs of breakdown are evident in a sense of having lost control over the files of a computer desktop or the email messages in the inbox.

People also experience information overload when they begin to fall behind in their plans for handling incoming information relating to a project or a decision to be made. Perhaps the information is in the form of paper job applications awaiting review for a position to be filled. Or perhaps the information relates to alternatives in the purchase of a new computer or digital camera. Up to a point, people continue to master the material though perhaps by taking shortcuts as flow increases skimming rather than reading in depth, for example. But as a certain point is reached, the whole process begins to break down and the degradation in a person's ability to keep up is far from graceful.

There are no easy, certain solutions to this personal breakdown in the processing of decision or project-related information. But here are some approaches to consider:

    Satisfice rather than optimize. The term ?atisfice was coined by Herbert Simon (Simon 1957) in reference to a strategy of searching only until an alternative meeting some minimal level or set of criteria is found. Satisficing is in contrast to optimizing which requires that all alternatives be considered in order to locate the very best one.

  • Triage (sort) candidates in to ?o? ?es and ?aybe categories. Triage is especially useful when attempting to find a set of acceptable candidates the players who will make a team, for example, or the papers to be accepted at conference. More concentrated focus can then be given to the ?aybe pile.

  • Sample and then optimize within this sample. Set a limit on the amount of information that is collected or the number of alternatives or candidates that are accepted for consideration. A job review process, for example, may specify that only the first n job applications will be considered. The sampling approach is especially useful in cases where 1. Criteria for selection are not well understood ahead of time (and only become apparent through a comparison of candidates). 2. There is reason to believe that the sample is unbiased and representative of the population as a whole.

7.3.4. Manage the information channel, not the information itself

A theme throughout this section's discussion of inflow is that management works best when focused on channels of information and strategies associated with their selection and the processing of their information rather than on the information itself. We can't help but attend and be distracted by the television if it's on. Willpower alone won't change things (and exercising willpower is itself a drain on our energy). Turn the television off instead of leaving it on in the background. Disable the television as a channel of information.

Likewise, we can't avoid using the availability heuristic. In many situations we have few other alternatives. However, we can try to select our channels of incoming information so that the information ?vailable gives a more balanced representation of the likelihoods we need to estimate.

Finally, with information overload discussed here as a breakdown in our ability to keep pace with incoming information focus needs to shift to the channel of incoming information and to overall strategies for its processing. Do we satisfice, triage or sample? This is a choice of strategy involving a consideration for the channel of information (or set of alternatives) as a whole.

?hen I am . . . traveling in a carriage, or walking after a good meal, or during the night when I cannot sleep; it is on such occasions that ideas flow best and most abundantly.?/b>

- Austrian composer Wolfgang Amadeus Mozart (1756 - 1791)

7.4. Staying in the flow

We have all, it is to be hoped, had occasional moments of supreme concentration and productivity where we feel as if we're working at a higher level as a super version of the person we normally are. No only do we get a great deal done but the quality of our work may be better, too. Even better, moments like these often referred to as being ?n the flow or ?n the zone are also restful and fun even blissful. Mihaly Csikszentmihalyi (Csikszentmihalyi 1991) writes:

It does not seem to be true that work necessarily needs to be unpleasant. It may always have to be hard, or at least harder than doing nothing at all. But there is ample evidence that work can be enjoyable, and that indeed, it is often the most enjoyable part of life.

Csikszentmihalyi identifies several characteristics of being in the flow:

  1. A challenging activity that requires skills.

  2. A merging of action and awareness.

  3. Clear goals and feedback. Dishwashing or sweeping can be a flow activity so can video games.

  4. Concentration on the task at hand.

  5. A feeling of being completely in charge.

  6. A loss of self-consciousness.

  7. A transformation of time, in both extremes: small actions go on for long time, but day goes by in no time at all.

    Each of these characteristics makes good intuitive sense. However, there is little beyond intuition or personal experience to suggest that any one characteristic is necessary to the experience of flow. Nor is there reason to believe that some combined realization of these characteristics is sufficient to reach a state of flow. Some characteristics such as challenging activity, concentration or clear goals and feedback seem to be ours to achieve through the actions we take. Others, like feeling completely in charge or losing self-consciousness do not seem so easy to control, other than possibly through a program of meditation.

