First, Do No Harm: Inequality in American Education Will Not Be Solved Online

January 21, 2013 at 1:44 am 31 comments

Ian raises a really important issue that I don’t think is being discussed enough.  I predict that computer science MOOC completers are even more white and male than in existing computing education.  Replacing more face-to-face CS courses with MOOCs may be reversing the hard-fought gains we’ve made through NCWIT and NSF BPC efforts.  I’ve asked both Udacity and Coursera about the demographics of their completers.  Coursera said that they don’t know yet because they simply haven’t looked.  Udacity said that it’s “about the same” as in existing face-to-face CS classes.

To address issues of inequality, we will have to do something different than what we are doing now, but we want to do something different that has better results.  We need to be careful that we don’t make choices that lead us to a worse place than we are now.

Here’s a concrete proposal: Any institution that belongs to NCWIT (or more significantly, the NCWIT Pacesetters program) that runs a MOOC for computer science and does not check demographics should have its membership revoked. (See Note.)  We should not be promoting computer science education that is even more exclusive.  We need new forms of computer science education that broaden participation.  At the very least, we ought to be checking — are we doing no harm? Are we advancing our agenda of broadening participation, or making it more exclusionary?

I wonder if the responsibility to check is even greater for public institutions.  Public institutions have a responsibility to the citizens of their state to be inclusive. Readers of this blog have argued that Title IX does not apply to academic programs, suggesting that there is no legal requirement for CS departments to try to draw in more women and minorities.  We in public universities still have a moral responsibility to make our courses and programs accessible.  If we choose to offer instruction via MOOCs, particularly as a replacement for face-to-face courses, don’t we have a responsibility to make sure that we are not driving away women and minorities?

The SJSU test will be run on “remedial” courses at one of the country’s most ethnically diverse universities, of which only 25 percent of the student population is white, and which is primarily comprised of minorities, first-generation college students, and commuting students. This is a population that has more likely been subject to underfunded primary and secondary schools and, generally speaking, a whole regime of distress, neglect, and bias compared to California residents who would attend Berkeley or UCLA. Put differently, the conditions that produced the situation that the Udacity deal is meant to solve, at least in part, was first caused by a lack of sufficient investment in and attention to early- and mid-childhood education.

In response, California could reinvest in public schools and the profession of secondary teaching. But instead, the state has decided to go the private paved surface and illumination services route — siphoning California taxpayer receipts and student tuition directly into a for-profit startup created, like all startups, with the purpose of producing rapid financial value for its investors. Just how much of those proceeds Udacity will hold onto is unclear. While the company has reportedly paid instructors in the past, it’s unclear if its new institutional relationships will support paid teaching or not. Coursera, Udacity’s primary competitor in the private MOOC marketplace, has managed to get faculty from prestigious institutions to provide courses for free, in exchange for the glory of a large audience and the marketing benefit of the host institution.

via Inequality in American Education Will Not Be Solved Online – Ian Bogost – The Atlantic.

Note: While I sit on the NCWIT Leadership Team, the opinions in this blog are my own.  They do not represent NCWIT’s policy. I shared this blog post with Lucy Sanders, CEO of NCWIT, and she made an interesting suggestion.  Some NCWIT Pacesetters are departments who may have little control over what their college, school, or university does.  If they must use MOOCs, because of decisions made higher in the administrative chain, then perhaps measuring the demographics of the completers might be a way of being a Pacesetter.

Entry filed under: Uncategorized. Tags: , , .

Teaching Programming To A Highly Motivated Beginner: The Difference between Anecdote and Data Where did CS PhD’s get their undergraduate degrees?

31 Comments Add your own

  • 1. BenK  |  January 21, 2013 at 5:32 am

    To believe that teaching better programming methods and practices, or computer science, to any specific demographic constitutes ‘doing harm’ to society is a pathology. While specifically providing certain kinds of training to violent criminals might constitute a harm, that does not operate at the level of typical demographics – and computer science is not presently among the kinds of training that are likely to be problematic.

    Offering effective teaching to all comers is a kind of liberality that may not solve the underlying issues which tear at various communities around the world – just as cures for cancers won’t.

