Education, privacy, and big data algorithms: Taking the persons out of personalized learning
First Monday

Education, privacy, and big data algorithms: Taking the persons out of personalized learning by Priscilla M. Regan and Valerie Steeves



Abstract
In this paper, we review the literature on philanthropy in education to provide a larger context for the role that technology company foundations, such as the Bill and Melinda Gates Foundation and Chan Zuckerberg Initiative, are playing with respect to the development and implementation of personalized learning. We then analyze the ways that education magazines and tech company foundation outreach discuss personalized learning, paying special attention to issues of privacy. Our findings suggest that competing discourses on personalized learning revolve around contested meanings about the type of expertise needed for twenty-first century learning, what self-directed learning should look like, whether education is about process or content, and the type of evidence that is required to establish whether or not personalized learning leads to better student outcomes. Throughout, privacy issues remain a hot spot of conflict between the desire for more efficient outcomes and a whole child approach that is reminiscent of John Dewey’s insight that public education plays a special role in creating citizens.

Contents

Introduction
Philanthropy in education
Tech foundation activities in K-12 education
Materials and methods
Results and discussion
Implications and conclusions

 


 

Introduction

A number of major foundations in the United States have been supporting the introduction of personalized learning systems into public education. Many of these systems — which are the focus of this paper — utilize software programs that rely upon the collection of personal data from children as they go about their daily work in and out of the classroom; algorithms then analyse the data to identify individuals’ learning needs and steer their learning activities. Other than the Carnegie Corporation, the private foundations who have been most supportive of personalized learning are those supported by the technology companies, including the Bill and Melinda Gates Foundation, the Chan Zuckerberg Initiative, and the Google Foundation [1]. These foundations not only have a vested interest in schools adopting the innovative technology they have developed but also have a vested interest in researching how young people respond to technological innovations and what new innovations might prove attractive.

These foundations are sometimes referred to as ‘new philanthropy’ or ‘venture philanthropy’ or ‘philanthrocapitalists’ to distinguish them from the more traditional foundations of the early twentieth century, funded by wealthy industrialists such as Andrew Carnegie. These new foundations usually create trusts with endowments that are designed to last and projects are funded from the endowment investments. Alternatively they may be created as limited liability corporations, which are permitted to invest in for-profit companies and to engage in political lobbying [2]. The rise of ‘new philanthropy’ in K-12 education raises a number of important issues. It presents yet another example of the transformation of the traditional role of public education as educating citizens to one of educating future workers and consumers, a contrast of liberal democracy with neoliberal democracy. However, we regard tech company foundation support for K-12 education as a new and significant phase in this trend because the activities of the tech company foundations advantage not just the workforce generally but also the revenue and political/social power of the companies themselves; this happens directly (through the sales of products and services) and indirectly through the ongoing inscription of young people into the emerging big data surveillance economy that is driven by the seamless and continuous collection of information about users as they go about their business on technical platforms. The activities of the edtech foundations are thus distinct from what is referred to as “data philanthropy” (Lev-Aretz, 2019; Taddeo, 2016) whereby private sector companies donate their data for an endeavor that is in the public interest, for the common good, or to advance solutions to social problems (e.g., donation of location data by a mobile provider to expedite humanitarian aid). Although the definition and categories of data philanthropy are being clarified, it is widely accepted to include the donation of sharing of data and in this case data is not being donated but collected and used by the companies to further a particular view of education, not one accepted as the public or common good, and the data remain in the hands of the private companies.

But most worrisome is that personalized learning represents a new phase of both consolidating and also using social and political power with three consequences. First, the use of personalized learning systems by young people is likely to desensitize them to surveillance of their activities and to messages targeted specifically to them, both of which are likely to further acceptance of the surveillance state (Zuboff, 2019). Second, this will lead to further weakening of a sense of the public in public education specifically (Dewey, 1916), as well as a sense of the public more generally (Dewey, 1927), and to exacerbating the negative effects of big data personalization on citizens’ ability to form a public to resist surveillance. Finally, given budgetary constraints, these private foundations offer a revenue source that few government jurisdictions can reject but that also allows for shrinkage of government control, further ‘hollowing out of the state’ (Rhodes, 1994).

Our interest in personalized learning systems flows from our ongoing research on privacy generally and specifically on privacy and youth. Interestingly, both students (Steeves, 2012a) and teachers (Steeves, 2012b) have raised concerns about a loss of privacy in the classroom due to the introduction of networked technologies, such as Internet-connected computer labs and wifi-enabled iPads; and existing privacy regulations have done little to constrain the collection of students’ data through those devices (Steeves, 2014; 2009). Many edtech innovations, particularly personalized learning systems, have now upped the ante by challenging a number of factors traditionally associated with privacy [3]. First are information privacy issues regarding collection and use of personal information without individual knowledge and consent, which in the U.S. is addressed in the Family Educational Rights and Privacy Protection Act. A second concern involves anonymity and protection of the confidentiality of data sets which become difficult in the era of big data when it is easier to reidentify individuals. A third concern involves surveillance or tracking of activities, such as a student’s learning process and location, which is a central activity in personalized learning systems. The ability of individuals to make their own choices, autonomy, is a fourth concern that is tested by personalized learning for both students and teachers. Possible discrimination among students by treating them differently based on factors that are masked in personalized learning systems constitutes a concern. Finally is the question of ownership of data about students which in the case of education should be the school but may also, or actually, be the edtech vendor.

As we began to investigate this issue, we identified two research traditions within the education community that were relevant to understanding interest in and attention to personalized learning systems. Education policy scholars, as discussed in detail below, have identified philanthropists as playing an increasingly important role in charting the course of education policy both in terms of setting priorities and in terms of influencing policy discussions. This has been true for policies such as educational standards, for measuring teacher effectiveness, and for training teachers. At the same time, education scholars have investigated the pedagogical implications of these policies and examined their results. Our analysis of the privacy implications in personalized learning systems builds upon the insights from these two research traditions investigating educational philanthropy, as well as our own research in privacy.

In order to provide a window for contextualizing and analyzing the range and roots of privacy issues related to personalized learning, this paper is based on a discourse analysis comparing the construction of personalized learning on tech foundation Web sites and in American trade magazines targeting teachers [4]. Although our analysis is focused on U.S. foundations and trade magazines, the U.S. market is particularly relevant to non-U.S. contexts since learning software developed by the major American tech companies has significant penetration in markets outside the U.S. Accordingly, an exploratory study of how personalized learning is talked about in the U.S. is likely to be relevant worldwide, as personalized learning software rolls out in the global marketplace. Specifically, we examine the construction of the meaning of learning/discovery and teaching, the role of learners and teachers, and the interplay of public and private within the educational sphere. We argue that, although there has been no formal recognition, personalized learning as conceptualized by foundations marks a significant shift away from traditional notions of the role of education in a liberal democracy and raises serious privacy issues that must be addressed.

The paper proceeds as follows. First, we review the literature on philanthropy in education which provides the larger context for the role that technology company foundations appear to seek in this space. Second, we examine the activities that technology company foundations have supported in K-12 education and categorize the trends and goals. Third, we explain the data and research methods we use in our analysis. Fourth, we analyze the ways that education magazines and tech company foundation outreach discuss personalized learning and their support for personalized learning applications. Fifth, we specifically examine the privacy issues that are raised by personalized learning and analyze how the tech company foundations and trade press discuss issues of privacy associated with personalized learning. Finally, we discuss the larger implications for the role of education in a liberal democracy raised by tech company foundation support for personalized learning.

 

++++++++++

Philanthropy in education

The policy space around K-12 public education is complex in nature, congested in terms of number of participating parties, and contested in terms of goals and priorities. Although foundations, such as the Carnegie Foundation and Annenberg Foundation, have long provided support for education generally and for K-12 public education specifically, research indicates that foundation giving to education in the U.S. has undergone a fundamental shift since the 1980s and 1990s. Earlier foundation giving was targeted within the traditional public school system while newer foundation giving has been more disruptive [5]. Newer foundations, in contrast, were early investors in organizations challenging traditional public education institutions, including Teach for America, and charter schools (Scott, 2009), and the jurisdictional control of the public sector on K-12 education [6]. At the same time, the federal government has become more actively involved in directing K-12 education, as demonstrated by the No Child Left Behind (2001), Every Student Succeeds (2015) and Race to the Top (2009) initiatives. Given the larger federal role, foundations are not restricted to working at the state and local levels and can play a more coordinated and focused role in education policy (Henig, 2013).

Although this trend is most prominent in the U.S., it is also seen in other countries, partly because many of the large foundations provide grants internationally. In an analysis of education policy and philanthropy in England, Ball and Junemann (2011) point out that modern forms of philanthropy are tied to corporate social responsibility and are interested in ‘giving’ to ‘outcomes.’ They quote from an interview with a corporate sponsor who said that his company wanted to ‘create an environment in which we can continue testing ideas and continue evolving and from which we can influence practice in other places.’ [7] Additionally, corporate philanthropists use ‘forms of business research and due diligence to identify or vet potential recipients of donations, and the use of metrics and other indicators to monitor the impacts and effects of donations on social problems.’ [8] These same strategies are seen in the U.S.

In one of the largest social science studies of modern foundation giving for K-12 education in the U.S., Reckhow and Snyder (2014) analyzed grant data from the 15 largest grant organizations funding K-12 education for the years 2000, 2005, and 2010 to determine how their activities have changed over this period. Interestingly, six foundations were among the largest in both 2000 and 2010, including two related to tech companies — Bill and Melinda Gates Foundation and Michael and Susan Dell Foundation [9]. Their analysis demonstrated that philanthropic funding has increasingly been directed to national advocacy and research groups, especially professional organizations that produce reports and recommendations and maintain a presence in Washington D.C. In the period 2000–2005, foundations were participants in policy debates on charter schools and common standards, and increasingly supported grants to non-traditional organizations such as charter schools and teacher training organizations rather than public sector institutions. Moreover, these largest grant organizations tend to support the same issues and have the same policy priorities, often funding the same projects and non-traditional organizations [10]. The increased activities aimed at influencing the political process were identified in earlier research as well (Reckhow, 2013; Scott, 2009; Thümler, 2011).

