A blast from the past

This post is a winding back the clock a few years to take another look at a correlation we noticed some time back. A correlation that is somewhat curious.

Ever since David and I started the Indicators project way-back-when, we have considered student clicks within the learning management system (LMS) as an indicator (and only an indicator) of student behavior. This was based on correlations between student clicks on the LMS and their resulting grade as per the following chart (updated recently).

Screen Shot 2015-12-01 at 10.27.40 AM

We have no way of knowing what a click means as I suspect some students are like me in that they randomly click around the place (as a way of procrastinating in my case). There’s a good paper talking about this from Jason Lodge and Melinda Lewis from 2012.

We know that clicks are meaningless to a large extent but they are one of the few unobtrusive indicators we can easily extract from the LMS. On average more clicks within the LMS == better grades. Yes I know! This is not universally true. It’s just an average. Our own 2012 paper suggested why this isn’t universally true.

One thing we did notice back then was not so much the quantity of clicks that the students were making, but the variety of different content items that they were clicking on. What happens if we consider a click by a student on a particular activity or resource as a connection and disregard how many times they click on that particular activity or resource?

So using the same dataset (n=34930) as the previous chart above, the following chart is showing the average number of connections for each student grade.

Screen Shot 2015-12-01 at 10.38.42 AM

To me, this looks very similar to the trend from the previous chart that showed clicks against grade. However what I did find very interesting is the average number of clicks that each student grade group made on each individual content item.

Grade group Average clicks per course element
HD 5.6
D 5.4
C 5.6
P 5.7
F 4.8

I found this interesting because of the apparent lack of variation between the different grades. Broadly speaking, each grade group makes roughly similar amounts of clicks on each activity and resource within Moodle. However the higher achieving students click on a larger proportion of the course activities and resources but don’t necessarily click more on each individual element. I guess I’m not surprised as, if we take a network perspective; the higher achieving students have a greater number of nodes in their network than the lower achieving students. Something that the SNA folk have known for some time.

 

 

Advertisements

More learning analytics ponderings

There are many publications and presentations espousing the potential of learning analytics to contribute to improved decision-making in education. For example:

“Basing decisions on data and evidence seems stunningly obvious, and indeed, research indicates that data-driven decision- making improves organizational output and productivity.”
(Siemens and Long 2011)

This quote was based on very interesting research that showed improved performance across a bunch of publicly listed companies that better utilized data-driven decision-making (Brynjolfsson, Hitt, and Kim 2011). I like the Siemens quote and the research it was based on for two reasons, one good and one bad. Firstly, I work in a regional Australian university and I know that higher education can do a lot better with regards to making decisions based on evidence. Secondly, the Siemen’s quote appeals to me because (a) I have great respect for his opinion and (b) there is a part of me that believes that better data and information will lead to better decisions.

Note that I used the work ‘believes’ deliberately in the previous sentence. Upon reflection, this is probably more of a faith thing than a scientific reality. The trouble is that better and particularly more detailed data does not necessarily lead to better sense-making and/or decision-making (Aaltonen 2008). Human beings are just not wired that way.

“We are a bricolage of cognition, emotion, intuition, information consumption, doubt and belief”
(Siemens 2006)

This highlights the importance of including human beings in the learning analytics cycle. Something that Doug Clow has previously noted:

“Previous work in the literature has emphasised the need for and value of human meaning-making in the process of interpretation of data to transform it in to actionable intelligence.”
(Clow 2014)

So even though we know that humans are essential in the information / learning analytics cycle, we also know that humans are bad at making ‘rational’ decisions based on data. We filter our observations of the world through our cognitive frameworks. Our frameworks are individual to each and every one of us and include things such as experience, intuition and instinct. Throw in that learning analytics is closely coupled with IT where the main considerations are likely to be precision, rigor and reproducibility rather than the human consumers (Norman 1993).

“The logic behind many investments in IT tools and big data initiatives is that giving managers more high-quality information more rapidly will improve their decisions and help them solve problems and gain valuable insights. That is a fallacy.”
(Marchand and Peppard 2013)

A long way of saying that I’m wondering if we need to spend more time on the recipients of the information rather than the data side of things?

 References

Aaltonen, Mika. 2008. “Multi-ontology, sense-making and the emergence of the future.” Futures 41:279-283. doi: 10.1016/j.futures.2008.11.017.