    Reply
    • 2. Mark Guzdial  |  January 21, 2013 at 7:46 am

      Ben, I’m not understanding your concern. I’m suggesting that adopting a pedagogical approach that excludes certain groups does them harm, and that use of MOOCs may be doing harm to our goals of broadening participation in computing. I probably was unclear. I did not mean to say that teaching computer science does society harm.

      Reply
      • 3. BenK  |  January 21, 2013 at 9:10 am

        ‘Do no harm’ is an element of medical practice that states that inaction is preferred when there is a risk of harming the patient. This is traditionally the result of physicians being paid for procedures or therapies; so, to preserve the medical profession, they should decline to do something (for pay) when it might actually be against the interests of the patient.

        Now, applied to the statements about MOOCs and gender/race relative rate ratios, you are saying that educators should not offer MOOCs at all, because it will lead to more white males being better educated computer programmers or computer scientists, thereby making society as a whole worse off.

        I am in disagreement with this on several levels; but the first is that I believe that as educators, we should want the programmers we do have, the computer scientists we do have, and so on, to be the best educated possible.

        Secondly, I do not believe society will correct educational imbalance from inside the classroom. The issues are too large and located elsewhere, fundamentally. The imbalances that are most detrimental are not those of deconstruction, derisively characterized as feminist logarithms and post-colonial Taylor expansions, but instead the absence of entire communities both poor and African-American, or poor and recent immigrant, or poor and rural, in the most prestigious, profitable and socially powerful professions that follow certain classroom experiences. These will only be corrected by addressing systematic lack of social capital, rates of incarceration and criminal activity, family and home education failures, safety and public health concerns (diabetes, diet, etc).

        Some people have oversold MOOCs based on their low up-front costs, suggesting that a lack of entry fees will provide educational access, training, attainment, and then access to prestigious professional careers. This is unlikely. The much greater cost of the education is the ability to think ahead, to plan for 5-10 hours of otherwise unoccupied time in a week for 15 weeks in a row, to know what to take and how to represent it on a resume, how to apply it to a career, and so on. Completion is not easy. It requires considerable resources going in; tuition is not the hard part of a good college education – people who rant about grade inflation at the top schools are often pretty delusional in this regard.

        The lack of social capital, long commutes, dependent parents, childcare, serious health issues, are many of the same issues which hobble attempts to enter and maintain high prestige, high reward careers; as well as professional educations; and they will impact MOOC completion. Perhaps less important will be the possession of criminal records (30% among African-American males) – but that is only one factor among many.

        Where educational policies do the most harm are where they are elements in a mandatory educational system. Choices that hurt boys, girls, Spanish-speakers, Asians, rural students, etc, are most detrimental when they are applied to everyone without exception. I’m not a big fan of mandatory public schooling anyway – but this is a larger issue. Public schooling (grade school) is the primary venue for mandatory and discriminatory policies as far as I can tell, but also institutions that provide mandatory certification (law school, medical school) can be a problem.

        With regard to MOOCs, let a thousand flowers bloom, even if the MOOCs do little to correct imbalances and merely better the education of existing professionals. We should not discourage the creation of MOOCs just because they are not necessarily helpful by themselves. Personally, I encourage employees who might otherwise not have time for a formal degree program to avail themselves of MOOCs as a form of on-the-job training, atomized somewhat less than the normally appropriate level, to provide opportunity. I can give them encouragement, flexibility, and continuity to assist them. As a compassionate employer, this may lead to employees ‘graduating’ from my employment with their new found job skills – but I serve society in this small way.

        I will stop here, hopefully to have some responses.

        Reply
        • 4. Mark Guzdial  |  January 21, 2013 at 9:34 am

          “You are saying that educators should not offer MOOCs at all, because it will lead to more white males being better educated computer programmers or computer scientists.” I’m sorry, Ben, but I can’t see where I say that. That’s certainly not what I meant in this blog post.

          I meant that we should check. Are CS MOOC completers even more white and male than in a face-to-face CS course? If that’s true, it’s worth investigating further. Is it because MOOCs are serving as a filter? Are there perceived barriers? Is there implicit bias? I’m not saying that any of these are the case. These are possible, though. The first step is to check.

