What are predictive analytics? And why does the idea create so much confusion?

The greatest barrier to the widespread impact of predictive analytics in higher education is adoption. No matter how great the technology is, if people don’t use it effectively, any potential value is lost.

In the early stages of predictive analytics implementations at colleges and universities, a common obstacle comes in the form of questions that arise from some essential misunderstandings about data science and predictive analytics.  Without a clear understanding of what predictive analytics are, how they work, and what they do, it is easy to establish false expectations.  When predictive analytics fail to live up to these expectations, the result is disappointment, frustration, poor adoption, and a failure to fully actualize their potential value for student success.

This post is the first in a series of posts addressing common misunderstandings about data science that can have serious consequences for the success of an educational data or learning analytics analytics initiative in higher education.  The most basic misunderstanding that people have is about the language of prediction. What do we mean by ‘predictive’ analytics, anyway?

Why is the concept of ‘Predictive Analytics’ so confusing?

The term ‘predictive analytics’ is used widely, not just in education, but across all knowledge domains. We use the term because everyone else uses it, but it is actually pretty misleading.

I have written about this at length elsewhere, but in nutshell the term ‘prediction’ has a long history of being associated with a kind of mystical access to true knowledge about future events in a deterministic universe.  The history of the term is important, because it explains why many people get hung up on issues of accuracy, as if the goal of predictive analytics was to become something akin to the gold standard of a crystal ball.  It also explains why others are immediately creeped out by conversations about predictive analytics in higher education, because the term ‘prediction’ carries with it a set of pretty heavy metaphysical and epistemological connotations.  It is not uncommon in discussions of ethics and AI in higher education to hear comparisons between predictive analytics and the world of the film Minority Report (which is awesome), in which government agents are able to intervene and arrest people for crimes before they were committed.  In these conversations, however, it is rarely remembered that Minority Report predictions were quasi-magical in origin, where predictive analytics involve computational power applied to incomplete information.

Predictive analytics are not magic, even if the language of prediction sets us up to think of it in this way.  In The Signal an the Noise, Nate Silver suggests that we can begin to overcome this confusion by using the language of forecasting instead.  Where the goal of prediction is to be correct, the goal of a forecast is to be prepared.  I watch the weather channel, not because I want to know what the weather is going to be like, but because I want to know whether I need to pack an umbrella.

In higher education, it is unlikely that we will stop talking about predictive analytics any time soon.  But it is important to shift our thinking and set our expectations along the lines of forecasting.  When it comes to the early identification of at-risk students, our aim is not to be 100% accurate, and we are not making deterministic claims about a particular student’s future behavior.  What we are doing is providing a forecast based on incomplete information about groups of students in the past so that instructors and professional advisors can take action. The goal of predictive analytics in higher education is to offer  students an umbrella when the sky turns grey and there is a strong chance of rain.

Ethical AI in Higher Education: Are we doing it wrong?

In higher education, and in general, an increasing amount of attention is being paid to questions about the ethical use of data. People are working to produce principles, guidelines and ethical frameworks. This is a good thing.

Despite being well-intentioned, however, most of these projects are doomed to failure. The reason is that, amidst talk about arriving at an ethics, or developing an ethical framework, the terms ‘ethics’ and ‘framework’ are rarely well-defined from the outset. If you don’t have a clear understanding of your goal, you can’t define a strategy to achieve it, and you won’t know if you have reached it if you ever do.

As a foundation to future blog posts that I will write on the matter of ethics in AI, what I’d like to do is propose a couple of key definitions, and invite comment where my assumptions might not make sense.

What do we mean by ‘ethics’?

Ethics is hard to do. It is one of those five inter-related sub-disciplines of philosophy defined by Aristotle that also includes metaphysics, epistemology, aesthetics, and logic. To do ethics involves establishing a set of first principles, and developing a system for determining right action as a consequence of those principles. For example, if we presume the existence of a creator god that has given us some kind of access to true knowledge, then we can apply that knowledge to our day-to-day life as a guide to evaluating right or wrong courses of action. Or, instead of appealing to the transcendent, we might begin with certain assumptions about human nature and develop ethical guidelines meant to cultivate those essential and unique attributes. Or, if we decide that the limits of our knowledge preclude us from knowing anything about the divine, or even ourselves, except for the limits of our knowledge, there are ethical consequences of that as well. There are many approaches and variations here, but the key thing to understand is that ethics is hard. It requires us to be thoughtful about arriving at a set of first principles, being transparent, and systematically deriving ethical judgements as consequences of our metaphysical, epistemological, and logical commitments.