    There may be no reliable set of steps we can follow to get into a state of flow. But there are certainly events and conditions that can take us out of a flow once we're there. Ringing phones, people at our doorstep demanding our attention, screaming babies, even the alerts that persistently pop up on our computer monitor (rather useless alerts informing us that our connection to email has been restored, for example) these events, always distracting, can also shake us out of any flow we've managed to achieve.

    Studies suggest, for example, that task interruption is built into the fabric of a typical work day and that people spend an average of only three minutes on a task before switching to another task (Czerwinski, Horvitz et al. 2004; Gonzalez and Mark 2004; Mark, Gonzalez et al. 2005). Sometimes the task is completed. Many times people ?elf-interrupt as they reach a stopping point for example, a point where more information such as an email confirmation from a co-worker is needed before the task can be continued. But people are also frequently interrupted by external events an average of four times an hour by one study (O'Connail and Frohlich 1995).

    To reduce self-interruptions, try to assemble all of the task-requisite information (including confirmations from co-workers) ahead of time. To reduce interruptions by others, clear the calendar, close the door, forward or mute the phone and turn off email. Reductions in interruptions surely increase the chances of entering a state of flow and, in any case, reduce that chances that, once ?n the flow we'll be untimely wrenched out of this state.

    ?he quality of the imagination is to flow and not to freeze.?/b>

    US philosopher, poet, essayist Ralph Waldo Emerson (1803 - 1882)

    7.4.1. Personal liquidity

    Flow suggests a liquid state. Water flows. But not always. When the temperature of water goes below 0 degrees Celsius (32 degrees Fahrenheit) a profound transformation occurs. Water freezes into ice. Flowing water, before freezing, becomes dense and flows more sluggishly.

    An analogous thing happens in other situations of flow when we replace temperature with measures of utilization as a percentage of capacity. Traffic flows at off hours when there are only a few vehicles on the roadway and percentage utilization is low. As the number of vehicles reaches and then exceeds capacity, traffic ?reezes and we have traffic jams. Traffic stops flowing. It is a solid. The flow of people leaving a building or stadium can also freeze when capacity is exceeded sometimes with terrifying and disastrous consequences.

    People may experience a sense of freezing with respect to the things in their houses. The house that once seemed to have so much space is now in a frozen state in which daily activities no longer flow as they used to. In a cluttered house, movement becomes difficult and re-arrangements (of furniture, the kitchen utensils in a drawer, the food in the refrigerator) become especially difficult. Item A can't be moved to place B because place B is already occupied by item B. But item B can't be moved to place C because Place C is also occupied and so on. Moving one item provokes a chain reaction of other movements. Worse, the frozen world of too much stuff becomes opaque. We can't find the one spice jar because so many other spice jars are in the way. We can't easily find one item in the refrigerator because so many other items are blocking its view. Freezing can also happen for the items in a duffle bag or toilet kit or a medicine cabinet.

    Freezing can happen, too, with respect to our schedule of workday meetings or our calendar of social events. If the schedule or calendar is filled beyond a certain level, any change short of canceling a meeting or event outright becomes extremely difficult to make. Moving one meeting can't be done without moving another meeting and so on. Freezing can happen with respect to the paper piles on a desktop or with respect to the icons of a computer desktop. Again, beyond a certain point, there is not enough free space left: any movement forces a chain of follow-on movements.

    A concept from economics, referred to as capacity utilization, seems to have direct relevance (Berndt and Morrison 1981). A company, or a country, has some maximum capacity with respect to the goods it is able to produce. But well before this maximum is reached as utilization rises above 80% or so signs of stress inflation for a country, increases in per unit cost for a company will begin to appear. At what levels of utilization of our desktop, our schedule, our hard drive, our ability to juggle different tasks or assignments do we begin experience an analogous stress in our ability to manage our lives and our information?

    ?well as near as possible to the channel in which your life flows.?/b>

    U.S. essayist, poet, naturalist Henry David Thoreau (1817 - 1862)

    7.5. Looking back, looking forward

    Issues of information flow arise for each of the senses in which information can be personal. This chapter has given special attention to the management of outflow of information about us and the management of inflow of information directed towards us.