There are at least three schools of thinking about the role of the new philanthropy in K-12 education. One school, vocal in social science academic scholarship, is that corporate foundation funding is not motivated by public interest but by self-interest in economic returns, either in the short-term by placing their hardware or software in schools or in the long-term by changing the nature of K-12 education to be more technology and data based, as well as disseminating ideas and values supportive of the social and market conditions they favor and that redound to their corporation’s bottom line (Abowitz, 2000; Boyles, 2005). Gurn (2016) identifies this as the ‘privatization or corporatization in public education,’ a discourse that Diane Ravitch (2010) and others (Boyles, 2005; Barkan, 2011) epitomize as the Billionaire Boys Club of the foundations sponsored by Gates, Broad, and Walton. In this school of thinking, these ‘venture philanthropists’ are seen as funding educational initiatives that express the values of the market, that involve routine contact between donors and recipients, and that enhance the commercial or public reputation of the donor [11]. Saltman (2010) notes that the expected returns on these investments are not financial per se but often accrue in terms of data, including tracking and monitoring results of new programs in order to improve programs. In the area of personalized learning, this is consistent with Zeide’s view of using data on student experience to ‘beta test’ edtech innovations [12]. These scholars are concerned that foundation investments will undermine more traditional goals of public education, such as fostering citizenship and moral values, and instead promote consumerism and materialism (Boyles, 2005; Cuban, 2004). Morsy’s (2015) analysis of survey results from 49 corporate foundations active in K-12 education philanthropy in late 2010 concluded that

‘The corporate philanthropists in this study report that their giving is meant to benefit the company primarily by training the future workforce and improving community relations. Taken together, these factors suggest that corporate foundations are less focused on actively promoting educational reform and challenging current educational policies and program but instead are interested in using philanthropy to address more immediate needs of the company, including marketing and establishing and maintaining positive relationships with the communities in which they operate.’ [13]

The second school of thinking is positive and welcoming to new philanthropy; this school is less represented in social science research and more in the trade and foundation press. Foundation dollars are viewed as ‘manna from above,’ with a focus on the benefits that will accrue to all involved, an optimistic sense of a bright future, and trust in the altruism of the business community. The picture that is presented is one of partners working together collaboratively to solve difficult problems with the active involvement of all [14]. As Selsky and Parker (2005) point out there is no recognition of the disparities in organizational logistics or the power dynamics that exist between schools and corporate-related foundations to challenge the notion of establishing a mutually beneficial relationship.

Finally, the third school of thinking about new philanthropy in K-12 education is skeptical but not entirely critical and adopts the position that corporate philanthropists can play a valuable role but need to be more accountable and transparent about what they fund and why they select certain programs and priorities. Gurn (2016), for example, cautions that ‘exercising skepticism and casting doubt on the presumed value of partnerships must not be conflated with a blanket presumption of exploitative motivations or deleterious outcomes’ [15]. Instead ‘educators need to critically and fairly assess the breadth of interests, aims, and effects of each partnership’ [16]. Hess and Henig (2015) similarly ‘believe that what we call muscular philanthropy can play an enormously healthy role ... this kind of giving can prove a valuable catalyst ... [but] also poses important questions about who gets to influence public decisions and how they should do so’ [17] and ‘brings with it a new level of civic responsibility’ [18].

 

++++++++++

Tech foundation activities in K-12 education

Before proceeding with our identification of activities in which tech foundations engage, some context on personalized learning and education technology (edtech) is briefly presented. Across the globe more focus is being placed on the promises offered by edtech because of increases in the costs of providing education and concerns about financial responsibility, heightened consideration of accountability and results, elevated awareness of the range of teacher skills and student learning styles and needs. Technology companies recognize the huge market offered by K-12 education and aggressively develop and market their products to school districts, schools, and teachers. The benefits of edtech innovations promise more sophisticated analyses of student learning and testing, more personalized learning, more effective delivery of educational materials, improved assessment, and more responsiveness to student needs. On the downside edtech applications and products raise the possibility of discrimination as a result of profiling and tracking of students, as well as uses of student information for a wider range of purposes.

Our review of articles in the scholarly publications and in the popular press identifies five types of activities in which tech foundations engage in the K-12 space. These activities are ones that were identified in previous research by education policy scholars (see in particular Rethkhow and Snyder [2014] and Reckhow and Tompkins-Strange [2015]) and parallel the activities that we found with respect to edtech and personalized learning — grants to schools adopting edtech; financial support to edtech companies; support for media coverage of edtech; funding for edtech evaluation research; and, funding for advocacy groups. Each of these is described below and examples are provided; it may also be instructive to examine selected Gates Foundation grants found on its 2017 IRS Form 990-PF tax forms and summarized in Table 1.

 

Table 1: Selected Gates Foundation grants: 2017 tax year..
GranteeType of activityAmount funded (rounded US$)
Common Sense MediaAdvocacy580K
Council of Chief of State School OfficesPeak association/advocacy10M
KhanCharter schools2.75M
National Alliance for Public Charter SchoolsAdvocacy2M
New Classrooms Innovation PartnersNon-profit organization that creates transformative innovations for K-12 education that enable personalized learning for every student, every day. Design innovative learning models and support implementation.5M
New ProfitResearch on personalized learning4.2M
New Schools FundNon-profit funding and supporting entrepreneurs2.6M
New Teacher ProjectNon-profit with a mission of ensuring that poor and minority students get equal access to effective teachers. Michelle Rhee’s organization.7M
New Venture FundNon-profit supporting innovative and effective public interest projects, including in education.2M
RANDResearch5M
SchoolzillaEdtech company2M
Student Achievement PartnersNon-profit organization that assembles educators and researchers to design actions based on evidence that will substantially improve student achievement. Worked on Common Core.3M
Summit Public SchoolsCharter schools7M
Unbounded LearningTeacher training4.4M
WestedResearch, development, and services agency, works with education5M
Zearn LearningEdtech company1.8M

 

The first activity in which tech foundations engage, and the one that receives the most attention in the popular, scholarly and professional press, involves grants to public schools for adoption of edtech applications, including personalized learning initiatives, or to educational initiatives to organizations working in public schools (such as Teach for America) or to organizations providing alternatives to public schools (such as charter schools) (Williamson, 2018).

The second entails grants or some form of funding support for edtech companies. This activity receives somewhat less coverage in the popular and professional press but is at times picked up by these publications. Likewise, some of the academic articles are tracking activities in this area. Some of this funding is given directly to edtech companies, for example Gates funding for Zearn Learning and Schoolzilla, and some of this funding is given to venture capital firms that fund edtech companies, for example New Venture Fund and New Schools Fund.

The third area of activity is tech foundation support for coverage of edtech, especially coverage in publications directed to education professionals. This activity receives coverage as disclaimers in the professional publications and some attention in the popular press, but is being tracked by education policy scholars and noted in some of the academic articles. A 2011 New York Times article reported on the activities of the Gates Foundation in advocating to overhaul education policies including creating new advocacy groups and funding some journalists and media organizations. It spent US$78 million on advocacy in 2009 [19]. Tech foundations also directly or indirectly provide funding for reports on the state of personalized learning (not social science research but more opinions about personalized learning) analyzing its potential. A 2010 report, ‘Innovate to Educate: System [Re]Design for Personalized Learning,’ was supported by the Council of Chief State School Officers, which receives Gates funding, and the Software and Information Industry Association (SIIA).

A fourth area of activity is tech foundation funding for research into studies evaluating the results of edtech applications, including personalized learning. For example, a 2017 Education Week article reported that the Bill & Melinda Gates Foundation and Chan Zuckerberg Initiative (CZI) are jointly funding a grant for New Profit, an organization for research on personalized learning (Herold, 2017). The Bill & Melinda Gates Foundation funded three RAND reports on personalized learning in 2017 [20]. The reports had the standard disclaimer that ‘The findings and conclusions contained within are those of the authors and do not necessarily reflect positions or policies of the Bill & Melinda Gates Foundation.’

The fifth area of activity is tech foundation funding for advocacy groups who work in K-12 education. This area receives almost no attention in education publications or the popular press but is increasingly attended to by education policy scholars (Greene, 2015). A review of the Web sites of groups that advocate on edtech and privacy issues reveals that large tech company foundations are major contributors. The Data Quality Campaign, which tracks and analyzes state privacy laws, reports on its Web site that its work among its six contributors are the Bill & Melinda Gates Foundation, Chan Zuckerberg Initiative, and Michael & Susan Dell Foundation. Two of the leading foundation sponsors of the Future of Privacy Forum are the Bill & Melinda Gates Foundation and the Chan Zuckerberg Initiative. The William and Flora Hewlett Foundation is one of the largest donors to the Center for Democracy and Technology (CDT), and the Silicon Valley Community Foundation Solidarity Giving Fund is also a main supporter. CDT has recently launched a student privacy project, particularly focused on edtech, which is ‘partially supported by the Chan Zuckerberg Initiative DAF, an advised fund of Silicon Valley Community Foundation.’ [21]

Related to this last activity is the ‘revolving door’ of people who move from the tech company foundations into government positions in K-12 education. Reckhow and Tompkins-Stange (2015) note that Arne Duncan, President Obama’s first Secretary of Education, received Gates funding as superintendent of Chicago Public Schools and that his assistant deputy for innovation and improvement was Jim Shelton, a former education program director at the Gates Foundation [22] who later moved to the Chan Zuckerberg Initiative and is now with an edtech venture capital firm. Boltodano (2017) points out that Arne Duncan’s first chief of staff, Margot Rogers, came from the Gates Foundation and her replacement, Joanne Weiss, came from the New Schools Venture Fund [23] and currently serves on the boards of LearnZillion and the National Alliance of Public Charter Schools.

 

++++++++++

Materials and methods

In order to explore the ways in which technology company foundations talk about personalized learning and the use of student data, we reviewed Reckhow and Snyder’s list of the 15 largest grant organizations funding K-12 education for the years 2000, 2005, and 2010 to identify the technology company foundations most active in the area (see above). There were three tech foundations on the list: Bill & Melinda Gates Foundation (the first largest donor in all three years); Michael and Susan Dell (fourth largest donor in 2010); and William and Flora Hewlett Foundation (#12 in 2000, #9 in 2005, and #8 in 2010) [24]. We selected them for analysis and then added two philanthropic organizations linked to tech companies, Google.org and the Chan Zuckerberg Initiative. The Chan Zuckerberg Initiativen is not a typical non-profit foundation but a limited liability corporation funded by $US1 billion in Facebook stock donated to the Foundation by Zuckerberg and Chan. Google.org also is not a typical non-profit foundation but is the charitable arm of Google, representing the philanthropic goals of the company, largely based on the importance of data and empirical-based programs, not the goals of wealthy founders or directors. It has a $US100 million annual grants budget that funds non-profits that align with Google’s core values and what it terms ‘regional impact challenges’ (Alba, 2016) [25]. Both Chan/Zuckerberg and Google.org are active in the edtech sphere and have heavily supported personalized learning initiatives. They are also excellent exemplars of the ‘new philanthropy’ identified in the literature.

In Fall 2018, we did a search on each of the five tech company related foundation websites for their activities in education, with specific attention to the K-12 level. On the Michael & Susan Dell Foundation, these were categorized as ‘data-driven education.’ [26] On the Bill & Melinda Gates Foundation, the overall education focus was ‘Ensure More Children and Young People Survive and Thrive,’ and the first priority area was ‘enhance education through education.’ [27] The Chan Zuckerberg Initiative defined its overall theme as ‘A Future for Everyone: Our mission is to find new ways to leverage technology, community-driven solutions, and collaboration to accelerate progress in Science, Education, and within our Justice and Opportunity Work.’ [28] The William and Flora Hewlett Foundation strategy in K-12 education is to ‘Empower and equip every young person to become an engaged and thriving participant in society.’ [29] Under Google.org’s banner of ‘Data-driven, human-focused philanthropy powered by Google,’ one of the three areas where ‘technology and innovation can move the needle’ is education [30]. We downloaded the text published on the first two levels of the Web sites that covered education to identify themes in how they characterize personalized learning related activities, especially as relates to privacy. We also tried to identify the activities that each of these are engaged in to support think tank and academic reports, as well as to advocate (directly or indirectly) for personalized learning related activities.

