Twenty-five years ago, in 1996, Thousand Oaks voters were included as an official part of preserving the city’s general plan, as ordinances were enacted requiring voter approval of important general plan changes before becoming effective.  Sponsored by the two political protagonists of the time, Linda Parks authored the Parks Initiative, a precursor to the successful Ventura County open space protection campaigns, while Andy Fox championed the city-sponsored Measure E.

The Parks Initiative, qualifying for the ballot as a citizens’ initiative and approved in June 1996, focused on the parks and open space lands within the city.  Under the initiative, lands categorized as ‘Parks, Golf Courses, Open Space’ “shall remain so designated… unless redesignated to another General Plan land use category by vote of the people”,[1] and “[w]henever the City Council adopts an amendment requiring approval by a vote of the people…, the City Council’s action shall have no effect until after such a vote is held and a majority of the voters vote in favor of it.” [2]

The language of the Parks Initiative provides clarity as to when the voters must be included for changes to the general plan:  if land is designated in a protected category and an amendment is proposed to designate that land otherwise, the amendment does not take effect until approved by the voters.

By defining the approvals that require voter concurrence plainly and specifically, the Parks Initiative established very effective protections.  The clarity of language ensures that landowners and the public know the process for making any such changes.   While not a prohibition, such a requirement lowers the likelihood of any such proposals being made, making for an effective measure for protecting open space.

Just weeks after enacting Parks’ open space initiative, the Council placed its own measure on the November 1996 ballot, which was approved by voters that fall.  The ordinance now known as Measure E also proved effective as a tool in curbing development pressures within Thousand Oaks, but not for the reasons Measure E advocates claim.  In contrast with the Parks Initiative, the measure went through evolving tactics for implementation, based upon how interpretable the language is and the circumstances that gave councilmembers opportunities to modify their interpretations.

Measure E requires that “[a]ny amendment which cumulatively provides a net increase in the maximum number of residential dwelling units which could be permitted under the proposed land use designation” or “[a]ny amendment which cumulatively provides a net increase in the land designated ‘commercial’” requires voter approval.[3]  The challenge:  what does “cumulatively provides a net increase” mean?

When Measure E was introduced to the Council, Mayor Fox, the measure’s primary advocate, provided a way forward in a memo for such “minor changes in the General Plan…if these changes might mean that the residential and/or commercial uses might be increased.  If the Council chose to do this, they would be required to do one of two things:  (1) concurrently reduce densities and commercial space elsewhere in the City or (2) submit the proposed General Plan changes which would increase the City’s ultimate population or commercial areas to the voters.”[4]

To grant flexibility under Measure E without seeking voter ratification, the City Council could, for example, increase acreage designated as commercial in one part of the city, as long as an equal or greater amount of acreage were decreased elsewhere, thus resulting in no net increase of commercial acreage.  If this were done, then according to Fox, the general plan amendment would not need voter approval to become effective.

Based on Fox’s 1996 memo, these offsetting shifts in residential density or commercial acreage were anticipated to occur “concurrently” – at the same time.  In the succeeding years after Measure E’s passage, no issues arose regarding the measure, as only amendments decreasing residential or commercial lands were approved by the Council.

The first true test of Measure E came in 2001, when former city councilmember and county supervisor Ed Jones sent a letter to the city representing Jerry Brodecky, owner of Thousand Oaks Toyota on Thousand Oaks Boulevard.  Jones proposed that a portion of Brodecky’s property, about 1.2 acres, be redesignated from “medium density residential” to “commercial” to accommodate the dealership’s expansion.  Noting that this could trigger Measure E, Jones proposed that “an alternative to a vote of the people be invoked in which it is demonstrated that an equivalent amount of land within the city has been rezoned by removing it from commercial since Measure E became law…”[5]  Jones went on to recommend that this could be accomplished with land that “has been or is in the process of being removed from commercial and use that acreage to demonstrate that there will be no net increase…”[6]

Regarding the request, City Attorney Mark Sellers advised the Council that “[i]f land has been removed from the ‘commercial’ designated [sic] since 1996, and no equivalent offsetting increases have occurred, there is a net reduction which can be reallocated.”[7]  A further memo from the City Attorney’s office advised that “[t]he Council controls how it will reallocate those past reductions in acreage of commercial uses.” [8] The Council initiated the general plan amendment for Brodecky Toyota in 2003,[9] but also established an Ad Hoc General Plan Review Committee to evaluate the “applicability of Measure E to banking of commercial acreage and residential density”.[10] The Ad Hoc General Plan Review Committee comprised two members of the City Council, Mayor Pro Tem Bob Wilson and Councilmember Claudia Bill-de la Peña, and two members of the Planning Commission, Commissioner Tom Glancy and myself.

In the committee’s report back to the Council, one of the interpretations included processing “a simultaneous increase/decrease in commercial acreage or residential density” as a “single General Plan amendment, [which] could involve several properties” and that “[t]he scope of ‘net’ change could be clearly assessed by comparing/balancing the offsetting loss/gain in acreage or density among the properties.”  The other interpretation involved creating a “bank” for any “reduced commercial acreage or the number of reduced residential units resulting from [a General Plan] amendment…approved since Measure E was adopted…” and that “[a] future request… to increase commercial acreage or residential units elsewhere within the City’s Planning Area could draw from the ‘bank,’ reallocating the reserved acreage or units to another site or sites… This option would permit greater flexibility… in that the acreage or units could be reallocated as needed, without going to the voters for approval.”[11]

Councilmember Bill-de la Peña and I supported the interpretation of evaluating each general plan amendment on its own, consistent with how Measure E was originally proposed assuming “concurrent” increases and decreases; Mayor Pro Tem Wilson and Commissioner Glancy backed the banking option, which aligned with the preferred interpretation from the City Attorney’s office, based on previous memos.  Given the language of Measure E and the split among the committee, I recommended that we propose options for interpretation to the Council, as opposed to presenting a split recommendation; we all agreed that these were the options for interpreting the measure, even if we didn’t agree on which interpretation was preferred.  The Council directed the City Attorney’s office to provide a deeper legal analysis, which was reported back to the Council the following year.