          Ben, have you read “Unlocking the clubhouse” by Jane Margolis and Alan Fisher? They describe the barriers, filters, and implicit bias at Carnegie Mellon University that prevented women from persisting in computer science there. I also recommend the work of Lecia Barker (such as this paper and this one) that describe factors at different University’s introductory CS courses that led to high rates of female attrition from CS programs. Or in contrast, check out the work at UCSD on changes to the course that led to improved retention.

          It is absolutely the case that our courses can drive away women and under-represented minorities, and changes to those courses can serve to improve retention. We can address some issues of inequality from within the classroom.

          Reply
          • 5. BenK  |  January 21, 2013 at 10:05 am

            Thank you for clarifying. I interpreted the post in the context of the title; if I wrote a post about how a cancer drug selectively killed less malignant cells and left more malignant cells alone, and titled the post ‘first do no harm,’ I would presume that I was saying we should no longer administer that drug.

            Of course it is worth investigating who completes MOOCs and what use the education is in the workplace; and if there are some MOOCs that serve well one population or another but miss a second large interest group – then we should probably create a new MOOC that serves that next group.

            In fact, there are two competing models for how to do this. First, one could look at the current interest and serve the largest interest groups first, then find the largest hole and fill that, using whatever criterion makes sense. I would normally cast this as ‘Programming for Computer Programmers’ followed by ‘Programming for Engineers,’ then mathematicians, economists, physicists, chemists, biologists, linguists, etc; but it might be more readily cast as ‘Upper-middle-class men from families of professionals’ and then ‘Upper-middle-class women from east and south asia’ – or by demographic category shows the next most interest. This makes sense because it captures the largest audiences first; it is perhaps relatively efficient.

            Along the same vein, one could presume that the highest participation categories are also the best served by current models, and skip directly to the third and fourth categories – but continue to walk down the interested groups by size, trying to capture the most interest up-front. Ultimately, though, this still amounts to serving already moderately interested parties better rather than addressing the most severe imbalance quickly.

            Another reasonable alternative is to create the first course for the most under-represented group by proportion, attempting to correct the greatest degree of imbalance first, and then stepping through degrees of imbalance. This would serve the least-served (most under-served) first. However, each group might be miniscule and difficult to reach at all, due to outside factors.

            A third reasonable alternative is to create the first course for the most under-represented by total size; to catch the most persons who are also under-represented. This would hopefully address the problem of having your first course attract no attendees (in the worst case scenario).

            I would think (perhaps wrongly) that mechanisms for demographic correctives already exist; but MOOCs are too new for them to have worked yet. Historically Black Colleges and the educational institutions for the deaf, for example, will probably produce MOOCs (or collaborate on their adaptations) soon enough; once producing a MOOC becomes an established credential for academic success honored at prestigious institutions, etc. The incentives need to align. One advantage is that MOOCs can aggregate demand that previously was too dispersed to sustain tailored courses – for instance, programming GIS software for ornithologists – or, GIS programming and interfaces for the visually impaired ornithologist.

            Reply
  • 6. rdm  |  January 21, 2013 at 8:08 am

    I think your prediction in your opening paragraph utterly ignores obvious issues like the international scope of MOOCs.

    Reply
    • 7. Mark Guzdial  |  January 21, 2013 at 8:19 am

      From the data I’ve seen, Raul, I’ll bet that completers are still mostly white and male. Charles Severance’s data show that the vast majority of his Coursera students come from the US and Europe. In the US and Europe, we do have inequality in who participates in computing. MOOCs are mostly a developed world phenomenon.

      It’s perfectly reasonable for those checking MOOC completer demographics to limit the scope to US students, or for a state university, to those in that state. I don’t think it’s necessary for my prediction, but it does make sense in terms of focusing on the student population for which there is a responsibility.

      Reply
      • 8. rdm  |  January 21, 2013 at 8:40 am

        Here’s a heatmap of initial signups from the one MOOC I participated in: http://batchgeo.com/map/2fef51e341887d6385a31e692a27aa99 — roughly speaking, a sixth of the initial signups were neither U.S. nor European in nationality.