Because we were interested in seeing if and how the narratives contained with the foundation Web sites were being replicated and/or resisted in the education sphere, we then collected every article that was published in Education Week for the five years from 2013 to 2017 that mentioned personalized learning. After consulting with several K-12 teachers and surveying publications addressing issues of relevance to teachers, we selected Education Week because of its generally recognized position within the field of K-12 education. Since it launched in 1981, it has been a leading U.S. commercial publication addressed to K-12 teachers and now has a print circulation of more than 50,000 a week plus 725,000 registered online subcribers. Owned by a non-profit and describing itself as non-partisan, it provides local, state and national news and analysis. It identifies itself as “America education’s newspaper of record.” [31] It receives funding from a number of foundations, including the Gates Foundation. It accordingly is a site where various actors involved in personalized learning, including, teachers, school administrators, developers, policy-makers and foundations, share their views. In addition, EdWeek intersects with all five activities in which tech foundations engage, as set out above. It both reports on grants to schools (first activity) and to edtech companies (second activity) to support the adoption of edtech, and accepts grants from edtech foundations to support its own coverage of edtech initiatives and products (third activity). It also reports on research studies evaluating these initiatives and products, as well as on the source funding for these studies (fourth activity). Although, in keeping with the general trend, it pays almost no explicit attention to the relationships between tech foundations and K-12 advocacy groups, it provides a window into the ‘revolving door’ between edtech companies, edtech foundations and government positions in K-12 education through its description of the work histories of individuals who are quoted in the articles.

The EdWeek articles were accessed through the libraries at the University of Ottawa and George Mason University. A content coding form developed by three coders (one Ph.D. student, one J.D. student and one professor) identified 521 articles relevant to edtech [32]. This dataset was then reviewed by a graduate student to extract those articles on personalized learning (N=44). The authors then read these articles to identify mentions of both personalized learning and foundations, and privacy. Each author read and analyzed each article and then we discussed our independent findings to ensure consistency and reliability.

Because we were interested in the ways in which Web site and EdWeek authors mobilized the various constructions identified in our thematic analysis — especially with respect to the role of teachers, students and parents, the central importance of data, and privacy with respect to personalized learning — we then analysed the data using Gee’s (2011b) discourse analysis toolkit. Discourse analysis is helpful because it starts from the assumption that ‘systems of meaning are caught up in ... socially defined practices that carry more or less privilege and value in society’ and, as such, the meanings articulated ‘cannot be considered neutral’ [33]. This perspective allowed us to link our ‘inquiry into meaning making ... to an exploration of [political and social] power’ [34] by ‘discovering the situation-specific or situated meanings of forms used in specific contexts of use.’ [35]

Since we were interested in the ongoing dialogue between foundations, educators, edtech companies, and policy-makers, we divided our data by author, and removed 15 articles by EdWeek authors who only wrote once on the subject of personalized learning [36]. That left 29 articles written by nine individual authors. Treating each article as the unit of analysis, we then asked the following questions of the data to determine how each author was positioning personalized learning in the broader discursive space where its meaning was (potentially) contested:

[W]hat specific meanings do listeners have to attribute to ... words and phrases given the context and how the context is construed?
Based on what was said and the context in which it was said, what needs to be filled in here to achieve clarity? What is not being said overtly, but is still assumed to be known or inferable?
[H]ow [are] words and grammatical devices [being] used to build up or lessen significance (importance, relevance) for certain things and not others?
[How do] the words and grammar being used in the communication connect or disconnect things or ignore connections between things?
[How are] words and various grammatical devices ... being used to build and sustain or change relationships of various sorts among the speaker, other people, social groups, cultures, and/or institutions?

The following section reports on our findings.

 

++++++++++

Results and discussion

In this section we identify and analyze the discourses on both the tech foundation Web sites and in Education Week articles regarding the role of foundations and a number of key aspects of personalized learning, including as you will see below: the role of teachers, students, and parents; the central importance of data; and, privacy and security issues. We begin with the language on the tech foundation Web sites.

Foundation Web sites

As noted above, we analyzed the text that appears on the first two levels of the five foundation Web sites (Dell, Gates, Hewlett, Chan Zuckerberg Initiative, and Google.org). Our goal here was two-fold: first, to compare the way they discuss personalized learning and the evidence for it, the roles of foundations, teachers, students, and schools, and privacy implications with the way that these topics are discussed in the Education Week articles; and second, to determine whether they are ‘singing from the same hymnbook’ in a way similar to what Reckhow and Tompkins-Stange (2015) found with respect to Gates’ and Broad’s advocacy grantmaking regarding teacher quality.

The foundations all have a view as to the role of education, what its shortcomings are, and how it can be improved. Gates sees education as the means to cross ‘the bridge to opportunity’ and states that a ‘high-quality education is a proven path to prosperity and participation in the American Dream.’ To quote directly: ‘In the United States, a high-quality public education is a bridge to opportunity — particularly when it comes to good jobs and career paths, social mobility, and personal growth and fulfilment.’ In terms of how to improve education, Gates wants ‘to expand the number of high-quality public charter schools, with a specific focus on improving outcomes for students with disabilities, including special needs students.’

The Chan Zuckerberg Initiative portrays American education as outdated, ‘established over a century ago, before the science of human development emerged as a field’ and highlights the importance of ‘accelerating the integration’ of human development into ‘a whole child approach to learning that accounts for the many different ways in which young people grow.’ Hewlett provides the first mention of the social/political importance of education as it ‘empower[s] and equip[s] every young person to become an engaged and thriving participant in society.’ Hewlett goes on to say: ‘Since our country works best when everyone has the chance to contribute, this waste of human potential weakens our society and democracy, leaving us less prepared to tackle the most pressing issues our nation faces today.’

With respect to personalized learning, all five of the foundations emphasize that there are differences in the ways student learn and the importance of ‘flexible learning opportunities’ (Hewlett), ‘the right experiences to help students learn’ (Dell), ‘a truly transformative, personalized learning experience’ (CZ), and ‘the right learning materials’ (Google), which leads to the importance of ‘real-time assessments for gauging student learning’ (Gates) and ‘formative data ... gathered as learning is happening ... in-the-moment use of data in the classroom’ (Dell). None of the five foundations, however provide a definition of what they actually mean by personalized learning instead describing the importance of data and differences.

Moreover, none of the five foundations offers actual evidence for the effectiveness of the innovations they are advancing although all discuss the importance of evidence. Dell, for example, speaks of a ‘proven approach to instilling problem-solving’ but does not explain further what this means or provide proof. Similarly, Gates posits ‘continuous improvement grounded in data and evidence’ and ‘evidence-based interventions’ but likewise goes no further in providing evidence. The Chan Zuckerberg Initiative notes it supports ‘the development of cutting-edge research on how people learn and the implementation of evidence-based practices that empower teachers to support each student according to their unique learning needs.’

Indeed, there appears to be an interesting side-step in the way that foundations discuss evidence — rather than relying on evidence for their innovations, they discuss the importance of gathering evidence for their innovations. Gates, for example, talks about ‘advancing research and development in support of new innovations,’ ‘invest[ing] in building the evidence for social and emotional learning,’ and ‘expand[ing] the evidence base and validat[ing] exemplars of interventions.’ Hewlett is interested in learning ‘what it takes to turn schools into places that empower and equip students’ and in ‘studying these efforts to share what is learned.’ In this side-step, the foundations reveal that not only are students learning — but foundations are learning as well, and students in this scenario are not only learners but research subjects beta-testing foundation-supported innovations. The recent CZ/Gates Initiative underscores this in the following: ‘The goal is to gain a better understanding of breakthrough work happening in these fields so that we, and many others across the field, can figure out how best to direct energies and resources to accelerate progress for students.’

The five foundations describe their roles in very similar, although not surprising, terms emphasizing partnerships and collaboration. Dell describes itself as ‘a funder, thought-partner, and collaborator,’ Gates notes working ‘with partners in the field,’ CZ collaborates ‘with students, educators, and families,’ Hewlett’s sees its role ‘to listen carefully to teachers, students, and parents, and equip them with what they need to share their stories,’ and Google connects ‘education innovators to catalytic funding.’

The foundations also portray teachers in very similar terms as lacking resources and knowledge that would make them more effective. Dell states: ‘Teachers are expected to differentiate teaching strategies to respond to student needs, but don’t often have the autonomy, tools, or skills in their classrooms to do so.’ Similarly, Hewlett notes that ‘Teachers lack the agency and resources they need to reach all their students, and schools are restricted in what, how, and when they can improve.’ The role of the foundations then is, as the Chan Zuckerberg Initiative states, to ‘empower teachers to support each student according to their unique learning needs.’ To that end, Gates invests in ‘ensuring that teachers and leaders have what they need to be successful — high-quality preparation, standards-aligned curriculum and tools, and professional learning opportunities.’ Dell points out that ‘Talented teachers with the right resources, information, skills and control of their classrooms, they are able to provide the most useful and relevant instruction to help their students achieve more.’ Google wants to give ‘teachers the latest skills and techniques.’

There is some variation in how Web sites voice their concerns about and views of children. A number of the foundations highlight concern about disadvantaged students and the need to focus attention on them. Gates, for example, speaks of ‘underserved students’ and its interest ‘in solutions that dramatically increase the number of black, Latino, and low-income high school students who have access to technology and engineering coursework, as well as credential-based pathways that give them access to good jobs.’ Google targets concern to students in ‘disadvantaged communities ... [who] lag behind their peers with more resources,’ a concern that is particularly directed globally. Beyond concern about disadvantaged students, the foundations’ narratives about students are somewhat distinct. Dell notes that ‘technology [is] a part of most students’ lives.’ The Chan Zuckerberg Initiative is interested in the ‘holistic needs of every child,’ and helping students cope with stress and forming healthy adult attachments. Hewlett emphasizes preparing students for ‘real life,’ a component of which is ‘a strong sense of agency and identity.’ Interestingly, parents are not often mentioned in the foundations’ overview, except as partners and part of the collaboration the foundations seek to form.

Perhaps most interesting in our review of foundations’ Web sites was the almost universal absence of any mention of privacy or the implications of collecting all this data on students’ learning and personal characteristics that would be a necessary component to implement personalized learning, as well as an outcome of that implementation. The only foundation to mention privacy is Dell, which states at the end of its overview: ‘While we believe data is a powerful tool in the classroom, we also believe it should be used in a safe and secure manner. Therefore, we work with leading organizations to improve data privacy policies and practices of education technology vendors, districts, and educators.’ The other foundation overviews mention neither privacy nor security — in effect, erasing this as a concern or implication of the vast amount of data that the kind of innovations they discuss would entail. The absence of this topic from their overviews is startling given the attention companies like Google and Facebook have been forced to pay to both privacy and security. It cannot plausibly be argued that they are unaware of privacy and security in the context of personalized learning, although it can be argued that they would prefer not to have attention drawn to these concerns.

Education Week articles

The EdWeek data set [see the Appendix for a list of articles in the data set] bifurcates into two, mostly separate, discourses. The first replicates the same themes we found in the foundation Web site materials. It consists of 14 articles written by eight authors, including senior EdWeek writers (the Executive Editor of Market Briefs, Associate Editor of Business and Technology Issues, Senior Writer for Digital Directions); all eight authors are explicitly assigned to cover ed-tech from a business perspective (e.g., ‘industry and innovation in K-12 education’, ‘ed-tech start-ups’, ‘marketplace trends’, and ‘education companies’). Their articles draw on a variety of players within the ed tech sector, and include quotes from industry reps, policy-makers, teachers, school administrators, parents, and students. For simplicity sake, we refer to this as the dominant discourse.