City Attorney Mark Sellers stepped down from his post after 21 years of city service in 2004,[12] and Amy Albano began her tenure as his permanent replacement in 2005.[13]  By mid-year, City Attorney Albano penned a formal eight-page legal opinion, guiding the city’s interpretation of Measure E.  In effect, the banking interpretation would stand, allowing previous reductions in residential density or commercial acreage to be used to balance future increases.

In her analysis, Albano said that the Council should not consider reductions as being stored in a “bank” but rather a baseline to judge an amendment against.[14]  She did argue, however, that some density reductions, even though approved after Measure E’s passage, shouldn’t be used to reduce the baseline, as these amendments were merely “ratifying current conditions” in the city and didn’t represent actual decisions by the Council to reduce density that could be reallocated later, since “to do so does not further the purpose of Measure E.” [15]  Specifically, two “housekeeping amendments” that “in total would reduce the commercial baseline by 4 acres and the residential dwelling unit baseline by 4,102 dwelling units”[16] were to be excluded as reductions eligible for reallocation.

The Council accepted Albano’s analysis, excluding these amendments merely “ratifying current conditions”, and as a result, the Council’s 2005 policy established a baseline of reductions for future reallocation of 1 acre for commercial use and residential density allowing for 368 units.    From this understanding, the Brodecky Toyota proposal was modified to reduce their commercial acreage request from 1.2 acres to 1 acre so that a popular vote was not needed.[17]

This implementation became the governing interpretation of Measure E for the next decade, establishing a political environment where only requests for minor development increases were considered acceptable.  Any sudden and sizable increases in residential development without voter approval would run counter to the stated intent of Measure E per the measure’s ballot argument that voters “should have the power to vote yes or no when significant changes to the General Plan are proposed.”[18]  Thousand Oaks had become a mature city where the annual development of residential units had slowed significantly; for example, only 137 residential development allotments were approved in 2005.    A decade later, creative interpretations of Measure E were revisited.

In 2015, former City Attorney Sellers, who helped draft Measure E, was now a member of the Thousand Oaks Boulevard Property Business Improvement District (BID), which represents a group that owns much of the property on Thousand Oaks Boulevard, including the City of Thousand Oaks. Sellers believed the measure should be reinterpreted to allow more development without a citywide vote.[19]

Regarding Albano’s memo guiding the Council’s 2005 implementation of Measure E, Sellers said in an interview with the Thousand Oaks Acorn that “(Albano) made an interpretation that (she said) was ‘in the spirit of Measure E.’ I and a number of other people feel that was a wrong interpretation.”[20]

“What we have to do is get the city to take a new look at it. The BID would obviously love to have the new interpretation,” Sellers said.  Under a new interpretation, Sellers said, “I believe the cap can be higher. In my 21 years as city attorney, I never saw a project that wasn’t underdeveloped (with) less homes than was contemplated by the zoning or the General Plan. Everything in this community has been underdeveloped. It’s a goal to use that flexibility.”[21]

Two years later, pursuing this newly conceived flexibility, the city conducted an analysis to identify areas where general plan changes could be made to create more units for reallocation and future development without seeking voter approval.  In the study, city staff identified 5,400 units that were “theoretically accessible for reallocation” as they were residential areas that “were built at a density that is within an entirely lower residential land use category.” [22] As an example, a development may have been designated on the general plan as medium density residential, allowing for 4.6 to 15 units per net acres; however, if the property were ultimately developed at 3 units per net acre, the general plan designation could be reduced to low density residential, which allows between 2 to 4.5 units per net acre, and the difference in residential densities can be used for reallocation.  The problem: such density reductions seem to be exactly the type the 2005 Council interpretation excluded from reallocation as “ratifying current conditions” – existing development at a lower residential land use category.

Although a Thousand Oaks Acorn editorial urged the Council to “[r]eject the amendment and stick with what the majority of residents feel they were promised: the right to vote ahead of the demands of developers”,[23] the Council moved ahead and amended the general plan to increase the available residential density balance by 1,088 units.[24]

Continuing the evolving interpretations of Measure E since its inception, the 2018 Council approvals raise further questions about whether the voters will be involved in the city’s general plan decision-making as intended; based on the differing political climates in 2005 and 2018, amendments of the same type – those “ratifying current conditions” – now seem to be treated differently.  Under the 2005 interpretation, the 2018 amendments would not be available for further allocation; however, if 1,088 units are indeed available for reallocation, this could open up the amendments excluded in 2005 to be available as well, leading to a potential increase of an additional 4,102 units.[25]

This could easily be interpreted as a “significant change” where the voters “should have the power to vote yes or no”[26] on such a proposal.  The general plan reductions approved in 2018 may not yet technically run afoul of the voter approval requirements; the real test will come when the Council entertains increases that attempt to tap into this arguably questionable reservoir of residential density.

The successes of voter approval measures are only as good as the text of the measures themselves.  If a measure’s language is clear and precise such as with the Parks Initiative, the voters’ will is protected; however, if the language is vague, the measure could potentially provide no protection at all, becoming dependent on the interpretations and decisions of a narrow council majority.