Brynjolfsson, Erik, Lorin M Hitt, and Heekyung Hellen Kim. 2011. “Strength in numbers: How does data-driven decisionmaking affect firm performance?” Available at SSRN 1819486.

Clow, Doug. 2014. “Data wranglers: human interpreters to help close the feedback loop.” Proceedings of the Fourth International Conference on Learning Analytics And Knowledge.

Marchand, Donald A., and Joe Peppard. 2013. “Why IT Fumbles Analytics.” Harvard Business Review 91 (1):104-112.

Norman, Donald A. 1993. Things that make us smart: Defending human attributes in the age of the machine: Basic Books.

Siemens, George. 2006. Knowing knowledge: Lulu. com.

Siemens, George, and Phillip Long. 2011. “Penetrating the Fog: Analytics in Learning and Education.” EDUCAUSE Review 46 (5):9.

 

A possible indicator of faddism?

I have been preparing a document for my PHD supervisors that makes mention of a paper we wrote last year. The paper titled “Three paths for learning analytics and beyond: Moving from Rhetoric to Reality” talks about the dangers associated with learning analytics and management fads, and how the hype around technological concepts can swamp deliberate and mindful adoption and implementation.

While looking at this paper I conducted a quick search on Google scholar, year by year using the search term “learning analytics”. While it’s not particularly scientific, the trend is interesting nonetheless.

Screen Shot 2015-10-30 at 11.02.54 AM

Peer review of a colleagues assessment item

This post is a quick peer review of a design based research proposal a colleague has developed for a unit in their masters where they are required to demonstrate peer review. Good luck Rebecca!

The proposal is centered on a short course, five weeks in duration that is offered to high school students so as to provide them with some insight into tertiary education. The course uses Conley’s model of college readiness to guide what is taught in the course, which includes the following facets of readiness:

  • Key cognitive strategies
  • Academic knowledge and skills
  • Academic behaviours
  • Contextual skills and awareness

It is clear that there are some issues associated with the short course as it stands now. Some of these issues resonated with me as they are not limited to just this short course. For example, one of the problems mentioned is linked with the dominant online course delivery mechanism in higher education, the learning management system (LMS). Rebecca points out that LMS delivered courses have transactional distance, are instructor led and have to be completed in an allocated timeframe. According to the proposal introduction, the style of teaching and learning afforded by the LMS is not constructivist, connectivist or conducive to learner autonomy and critical thinking. All sentiments I agree with to some extent. However, given the dominance of LMS as the way that eLearning is delivered in higher education (Coates, James, & Baldwin, 2005), and given that this is a preparatory course for high school students, it seems appropriate that future students gain some experience with this medium, warts and all.

The proposal mentions moving towards a “more heutagogical approach”, I assume to offset some of the limitations associated with LMS based eLearning. I’m no expert at Heutagogy but I must admit that the idea of self determined learning is very attractive in comparison to the current approaches to eLearning. The proposal also considers the student cohort, many of who live in rural or remote areas, and come from low socio-economic backgrounds. This can correspond to limited access to technology, such as reliable broadband internet connections.

The proposal describes three research questions:

  1. Does restructuring the Preparation for Success in Health course to include a Heutagogical approach, allow students to collaborate, critically reflect and provide feedback in an open online environment?

  2. Does shifting the knowledge acquisition into the students’ hands mean they will access a wider variety of sources of information, including health professionals to answer their questions and build on their own ideas of what appropriate knowledge is?

  3. Will the students engage in the Preparation for Success in Health course more authentically if allowed to be more self-directed in their approach to learning, thus engaging in deeper cognitive learning?

The proposal’s literature review hinges upon student centred learning and suggests that Heutagogy might be an appropriate framework for digital age learning. In a Heutagogical approach, the learners are highly autonomous and the focus is on building the learner’s capacity to learn. Experiential and reflective learning is preferred over ‘transmissive’ approaches. On the surface at least, Heutogogy seems to have a lot in common with the personal learning environment literature from a while back. Even the graduate/generic attributes folk talk about some of this albeit from a different perspective, as do the problem based learning folk. The double-loop learning approach described in the proposal interests me as it links nicely to my PHD around complex adaptive systems. Non-linear learning, to me, is a better match for how people really learn from an anthropological perspective, yet the dominant socio-technical approach is very linear (IMHO).

Some feedback on the proposal:

Very interesting and worthwhile proposal. I’ll be disappointed if it doesn’t happen. Have you considered an internal SOLT grant?