The second discourse appears almost exclusively in 14 articles written by Benjamin Herold, a staff writer who came to EdWeek from public radio and who covers ‘ed-tech, newsroom analytics, digital storytelling and Philadelphia’. As in the dominant discourse articles, these articles refer to the opinions and perspectives of a diverse group of players within the sector, and include quotes from industry reps, policy-makers, teachers, school administrators, parents, and students. This second set of articles challenges the assumptions made within the dominant discourse and presents an alternative construction of personalized learning. There are also two fragments from the articles contained in the dominant discourse sub-set that make passing reference to alternative constructions of the role of teachers and of privacy, respectively. We refer to the articles by Herold and these two fragments as the alternative discourse.

Dominant discourse

Unlike the alternative discourse, which expressly engages with dominant constructions of edtech in general and personalized learning in particular, the dominant discourse is almost entirely self-referential. The only time an article in the dominant discourse sub-set expressly refers to an alternative construction of personalized learning is in Executive Editor Kevin Bushweller’s introduction to the EdWeek’s special report on ed-tech. That report includes seven articles in favour of edtech — including one by a Senior Program Officer for the Bill & Melinda Gates Foundation and one by a think tank that aims to use personalized learning to transform the traditional ‘factory-model school design’ (article 22) of education — and one article by Herold entitled ‘Investments in Personalized Learning Rise, But Research on Its Impact Is Lacking’ (article 20). In the introduction to the report, Bushweller makes passing reference to Herold’s reported finding that research evidence to support comprehensive personalized learning is ‘thin, at best’, but then immediately frames this concern within a broader set of ‘challenges’ that must be met so we can get to the ‘next generation of personalized learning initiatives’ (article 20). In this way, the dominant discourse remains untroubled and Herold’s article is bracketed by the majority of articles that celebrate the potential of personalized learning.

This tendency to bracket or dismiss critique is common throughout the dominant discourse articles. Parties with concerns are overwhelmingly characterized as ‘skeptics’ who ‘pushback’ against the ‘optimistic vision for innovation’ inherent in personalized learning (article 7). On the flip side of the coin, parties who embrace personalized education are celebrated for ‘believing’ in technology. For example, teachers who are confronting the ‘messy, hectic’ ways in which personalized learning changes learning, ‘[e]ven so, see it as an exciting step in the right direction, because they believe it will improve student engagement and build the skills students need to succeed’ (article 7). Teachers who take this ‘higher-level approach’ and ‘really embrace technology’ will also ‘embrac[e] student-centred pedagogy’ in new and ‘transformative’ ways (article 12).

Education system: The transformation embodied in this ‘exciting step in the right direction’ (article 7) is necessary because — as in the foundation discourse — the education system is assumed to be broken. The articles are full of references to the failures of the ‘traditional drill-and-kill or lecture models’ (article 18) that are the hallmark of ‘traditional egg-crate factory-model school design’ (article 32), and the need to ‘disrupt and transform the ‘factory model’ of education’ (article 24). Moreover, this starting assumption is represented as being shared across sectors. The clearest indication of this is found in Molnar’s article entitled ‘Student-Centered Learning Top of Mind for Companies’ (article 36), where:

  • a school principal warns ‘[w]e have a lot of unlearning to do in schools’ because ‘to do what you always did’ is ‘the wrong approach’ (article 36);
  • a school board member asks, ‘how much are your efforts going to be put into the 20th-century model — improving upon things that don’t work? Or will you start investing significantly in thinking about where [personalized learning] is going?’ (article 36);
  • and edtech company representatives conclude that, ‘Old assumptions about what schools really need will be called into question’ so the ‘industrialized system’ of hierarchical classrooms can be transformed (article 36).

Personalized learning: Personalized learning technology is constructed as a ‘natural’ (article 22) solution to the problem of the broken education system, because the ‘technology-driven’ (article 36) feedback loop between curriculum and assessment will bring about ‘a sea change in learning’ (article 37) and a ‘whole culture shift in innovation and innovation practices for teaching’ (article 12). Moreover, the change is constructed as inevitable: personalized learning ‘is just going to become the way of the future and the way we operate ... Technology will just be baked into the DNA of [teachers’] work’ (article 12).

However, the meaning of personalized learning itself is ‘still largely ill-defined’ (article 6). Indeed, the definitions offered throughout the dominant discourse focus less on what the technology is or does, and more on what it is expected to accomplish. For example, various articles offered the following definitions of personalized learning as technology that will:

  • ‘design teaching and learning around students’ distinctive academic needs, and even their personal interests’ (article 5);
  • ‘tailor instruction to individual students’ strengths and weaknesses’ (article 5);
  • ‘essentially customiz[e] learning to each student’s strengths, weaknesses, and personal interests‘ (article 6);
  • ‘customize the educational process’ (article 11);
  • ‘[tailor] the learning experience ... to the individual student based on their specific learning needs’ (article 12);
  • ‘By definition ... allo[w] students to move at their own pace through material’ (article 34)

How personalized learning will do this — and exactly what data practices will drive the engine — is rarely part of the discourse and, when it is, little attention is paid to the nuts and bolts of how students’ personal information is collected and analysed. In one of the three main fragments that discuss data, a tech teacher notes that, ‘Products for personalized learning generally produce a lot of data’ and ‘I need the data to help make instructional decisions’ so it is important for teachers to figure out how to ‘make sense of’ that data (article 34). However, there is no discussion about the personal nature of the data, or how it is collected or processed. Similarly, in the second fragment, a policy-maker notes that personalized learning ‘involves data, formative assessments, dashboards’ but the role of that data is presented as unfamiliar territory because ‘we’re just moving into that phase now’ (article 12). In the third fragment, an edtech representative includes ‘learning analytics, mapping educational requirements, collecting valuable metadata, and tracking how it is used’ on a list of ‘key technological features that edtech companies will need to pay close attention to in the years ahead’, pushing difficult questions about information practices into the future (article 36).

In the meantime, the main emphasis in the discourse is placed on removing barriers to implementation. Challenges include insufficient funding, inconsistent implementation practices, and lack of vision on the part of schools who may not have the ‘far-reaching vision’ (article 6) required for success. But the biggest barrier identified in the sub-set is teachers who do not use technology. As in the foundation discourse, these teachers are presented as overwhelmed, under-skilled, and out-of-date. Teachers accordingly need help from technology to sort through the ‘dizzying array of options these days’ (article 37) because ‘millions of years of evolution have not prepared us to read streams of data’ (article 36).

Teachers: However, technology alone cannot fix the problem these teachers pose because it is ‘teachers’ beliefs about technology [that] hinder its effective use in the classroom, even with high-quality professional development and other supports’ (article 12). These beliefs ‘[hold] teachers back in terms of using technology in meaningful ways. We are seeing more effective uses of technology than we were, but sadly there are still pockets of very ineffective use’ (article 12). If these beliefs are not challenged, and teachers fail ‘to be committed to transforming the ways they teach ... personalized learning will never evolve beyond a buzzword’ (article 5).

Transformation is possible when, echoing the language used in the foundation discourse, teachers, government officials, and technology firms collaborate and work in partnership. For example, when ‘the technology team and the teaching and learning group work hand-in-hand [t]he result has been a collaborative approach to incorporating edtech into the district at all levels ... to modernize schools’ (article 12). From this perspective, teachers are not experts in the education process equipped to make decisions about how and when to use edtech; instead, they must embrace the fact that, because of technology, ‘they don’t need to know it all. They’re not the experts’ (article 12). Expertise resides in the edtech itself. Even teachers’ ‘intuition about their students’ academic and social issues’ (article 36) will be ‘replicate[d] ... via technology’ (article 36). Although there is a recognition that this kind of transformation is ‘disruptive’, it is to be embraced because ‘Change is hard ... This is what innovation does’ (article 6).

Students: Given the prominent role played in the discourse by promises that edtech will advance student-centred learning, it is notable that students are absent from almost all of the articles in the sub-set. When individual students appear, it is to celebrate their innovative use of technology. For example, in the article entitled ‘The Challenge: Turning State Law Into Classroom Reality’, Willem, ‘a senior with shoulder-length, wavy brown hair and glasses,’ is praised for designing a circuit board for a self-playing saxophone and for turning a toy car into a child-sized wheelchair (article 7). Although Willem is presented as exceptional — ‘the reality is most teenagers are not blessed with Willem’s motivation, intellectual curiosity, and technical skills — at least not at this point in their lives’ — the article concludes that ‘personalized learning can work for all students’ because it is ‘a powerful tool for analyzing the individual strengths and weaknesses of all students’ and, ‘If the kid gets the right project, they don’t stop thinking about it’ (article 7).

At the same time, the success of personalized learning is presented as the responsibility of the neo-liberal student at the heart of the endeavor. The technology will work if students take ‘greater responsibility’ (article 36) and ‘ownership of their own learning by setting their own goals’ (article 34) so they can ‘drive class projects’ (article 34) and ‘design their own path through the learning process’ (article 37). The burden placed on the student in the dominant discourse is actually quite high. In a world where edtech representatives ‘don’t think that [content] is the most important part’ (article 12), learning is not about exploring a curated set of resources with the help of a highly skilled teacher. It is instead, ‘inquiry-driven education in which students ask powerful questions that no one knows the answer to’ (article 36). The student is therefore exhorted to demonstrate ‘maturity’ and exercise choice.

Ironically, since the algorithm is designed to decide what content a student should be learning and indeed selects that content from a library of curricular resources, the student’s choices with respect to what to learn next and how to learn it are taken away from him or her and made by the technology on the student’s behalf. Personalization accordingly takes on a perverse meaning. Although the discourse acknowledges that the person at the heart of the process is complex and unique — ‘There are all these factors you have to consider: culture, family background, interests, learning strengths, social-emotional development. We could go on and on ... We’re talking about a human being’ — it also maintains that that person can best be known as a data stream flowing through edtech ‘products that provide ... more in-depth insights into ... students’ in order to steer their learning (article 34).

Effectiveness of personalized learning: Whether or not this will work is contested within the discourse, but only by non-believer teachers and parents whose ‘skepticism’ has to be ‘overcome’ (article 7). Suggestions that schools should first look for independent research that edtech leads to better learning outcomes before implementing it show a lack of vision: ‘The problem ... is that administrators want numbers; they want evidence that this is going to work before they adopt things. ... Skeptics will point that out and say, ‘Where’s the evidence’?’ (article 12).

Even when research is available, positive results that support the product are given prominence in the discourse and counter indications are ignored, as in this excerpt from the article entitled ‘Numbers to Watch’:

Promising Evidence? Over a two-year period, student achievement grew most in math and reading where personalized learning practices were used, compared to more traditional classrooms, according to a RAND study of initiatives funded by the Gates Foundation. The National Education Policy Center questioned whether the study contained ‘promising evidence’ as claimed, noting only ‘limited evidence’ of such promise. Still, the practice of engaging students in analyzing their own data was consistently related to positive outcomes (article 35).