In the end, the effectiveness of Measure E as a growth control instrument has merely paralleled the desire of the measure’s advocates to control growth.  If a Council majority chose to argue against a particularly intense development, opponents could point to Measure E as a politically convenient limitation.  However, if the Council wanted to find ways to allow for growth, creative ways to navigate through Measure E’s language are explored, interpreting away the right to approve significant changes the voters thought they were guaranteeing by passing Measure E.

When restrictions are placed on the Council on what they can and cannot do, they can come in two forms:  restrictions that are so tight that no one tests the measure, and restrictions that are so loose that they can always be interpreted around when convenient.  When it comes to including the voters in city decisions, the Parks Initiative qualifies as the former, and Measure E serves as the latter.

As we consider significant changes to the city’s general plan, we should keep the rightful place of Thousand Oaks voters in approving these changes top of mind.  Honoring the voters’ intent is critical to healthy democratic governance.

===================

[1] Thousand Oaks Municipal Code, Section 9.2.204(b)

[2] Thousand Oaks Municipal Code, Section 9.2.204(d).  Exceptions exist in subsection (c) if the redesignation is deemed necessary to avoid an unconstitutional taking of a private landowner’s property.

[3] Thousand Oaks Municipal Code, Sections 9.2.203(b)(2) and (3), as listed.  However, research for this book indicates that the ordinance approved by the Council codifying Measure E contains errors.  The current ordinance listed in the Municipal Code appears to be an older version of Measure E and not the version presented to the voters for approval.

[4] Memo to City Council from Andrew P. Fox, Mayor, “Subject:  Growth Control,” April 22, 1996.

[5] Memo to City Council from Mark G. Sellers, City Attorney, “Subject:  Request of Mr. Brodecky on Thousand Oaks Toyota Expansion,” June 5, 2001.  Attached letter from Ed Jones.

[6] Ibid.

[7] Memo to City Council from Mark G. Sellers, City Attorney, “Subject:  Request of Mr. Brodecky on Thousand Oaks Toyota Expansion,” June 5, 2001.

[8] Memo to Chris Ronneberg, Associate Planner, from Nancy Kierstyn Schriener, Assistant City Attorney, “Subject:  LU 2002-226/Brodecky – Applicability of Measure E,” August 20, 2003.

[9] Minutes of the Thousand Oaks City Council, September 2, 2003.

[10] Minutes of the Thousand Oaks City Council, September 16, 2003.

[11] Memo to City Council from Ad Hoc General Plan Review Committee, “Subject:  1. Applicability of Measure E Limitations on General Plan Amendments that increase Commercial Acreage or Residential Density. 2.  Special Use Permits for Certain Commercial Uses in all Commercial Zones,” June 22, 2004.

[12] Minutes of the Thousand Oaks City Council, January 27, 2004.

[13] Minutes of the Thousand Oaks City Council, December 14, 2004.

[14] Memo to City Council from Amy Albano, City Attorney, “Subject:  Applicability of Measure E Limitations on General Plan Amendments That Increase Commercial Acreage or Residential Density,” June 14, 2005.

[15] Memo to City Council from Amy Albano, City Attorney, “Subject:  Measure E Opinion,” June 6, 2005.

[16] Ibid.

[17] Minutes of the Thousand Oaks City Council, July 26, 2005.

[18] Argument in Favor of Measure “E”, Sample Ballot and Voter Information Pamphlet, County of Ventura, City of Thousand Oaks, General Election, November 5, 1996.

[19] Bitong, Anna, “Measure E:  under the microscope,” Thousand Oaks Acorn, February 5, 2015.

[20] Ibid.

[21] Ibid.

[22] Memo to Andrew P. Powers, City Manager, from Mark A. Towne, Community Development Director, “Subject:  Measure E Residential Baseline,” October 24, 2017.

[23] “Give residents the vote they voted for,” Thousand Oaks Acorn, April 19, 2018.

[24] Minutes of the Thousand Oaks City Council, April 24, 2018.

[25] Memo to Andrew P. Powers, City Manager, from Mark A. Towne, Community Development Director, “Subject:  Measure E Residential Baseline,” October 24, 2017.

[26] Argument in Favor of Measure “E”, Sample Ballot and Voter Information Pamphlet, County of Ventura, City of Thousand Oaks, General Election, November 5, 1996.

You might think journalism and data science don’t really go together, but on that, I differ. Below are some thoughts on the topic and lessons we can draw from data science on how to make journalism better and more effective in these times.

To begin, my background isn’t journalism, but it is science. I believe that the goals of true journalism are the same as science — get to the truth about what is happening and inform others about that truth.

In decision theory, a topic that I enjoy and in which I have deep practical experience, we learn that there are two primary factors that lead to the decisions we make. One is the most likely explanation for the facts we observe, and the other is our own cost-benefit analysis about what to do as a result.

The first element — finding the most likely explanation for the facts — requires us to be objective about the data we have in front of us. It also requires that we are open to alternative explanations (usually referred to as hypotheses).

The data lead us to the truth, but only when we are asking appropriate questions. Some facts or data may look disparate and unintelligible when viewed collectively, but things can make perfect sense if they are put in the right context. This is where epiphanies can happen — when the most likely explanation for the data finally becomes apparent.

However, some hypotheses might fit the data perfectly yet aren’t at all robust to new information through repeated and independent examination. In data science, we call this a solution that overfits the data. In journalism, this might be called a conspiracy theory — an overly complicated explanation that likely wouldn’t hold up when exposed to new facts.

The truth comes out when the explanation we find is consistent through repeated and independent examination. This holds in science, in journalism, in criminal investigations, or in any other endeavor whose goal is actually to seek the truth.