What ethics is NOT, is a set of unsystematicly articulated opinions about situations that make us feel uneasy. Unfortunately, when we read about ethics in data science, in education, and in general, this is typically what we end up with. Indeed, the field of education is particularly bad about talking about ethics (and of philosophy in general) in this way.

What do we mean by a ‘framework’?

The interesting thing about the language of frameworks is that it has the potential to liberate us from much of the heavy burden placed on us by ethical thinking. The reason for this is that the way this language is used in relation to ethics — as in an ‘ethical framework’ — already presupposes a specific philosophical perspective: Pragmatism.

What is Pragmatism? I’m going to do it a major disservice here, but it is a perspective that rejects our ability to know ‘truth’ in any transcendent or universal way, and so affirms that the truth in any given situation is a belief that ‘works.’ In other words, the right course of action is the one with the best practical set of consequences. (There’s a strong and compelling similarity here between Pragmatism and Pyrrhonian Skepticism, but won’t go into that here…except to note that, in philosophy, everything new is actually really old).

The reason that ethical frameworks are pragmatic is that they do not seek to define sets of universal first principles, but instead set out to establish methods or approaches for arriving at the best possible result at a given time, and in a given place.

The idea of an ethical framework is really powerful when discussing the human consequences of technological innovation. Laws and culture are constantly changing, and they differ radically around the globe. Were we to set out to define an ethics of educational data use, it could be a wonderful and fruitful academic exercise. A strong undergraduate thesis, or perhaps even a doctoral dissertation. But it would never be globally adopted, if for no other reason than because it would rest on first principles, the very definition of which is that they cannot themselves be justified. There will always be differences in opinion.

But an ethical framework CAN claim universality in a way that an ethics cannot, because it defines an approach to weighing a variety of factors that may be different from place to place, and that may change over time, but in a way that nevertheless allows people to make ethical judgments that work here and now. Where differences of opinion create issues for ethics, they are a valuable source of information for frameworks, which aim to balance and negotiate differences in order to arrive at the best possible outcome.

 

Laying my cards in the table (as if they weren’t on the table already), I am incredibly fond of the framework approach. Ethical frameworks are good things, and we should definitely strive to create an ethical frameworks for AI in education. We have already seen several attempts, and these have played an important role in getting the conversation started, but I see the language of ‘ethical framework’ being used with a lack of precision. The result has been some helpful, but rather ungrounded and unsystematic sets of claims pertaining to how data should be used in certain situations. These are not frameworks. Nor are they ethics. They are merely opinions. These efforts have been great for promoting public dialogue, but we need something more if we are going to make a difference.

Only by being absolutely clear from the outset about what an ethical framework is, and what it is meant to do, can we begin to make a significant and coordinated impact on law, public policy, data standards, and industry practices.

How ed tech marketers are bad for higher education

A lot of ed tech marketers are really bad. They are probably not bad at their ‘jobs’ — they may or may not be bad at generating leads, creating well-designed sales material, creating brand visibility. But they are bad for higher education and student success.

Bad ed tech marketers are noisy. They use the same message as the ‘competition.’ They hollow out language through the use and abuse of buzz words. They praise product features as if they were innovative when everyone else is selling products that are basically the same. They take credit for the success of ‘mutant’ customers who — because they have the right people and processes in place — would have been successful regardless of their technology investments. Bad marketers make purchasing decisions complex, and they obscure the fact that no product is a magic bullet. They pretend that their tool will catalyze and align the people and processes necessary to make an impact. Bad marketers encourage institutions to think about product first, and to defer important conversations about institutional goals, priorities, values, governance, and process. Bad marketers are bad for institutions of higher education. Bad marketers are bad for students.

Good marketing in educational technology is about telling stories worth spreading. A familiar mantra. But what is a story worth spreading? It is a story that is honest, and told with the desire to make higher education better. It is NOT about selling product. I strongly ascribe to the stoic view that if you do the right thing, rewards will naturally follow. If you focus on short-term rewards, you will not be successful, especially not in the long run.