And ultimately, when faced with hard numbers that suggest personalized learning is not effective, the dominant discourse falls back on the need to believe in the technology. For example, when Vermont test scores dropped after implementation, the Secretary of Education ‘is not hitting the panic button. She believes Vermont needs to stay the course’ (article 6).

Role of foundations: The voice of foundations is expressly included in the sub-set through an article entitled ‘Investing in the Promise of Quality Personalized Learning’ (article 33) written by a Senior Program Officer for the Bill & Melinda Gates Foundation as part of the EdWeek Special Report on Personalized Learning. This article directly reproduces the foundation discourse in ways that resonate strongly with the dominant discourse in EdWeek. It also positions the Gates Foundation as a foundation ‘making investments to bring personalized learning approaches to scale — so they can reach the students who need them most’ (article 33).

The emphasis on investment resonates with the ways in which foundations are inserted into the dominant discourse, as organizations that ‘support’ personalized learning initiatives through donations of money. The dollar figures play upon the implicit understanding that public education is under-funded; given that understanding, foundation money is constructed as a necessary fix. However, the Gates Foundation and the Chan Zuckerberg Initiative are given a special place in the discourse, both quantitatively and qualitatively. First, they are covered much more thoroughly. Although there are 15 foundations and LLCs mentioned throughout the sub-set, Gates, Chan Zuckerberg Initiative, and the Carnegie Corporation are the only ones that appear more than once (in 13 articles each for Gates and the Chan Zuckerberg Initiative, and in nine articles for Carnegie). Second, unlike Carnegie which is mentioned in passing as a funder [37] (e.g., ‘The school, a recipient of a Carnegie Corporation Opportunity by Design grant’), the tech foundations are set apart both by highly personalized portrayals of their principals, including a number of quotes from Gates and the Chan Zuckerberg Initiative, and by the fantastical dollar figures attached to their names:

All are part of a new, multi-pronged effort by Facebook founder and CEO Mark Zuckerberg and his wife, pediatrician Priscilla Chan, to use their massive fortune to reshape public education with technology (article 31).
In December, the couple announced they will eventually give 99 percent of their Facebook shares — worth an estimated $45 billion — to a variety of causes, headlined by the development of software ‘that understands how you learn best and where you need to focus’ (article 31).
‘We think that personalized learning makes sense,’ Zuckerberg told Education Week ... (article 31). [38]

In the dominant discourse, Gates, Zuckerberg, and Chan are accordingly represented as visionaries whose largess is helping reshape education through philanthropy, mirroring the discourse found on the foundation Web sites.

Alternative discourse

Unlike the dominant discourse, which is almost entirely self-referential, the alternative discourse expressly engages with the dominant discourse to unpack and challenge its assumptions in ways that re-root the dialogue in older notions of the role of public education in a democracy.

Effectiveness of personalized learning: From the first article in 2013 forward, the alternative discourse calls for peer reviewed research to determine if personalized learning can help teachers meet curricular goals. This call is expressly rooted in ‘lingering disagreements over how computer-adaptive assessment should function’ (article 17), given evidence that learning technologies in the past had negatively affected learning outcomes for students with disabilities. In other articles, the debate over personalized learning is positioned as the latest iteration of the ongoing tension that began in the ‘early part of the 20th century’ (article 24) when John Dewey’s whole child reforms were contested by reformers interested in increasing efficiency, and just one more in a long line of debates over ‘how children learn, the proper role of teachers, and who gets to decide how public education is organized’ (article 23). This historical context resists the futuristic narrative of newness that is so present in the dominant discourse, and refocuses the discussion on broader and longstanding questions about education as a public good.

In addition, the call for more research on the potential effects of personalized learning is used to attack the aspirational nature of the definitions contained in the dominant discourse. There are numerous references to the ‘lack of clarity’ (article 23) around definitions of personalized learning, which are almost always followed by a statement that ‘the research evidence behind ‘personalized learning’ remains thin’ (article 20). For example, an article entitled ‘Popularity of Ed Tech Often Not Linked To Products’ Impact’ (article 21), reports on a study (funded by the Gates Foundation) that concludes that, of the ‘29 digital learning tools funded by the Bill and Melinda Gates Foundation ... Most had no statistically significant impact on student outcomes’ (article 21). As such, personalized learning, however it is understood, has yet to deliver on its promised ‘hope’ of improving learning and should therefore be approached with caution.

But this caution is not just a question of tweaking the technology to improve it. Contrary to the dominant discourse which sees edtech as the solution for a broken education system, the alternative discourse positions edtech as an unproven strategy that potentially diverts money from a public education system that would work if it was properly funded. For example, given the question marks that accompany personalized learning, ‘critics argue schools would be much better off investing in proven strategies that rely on increased teacher-to-student interaction, such as smaller class sizes’ (article 20) because the ‘proven strategies’ (article 20) of traditional education are known to work well, even in schools that have implemented edtech:

One major study found no significant effects on achievement [with personalized learning], with many students saying they felt they learned more when they worked directly with a teacher in a more traditional manner. That combination of hype and uncertainty has led some observers to voice concern that algorithm driven playlists are just another technology that K-12 schools are embracing without adequate scrutiny or regard for possible unintended consequences. (article 27).

From this perspective, education is not broken, but the untrammeled adoption of personalized learning technologies may break it. Edtech is accordingly ousted from being the ‘solution’ (as it is in the dominant discourse) to being a ‘problem’ (in the alternative discourse) which ‘rais[es] tough questions about whether many ed-tech vendors’ emphasis on quickly bringing their products to scale is actually hampering the larger goal of improving schools’ (article 21).

Public value of education: The public-private divide is key to this move. The alternative discourse concedes that private sector pursuits of profit are fine in the marketplace but resists the application of those same market logics in the educational sphere, arguing that ‘the dominant Silicon Valley business approach of seeking to quickly gain as many users as possible ... [is] particularly ill-suited for schools’ (article 21). Although the adoption of edtech in the public sector by ‘School and district leaders [has] helped turn personalized learning into a multi-million dollar market’ (article 20), the alternative discourse seeks to problematize this marketization precisely because education is a public good, not a private good. As one teacher says, ‘Equating usage with value is fine for a consumer product that users are spending their own time on ... But students are not volunteers, and we’re devoting instructional time to products when we don’t know whether they work or not’ (article 21). Another ‘veteran teacher’ puts it this way: ‘This whole thing is coming from the tech industry, which doesn’t understand that what kids need is someone to love them and get excited about them ... I’m not aware of any research that says sticking a child in front of a computer for hours on end does them any good’ (article 23).

This conflict between private sector goals and public sector values is most pronounced in the discussion of evidence. The following fragment is illustrative:

The edtech sector has been focused on the notion [of personalized learning] for roughly half a decade. While companies have generated hundreds of products and a smattering of new school models are showing promise, there is little large-scale evidence that the approach can improve teaching and learning or narrow gaps in academic achievement. Many in Silicon Valley, including Zuckerberg, don’t seem to mind. ‘We don’t know for certain that it’s going to work,’ he said. ‘All we can really hope to do is provide an initial boost and try to show that this could work as a model, and hopefully it gets its own tailwind that carries it towards mainstream adoption’ (article 31).

From the perspective of the alternative discourse, the entrepreneurial willingness to sell technology that may — or may not — improve learning outcomes sits poorly with the democratic function of public education, which are articulated as ‘making kids literate, preparing them for the workplace, passing on the moral values of community’ (article 24). This public nature of education accordingly grounds resistance to the ‘transformative’ promises of personalized learning by reasserting the value of ‘proven strategies’ (article 20): ‘The socialization that schools must perform, that’s where a lot of the stability in classroom practices comes from. So what I see [in personalized learning] is generally not transformative’ (article 24).

In order to fill in the epistemological gap left by the aspirational approach of the private sector, the alternative discourse seeks to define personalized learning by focusing on the practices that accompany personalized learning in the classroom. This definition encompasses both ‘how the technology works’ (article 27) (including the use of diagnostic assessments and other user data to generate ‘learner profiles’ which are then updated by algorithms that ‘decide’ which of the 9,000 lesson plans that have been pre-created by ‘content partners’ will be ‘served up’ to the student) and how students and teachers are inserted into the process during a regular school day (article 27). For example, when students arrive at a class, they are ‘point[ed] to the section of the room where they spend the period’ by ‘large flat-screened TVs’; at the end of the class, they fill out ‘exit tickets’ that go directly to the edtech company (article 27). The algorithm uses this information to develop a preview schedule for the next day, which is vetted by the edtech company’s ‘human scheduling team’ and then sent to the ‘on the ground teachers [who] get the final say’ (article 27).

Teachers: Although the teacher’s role appears minimal in this description (which appears in the article as a series of quotes from edtech documents), the emphasis on the praxis of education reinserts the teacher as the expert who is able to ‘get the final say’ and decide whether or not the plan generated by the tool is relevant to curricular goals. Implicitly, the teacher is also the person in the room who stands juxtaposed against the ‘large flat-screened TVs’ and ‘exit tickets’ that dominate the description of students’ interactions with technology. This reinsertion of the teacher as expert shifts the focus away from the importance of ‘scaling’ product to the ‘significant unresolved pedagogical tensions’ that surround ‘the appropriate role for software in the classroom, how much autonomy is best for student learning, and the challenge of maintaining high standards and social interaction when every student is pursuing his or her own path. Too often ... [edtech] proponents gloss over such concerns’ (article 29). Teachers, however, have the expertise to address the issues in an informed way and make the call.

Once the teacher is back in the metaphorical room, it is easier to concentrate on the impact of personalized learning on the social interactions at the heart of education. As one school principal notes, ‘Good teaching is always good teaching ... [personalized learning] is just a tool that helps teachers be more targeted’ (article 27). However, targeting may be what makes it harder for teachers to teach well: ‘When the District of Columbia school system tried [personalized learning] at one of its middle schools, for example, experienced teachers ended up feeling limited by the technology, because they wanted the freedom to plan more than one day ahead’ (article 27). The district director of education accordingly dropped the program because, ‘When the rubber met the road, there were just so many little practical challenges to making it work’ (article 27). Teachers in another district similarly dropped edtech because it created:

... pressure to move students along through the curriculum and toward graduation, whether or not they had actually mastered the material they were learning. And some teachers described difficulties with getting students to attend school, complete their work, and turn assignments in on time. Together, such factors in some cases contributed to an over-reliance on software and online curriculadespite a widespread worry that the digital tools available weren’t providing solid learning experiences for students. (article 25).

The alternative discourse supports this reclamation of teachers’ expertise by expressly playing with the skeptic/believer dichotomy introduced in the dominant discourse. This is most apparent in an article entitled, ‘Edtech Skeptic Finds a New Perspective’ (article 24). In this article, Herold interviews a well-known ‘edtech skeptic’ teacher who has visited a small number of foundation-supported schools to evaluate personalized learning initiatives. When asked if he is now ‘less skeptical,’ the interviewee first rejects the dominant discourse’s claim that edtech has the power to transform education: ‘‘Transformative’ is the language of hype, of vendors, promoters, and technologically driven reformers. I am not advocating for that ...’ (article 24). He then focuses not on the technology itself but on the practices of teachers who were ‘regularly and easily integrating technology ... [after asking] ‘What are the learning goals of this lesson?’ and ‘When and how can I use these technologies to best achieve these goals? — I saw that was going on and was very impressed’ (article 24). When Herold returns to the original question, ‘So have you gone from edtech skeptic to believer?,’ the interviewee responds, ‘I wouldn’t go that far’ (article 24). ‘Skepticism is in my DNA’ (article 24). The implication is that, as a teacher, it is his job to be skeptical and judge if and how to use technology to advance learning.