The second element — evaluating costs and benefits — is different for each decision maker. We can conflate these two elements in our decision making — letting our desire for one outcome cloud our interpretation of the facts in a way to justify the decision.

Much like how a rainbow color scale can fool the eye into seeing sharp gradients in scientific data, journalism has the potential to color the story at the expense of the facts themselves. This can become propaganda (if used intentionally) or can serve the journalistic form itself at the expense of the truth.

Much has also been made about press “objectivity”. From a recent thread (https://twitter.com/KlasfeldReports/status/1011334068275961856) by Adam Klasfled (@KlasfeldReports), referencing work by Walter Lippmann, “neutrality” — a device to convince the reader of one’s accuracy or fairness — is not the same as “objectivity”.

We should strive — whether in journalism or in science — to seek the most likely explanation for the facts and separate out any approaches for assessing the right decision to make as a result. Facts and values should be separate.

In a lesson from decision theory, two decision makers can agree on the facts, and even on the likelihoods of certain explanations, yet *disagree* on the decision to be made. Each decision maker has different tolerances for error or differing ways to assess costs and benefits — in other words, different values.

One example: Sales and Engineering teams within a company can agree on a product and what’s needed to improve it. The Sales team will want to ship the product sooner than the Engineering team, since Sales wants to increase sales (so delay is bad) and Engineering wants to improve the product (so delay is good).

People will still have disagreements because they apply different values to the decision making situation. However, we should *all* strive to make those evaluations on the same set of facts.

That said, that *doesn’t* mean that if there are two possible decisions, there must be given equal treatment. While reporters, investigators, and scientists (and basically all humans) have bias, the lack of equal treatment is not evidence of bias in and of itself.

Some explanations for the facts are not likely, and thus should not really be given much reporting weight. If a plausible explanation for the facts is 10x (or 1000x) more likely than a conspiracy theory, treating each equally unduly injects the values of the conspiracy theorist.

Conspiracy theorists have reasons for choosing the conspiracy over the far more likely plausible explanation — self-rationalization, camaraderie with other true believers, or doubling down as a self-defense of previous choices.

However, journalism should not unwittingly conflate these value judgments in their reporting.

Again, one example is that the government may be actively lying to the public and the press about facts. The free press has an obligation to find the truth and confront the government for how it operates.

Spin around the edges has been practiced by both Democratic and Republican administrations, moving back and forth between liberal and conservative, but within the confines of normality. In science, this might be described as working within the linear regime.

Previous rules of thumb — such as “if the President says it, it’s news” or “we present what he says, the viewers can decide whether it’s true” — can work as effective operating principles in the linear regime.

However, what is happening now has broken past the linear regime. The President and his staff are lying — verifiably, directly and without remorse — on a daily basis. With this, we have entered the nonlinear regime, so our approach to coverage must evolve to remain objective to the truth.

In science, we may have had an experimental setup that was working just fine, maybe with some hiccups and tweaks along the way. However, if suddenly the data start to look very different, we should question what is happening with our setup and look for other explanations.

This doesn’t mean that we give up on our scientific objectivity. It does mean, though, that we can’t interpret the data the way we used to, since conditions on the ground have changed. We have to apply our same principles, but to a new problem. The same goes for journalism.

If we focus on the goals of journalism and the new facts on the ground, we will get to a better objective explanation of the facts and serve journalism and its audience better in these changing times.

With the increasing speed of information coming at us, how do we know what’s true and what’s not, or even worse – what’s fake?

Figuring out what’s true and false is tough, and then understanding what to do about it can be even tougher.  But we should recognize one aspect between lies and the truth.

Lies spread faster.  Here’s why.

In the information age, we’re understanding our world more and more in terms of information and data.  We’re also recognizing that we are (and always have been) making decisions based on this information – even to the point of automating many of these decisions.

We are explainers, building internal models in our head for describing what we see in the world, and our decisions are based upon what we come up with as that explanation.

Some explanations are good at describing the data, and others are not.  However, the most likely explanation – the truth underlying it all – does well at explaining what we’ve seen and anticipating what’s to come.

One example of a bad explanation is the conspiracy theory; they are easy to create, and there seem to be so many of them.  In data science, we call this phenomenon overfitting, where an overly complicated explanation for the data is proposed that (1) seems to fit the data we have perfectly and (2) is highly likely to fall apart with any new information.

On the other hand, a truthful explanation needs to meet two criteria:  (1) faithfully explain what we have observed and (2) be consistent any future observations.  In a way, truthful explanations are consistent throughout, as well as have the power of prediction.

False explanations spread faster than the true ones, because they are easier to come up with and aren’t burdened by being consistent with new information.   The false explanations will be created and disseminated quickly, because they only need to meet the first criterion.

When we search for the truth, we strive to meet both criteria, and that takes time.  Plus, the more certain we want to be about the truth, the longer it takes.

When we don’t care about certainty of the truth, this incentivizes weak or false explanations which can spread faster.  The intentionally false explanation – the lie – comes from malevolent intent to force a decision before the right explanation can be found.

The goal of the lie is of course not the truth, but a quicker than expected decision (usually by someone else) that might not otherwise be made if the truth were actually known.

There are times, such as in business and entrepreneurship, when quicker decisions are preferable and have positive intents.  In these cases, the point of the decision is to act quickly, gaining feedback needed in order to make better subsequent decisions.  “Perfect is the enemy of the good” is a mantra used to characterize this approach.

However, in areas such as science, criminal proceedings, or investigative journalism, where we are trying to find out what the truth really is, there is a bigger penalty for getting it wrong, so it’s important to gather the best information before the decision is rendered.