  • The chosen methodology is design based research or DBR. This needs to be unpacked more comprehensively. It is not clear to me how the implementation plan links with the methodology. DBR does seem to be driving the methodology behind implementation plan, but it’s not explicit.
  • I’m not sure about the three research questions. All together they describe a very broad scope that includes a huge body of literature. My advice would be to narrow the scope somewhat. The first research question is great, and, in my mind, enough.

More specifically:

  • Conley’s model needs to be unpacked in the introduction.
  • Heutagogy is introduced in the introduction without introduction. I would suggest that describing the problem in the introduction, without Heutagogy, is the way forward. Then introduce it as an alternative framework in the body of the proposal.
  • The proposal could benefit from an abstract/paragraph/exec summary to help set the scene for the introduction section.
  • The proposal touches on constuctivism, connectivism and technology. Might be just my pattern-entrainment but this could be a nice way to articulate how this proposal is different and is challenging the status quo. Some of the technological issues associated with the LMS, David has unpacked here.

All in all, a very interesting and worthwhile proposal. Well done and good luck.

References

Coates, H., James, R., & Baldwin, G. (2005). A critical examination of the effects of learning management systems on university teaching and learning. Tertiary education and management, 11(2005), 19-36.

The ‘wickedness’ of student attrition and retention

Earlier this year Dr Celeste Lawson and I wrote a couple of papers (still in review) about student attrition in Australian Higher Education. In the first paper we looked at the actual nature of the student attrition problem while in the second, we looked at the approaches that universities took in their attempts to address student attrition. This post is focusing on the first paper in which we questioned the way that universities conceptualized their student attrition problems. Note: I should add that referring to student attrition as a problem is probably wrong. It creates a negative impression of the situation and and also implies that it has to be solved. This is perhaps the wrong way to think about student retention and attrition.

“Reasons for student non-completion are complex”

(Maher & Macallister, 2013)

We analyzed a survey of students who started, but failed to finish their degrees. The results were pretty much what folk in higher education have come to expect. Students leave for reasons such work commitments, family commitments, financial problems, personal problems, health problems and so on; The usual array of reasons that are found throughout the student attrition literature.

If we consider student attrition as a problem within a linear (causal) system, (which as a sector we tend to do) these factors can be addressed systematically within the organizational hierarchy. For example, many students mentioned work commitments as a significant factor in their decision to leave the university. The typical university response would be to perhaps develop an instructional time-management module for new students; or allocate a learning support person who can help students with their study load; or provide a service whereby students could receive advice on how to better balance their work-study life. All of which are valid responses if the problem was single dimensional.

We conducted a content analysis on the free text comments that students made within these surveys and looked closely at the factors that led to attrition. We found that it was the accumulation of factors that led to the students dropping out, and not single reasons. The complex interplay between a range of factors, and the students’ context ends in their premature departure from university. The following diagram from our paper attempts to visualize this by showing relationships between attrition factors. Note that the strength of the line between the factors indicators the frequency in which the factors appeared together:

Screen Shot 2015-10-14 at 8.14.27 AM

What is not shown here (and is perhaps an avenue for future research) is that a similar diagram showing weighted interactions between contributing factors, could be developed for individual students. So from a university perspective, we have, not a single or even a series of issues to address, but a complex network of context dependent issues. Many of which are beyond our ability to address, or even perceive. Add to this that even the small subset of contributing factors shown above, have dependencies at multiple levels. For example, a student might identify as struggling with a financial situation that could have been externally triggered at a local, regional or national level.

It appears we have a complex web of inter-related and temporal factors that can contribute to a student withdrawing from their studies. We describe this in the paper as a wicked problem, which I have mentioned before. Wicked problems are difficult to define, have many inter dependencies, are multi-causal, unstable and socially complex (Briggs, 2007). Importantly, traditional bureaucracies with their vertical silos are unable to tackle wicked problems that are ambiguous and lack clarity. Traditional bureaucracies are also risk adverse which can inhibit the innovation, experimentation or bricolage needed to address wicked problems (Briggs, 2007).

the social complexity of wicked problems as much as their technical difficulties that make them tough to manage
(Camillus, 2008)

Where to from here is the million-dollar question although there are some ideas in the literature about tackling wicked problems that require exploring. Two in particular grabbed my attention given my interest in complex adaptive systems:

  • Involve stakeholders, document opinions and communicate, especially horizontally. This appears to align with the complexity thinking around ongoing ethnographic collection and growing the network conduits between agents. There are also some links to the self-assertive and integrative paper that David has mentioned.
  • Focus on action. Something I’ve been banging on about recently in regards to learning analytics. Detailed planning and analysis are of little use in complex systems, or in this case, with wicked problems, as the future systems states cannot be predicted due to unknowable effects stemming from interaction between agents. Take a number of small-scale actions and monitor for emergence and repeat. As opposed to upfront planning and analysis, then a single course of action. Safe-fail probes is the term that Snowden uses, and it makes a lot of sense (no pun intended).