Privacy: Interestingly, this productive skepticism is mobilized most effectively in the articles that examine the potential impact of personalized learning on student privacy. The importance of privacy as a disruptive narrative is underlined in 2014 when it is described as a key issue that ‘cuts to the heart of the tensions that define the digital learning revolution’ (article 19). However, it is in 2017 when Data & Society, a research institute examining social and cultural issues related to data-centric technology, releases its report on the InBloom controversy that the issue gains real traction.

InBloom was a personalized learning initiative created in 2013 with US$100 million from the Gates Foundation and support from the Carnegie Corporation. It collected over 400 items of information — such as social security numbers, details of intimate family relationships (e.g., ‘foster parent’ or ‘father’s significant other’) and reasons for leaving schools (e.g., ‘withdrawn due to illness’ or ‘leaving school as a victim of a serious violent incident’) — from tens of thousands of students. It then used algorithms to mine the data to generate reports on things such as the extent to which an individual student ‘actively participates,’ ‘shows enthusiasm,’ and ‘resists distractions’ (Singer, 2014). InBloom was forced to disband when parents mounted protests over the informational practices at the heart of the initiative (Singer, 2014).

The discussion of InBloom plays an important role in the alternative discourse both as a key point of resistance and as a way of problematizing the dominant discourse. Interestingly, it does this by reinserting parents directly into the debate; it is notable that this is the only context in either the alternative or the dominant discourse where this occurs. For example, the statement that ‘parent activists see InBloom and like-minded efforts through a lens of potential harm’ (article 28) resonates with the alternative claim that education is not broken and that wholesale adoption of edtech could have negative consequences for learning. Moreover, parental concerns are mobilized to problematize the aspirational claims of edtech proponents:

If you start out with the assumption that inBloom was a revolutionary tool that would transform education, be more secure than existing tools used by states and districts, and be fully transparent, then of course you come to the conclusion that the opposition of parents was ‘irrational,’ [parent activist] Haimson said. ‘But all of those suppositions are completely wrong.’ (article 28).

The parents’ perspective also provides a moment to unpack the ‘sheer volume of information collected from children in school, especially amid the recent push to better understand students behaviors, feelings and mindsets’ (article 29) and the ‘large-scale student data sharing’ (article 29) at the heart of personalized learning. It does this by drawing attention to the ways in which these practices negatively affect the persons at the heart of the educational enterprise: students. Not only does this collection and sharing threaten to ‘shackle us to our past, denying us due credit for our ability to evolve, grow, and change’ (article 19); it opens up children to discrimination. By focusing on who the data is about, parents are able to reinsert the ‘persons’ — who are absent in the dominant discourse — back into the ‘learning’ in order to support demands for more individual, and conversely less corporate, control over young people’s life chances. As parent Karen Effrem, ‘the president of ... an advocacy organization that supports parents’ right to control their children’s education’ says, ‘We’re sacrificing our children’s privacy, and we’re allowing corporations to make potentially life-changing decisions about our kids, all for technology that doesn’t actually help them’ (article 29).

The gap between the dominant and alternative discourses is explicitly acknowledged throughout the discussion on privacy. Indeed, it is the ‘polarized’ nature of the views of proponents, on the one hand, and parents, on the other, that strongly suggests that edtech continues to be a contested space. This creates room for more radical definitional attacks that construct personalized learning as ‘little more than breaking knowledge and ideas down into ‘itty-bitty parts’’ (article 29), then using extrinsic rewards to ‘march kids through a series of decontextualized skills they had no meaningful role in choosing’ that can only lead to a ‘reductionist type of education’ (article 29). These definitions are then mobilized to challenge the ‘naturalness’ of dominant claims about the transformative potential of personalized learning in later articles, such as ‘The Case(s) Against Personalized Learning’ (article 29), and to strengthen resistance by positioning parents as allies of teachers who are seeking to use their expertise to advance the best interests of students:

Bill Gates and Mark Zuckerberg are backing it with hundreds of millions of dollars. States from Florida to Vermont have adopted supportive laws and policies. And school districts across the country are embracing this emerging education trend. But as ‘personalized learning’ takes root, it’s also coming under greater scrutiny. Leading researchers say their work does not support the most enthusiastic claims being made by personalized-learning supporters. Education experts are raising questions about implications for teaching and learning. Tech-industry critics are sounding alarms about Silicon Valley’s growing influence over public schools. And a small but vocal coalition of parents and activists from across the political spectrum deride the term ‘personalized learning’ as an Orwellian misnomer for replacing teachers with digital devices and data-mining software. (emphasis added) (article 29).

Polarization also makes it easier to explicitly raise concerns about the role of foundations in the personalized learning enterprise. Early attempts to do this are formulated around references to rich donors; this is, in effect, the other side of the coin that the dominant discourse uses to valorize the actions of the Gates and the Chan Zuckerberg Initiative as super-successful:

For rich donors, the upside is more levers to pull when trying to change the world. For everyone else, the downside is that these new structures further blur the lines between business and philanthropy and partly circumvent the regulations that have governed charitable giving for decades. (article 31)

However, it is difficult to give this critique teeth precisely because foundation operations are so opaque: ‘It’s hard to tell exactly what these new donors emerging from the tech sector are doing, because so much is in flux and they’re not very transparent’ (article 31). Although there is an ongoing attempt to unpack the ‘tangled web of overlapping entities’ that allow tech foundations to ‘aggressively push’ or ‘boost’ the idea of personalized learning, it is relatively anemic because of the lack of detail (article 31). Accordingly, the early discourse falls back on general claims that ‘any time super-empowered people with lots of money try to influence how the rest of us educate our children, we have a right to know what they’re up to’ (article 29).

However, this is changed by the controversies around InBloom and Facebook’s possible role in election interference, both of which revolve around technologically-enabled privacy invasions. These controversies create discursive opportunities to shift the narrative to openly challenge the status quo. As a 2017 article puts it:

... educators and the public would be foolish to not carefully consider the tech sector’s influence in public schools, especially given its recent stumbles ... Billions of public dollars are at stake ... So are big questions about the fundamental nature of schooling: How do we believe children learn? Who should decide what students need to know and get to experience? How will we determine what they’ve learned? ‘We need to open up a bigger debate about whether we really want Silicon Valley establishing this new model of data-driven schooling,’ Williamson said. ‘These are people whose vision for reforming public education puts their own industry in charge.’ (article 31).

The alternative discourse accordingly seeks to challenge the unproblematic nature of tech foundation largesse, by linking support of personalized learning to lobbying at the federal level (for tech-friendly regulations) and at the local level (for levers to encourage districts to purchase product). Together, this suggests that tech foundations ‘are definitely setting up a model where [they] can push on multiple sides of an issue’ (article 31), especially given the deep network connections between the tech industry, the non-profit sector and government [39]. As such, this ‘new crop of young donors who made their fortunes in technology’ is seeking to ‘evolving into a different kind of funder’ to create a ‘top-down approach’ to avoid ‘a buzz-saw of community opposition.’ (article 31).

Contested meanings

Our analysis of the five tech foundation Web sites and EdWeek articles suggests that competing discourses on personalized learning revolve around contested meanings for five key concepts. First, expertise is positioned as sitting either in the hands of technocrats who design edtech or in the hands of teachers who are educational experts. Second, the meaning of self-direction and choice is highly contested. On the one hand, self-direction is situated in algorithms that use data to know students better than teachers can. From this perspective, students are streams of information that can be captured and acted upon through algorithmic analysis; their education is therefore ‘personalized’ when the technology directs them to their next learning task based on the data they generate as they go about learning. On the other hand, students are positioned as autonomous persons who are being raised to be citizens. Education therefore requires helping them learn how to exercise choice by creating opportunities for experiential learning.

Nestled within this are competing notions of what it means to learn. The dominant discourse constructs learning as being about process, not content. Given the dominant discourse’s underlying frame of ‘newness’, it is impossible to decide what content is relevant: ‘Educators will need help sorting through newly available information, because ‘millions of years of evolution have not prepared us to read streams of data’’ (article 36). From the perspective of the alternative discourse, learning is all about content, and the teacher’s role is one of curating that content to provide students with access to materials that will support high quality learning opportunities.

The meaning of the word ‘evidence’ is also contested. For the dominant discourse, the algorithms that drive personalized learning technologies are evidence of progress, because they are objective markers of the knowledge needed to properly steer each student’s learning. The alternative discourse instead looks for peer reviewed research to determine whether or not learning outcomes are being met.

Finally, each discourse calls upon a different conception of learning, harkening back to the ongoing twentieth century debate between education as efficient learning and the whole-child approach of John Dewey.

 

++++++++++

Implications and conclusions

Our analyses of the tech foundation Web sites and the Education Week articles identifies discourses that are very similar to those evident in earlier examinations of discourses regarding philanthropy in education — leading with a discourse that is largely defined and promoted by the tech foundations themselves, followed by the emergence of an alternative discourse that engages critically with the tech foundation discourse, raising questions about the claims and highlighting the possibility of self-interest on the part of the tech foundations. For example, Srivastava and Oh (2010) identified ‘two particular discourses — one resulting from the macro-policy backdrop for education finance, and the other entrenched in an uncritical ideological acceptance of a logic of neutrality, and the efficiency and effectiveness of partnerships and philanthropy.’ [40] The discourse on ‘philanthropy and private foundations rests on stressing the positive ideals of social service and giving without a broader understanding of potential changes to the responsibilities of the state in sectors such as education, which are typically entrusted to it because of their association with the fulfilment of fundamental human rights. This uncritical acceptance of an expanded role for private foundations is helped by an ideological meta-narrative fusing partnership and philanthropy.’ [41] Similarly, Gurn (2016) reviewed academic, professional, and government literature on the philanthropic education activities of corporate-related foundations and identified two discourses [42]. The first, or dominant discourse he labels Corporate Altruism, where funds are considered ‘manna from heaven.’ Literature embracing this discourse has the following features: focus on advantages; focus on what works and organizational efficiencies; limited attention of analysis of asymmetrical power dimensions; and an assumption that schools are broken. The counter or non-dominant discourse he labels Corporate Profiteering. Literature adopting this discourse has the following characteristics: focus on revealing the contradictions in socio-political dimensions; an emphasis on the unequal power in public-private partnerships; less attention on practical questions; and, a somewhat romanticized view of public schools and negative view of corporations.

Our analysis of the activities of tech company foundations and the discourses that are emerging around personalized learning also provides concrete evidence that the tech foundations continue to follow the same playbook, with similar goals to their earlier K-12 initiatives including school choice, charter schools, teacher accountability, and effectiveness. Tech company foundations are once again active in framing issues about personalized learning, supporting advocacy organizations and research studies, and funding a range of edtech companies and initiatives in the area of personalized learning. Their voices in the landscape of K-12 education are particularly loud, amplified by their grant dollars and their networks of influence. The import of their policy influence in the area of personalized learning raises many of the same questions that education policy scholars have identified in the past — are they able to speak more loudly, more effectively and in more venues about the positive potential of personalized learning than those who have questions about the implications and effectiveness of personalized learning?