In these situations, beware of those who seek to hide the truth.  Those who seek to obfuscate – lie – are not in search for the truth.  They are trying to force an outcome that would be different than what would occur if the truth were uncovered.

Unfortunately, we are in a time where the lies have turned into deliberate attacks on the search for truth itself.

We’re seeing it with science with attacks on findings relating to climate change and theories of evolution.

We’re seeing it with lawful investigations into foreign influence and potential criminality.

We’re seeing it with attacks on the free press, investigative journalism, and the First Amendment itself.

These institutions seek truth and need a manner of protection in order to allow them to get to that truth.

Why?  Because, as we’ve discussed, getting to the truth – finding the best explanation for the facts – takes more time than finding and promulgating less valid (or likely false) explanations.  Thus, we find ourselves engaged with defending against these false explanations right at the time when we’re searching for the best one.

There’s no perfect solution, since getting ahead of the lies isn’t possible.  However, we can build up our systems to combat against the lies better.  Here are several things we can do:

Let the data be collected

Whether in science or journalism or investigations, we shouldn’t be afraid of collecting more independent information.  The truth comes out through repeated and independent examination, and we should strive to learn what the facts are.

Differences about what to do in light of the truth may still exist, but let our differences result from an acknowledged difference in values, not our collective lack of understanding or in differing sets of facts.

Be open about how the data was collected

Here lies the real challenge.  In science, this is easier to accomplish – we need to make sure everyone analyzing the results knows where the information came from.  There is no real value in hiding information about the sources providing the data.

However, in investigatory or journalistic work using confidential or unnamed sources, this becomes more difficult.  There is a real penalty for the sources being as transparent as possible.

The world proves itself to be a dangerous place, and there are interests out there who don’t want information to be known.  Silencing the source is in their interest, so transparency, while desired in an altruistic environment in search for the truth, could put people in danger.  The costs of not being transparent about information can sometimes be outweighed by the costs of harm that could come to those sources.

The costs don’t come from our desire to be transparent.  The costs come from bad actors (in autocratic countries, this is the government, and in free societies, these are liars and criminals) who wish to prevent truth from being exposed to hide their bad acts (and potentially give the impression that they are really good).

Striking that balance is key, but hopefully we will trend toward transparency and truth.

Protect the truthseekers

The scientists, investigators, and journalists who care about the truth need our protection.  We can’t leave them open to attacks from those with other nefarious agendas.

Sometimes they get it wrong – those in search for the truth would readily admit that – but seeking the truth will uncover those wrongs, with understandable explanations for what led to the original errors.

Society benefits by having truthseekers, so society needs to protect them.

Allow freedom to question the explanations

We don’t question “the data” – we can question the hypothesis for why the data look that way.

For example, in investigations, DNA of the defendant at the crime scene can be powerful evidence of guilt.  One explanation is that the defendant committed the crime and their DNA was left at the scene.  Might there, however, be other explanations?

Of course.  There could be prosecutorial misconduct or planting of evidence or a problem with chain of custody.  There are many possible explanations for the data – our job is to seek the most likely explanation for the facts.

It’s not the data that are bad; it’s the explanation for what the root cause is.

We need the ability to question the explanations we are hearing.  In fact, we should strongly encourage such questioning.  A vigorous discussion of the facts and how likely our explanations are to be the right ones is important in getting to the truth.

We must protect freedom of thought and expression because it’s necessary for seeking the truth.

Develop new technologies that help to close the gap between when falsehoods start and the truth is revealed

Again, by its nature, the truth takes time to be revealed, so the lies will always come first.  That said, we can best enable the truth’s unveiling through technology.  We need approaches that focus on connecting available information into the most likely explanation.   We need a way to shrink the time gap between when the lies start being disseminated and when the truth can catch up.

Fundamentally, if we care about the truth, we should allow the search for truth to occur.

We must support the search for truth, the approaches of finding and evaluating that truth, and the institutions, such as science, investigations, and journalism, that are in that business.

FAKE NEWS!

It’s amazing that we’ve now had our collective awareness heightened to the problem of fake news.  I get frustrated at times with the sheer nonsense that seems to swim in the public consciousness, but in search of what I can do about it, I figured I’d share something that happened to me recently.

I have many people in my Facebook feed that are professional, fun, and have real character, but there are some that carry a bit of an aggressive and intolerant flavor.  They’re Facebook friends for other reasons, and, while I could easily unfriend these folks, I resist the urge so that I remain open to what they have to say or what they share.  People have a variety of perspectives, and I feel it’s important to understand them, even if the purpose is to counter them with better information.  That said, it can be a bit trying at times, and here’s one of those times.

This image was spread widely via Facebook (29,218 shares at the time of this writing), finding its way into my own feed.  The clear implication was that it is unfair for Ivanka Trump to be criticized for what she wears, when Michelle Obama, who is lauded for her fashion sense, wore something ugly in an official setting.

Some of the less incendiary comments from the thread include:

• but it is ok for Moochell to dress like a pinata. she was a disgusting first lady

• Is there anything about the Trump family the press will not complain about. Frankly I am tired of their [expletive deleted] complaints. The press is awful

• …Michelle’s dress is horrid!! ”””

Images like these seem to spread so easily via Facebook Shares or Twitter retweets, potentially reinforcing any biases held by the people that view them.   The assumption that most of us hold is that the images we see are genuine, but in this case, it’s worth questioning what our own eyes see.

When I first saw this image, it didn’t seem right to me.  Did Michelle Obama really wear this dress?  It didn’t seem consistent with what I’ve understood.  I tried to evaluate the information with a critical eye, questioning whether it makes sense.  Sure enough – with a little investigating, I found that this image of the Obamas was altered somewhere along the way to superimpose the bullseye dress on the First Lady.