It is safe to say that there are no silver bullets when it comes to student attrition. However, I believe there is scope to start thinking about and tackling attrition differently. Attrition is a complex multi-causal issue that the sector continues to try and address using SET mindsets and methods. I’m saying we need to think about it differently, and perhaps engage in some BAD practices.

References

Briggs, L. (2007). Tackling wicked problems: A public policy perspective. Canberra: Australian Government, Commonwealth of Australia.

Camillus, J. C. (2008). Strategy as a Wicked Problem. Harvard Business Review, 86(5), 98-106. Retrieved from http://ezproxy.cqu.edu.au/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=bth&AN=31730150&site=eds-live&scope=site

Maher, M., & Macallister, H. (2013). Retention and attrition of students in higher education: Challenges in modern times to what works. Higher Education Studies, 3(2), p62.

Is learning analytics hamstrung from the outset?

Over the last six months I’ve been writing about student attrition and retention with a colleague from work. We’ve submitted a couple of journal articles that are currently in review about how universities continue to misinterpret the nature of their student attrition issue. To cut a long story short, we argue that attrition is a wicked problem, a problem in complexity. Climate change, deforestation and geopolitical conflicts are examples of wicked problems where there are no single solutions or even any obvious paths towards solution. Conceptualizing student attrition as a wicked problem occurring within non-linear, complex systems changes how we approach these types of issues. As an aside, even classifying attrition as a problem (or issue) the wrong way to think about it. It’s more like a symptom of a network of problems, a network where we can’t possible know what or where most of the nodes are.

While writing these papers I saw some similarities between how universities are approaching student attrition and how they are approach learning analytics adoption. In both cases they have mis-specified the nature of the organization. Their approaches are based on assumptions that the organisations are machine-like.

“Managers want workers to respond predictably to incentives and to accomplish goals defined by managers and to do this with little deviation from plans that management has developed to improve performance”(McDaniel, 2007)

The Machine

The machine like model of organizations is associated with management approaches based on command, control and planning (McDaniel, 2007). This is a valid approach for managing in a linear, stable environment where future states can be anticipated. In fact these approaches depend on the ability of managers and workers to forecast future system states (McDaniel, 2007). However, if we view organisations and the environments in which they operate as complex adaptive systems, machine-model management no longer works. It is simply not possible to predict future states when the systems are made up of agents that are information processors with the capacity to modify their behavior based on information they receive (J. Holland, 2006; J. H. Holland, 1995). An important contrast between viewing an organization as a machine or as a complex adaptive system is the diversity of the agents within the system. Complex adaptive systems encourage diversity whereas the machine model tends to favor agent homogenization (McDaniel, 2007).

“Participation of clinicians in hospital strategic decision making is more helpful in terms of bottom line performance than the participation of middle managers”
(Ashmos, Duchon, McDaniel Jr, & Huonker, 2002)

“If we want workers to be able to improve performance in the face of unknowability, we must invest in efforts to help them make sense of the world in a way that enables the organisation to take action and to learn about the world from the actions that are taken”
(McDaniel, 2007)

So we appear have a mis-interpretation of the actual nature of organisations and the environments in which they operate. Organisations that are composed of information processing agents that change and adapt with the information they receive. And the rapid spread of hype around learning analytics, fueled by commercial entities, aimed at assisting decision-makers at many different levels of the academy. While I believe that learning analytics has enormous potential, I can’t help wondering if the fundamentally misunderstood nature of organisations is going to be its greatest limiting factor, especially when we are talking about information/action cycles.

 References

Ashmos, D. P., Duchon, D., McDaniel Jr, R. R., & Huonker, J. W. (2002). What a mess! Participation as a simple managerial rule to ‘complexify’organizations. Journal of Management studies, 39(2), 189-206.