Although this question is one of import in any area of K-12 education, we would argue it is particularly troublesome in an area of K-12 education that not only involves education and learning of students but that also entails student privacy and the social construction of and expectations about data and surveillance. As the title of this paper indicates, we are struck by what we see as the trend to ‘take persons out of personalized learning.’ The emphasis in the dominant discourse on the tech foundation Web sites and in the Education Week articles is on data — technical and abstract — and fails to even acknowledge that this data is about the learning habits, potential, and inclinations of children. And that the push is to extend collection of data on children beyond learning per se into data on their social, family, friends, health, economic situation because all of these areas affect their learning potential.

Further research on the role of the tech company foundations in the area of personalized learning is warranted given our findings but our findings do provide a baseline and something of a timeline for further investigation. End of article

 

About the authors

Priscilla M. Regan is a Professor in the Schar School of Policy and Government at George Mason University in Fairfax, Virginia.
Send comments to: pregan [at] gmu [dot] edu

Valerie Steeves is a Full Professor in the Department of Criminology at the University of Ottawa in Canada.
E-mail: vsteeves [at] uottawa [dot] ca

 

Acknowledgements

The research for this paper was funded by the eQuality Partnership Grant from the Social Science and Humanities Research Council of Canada. For information on the grant, please see: http://www.equalityproject.ca/. An earlier version of this paper was presented at the Computers, Data Protection and Privacy Conference in Brussels in January 2019. The authors wish to acknowledge the invaluable research assistance of Bay Jaber, J.D. student at the University of Ottawa and Elsa Talat Khwaja, Ph.D. student at George Mason University.

 

Notes

1. The Carnegie Corporation is the exception to the rule since it is not connected to a technology firm. Moreover, it has focused on funding education initiatives since it was created in 1911. More recently, it has provided support for personalized learning initiatives but typically in partnership with one of the tech foundations.

2. The Bill and Melinda Gates Foundation, for example, fits this model. These foundations are generally classified as non-profit foundations (as U.S. organizations, they have been given what is known as a 501(c)(3)s tax designation that entails disclosure requirements and limitations on activities). The Chan Zuckerberg Initiative was established as a limited liability corporation (LLC) originally funded with US$1 billion of Facebook stock and the LLC invests for profit in enterprises that fit the social goals of the LLC (Levine, 2016, p. 1). Unlike non-profit foundations, LLCs are legally permitted to directly invest in for-profit companies and to engage in political lobbying and donations; moreover, the extent to which the group is legally required to publicly report on its activities is limited. In either case, a foundation or an LLC provides a way for super wealthy individuals to shield money from taxes and to give tax-free to those causes which it views as important.

3. Regan and Jesse, 2019, pp. 169–173.

4. For example, Google Classroom is used by 60 million students and teachers in countries around the world (Desson, 2018). In addition, U.S. foundations are endowing projects across the globe (Armour-Jones, n.d.). Accordingly, an exploratory study of how personalized learning is talked about in the U.S. is likely to be relevant worldwide, as personalized learning software rolls out in the global marketplace.

5. Snyder, 2015, pp. 30–31.

6. Reckhow and Snyder, 2014, p. 187.

7. Ball and Junemann, 2011, p. 649.

8. Ibid.

9. Reckhow and Snyder, 2014, p. 188.

10. Reckhow and Snyder, 2014, pp. 188–190.

11. Gunn 2015, p. 7.

12. Zeide, 2017, p. 516.

13. Morsy, 2015, p. 1,526.

14. Gurn, 2016, pp. 3–4.

15. Gurn, 2016, p. 11.

16. Gurn, 2016, p. 12.

17. Hess and Henig, 2015, p. 2.

18. Hess and Henig, 2015, p. 8.

19. Dillon, 2011; Baltodano, 2017, p. 152.

20. “Informing progress: Insights on personalized learning implementation and effects,” at https://www.rand.org/pubs/research_reports/RR2042.html (July 2017); “Observations and guidance on implementing personalized learning,” at https://www.rand.org/pubs/research_briefs/RB9967.html (11 July 2017); “How does personalized learning affect student achievement?” at https://www.rand.org/pubs/research_briefs/RB9994.html (7 December 2017).

21. https://cdt.org/press/elizabeth-laird-to-lead-cdts-expanded-student-privacy-advocacy/.

22. Reckhow and Tompkins-Stange, 2015, pp. 64–65.

23. Boltodano, 2017, p. 153.

24. We originally included the Silicon Valley Community Foundation (#15 in 2005, #5 in 2010) but found that there was little text on its Web site for analysis.

25. Google only files annual tax-related financial information for the Google Foundation, which is only a part of Google.org’s total giving (Alba, 2016).

26. https://www.msdf.org/initiatives/data-driven-education/.

27. https://www.gatesfoundation.org/What-We-Do/US-Program/K-12-Education.

28. https://chanzuckerberg.com/education/.

29. https://hewlett.org/strategy/deeper-learning/.

30. https://www.google.org/our-work/education/.

31. https://www.edweek.org/media/edweek_info.pdf.

32. This dataset was used in a related analysis of the representation of edtech in general (see Regan and Bailey [2018]). There were a total of 6,628 articles in EdWeek for the time surveyed with 521 relevant to edtech, 110 discussing privacy, and 44 on personalized learning.

33. Ciechanowski, 2012, p. 300.

34. Ibid.

35. Gee, 2011a, p. 24.

36. These were either short, informational articles or articles that included peripheral references to personalized learning.

37. There is also one article devoted to Cognitive Tutor Algebra 1 which was developed for profit by Carnegie Learning Company.

38. See also, ‘Bill Gates Announces $1.7 Billion In New K-12 Investments — Bill Gates, the billionaire co-founder of the foundation, delivered the news in a speech’ (Vara-Orta, 2017).

39. For example, some articles in the alternative discourse make passing references that show how individuals move between sectors (e.g., ‘Karen Cator, a former Apple executive who previously managed the U.S. Department of Education’s office of educational technology ... who currently heads the Washington-based nonprofit Digital Promise’); and the financial relationships between foundations, program evaluators, and EdWeek itself (e.g., ‘The digital learning tools analyzed by SRI were awarded ... grants by the Gates Foundation in 2010. The foundation also funded SRI as an independent contractor to track and evaluate the products’ progress over the following two years. (The Gates Foundation also helps support Education Week’s coverage of the implementation of college- and career-ready standards and the use of personalized learning)’).

40. Srivastava and Oh, 2010, p. 460.

41. Srivastava and Oh, 2010, p. 462.

42. Gurn, 2016, p. 11.

 

References

K.K. Abowitz, 2000. “Democratic communities and business/education ‘partnerships’ in secondary education,” Urban Review, volume 32, number 4, pp. 313–341.
doi: https://doi.org/10.1023/A:1026455708484, accessed 23 October 2019.

D. Alba, 2016. “Google.org thinks it can engineer a solution to the world’s woes,” Wired (8 March), at https://www.wired.com/2016/03/giving-google-way/, accessed 6 January 2019.

S. Armour-Jones, n.d. “Foundation maps for media funding: Introducing a new tool for unlocking the power of media grants data,” Media Impact Funders, at https://mediaimpactfunders.org/wp-content/uploads/2016/06/Map-Report_final-to-print.pdf, accessed 23 October 2019.

S.J. Ball and C. Junemann, 2011. “Education policy and philanthropy — The changing landscape of English educational governance,” International Journal of Public Administration, volume 34, number 10, pp. 646–661.
doi: https://doi.org/10.1080/01900692.2011.583773, accessed 23 October 2019.

M.P. Baltodano, 2017. “The power brokers of neoliberalism: Philanthrocapitalists and public education,” Policy Futures in Education volume 15, number 2, pp. 141–156.
doi: https://doi.org/10.1177/1478210316652008, accessed 23 October 2019.

J. Barkan, 2011. “Got dough? How billionaires rule our schools,” Dissent, volume 58, pp. 49–57, and at https://www.dissentmagazine.org/article/got-dough-how-billionaires-rule-our-schools, accessed 23 October 2019.

D.R. Boyles (editor), 2005. Schools or markets? Commercialism, privatization, and school-business partnerships. Mahwah, N.J.: L. Erlbaum Associates.

K. Ciechanowski, 2012. “Conflicting discourses: Functional linguistic and discourse analyses of Pocahontas texts in bilingual third-grade social studies,” Journal of Literacy Research, volume 44, number 3, pp. 300–338.
doi: https://doi.org/10.1177/1086296X12450699, accessed 23 October 2019.

L. Cuban, 2004. The blackboard and the bottom line: Why schools can’t be businesses. Cambridge, Mass.: Harvard University Press.

C. Desson, 2018. “As Google for education enters classrooms across Canada, some parents are asking to opt out,” CBC Radio? (11 June), at https://www.cbc.ca/radio/spark/401-google-for-education-1.4694935/as-google-for-education-tools-enter-classrooms-across-canada-some-parents-are-asking-to-opt-out-1.4694939, accessed 23 October 2019.

J. Dewey, 1927. The public and its problems. New York: Holt.

J. Dewey, 1916. Democracy and education: An introduction to the philosophy of education. New York: Macmillan.

S. Dillon 2011. “Behind grass-roots school advocacy, Bill Gates,” New York Times (21 May), at http://www.nytimes.com/2011/05/22/education/22gates.html, accessed 23 October 2019.

J.P. Gee, 2011a. “Discourse analysis: What makes it critical?” In: R. Rogers (editor). An introduction to critical discourse analysis in education. Second edition. New York: Routledge, pp. 23–45.

J.P. Gee, 2011b. How to do discourse analysis: A toolkit. New York: Routledge.

J.P. Greene, 2015. “Buckets into another sea,” In: F.M. Hess and J.R. Henig (editors). The new education philanthropy: Politics, policy, and reform. Cambridge, Mass.: Harvard Education Press, pp. 11–28.

A.M. Gurn, 2016. “Courting corporate philanthropy in public education: Multi-disciplinary literature review of publicprivate partnerships (PPPs) in urban public schooling,” SAGE Open (19 April).
doi: https://doi.org/10.1177/2158244016635714, accessed 23 October 2019.

J.R. Henig, 2013. The end of exceptionalism in American education: The changing politics of school reform. Cambridge, Mass.: Harvard Education Press.

B. Herold, 2017. “Gates, Zuckerberg teaming up on personalized learning,” Education Week, volume 36, number 36 (20 June), p. 7.

F.M. Hess and J.R. Henig, 2015. “Introduction,” In: F.M. Hess and J.R. Henig (editors). The new education philanthropy: Politics, policy, and reform. Cambridge, Mass.: Harvard Education Press, pp. 1–10.

Y. Lev-Aretz, 2019. “Data philanthropy,” Hastings Law Journal, volume 70, pp. 1,491–1,545, and at http://www.hastingslawjournal.org/wp-content/uploads/70.6-Lev-Aretz.pdf, accessed 23 October 2019.

M. Levine, 2016. “Andrew Carnegie and Bill Gates: Philanthropy and oligarchy, then and now,” American Journal of Orthopsychiatry, volume 86, number 1, pp. 1–9.
doi: https://doi.org/10.1037/ort0000150, accessed 23 October 2019.