I could have stopped there, satisfied with my own discovery of the truth, but I decided to inform others, sending a comment to the Facebook thread focusing on the facts that the Michelle Obama image was faked.  My premise:  in order to have a better informed view of the world, the people in this thread needed to be presented with facts that (1) were true and (2) countered their current view.

Here was my post:

=====

At risk of receiving a lot of intemperate responses, this post was shared in my FB feed so given the epidemic of spreading fake news, it’s worth a reply to get some facts in the mix.

The image of Michelle Obama is fake – it was photoshopped from the attached image from 2009, where the Obamas were on their way to award a posthumous Medal of Honor to a soldier killed in Afghanistan, U.S. Army Sergeant First Class Jared C. Monti. (https://www.gettyimages.com/…/obama-awards-medal-of…)

Discourse is important and it’s important to share our views and opinions, but they need to be based on the facts. None of us benefit otherwise.

Facts matter. We should exercise our freedom of expression, but let us all agree to seek the truth and honor the facts, so that our views can be better informed.

=====

If you are already predisposed to see the Obamas negatively, the doctored photo is information that, upon a quick glance, confirms your understanding, therefore you might determine that it is “the truth”.

However, this violates what I believe to be a more scientific definition of truth — consistency under repeated and independent examination, and this phenomenon, which is referred to as confirmation bias, becomes a violation of independent examination.

The values people hold (“I can’t stand those Obamas…”) get in the way of the information presented to them (“the image you’re viewing, which is reinforcing your negative view of the Obamas, was faked”).  For us to seek the truth accurately, we need to be more open to the information presented to us, while simultaneously being critical of whether the information itself is consistent with other independent information.

This is truly the challenge of our time, that of arming us all with the ability to review information/data critically and make beneficial and well-informed decisions.  We are being inundated with more and more data, and much of our society’s decision making is being automated.  Much of what we’re experiencing with hand-coded bots, trolls, and disinformation campaigns are human-driven, but they really are decision-driven – automated technologies will be able to create the same kinds of chaos, only at scale, unless we understand and approach differently how we make better informed decisions in this ocean of information.

Given these challenges, here are three things we can do now (and learn to be better at):

1)  Evaluate the data – not just the source

To find the actual truth – the one which serves as the best explanation for what we observe – we can’t trust the source of information without question.  We can’t just assume that the New York Times is accurate because they are a trusted source for us.  For those that trust the Washington Post, there are others that trust Fox News.  If we only focus on the sources, we keep ourselves in our own information tribes and don’t get at the truth.

When we blindly believe what a trusted source tells us, we open ourselves open to confirmation bias, or worse, manipulation and coercion.  All of these effects lead to bad decision making, some with more severe impacts than others.

To counter this, we need to analyze the information, reporting, and data critically, and seek some manner of independent confirmation to see that all the information we receive is consistent with the explanation at hand.

2)  Engage – Conversation and communication is needed

Admittedly, it’s easier to just turn off the noise machine and speak only with those who already share our views.  But, if we are to be stronger as a nation, or even a global community, we need to share our perspectives, encourage a vigorous back-and-forth regarding our views, and jointly agree to hold that discourse in the presence of facts and in search of the truth.  The alternative is to disengage, leading us to destruction of “the other side” and their points of view for our own survival.  This is a natural tendency throughout all of human history, resulting in chaos, war, and worse, so tough work of engagement and striving for the truth is critical.

Again, as we start to rely upon automated decision making algorithms and artificial intelligence, these information processing biases will not go away – they will only become part and parcel of the automated information flows, so it’s an important phenomenon of which to be aware.

The sharing of knowledge, information, and perspectives is what is needed to get to a more accurate version of the truth and the best explanation for what happens in our world.

3)  Focus on the facts and choose to seek the truth

Facts matter, but values can differ.  Two people can, in fact, see the same information and agree on the root causes for what they see, and yet make two different decisions regarding what to do about it.

Even though different groups may hold contrasting values regarding what our decisions should be, such as Democrats and Republicans in Congress, both sides should agree to seek a common set of facts upon which to based our collective decision making.  It’s OK to disagree on the decisions, but we shouldn’t disagree on the facts.

If we focus on our tribes, we won’t seek truth; we’ll only seek “our” side winning.  If we seek the truth, ultimately, in the long run, we all win.

There is indeed much promise for data science, where predictive models and decision engines can target skin cancer in patient imagery, presciently recommend a new product that piques your interest, or power your self-driving car to evade a potential accident.

However, promise requires much effort for it to be realized. It takes a lot of work and brand new engineering disciplines that are not yet mature or even employed on a wide scale. As there is greater recognition of the value of data science, and the generation of data is increasing at exponential rates, this engineering effort is starting and will grow beyond its adolescence soon.

This is why we are at the advent of a new engineering discipline that can truly realize the promise of data science – a discipline that I call “analytics engineering”.

Consider the parallels.

Computer science became an established academic discipline in the 1950s and 1960s, and serves as a foundation for today’s technology industry, articulating the theory and application of algorithms, computation, logic, and information representation in building real computing devices. The applications of computer science have included code breaking in World War II, the creation of ENIAC and the IBM mainframe.

However, something happened in the 1970s to democratize the development of software – the announcement of the Altair. At the time, Intel sold a microprocessor for $10,000, but the Altair was only $400. At this price, the microcomputer became accessible to individuals – geeks who wanted to build their own computer. Club started meeting in Silicon Valley, such as the Homebrew Computer Club and the Altair Users Group, to show what could be done with these computers and how they could be programmed.