Holland, J. (2006). Studying Complex Adaptive Systems. Journal of Systems Science and Complexity, 19(1), 1-8. doi:10.1007/s11424-006-0001-z

Holland, J. H. (1995). Hidden order : how adaptation builds complexity / John H. Holland: Reading, Mass. : Addison-Wesley, c1995.

McDaniel, R. R., Jr. (2007). Management Strategies for Complex Adaptive Systems: Sensemaking, Learning, and Improvisation. Performance Improvement Quarterly, 20(2), 21-41. Retrieved from http://onlinelibrary.wiley.com/store/10.1111/j.1937-8327.2007.tb00438.x/asset/j.1937-8327.2007.tb00438.x.pdf?v=1&t=h672itdt&s=a7c341c21237351ad995bdb34074c98db94bb026

A little about sensemaking

In my previous post I described situation awareness as it applies to learning analytics in complex adaptive systems. The aircraft analogy I used, compared the black-box flight recorder with the cockpit instrumentation to differentiate the role of real-time sensemaking with retrospective analysis. As David commented, learning and teaching is more complex than flying an aircraft, which means the instrumentation and the black-box need to be more configurable and adaptable than the analogy would suggest. Irrespective of the analogy, my suspicion is that a majority of learning analytics projects are too focused on retrospective data analysis. This analysis has limited value in complex contexts when there is a need to act and adapt in the here-and-now.

That said, whether the learning analytics data is retrospective or real-time, they are both used by humans to make sense of something. Sensemaking is a well-researched phenomenon and I think it has the potential to help us improve learning analytics. Even my favorite learning analytics definition alludes to the important role that sensemaking has:

“Learning analytics is the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs”

Complexity

What is sensemaking?

Sensemaking refers to how we structure the unknown so as to be able to act in it (Weick, 2005). Sensemaking involves coming up with a plausible understanding, a map, of a shifting world; testing this map with others through data collection, action, and conversation; and then refining, or abandoning, the map depending on how credible it is (Ancona, 2010). Action is not a separate or later step in sensemaking, but is an integral part of it (Ancona, 2010). This aligns neatly with how agents act in complex adaptive systems whereby they probe, sense and respond. The unanticipated and unintended consequences of acting within a complex adaptive system make upfront analysis less valuable than typical organizational ways of doing things would suggest. (Weick, 2005) summarises this nicely when he says:

“ To work with the idea of sensemaking is to appreciate that smallness does not equate with insignificance. Small structures and short moments can have large consequences”

I mention organisations (such as universities) deliberately because the risk averse SET mindset drives approaches (such as those involving learning analytics) and are based on upfront analysis and one-off projects (I would say at the expense of sensemaking). The sense-making and decision-making models that I would associate with the SET mindsets are outmoded models based on linear and stable environments (Mika, 2008). In these models, analysis and upfront design make rational sense, but they do not match the unstable and turbulent contexts that we see today.

One of things that I notice with learning analytics is that data from information systems receives most, if not all the focus. According to (Ancona, 2010) when sensemaking, you “seek out and combine many different types of data. This includes system data and narrative of people involved”. The area of narrative is an area that I’m quite interested in, an interest learned from Dave Snowden’s Cynefin podcasts from sometime ago. I just wonder if the learning analytics community is a little too focused on data from information systems when there is an untapped human sensor network available?

The following are some other interesting quotes I found while scanning the sensemaking literature that I have to consider further:

“Failure is part of sensemaking”
(Ancona, 2010)

People create their own environments and are then constrained by them.
(Ancona, 2010)

“much of the effort to design information technology to support cognition in organizations has not addressed its distributed quality”
(Boland, 1994)

“Sensemaking is about the interplay of action and interpretation rather than the influence of evaluation on choice”
(Weick, 2005)

“ignorance and knowledge coexist, which means that adaptive sensemaking both honors and rejects the past”
(Weick, 2005)

References

Ancona, D. Framing and Acting in the Unknown.

Boland Jr, R. J., Tenkasi, R. V., & Te’eni, D. (1994). Designing information technology to support distributed cognition. Organization science, 5(3), 456-475.

Mika, A. Multi-ontology, sense-making and the emergence of the future. Futures, 41, 279-283. doi:10.1016/j.futures.2008.11.017

Weick, K. E., Sutcliffe, K. M., & Obstfeld, D. (2005). Organizing and the process of sensemaking. Organization science, 16(4), 409-421.