Here are three characteristics of educational technology stories worth telling:

  1. Giving credit where credit is due – it is wrong for an educational technology company (or funder, or association, or government) to take credit for the success of an institution. Case studies should always be created with a view to accurately documenting the steps taken by an institution to see results. This story might feature a particular product as a necessary condition of success, but it should also highlight those high impact practices that could be replicated, adapted, and scaled in other contexts regardless of the technology used. It is the task of the marketer to make higher education better by acting as a servant in promoting the people and institutions that are making a real impact.
  2. Refusing to lie with numbers – there was a time in the not-so-distant past when educational technology companies suffered from the irony of selling analytics products without any evidence of their impact. Today, those same companies suffer from another terrible irony: using bad data science to sell data products. Good data science doesn’t always result in the sexiest stories, even it it’s results are significant. It is a lazy marketer who twists the numbers to make headlines. It is the task of a good marketer to understand and communicate the significance of small victories, to popularize the insights that make data scientists excited, but that might sound trivial and obscure to the general public without the right perspective..
  3. Expressing the possible – A good marketer should know their products, and they should know their users. They should be empathetic in appreciating the challenges facing students, instructors, and administrators and work tirelessly as a partner in change. A good marketer does not stand at the periphery.  They get involved because they ARE involved.  A good marketer moves beyond product features and competitive positioning, and toward the articulation of concrete and specific ways of using a technology to meet the needs of students, teachers, and administrators a constantly changing world.

Suffice it to say, good marketing is hard to do. It requires domain expertise and empathy. It is not formulaic. Good educational technology marketing involves telling authentic stories that make education better. It is about telling stories that NEED to be told.

If a marketer can’t say something IMPORTANT, they shouldn’t say anything at all.

Does Student Success start with Diversity in Higher Ed Administration?

Twitter has finally begun to add tools to mitigate harassment.

Harassment on Twitter has been a huge problem in recent years, and the amount of poor citizenship on the platform has only increased post-election. Why has it taken so long to respond? On the one hand, it is a very hard technical problem: how can users benefit from radical openness at the same time as they are protected from personal harm? In certain respects, this is a problem with free speech in general, but the problem is even greater for Twitter as it looks to grow its user base and prepare for sale. On the other hand, Twitter insiders have said that dealing with harassment has simply not been a priority for the mostly white male leadership team. Diversity is famously bad at Twitter. A lack of diversity within an organization leads to a lack of empathy for the concerns of ‘others.’ It leads to gaps in an organization’s field of vision, since we as people naturally pursue goals that are important to us, and what is important to us is naturally a product of our own experience. Values create culture. And culture determines what is included and excluded (both people and perspectives). Read more

Student Success and Liberal Democracy

The political environment in the United States has increasingly highlighted huge problems in our education system. These problems, I would argue, are not unrelated to how we as a country conceptualize student success. From the perspective of the student, success is about finding a high-paying job that provides a strong sense of personal fulfillment. From the perspective of colleges and universities, student success is about graduation and retention. From the perspective of government, it’s about making sure that we have a trained workforce capable of meeting labor market demands. For all of the recent and growing amount of attention paid to student success, however, what is woefully absent seems to be any talk about the importance of education to producing a liberal democratic citizenry. In the age of ‘big data,’ of course, part of this absence may be the fact that the success of a liberal education is difficult to measure. From this perspective, the success of a country’s education system cannot be measured directly. Instead, it is measured by the extent to which it’s citizens demonstrate things like active engagement, an interest/ability to adjudicate truth claims, and a desire to promote social and societal goods. Now, more than any time in recent history, we are witnessing the failure of American education. In the US, the topic of education has been largely absent from the platforms of individual presidential candidates.  This is, perhaps, a testament to the fact that education is bad for politics.  Where it has been discussed, we hear Trump talk about cutting funding to the Department of Education, if not eliminating it entirely. We hear Clinton talk about early childhood education, free/debt-free college, and more computer science training in k-12, but in each of these cases, the tenor tends to be about work and jobs rather than promoting societal goods more generally.

But I don’t want to make this post about politics. Our political climate is merely a reflection of the values that inform our conceptions of student success. These values — work, personal fulfillment, etc — inform policy decisions and university programs, but they also inform the development of educational technologies. The values that make up our nation’s conception of ‘student success’ produce the market demand that educational technology companies then try to meet. It is for this reason that we see a recent surge (some would say glut) of student retention products on the market, and relatively few that are meant to support liberal democratic values. It’s easy to forget that our technologies are not value-neutral. It’s easy to forget that, especially when it comes to communication technologies, the ‘medium is the message.’