Hackers took hold of an industry and an explosion of innovation ensued. Steve Jobs and Steve Wozniak had formed Apple, Bill Gates and Paul Allen had launched Microsoft, and the personal computer was born.

Eventually, as industries were created and matured, a strong engineering discipline came to computer programming – a discipline we now refer to as software engineering.

Data Science includes the analysis of data and applying solid approaches to gain meaning or insight from the data. There are several fundamentals to the field of data science, which I’ve elaborated on before.

That said, the maturity of data science as a discipline is following a similar trajectory as that of software. Top universities such as Columbia, Cornell, and University of California-Berkeley now offer programs and degrees in Data Science, establishing the academic discipline.

With prototyping languages such as R and Python, which are downloadable free, anyone can literally start programming, working with data, and applying data science principles. The barrier to entry for becoming a data scientist is now nearly zero.

However, just because someone can do something doesn’t mean they can do it well. Becoming a true practitioner is important, and learning the disciplines of a craft through experience and hard work is a must. Additionally, firms that leverage data science capabilities cannot afford to deploy an 24/7 operational capability based on a model developed for free on someone’s laptop. More engineering specialty is required, which is where the industry is heading, just like software and other engineering disciplines.

Throughout the history of innovation, this maturity curve has followed a common path, being part of a great surge in capability and creativity, supported by solid engineering practice. With mechanical engineering, the Industrial Revolution was launched. Electrical engineering led to advancements such as electricity, radio, and television. Of course, software engineering has dawned the age of computing and the internet. What will analytics engineering bring? Possibly what’s needed to support the age of artificial intelligence.

Analytics Engineering encompasses a key set of specialties that are not yet in common practice. There is a promise for data science that all one needs to do is “just push a button and the models get developed”. Others say that there are many different models we could try, so if we try a thousand different models on the data, we’ll evaluate all of them and “pick the best one”.

These approaches are symptoms of technology hype, so we should take them with a grain of salt.

For example, even after many years of developing computer applications, software even today isn’t written with just a “push of a button”. Engineering practices are still needed (and need to be followed) for quality software to be shipped. Sure – hackers can prototype something quickly and demonstrate truly innovative capabilities. However, for this to scale, be reliable, and ultimately operational without frequent failures, engineering disciplines need to be employed.

In this new age, true analytics engineering disciplines are what is needed, tailored to the needs of analytics and decision modeling.

Data Science isn’t magic, and never will be.  Yet, more focused analytics engineering disciplines can be developed to become part of decision model development and improvement moving forward.  The promise of data science, machine learning, and artificial intelligence will depend on this trend, which makes this an exciting time for the industry.

Why is this important for data science? Imagine a ROC curve where the false positive rate is very low, say 1% at an acceptably high true positive rate.

Are we satisfied?  Consider the case where a decision model, say to identify high risk customers in a financial institution, is run on a database of 1 million customers.  A false positive rate of 1% would still yield a database of 10,000 customers that would need to be reviewed, and purely in error, since these are falsely flagged as high risk customers.  When you are working at scale, with millions if not billions of tests being run through our decision models, the performance of these models needs to demonstrate incredibly low false positive rates to be worth using.

As Analytics Engineering matures, here are some of the developments that we can expect:

– New metrics will be developed to compare model performance in more accurate ways, superseding effective yet crude metrics such as Area Under the Curve (or AUC).

– New analysis techniques will be leveraged to focus on insights gained from the tails of statistical distributions, which are the true drivers of false positive rates in decision models.

– Tools and technologies will be created and matured to manage models, control versions, and track audit changes in model development and deployment.

– Standards, similar to CMMI or Agile in the software engineering world, will be developed and gain traction to provide for more explicit best practices around the creation, management, and engineering of decision models.

Companies such as Netfilx, Tesla, Apple, Amazon, Google, Facebook, and others are already developing these disciplines in-house, as the success of their respective business models demand this advancement.  However, other businesses will need to leverage these capabilities soon to keep pace.

It’s an exciting time to recognize and help define what this new engineering discipline will become.  For data science, it’s currently like the Wild West of old – wide expanses, plenty of room to “stake your claim”, and a rush to “get in on” the field that is hot.  That said, we aren’t all cowboys and the West is being now tamed.

Welcome to the advent of Analytics Engineering.

Two of the biggest buzzwords in our industry are “big data” and “data science”. Big Data seems to have a lot of interest right now, but Data Science is fast becoming a very hot topic.

Below are my thoughts for an outline for such a set of fundamentals:

Fundamentals of Data Science

Introduction

The easiest thing for people within the big data / analytics / data science disciplines is to say “I do data science”. However, when it comes to data science fundamentals, we need to ask the following critical questions: What really is “data”, what are we trying to do with data, and how do we apply scientific principles to achieve our goals with data?

– What is Data?
– The Goal of Data Science
– The Scientific Method

Probability and Statistics

The world is a probabilistic one, so we work with data that is probabilistic – meaning that, given a certain set of preconditions, data will appear to you in a specific way only part of the time.  To apply data science properly, one must become familiar and comfortable with probability and statistics.