        Also… if there’s really a subject matter specific demographic bias in MOOCs — if computer science is [relatively speaking] less interesting for non-white, non-male students in an MOOC context than [for example] electronic engineering — that probably says something about either the subject matter or the way it’s being presented?

        Reply
        • 9. Mark Guzdial  |  January 21, 2013 at 8:45 am

          Signups are the not the same as completers. There are lots of ways that a domain may drive away certain demographic groups. The goal in broadening participation in computing is to make the education more exclusive. We can’t address the issue if we don’t know that there’s an issue — we ought to be checking demographics on MOOC completers.

          Reply
          • 10. rdm  |  January 21, 2013 at 8:55 am

            I understand that. But without significant evidence, that’s an uncertainty and not a bias.

            That said, in that case a student in India corresponded with me (we had encountered each other online a few years previous), and he had more time to invest in the class than I did and he did better than myself. So I have a personal bias which leaves me puzzled at your perspective.

            Reply
            • 11. Mark Guzdial  |  January 21, 2013 at 9:07 am

              It’s uncertainty about a potential bias. If we have an explicit goal to combat bias, shouldn’t we be checking?

              The real challenge of MOOCs is getting our head around the numbers. There are stories about individual MOOC takers around the world, but they are unlikely to be representative of 5K (completers) or 100K (signups). Similarly, 5K completers is a CS class may not be all that significant, if most of those 5K knew the content already, or it’s part of the same 5K who would be studying CS in some other forum without a MOOC. We just don’t know much about MOOCs yet.

              Reply
              • 12. rademi  |  January 21, 2013 at 9:40 am

                We should be investigating, yes. But calling for action before the investigation begins? And expecting someone else to do our investigating for us? Those seem like bad moves.

                Another thing we need to keep in mind here is economies, including economies of scale. For example, if the student’s cost for participating in an MOOC is 1e-5 the cost of participating in a regular class and if the completion rate on an MOOC is 1e-2 that of a regular class, and the only value of an MOOC is completing it, that MOOC will still be a good value for a lot of students (those where the time invested can be neglected, for example).

                But it’s also not the case that the only value of an MOOC comes from completing it. Personally, in the one MOOC I participated in, I only completed the first 3 (biweekly) projects before I ran into time conflicts (with work — note that I also had vacation I was not able to take advantage of, this was a busy year for me) which prevented me from going further. But I learned useful things from those projects. I would certainly have learned more if I could have stayed with the class, but I think it’s important to realize that incompletes can still benefit.

                Reply
                • 13. Mark Guzdial  |  January 21, 2013 at 9:44 am

                  I thought that the only “action” I was calling for in this post was to gather data. It will take action to gather that data — Coursera and Udacity currently do not offer demographic data, even to those offering the MOOC.

                  Reply
                  • 14. rademi  |  January 21, 2013 at 10:52 am

                    This — “Here’s a concrete proposal: Any institution that belongs to NCWIT (or more significantly, the NCWIT Pacesetters program) that runs a MOOC for computer science and does not check demographics should have its membership revoked.” — seems like a call to action.

                    Reply
  • 15. Don Davis (@gnu_don)  |  January 21, 2013 at 10:46 am

    It seems that a significant point of the original post was if MOOCs are only serving to help affluent white males – then they are not as promising as perhaps hoped. No one has said that affluent white males should not learn CS – however, they have not been substantively demographically underrepresented.

    Other groups are substantively underrepresented in CS (NSF, 2011). This happens for a myriad of reasons (Varma, 2006, 2007). – often those (e.g. Cohoon, 2007; Goode, 2008) are not accepted or acted upon by CS educators (as previously mentioned in this blog).

    So then – what is the point? There are serious issues of CS underrepresentation that are having real effects (e.g. Margolis et al., 2008; Tornatzky et al., 2002) on society. Is it not important for educators and researchers to spend time and energy on efforts to help those underrepresented in CS rather than helping a demographic that’s doing fine?