    • Managing the outflow. The best place to exercise control is at points where information about us (e.g., our credit card number or email address) is about to flow outward from regions of the PSI we control to regions under the control of others (such as companies we do business with on the Web). We need several kinds of help: 1. help to articulate privacy preferences that establish who can see what information about us under which circumstances. 2. help to interpret corresponding company policies concerning the handling of personal data as these are posted to the Web and elsewhere and to detect mismatches between company policies and our privacy preferences. 3. help to determine which companies, which web sites are following their policies as posted and which are not.

    • Managing the inflow. Large amounts of time, money and creativity are being invested in ways to capture our attention. We are wired to respond to peripheral motion, noises, pop-ups and sudden changes in scene. Our best means of controlling inflow is not at the level of individual event but rather at the level of the channel through which information presses in upon our resources. Turn off the TV, unplug the phone, disable notifications of incoming email. Also, manage the balance of incoming information since this information impacts our impressions of the world around us as surely as our direct sensory experiences of the world do.

    • Staying in the flow. There may be no reliable set of steps we can follow to get into a state of flow. But there are steps we can take to reduce the chances of being pulled out of our ?one once we're there. Create times of the day when flows of incoming information are blocked or diverted (e.g., to an answering machine or an out of office message). Make sure that the requisite information is at hand. Also, recognize that some of amount of information assembly and assimilation may be a necessary pre-condition.

    A dominant theme of the chapter is that our focus needs to be on the channels of information as these impact the inflow and outflow of information. Absent control of the channels and associated strategies for dealing with the information on these channels, we're easily pushed and pulled by individual information items and events. If the television is on, we'll watch. If the phone rings, our first inclination is to answer.

    P3P has been discussed as an effort to establish a common language for the expression of personal privacy preferences and the expression of provider (e.g. web site) privacy policies. Tools are under development to support the expression of preferences and policies, their comparison for conflicts or disparities and also to monitor provider compliance in meeting the conditions of the privacy policy they espouse.

    We can also hope for tools that learn from us to elaborate upon our policy for what gets in and what goes out as new cases are encountered. The management of information flow, inflow and outflow, is one kind of meta-level activity that, given the right tool support, can improve if given a more incidental, incremental, integrative treatment (as outlined in Chapter 3). Rules for deciding what information to let out do not spring from our minds fully formed. These are more easily, more accurately abstracted from a consideration of many cases that arise naturally or are possibly are manufactured.

    However, there is no need to wait for these tools nor is there a need to be totally dependent upon P3P or other proposed standards. Each of us can articulate our own privacy plan so that we begin to gain a measure of independence from the push and pull of daily events. If we're getting to much email already, for example, then perhaps we make it a habit to routinely decline (uncheck) offers to keep us informed of ?xciting new offers?

    Moreover, effective management of information flow incoming, outgoing and as this relates to the flow of work means thinking about more than just privacy and protection. Flow is not just about what we stand to lose if our privacy and security is compromised. Flow is also about the many ways we benefit from the effective dissemination of information. The management of flow is about privacy and protection but it is also about power and projection. The use of our information is a means of affecting desired changes in our world.

    But how do we know whether our plan for the management of information flow is working? For that matter, how do we know which parts of our personal information space need our attention most urgently? Earlier the chapter talked about achieving a ?alanced diet of incoming information. But how do we go about measuring our information intake? These are topics for the next chapter.

    7.6. (Sidebar) What now for you and me?

    We don't have to wait for the future. There are many choices we can take now to better protect our privacy and to better control the flow of personal information in each of the six senses of personal. A sampling of some of these choices is listed in Error! Reference source not found. Most steps are obvious and straightforward but easily overlooked even so.


    1. See the article on computer security in the Wikipedia for an excellent introductory discussion of computer security issues

    2. #6 can also be contrasted with #3 information directed towards us. Both are about the inflow of information. #3 is information as ?ushed to us from other people (and organizations); #6 is information that we ?ull towards us. Even when the initiative is ours, even when we're doing the pulling, we can still make mistakes. We can stumble onto offensive information we would rather not experience. Even more likely, our children are at risk of doing this as they browse the Web.

    3. Pliny the Elder mentions the belief in ?atural History? first published in 77 CE.

    4. Springer, P., ?un on Privacy: 'Get Over It'? Wired News, Jan. 26, 1999.,1283,17538,00.html

    5. Allegheny County, Pennsylvania.

    6 See, for example,