– The Two Characteristics of Data
– Examples of Statistical Data
– Introduction to Probability
– Probability Distributions
– Connection with Statistical Distributions
– Statistical Properties (Mean, Mode, Median, Moments, Standard Deviation, etc.)
– Common Probability Distributions (Discrete, Binomial, Normal)
– Other Probability Distributions (Chi-Square, Poisson)
– Joint and Conditional Probabilities
– Bayes’ Rules
– Bayesian Inference

Decision Theory

This section is one of the key fundamentals of data science.  Whether applied in scientific, engineering, or business fields, we are trying to make decisions using data.  Data itself isn’t useful unless it’s telling us something, which we’re making a decision about what it is telling us.  How do we come up with those decisions? What are the factors that go into this decision making process?  What is the best method for making decisions with data?  This section tells us…

– Hypothesis Testing
– Binary Hypothesis Test
– Likelihood Ratio and Log Likelihood Ratio
– Bayes Risk
– Neyman-Pearson Criterion
– Receiver Operating Characteristic (ROC) Curve
– M-ary Hypothesis Test
– Optimal Decision Making

Estimation Theory

Sometimes we make characterizations of data – averages, parameter estimates, etc.  Estimation from data is essentially an extension of decision making, a natural next section from Decision Theory.

– Estimation as Extension of M-ary Hypothesis Test
– Unbiased Estimation
– Minimum Mean Square Error (MMSE)
– Maximum Likelihood Estimation (MLE)
– Maximum A Posteriori Estimation (MAP)
– Kalman Filter

Coordinate Systems

To bring various data elements together into a common decision making framework, we need to know how to align the data.  Knowledge of coordinate systems and how they are used becomes important to lay a solid foundation for bringing disparate data together.

– Introduction to Coordinate Systems
– Euclidian Spaces
– Orthogonal Coordinate Systems
– Properties of Orthogonal Coordinate Systems (angle, dot product, coordinate transformations,
etc.)
– Cartesian Coordinate System
– Polar Coordinate System
– Cylindrical Coordinate System
– Spherical Coordinate System
– Transformations Between Coordinate Systems

Linear Transformations

Once we understand coordinate systems, we can learn why to transform the data to get at the underlying information.  This section describe how we can transform our data into other useful data products through various types of transformations, including the popular Fourier transform.

– Introduction to Linear Transformations
– Properties of Linear Transformations
– Matrix Multiplication
– Fourier Transform
– Properties of Fourier Transforms (time-frequency relationship, shift invariance, spectral
properties, Parseval’s Theorem, Convolution Theorem, etc.)
– Discrete and Continuous Fourier Transforms
– Uncertainty Principle and Aliasing
– Wavelet and Other Transforms

Effects of Computation on Data

An often overlooked aspect of data science is the impact the algorithms we apply have on the information we are seeking to find. Merely applying algorithms and computations to create analytics and other data products has an impact on the effectiveness data-driven decision making ability.  This section take us on a journey of advanced aspects of data science.

– Mathematical Representation of Computation
– Reversible Computations (Bijective Mapping)
– Irreversible Computations
– Impulse Response Functions
– Transformation of Probability Distributions (due to addition, subtraction, multiplication,
division, arbitrary computations, etc.)
– Impacts on Decision Making

Prototype Coding / Programming

One of the key elements to data science is the willingness of practitioners to “get their hands dirty” with data.  This means being able to write programs that access, process, and visualize data in important languages in science and industry. This section takes us on a tour of these important elements.

– Introduction to Programming
– Data Types, Variables, and Functions
– Data Structures (Arrays, etc.)
– Loops, Comparisons, If-Then-Else
– Functions
– Scripting Languages vs. Compilable Langugages
– SQL
– SAS
– R
– Python
– C++

Graph Theory

Graphs are ways to illustrate connections between different data elements, and they are important in today’s interconnected world.

– Introduction to Graph Theory
– Undirected Graphs
– Directed Graphs
– Various Graph Data Structures
– Route and Network Problems

Algorithms

Key to data science is understanding the use of algorithms to compute important data-derived metrics.  Popular data manipulation algorithms are included in this section.

– Introduction to Algorithms
– Recursive Algorithms
– Serial, Parallel, and Distributed Algorithms
– Exhaustive Search
– Divide-and-Conquer (Binary Search)
– Gradient Search
– Sorting Algorithms
– Linear Programming
– Greedy Algorithms
– Heuristic Algorithms
– Randomized Algorithms
– Shortest Path Algorithms for Graphs

Machine Learning

No data science fundamentals course would be complete without exposure to machine learning.  However, it’s important to know that these techniques build upon the fundamentals described in previous sections.  This section gives practitioners an understanding of useful and popular machine learning techniques and why they are applied.

– Introduction to Machine Learning
– Linear Classifiers (Logistic Regression, Naive Bayes Classifier, Support Vector Machines)
– Decision Trees (Random Forests)
– Bayesian Networks
– Hidden Markov Models
– Expectation-Maximization
– Artificial Neural Networks and Deep Learning
– Vector Quantization
– K-Means Clustering

Richard Feynman

Richard Feynman is one of the greatest scientific minds, and what I love about him, aside from his brilliance, is his perspective on why we perform science.   I’ve been reading the compilation of short works of Feynman titled The Pleasure of Finding Things Out, and I recently came across a section that really hit home with me.

The first was a small article by Brian Dumaine about the work being done at Applied Proteomics to identify cancer before it develops.  At Applied Proteomics, they use mass spectroscopy to capture and catalog 360,000 different pieces of protein found in blood plasma, and then let supercomputers crunch on the data to identify anomalies associated with cancer.  The company has raised $57 million in venture capital and is backed by Microsoft co-founder Paul Allen.  You can read the first bit of the article here.

The second is from the Word Check callout, showing how access to information is making the word a better place:

wasa: Pronounced [wah-SUH]

(noun) Arabic slang:  A display of partiality toward a favored person or group without regard for their qualifications.  A system that drives much of life in the Middle East — from getting into a good school to landing a good job.

But on the Internet, there is no wasa.

– Adapted from Startup Rising: The Entrepreneurial Revolution Remaking the Middle East by Christopher M. Schroeder