    Cohoon, J. P. (2007). An introductory course format for promoting diversity and retention. Paper presented at the. In Proceedings of the 38th SIGCSE Technical Symposium on Computer Science Education (pp. 395–399). Covington, Kentucky, USA.
    Goode, J. (2008). Increasing diversity in K-12 computer science: Strategies from the field. SIGCSE Bulletin, 40(1), 362–366.
    Margolis, J., Goode, J., Holme, J. J., & Nao, K. (2008). Stuck in the shallow end: Education, race, and computing. Cambridge, MA: The MIT Press.
    National Science Foundation, Division of Science Resources Statistics. (2011). Women, Minorities, and Persons with Disabilities in Science and Engineering: 2011 (Special Report NSF 11-309). Arlington, VA.: National Science Foundation, Division of Science Resources Statistics. Retrieved from http://www.nsf.gov/statistics/wmpd/.
    Tornatzky, L. G., Macias, E. E., & Jones, S. (2002). Latinos and information technology: The promise and the challenge. Claremont, CA: The Tomás Rivera Policy Institute.
    Varma, R. (2006). Making computer science minority-friendly. Communications of the ACM, 49, 129–134.
    Varma, R. (2007). Women in computing: The role of geek culture. Science as Culture, 16(4), 359–376. doi:10.1080/09505430701706707

    Reply
    • 16. rademi  |  January 21, 2013 at 10:58 am

      Yes… if we are trying to solve the problem of educating underprivileged people (and I think we are), then one of our major priorities should be creating efficient (low-cost) access mechanisms. And if the costs are a fraction of a cent per student, we should beware of imposing added costs of a similar magnitude — I think that costs of that magnitude should be a voluntary effort, not an enforced effort.

      That said, this cannot be the only effort, if we want to actually achieve that goal. MOOCs only work for people with adequate network access, and only engage people who already have sufficient interest and hope of success.

      So… I think MOOCs have a role to play, and they can indeed contribute to our knowledge of some of the larger issues. But I wouldn’t put very many eggs in this basket.

      Reply
      • 17. Don Davis (@gnu_don)  |  January 21, 2013 at 12:07 pm

        Rademi, you’ve raised important points. Mightn’t MOOCs be used to generate interest?

        Reply
    • 18. Bri Morrison  |  January 21, 2013 at 11:14 am

      I think what you really want to investigate is the percentage of sign-ups versus completers by gender. You can’t condemn a MOOC for addressing equality if the underrepresented students never sign up. What (I think) you want to investigate is if the MOOCs drive the underrepresented groups away. My prediction is that once you control for those leaving due to time issues, wasn’t what I thought it was, etc., you won’t have enough of each underrepresented group to get statistical significance.

      It’s a motivation issue…how can you teach those who don’t / won’t sign up?

      Reply
  • 19. Don Davis (@gnu_don)  |  January 21, 2013 at 12:05 pm

    Placing the blame on the learners e.g. “they are not motivated enough” is a way of avoiding responsibility for affecting systemic change (Brookfield, 2005). It is an approach that provides a description of a problem without identifying any actionable causal factors (Bijou, 1970; Skinner, 1974).

    Researchers and educators have identified a problem – underrepresentation – and the goal is to amend it. Moreover, researchers have indicated that this underrepresentation represents a vicious cycle – underrepresented students do not have opportunities to meaningfully engage with technology and CS and consequently do not pursue studies in such (Margolis et al. 2008).

    As educators – our job is to affect change – to help people improve students’ lives, to improve society for the benefit of all. If our job was merely to maintain the status quo – what would be the point?

    The question – how can you teach / motivate those who don’t sign up? Students must have interest to sign up and then have that interest supported. There have been several efforts in this direction, e.g. kids summer camps, CS0 classes, with varying levels of success. The question posed has and can form the basis for a lengthy discussion. Suffice it to say – it will be necessary to boost interest and perceived relevance of CS. To accomplish this students should:
    1. be afforded the opportunity to engage meaningfully with technology and the material (Margolis et al., 2008)
    2. have opportunities to interact with meaningful role models – preferably of similar backgrounds – who can explain the importance of CS to the community and their lives (Kuperminc et al., 2008; Ladson-Billings, 1995; Suárez-Orozco et al., 2011)

    Reply
    • 20. BenK  |  January 21, 2013 at 12:31 pm

      Placing the blame on the professors, universities, websites – not enough people who are working two jobs to ‘make ends meet’ are showing up for your on-line course on computer science – is also unreasonable.