What can educational technology companies do to meet market demands (something necessary to survival) while at the same time being attuned to the larger needs of society? I would suggest three things:

  1. Struggle. Keeping ethical considerations and the needs of society top of mind is hard.  For educational technologies to acknowledge the extent to which they both shape and are shaped by cultural movements produces a heavy burden of responsibility.  The easy thing to do is to abdicate responsibility, citing the fact that ‘we are just a technology company.’  But technologies always promote particular sets of values.  Accepting the need to meet market demand at the same time as the need to support liberal democratic education can be hard. These values WILL and DO come into conflict. But that’s not a reason to abandon either one or the other.  It means constantly struggling in the knowledge that educational technologies have a real impact on the lives of people.  Educational technology development is an inherently ethical enterprise.  Ethics are hard.
  2. Augment human judgment.  Educational technologies should not create opportunities for human beings to avoid taking responsibility for their decisions.  With more data, more analytics, and more artificial intelligence, it is tempting to lean on technology to make decisions for us.  But liberal democracy is not about eliminating human responsibility, and it is not about making critical thinking unnecessary.  To the contrary, personal responsibility and critical thinking are hallmarks of a liberal democratic citizen — and are essential to what it means to be human.  As tempting as it may be to create technologies that make decisions for us because they can, I feel like it is vitally important that we design technologies that increase our ability to participate in those activities that are the most human.
  3. Focus on community and critical thinking.  Creating technologies that foster engagement with complex ideas is hard.  Very much in line with the ‘augmented’ approach to educational technology development, I look to people like Alyssa Wise and Bodong Chen, who are looking at ways that a combination of embedded analytics and thoughtful teaching practices can produce reflective moments for students, and foster critical thinking in the context of community.  And it is for this reason that I am excited about tools like X-Ray Learning Analytics, a product for Moodle that makes use of social network analysis and natural language processing in a way that empowers teachers to promote critical thinking and community engagement.

The Trouble with ‘Student Success’

I’m increasingly troubled by ‘student success,’ and am even somewhat inclined to stop using the term entirely.

The trouble with ‘student success,’ it seems to me, is that it actually has very little to do with people. It’s not about humans, but rather about a set of conditions required for humans to successfully fill a particular role: that of a student.

So, what is a student?

A student (within the context of higher education, and as the term is employed within student success literature) is someone who is admitted to an institution of higher education, is at least minimally retained by that institution (many colleges and universities require at least 60 non-transferred credit hours in order to grant a degree), and graduate with some kind of credential (at least an Associate’s degree, but preferably a Bachelor’s). The student is the product of higher education. It is the task of colleges and universities to convert non-students into students (through the admissions process), only to convert them into a better kinds of non-students (through the graduation process). The whole thing is not entirely different from that religious process whereby an individual must first be converted from an a-sinner (someone who doesn’t grasp what sin is) into a sinner (they need to learn what sin is, and that they have committed it) in order to be transformed into a non-sinner through a process of redemption.

The language of ‘student success’ assumes that ‘being a student’ is an unmitigated good. But being a student is not a good in itself. The good of being a student is a direct consequence of the fact that being a student is requisite for attaining other higher goods. Having been a successful student is necessary in order to become a good worker. From the perspective of the individual, having been a successful student translates into being able to get a better job and earn a higher salary. From the perspective of a nation, a well-educated populace translates into an ability to meet labor demands in the service of economic growth. If this is the end of being a student, then, shouldn’t we talk about ‘Worker Success’? Replacing ‘student-‘ with ‘worker-‘ would retain every feature of ‘student success,’ but with the advantage of acknowledging that a post-secondary degree is not an end in itself, but is rather in the service of something greater. It would be more honest. It might also have the effect of increasing graduation rates by extending the horizon of students beyond the shoreline of their college experience and out toward the open sea of what will become something between a job and a vocation.

But I find the idea of ‘worker success’ still troubling in the same way as ‘student success.’ As with ‘student success,’ ‘worker success’ speaks to a role that humans occupy. It refers to something that a person does, rather than what a person is. As with being a successful student, being a successful worker implied having satisfactorily met the demands of a particular role, a set of criteria that come from outside of you, and that it is incumbant upon you to achieve. A successful student is someone who is admitted, retained, and graduates and so it is unsurprising that these are the measures against which colleges and universities are evaluated. A successful institution is one that creates successful students. Pressure is increasingly being put on institutions to ensure that students find success in career, but this is far more difficult to track (great minds are working on it). A successful worker is one who earns a high-paying job (high-salary serving as a proxy for the amount of value that a particular individual contributes to the overall economy).

What if we were to shift the way that we think about student success, away from focusing on conditional and instrumental goods, and instead toward goods that are unconditional and intrinsic? What if we viewed student success, not as an end in itself, but rather as something that may or may not help human beings contract their full potential as human beings? Would it mean eliminating higher education as it is today? I don’t think so. I’m not a utopian. I readily understand the historical, social, cultural, and material conditions that make school and work important. To the contrary, shifting out perspective toward what makes us human may in fact serve to underline the importance of an undergraduate education, and even of that piece of paper they call a degree. To the extent that an undergraduate education exposes minds to a world of knowledge, at the same time as it provides them with an opportunity to earn a good wage means that they are freed from the conditions of bare life (i.e. living paycheck to paycheck) and can commit their energies to higher order pursuits. Considered in this way, the importance of eliminating achievement gaps on the basis of race. ethnicity, gender, income, etc is also increased. For these groups who have been traditionally underserved by higher education, what is at stake in NOT having a post-secondary credential is not just a wage, but also perhaps their potential as human beings. At the same time as it make higher education more important, considering the student journey from the perspective of human success also opens up legitimate alternative pathways to formal education through which it is also possible to flourish. Higher education might be a way, but it might not be the way. And that should be okay.

I don’t know what this shift in perspective would mean for evaluating institutions. As long as colleges and universities are aimed at producing student-graduates, their reason for being is to solve a tactical problem — “how do we admit and graduate more students” — and they can be evaluated empirically and quantitatively by the extent to which they have solved the problem. The minute that colleges and universities start to reconceive their mission, not in terms of students, but in terms of humans, their success becomes far more difficult to measure, because the success of students-as-humans is difficult to measure. By thinking of education as a way of serving humans as opposed to serving students, our task becomes far more important, and also far more challenging.

But since when were the Good and the Easy the same thing?

The Costs of Privacy

In November 2012, in response to threats of expulsion from John Jay Science & Engineering Academy on account of her refusal to wear a mandatory RFID badge, Andrea Hernandez filed a law suit against San Antonio’s Northside Independent School District. If she continues to refuse even to wear an RFID-disabled badge–an accommodation sanctioned by a federal district judge who ruled against her–Hernandez will be placed in Taft High School beginning in September 2013, the public school to which she would normally be assigned.

In refusing to wear even an RFID-disabled badge, Hernandez’s case seems to have lost its ‘bite’ (it’s difficult to justify her appeal to religious freedom once tracking mechanisms are disabled). In spite of the fact that her concerns were ultimately voiced in terms of an interest in preserving religious freedom, however, the case nonetheless draws attention to the potential costs of privacy.

As elite institutions increasingly adopt comprehensive analytics programs that require students to give up their privacy in exchange for student success, are they also strongly contributing to a culture in which privacy is no longer valued? A robust analytics program requires every student to opt-in (i.e. students are not given the option of opting out). If analytics programs are seen as effective mechanisms to increase the chances of student success, and such programs are effective only to the extent that they gather data that is representative of their entire student body, and, as such, consenting to being tracked is made a condition of enrollment at the most elite universities (universities with the resources necessary to build and sustain such programs), then students must ask what it is that they value more: an education at a world-class institution (and all of the job prospects and other opportunity that such an education affords), or the ability to proverbially click ‘do not track.’ My suspicion is that, if explicitly given the choice, the vast majority of students are willing to give up the latter for the former, a symptom of our growing acceptance of, and complacence toward, issues of electronic privacy, but perhaps also an indication that a willingness to sacrifice privacy for success increasingly forms a key part of the ‘hidden curriculum.’

Erasing Privacy

[Image Creative Commons licensed / Flickr user Alan Cleaver]

(Interestingly, in addition to gathering data from Learning Management and operational systems, universities also regularly collect data from student id card swipes. This data can easily be mobilized as part of a kind of ‘card-swipe surveillance’ program, as in fact has been done by Matthew S. Pittinksy (co-founder of Blackboard) at Arizona State University. According to Pittinsky, tracking card-swipe behavior can allow an institution to effectively map a student’s friend group, determine their level of social integration, and predict their chances of attrition.)