L. Morsy, 2015. “Corporate philanthropic giving practices in U.S. school education,” Voluntas, volume 26, number 4, pp. 1,510–1,528.
doi: https://doi.org/10.1007/s11266-014-9474-x, accessed 23 October 2019.

D. Ravitch, 2010. The death and life of the great American school system: How testing and choice are undermining education. New York: Basic Books.

S. Reckhow, 2013. Follow the money: How foundation dollars change public school politics. New York: Oxford University Press.

S. Reckhow and M. Tompkins-Stange, 2015. “‘Singing from the same hymnbook’ at Gates and Broad,” In: F.M. Hess and J.R. Henig (editors). The new education philanthropy: Politics, policy, and reform. Cambridge, Mass.: Harvard Education Press, pp. 55–78.

S. Reckhow and J.W. Snyder, 2014. “The expanding role of philanthropy in education politics,” Educational Researcher, volume 43, number 4, pp. 186–195.
doi: https://doi.org/10.3102/0013189X14536607, accessed 23 October 2019.

P.M. Regan and J. Jesse, 2019. “Ethical challenges of edtech, big data and personalized learning: Twenty-first century student sorting and tracking,” Ethics and Information Technology, volume 21, number 3, pp. 167–179.
doi: https://doi.org/10.1007/s10676-018-9492-2, accessed 23 October 2019.

P.M. Regan and J. Bailey, 2018. “Big data, privacy, and education applications,” paper presented at the Eighth Biannual Surveillance Studies Network Conference (7-9 June, Aarhus, Denmark).

R.A.W. Rhodes, 1994. “The hollowing out of the state: The changing nature of the public service in Britain,” Political Quarterly, volume 65, number 2, pp. 138–151.
doi: https://doi.org/10.1111/j.1467-923X.1994.tb00441.x, accessed 23 October 2019.

K.J. Saltman, 2010. The gift of education: Public education and venture philanthropy. New York: Palgrave Macmillan.

J. Scott, 2009. “The politics of venture philanthropy in charter school policy and advocacy,” Education Policy, volume 23, number 1, pp. 106–136.
doi: https://doi.org/10.1177/0895904808328531, accessed 23 October 2019.

J.W. Selsky and B. Parker, 2005. “Cross-sector partnerships to address social issues: Challenges to theory and practice,” Journal of Management, volume 31, number 6, pp. 849–873.
doi: https://doi.org/10.1177/0149206305279601, accessed 23 October 2019.

N. Singer, 2014. “InBloom student data repository to close,” New York Times (21 April), at https://bits.blogs.nytimes.com/2014/04/21/inbloom-student-data-repository-to-close/, accessed 23 October 2019.

J.W. Snyder, 2015. “How old foundations differ from new foundations,” In: F.M. Hess and J.R. Henig (editors). The new education philanthropy: Politics, policy, and reform. Cambridge, Mass.: Harvard Education Press, pp. 29–54.

P. Srivastava and S.-A. Oh, 2010. “Private foundations, philanthropy, and partnership in education and development: Mapping the terrain,” International Journal of Educational Development, volume 30, number 5, pp. 460–471.
doi: https://doi.org/10.1016/j.ijedudev.2010.04.002, accessed 23 October 2019.

V. Steeves, 2014. “Young Canadians in a wired world, phase III: Online privacy, online publicity,” MediaSmarts, at http://mediasmarts.ca/ycww/online-privacy-online-publicity, accessed 23 October 2019.

V. Steeves, 2012a. “Young Canadians in a wired world, phase III: Talking to youth and parents about life online,” MediaSmarts, at http://mediasmarts.ca/ycww/talking-youth-parents-about-life-online, accessed 23 October 2019.

V. Steeves, 2012b. “Young Canadians in a wired world, phase III: Teachers’ perspectives,” MediaSmarts, at http://mediasmarts.ca/ycww/teachers-perspectives, accessed 23 October 2019.

V. Steeves, 2009. “Online surveillance in Canadian schools,” In: T. Monahan and R.D. Torres (editors). Schools under surveillance: Cultures of control in public education. New Brunswick, N.J.: Rutgers University Press, pp. 87–103.

M. Taddeo, 2016. “Data philanthropy and the design of the infraethics for information societies,” Philosophical Transactions of the Royal Society A, volume 374, number 2083 (28 December), at https://royalsocietypublishing.org/doi/full/10.1098/rsta.2016.0113, accessed 23 October 2019.
doi: https://doi.org/10.1098/rsta.2016.0113, accessed 23 October 2019.

E. Thümler, 2011. “Foundations, schools and the state: School improvement partnerships in Germany and the United States as legitimacy-generating arrangements,” Public Management Review, volume 13, number 8, pp. 1,095–1,116.
doi: https://doi.org/10.1080/14719037.2011.619065, accessed 23 October 2019.

F. Vara-Orta, 2017. “Bill Gates announces $1.7 billion in new K-12 investments,” Education Week, volume 37, number 9 (19 October), p. 1, 15.

B. Williamson, 2018. “Silicon startup schools: Technocracy, algorithmic imaginaries and venture philanthropy in corporate education reform,” Critical Studies in Education, volume 59, number 2, pp. 218–236.
doi: https://doi.org/10.1080/17508487.2016.1186710, accessed 23 October 2019.

E. Zeide, 2017. “The limits of education purpose limitations,” University of Miami Law Review, volume 41, pp. 493–526, and at https://lawreview.law.miami.edu/wp-content/uploads/2017/03/Zeide-Final_p493.pdf, accessed 23 October 2019.

S. Zuboff, 2019. The age of surveillance capitalism: The fight for a human future at the new frontier of power. New York: Public Affairs.

 

Appendix: Education Week articles

  1. K. Ash, 2013a. “Customized spaces widening,” Education Week, volume 32, number 32, pp. S32–S34.
  2. K. Ash, 2013b. “Idaho pilots effort to integrate Khan Academy videos,” Education Week, volume 32, number 24, p. 9.
  3. D. Burnette, 2017. “States waive regulations, create innovation zones,” Education Week, volume 37, number 12, pp. 12–14.
  4. K. Bushweller, 2016a. “National Education Technology Plan: How schools are using Technology and what they envision for the future,” Education Week, volume 35, number 17, p. S1.
  5. K. Bushweller, 2016b. “Personalized twists, turns,” Education Week, volume 36, number 9, p. 4.
  6. K. Bushweller, 2017a. “Taking a hard look at a movement,” Education Week, volume 37, number 12, p. 3.
  7. K. Bushweller, 2017b. “The challenge: Turning state law into classroom reality,” Education Week, volume 37, number 12, pp. 18–21.
  8. S. Cavanagh, 2013. “Federal efforts aim to bridge educational technology, learning sciences,” Education Week, volume 32, number 15, p. 7.
  9. S. Cavanagh, 2016. “Students pinpoint what they need,” Education Week, volume 36, number 9, p. 7.
  10. M. Davis, 2013. “English-learners going digital,” Education Week, volume 32, number 32, p. 28.
  11. M. Davis, 2016. “Checking up on the current status of personalized learning pioneers,” Education Week, volume 36, number 9, p. 29.
  12. M. Davis, 2017. “5 Ed-tech experts weigh in,” Education Week, volume 36, number 35, pp. 12–13.
  13. M.R. Davis and L. Loewus, 2017. “Students tell their tales of new ways of learning,” Education Week, volume 37, number 12, pp. 25–28.
  14. L. Doran, 2016. “Promising state policies for personalized learning,” Education Week, volume 35, number 32, p. 5.
  15. Education Week, 2016. “Data dive,” Education Week, volume 36, number 9, p. 5.
  16. R. Flanigan, 2017. “Case studies: Lessons from 3 schools.,” Education Week, volume 37, number 12, pp. 22–24.
  17. B. Herold, 2013. “Adaptive testing gains momentum, prompts worries,” Education Week, volume 32, number 36, p. 18.
  18. B. Herold, 2014a. “‘Playlists’ tailor curriculum,” Education Week, volume 33, number 25, p. 30.
  19. B. Herold, 2014b. “Balancing privacy and innovation,” Education Week, volume 34, number 9, pp. S8, S11.
  20. B. Herold, 2016a. “Investments in personalized learning rise, but research on its impact is lacking,” Education Week, volume 36, number 9, pp. 20–21.
  21. B. Herold, 2016b. “Popularity of ed tech often not linked to products’ impact,” Education Week, volume 35, number 30, p. 1.
  22. B. Herold, 2017a. “‘Playlists’ re-envisioned,” Education Week, volume 36, number 26, p. 23.
  23. B. Herold, 2017b. “6 key insights: RAND Corp. researchers talk personalized learning,” Education Week, volume 37, number 12, pp. 10–11.
  24. B. Herold, 2017c. “Ed-tech skeptic Larry Cuban finds a new perspective,” Education Week, volume 36, number 20, pp. 1, 11.
  25. B. Herold, 2017d. “From theory to practice, hurdles for personalized learning,” Education Week, volume 37, number 7, p. 11.
  26. B. Herold, 2017e. “Gates, Zuckerberg teaming up on personalized learning,” Education Week, volume 36, number 36, p. 7.
  27. B. Herold, 2017f. “Learning via ‘playlists’,,” Education Week, volume 36, number 26, pp. 19–22.
  28. B. Herold, 2017g. “Post-mortem on inBloom reignites data-sharing debates,” Education Week, volume 36, number 21, p. 7.
  29. B. Herold, 2017h. “The case(s) against personalized learning,” Education Week, volume 37, number 12, pp. 4–5.
  30. B. Herold, 2017i. “Under Trump, ed-tech leadership is big question mark,” Education Week, volume 36, number 30, pp. 1, 16.
  31. B. Herold and M. Riser-Kositsky, 2016. “Facebook CEO bets on personalized learning,” Education Week, volume 35, number 23, p. 1.
  32. M. Horn and H. Staker, 2014. “For blended learning, look beyond the technology,” Education Week, volume 34, number 14, p. 22.
  33. H.B. Jones, 2016. “Investing in the promise of quality personalized learning,” Education Week, volume 36, number 9.
  34. M. Molnar, 2016a. “8 ‘red flags’ to look for in products,” Education Week, volume 36, number 9, p. 29.
  35. M. Molnar, 2016b. “Numbers to watch,” Education Week, volume 36, number 9, p. 14.
  36. M. Molnar, 2017. “Student-centered learning top of mind for companies,” Education Week, volume 36, number 34, p. 9.
  37. M. Molnar and L. Loewus, 2017. “The evolving world of curriculum,” Education Week, volume 36, number 26, pp. 5–7.

 


Editorial history

Received 4 May 2019; revised 26 August 2019; revised 25 September 2019; accepted 25 September 2019.


Copyright © 2019, Priscilla M. Regan and Valerie Steeves. All Rights Reserved.

Education, privacy, and big data algorithms: Taking the persons out of personalized learning
by Priscilla M. Regan and Valerie Steeves.
First Monday, Volume 24, Number 11 - 4 November 2019
https://firstmonday.org/ojs/index.php/fm/article/view/10094/8152
doi: http://dx.doi.org/10.5210/fm.v24i11.10094





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2019. ISSN 1396-0466.