      Reply
      • 21. Mark Guzdial  |  January 21, 2013 at 12:55 pm

        Ben, you’re jumping to a conclusion. If there are few women and minorities completing MOOCS, then it’s because they’re working two jobs to make ends meet? If there are few women and minorities completing MOOCs, there’s nothing that professors, universities, and websites can do about it? Let’s first just find out — who is completing MOOCs? That’s all I’m asking for in this post, and I’m placing the responsibility on the backs of the universities offering the courses.

        Reply
        • 22. BenK  |  January 21, 2013 at 2:01 pm

          I was replying to Don; my experience with MOOC data is that the mean age is in the mid-30’s. How can we address the balance of representation across the economic divides in the late 20s and the 30s? How can we engage women and minorities – well out of the age when summer camp is any sort of solution? When the issue is career advancement and work-life balance, but life takes a large chunk of attention in part because of a lack of career advancement, and work takes extra time because of the need to compensate for lower wages?

          Getting college students to switch majors is different. Maybe college faculty are out of touch with the workforce. How to advance early/mid career potential professionals or unstick a stalled career? How to get someone with lower income to approach or exceed the median?

          In short, how to get people to show up who simply don’t have the time and sustained attention to make education a priority? After all, if an attractive educational offering is available, the people who show up the most are those who have the time and energy… This is based on limited MOOC demographic data as well as personal experiences in the workplace and continuing education.

          Reply
          • 23. Don Davis (@gnu_don)  |  January 23, 2013 at 9:34 am

            BenK, those are important questions. As I read the original post, the intent was to highlight that MOOCs will not (necessarily) provide the systemic / societal equity in education everyone is looking for. I sought to highlight some “whys”. You are, of course, quite right about the older ages of MOOC participants and concomitant difficulties of representation.
            Though I have taught adults, my work (teaching & research) has been primarily with high school and undergrad students. I don’t know that the balances can be righted for older students. By the time students are in their late 20s and early 30s, the majority will have established trajectories that make such shifts difficult if not near impossible.
            I think the questions you ask raise – fundamental underlying issues. These are non-rhetorical questions:
            What is the frame of reference for discussing MOOCs – are we talking about helping students move from the bottom income quartile and escape generational poverty or are we talking about unsticking careers?
            How much systemic/societal change can be affected with older (>25) students?

            Reply
            • 24. BenK  |  January 24, 2013 at 7:17 am

              Don; I had already been thinking along these lines – what is the population I am principally interested in, and what impact do I seek to have?
              For example, it is my perception that many people are focused on ‘glass ceilings’ separating the (economic/social) 0.1% from the 0.001% or similar; there are many arguments about why this matters at all for the rest of society, but in truth, I am not much interested. On the other end of the spectrum, many people are serving the lowest 10%, either nationally or globally. In these cases, food security, substance abuse, housing and outright daily violence are major issues. Literacy may be a focus. I have sympathy for this challenge; it seems sometimes overwhelming. Involvement with the military sometimes touches on the issues of the lowest 10% nationally; because people in this group do enlist and seek stable employment and education through that avenue. If MOOCs (particularly computer science, computational biology, and so on) are going to have an impact on the lowest 10%, they will probably do so through organizations like the military – or McDonalds, perhaps, which employ very large numbers of economically disadvantaged persons and may affordably enhance their job training programs and educational programs by systematically rewarding completion/achievement in MOOCs (I’m working on that; it addresses some of the ‘motivation’ issues that are raised).
              But the focus of many educators seems to be moving the lower quartile into the upper half or quartile over the course of a generation. If so, it makes sense to start in early education and move through college admissions. This is not at all my main focus. My main focus is early and mid-career white collar employees (post-college through about 50 years of age; with modestly technical backgrounds), most of whom are attempting to balance work and family life, but are somewhat unstable financially in the context of this difficult economy; to help them attain a more stable employment, a more fulfilling work life, to advance into an increasingly independent professional capacity. They are likely in the third quartile already, but may not ever attain the upper quartile by income. Training and education will help them tread water demographically and manage a flexible career path without being overtaken by younger, less expensive workers.
              To achieve this goal (a worthy goal, I believe), MOOCs may be a powerful element of flexible continuing education.

              Reply
              • 25. rdm  |  January 24, 2013 at 8:27 am

                Of course, McDonalds is hardly the only employer of low income people. Also, there’s a significant population with no employment (and, in recent years, this group can include highly trained individuals – employment requires an employer). All this is somewhat out of scope for MOOCs, but not entirely [it’s entirely possible to lose your job after you acquire a computer, and it’s also possible to retain internet access even while unemployed – unemployment insurance is one example of how this might happen].

                That said, when dealing with low income and unemployed, we should expect significant issues with pre-requisites and study skills. These are not insurmountable hurdles, and might sometimes be addressed by task-oriented documentation. But writing clearly for this varied of an audience can be a daunting task.

                It’s probably also worth keeping in mind that schedule constraints might weigh at least as heavily on this group as on any other. A lack of money often equates to needing to invest time to solve issues that could sometimes be solved cheaply and easily. [Yes, some unemployed are idle, but this is not the only state of affairs.]

                And this brings up another issue: is there any reason for an MOOC to be tied to a school schedule? I feel that there are some reasons here, especially when the MOOC is being generated. But it should also be possible to leave a trail of breadcrumbs, so that later students can take advantage of MOOC materials even after the original class has long since disappeared. There are some issues here [you will no longer have the support of the original class, and grading becomes a problem] but you can automate quizes some assignments allow automated tests and perhaps it’s possible in some cases to create a community which can coach people.

                Anyways, we should not expect that what we are seeing now will remain the state-of-the-art for MOOCs.

                Reply
                • 26. BenK  |  January 24, 2013 at 9:06 am

                  Well, I didn’t want to single out the military either. The services and service industries are both important on-ramps to employability for persons of low social capital. These people have many constraints and challenges, but some employers go out of their way to provide training, sometimes even during the wage earning day. Certain MOOCs might facilitate that.

                  Reply
    • 27. Bri Morrison  |  January 22, 2013 at 9:59 am

      Don, I agree with you completely. I am not indicting the learner for a lack of motivation, I’m raising the point that many underrepresented groups do not have enough *interest* in CS to sign up / persist through a MOOC. With face-2-face classes, I can actively recruit, cajole, encourage, serve as a role model, etc. for underrepresented groups. None of that (that I can see) can be accomplished with a “Here’s an opportunity, sign up if you want” MOOC.
      So the question becomes, how do we accomplish your final two points (early meaningful interaction and role models) that will encourage/motivate the underrepresented to apply themselves to a CS MOOC? Then we can measure if the MOOC retains and stimulates that interest or drives them away.

      Reply
  • 28. Don Davis (@gnu_don)  |  January 23, 2013 at 9:18 am

    Sorry Bri, I greatly appreciate your point. I thoroughly agree with you. I’ve heard “motivation” used so frequently to excuse systemic problems that I had thought it was being used that way. As to the role models and meaningful interaction… One method would be to have such role models physically visit with students… but this doesn’t scale well and is expensive / time consuming to reproduce. Building on the literature, I posit that video games (or similar environments) could be used to support this. [I’ve put together some design recommendations (and a prototype) based on implicit association, relevant pedagogy, and CS underrepresentation research.] While I don’t think that such media interventions can suffice alone – they may offer sunstantive support to ongoing efforts.

    Reply
  • 29. caitlynpickens  |  February 4, 2013 at 11:37 am

    Reblogged this on computing education.

    Reply
  • […] new study supports the concern that MOOCs are a particularly poor fit for underprepared students, the ones most likely to be taking reme…. It relates to the issues raised yesterday about the difficulty of covering all aptitudes and […]

    Reply
  • […] computing education available to women, my guess is that we could be required to make change.  For example, could we be forced to give up MOOCs as a discriminatory practice, since MOOCs have a measurable discriminatory […]

    Reply

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Trackback this post  |  Subscribe to the comments via RSS Feed


Recent Posts

January 2013
M T W T F S S
« Dec   Feb »
 123456
78910111213
14151617181920
21222324252627
28293031  

Feeds

Blog Stats

  • 1,268,139 hits

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 4,565 other followers

CS Teaching Tips


%d bloggers like this: