CNDLS Senior Scholar Dr. Phil Long Shares: How a Background in Biology Informs Work in Learning Analytics

  We are thrilled to welcome Dr. Phillip Long as a CNDLS Senior Scholar. Previously the Chief Innovation Officer for Project 2021 & the Associate Vice Provost at the University of Texas, Austin, Long’s work in the learning sciences focuses on emerging technologies and our cognitive interactions with them. This is the second post in our series highlighting his background in Biology and how it informs his work in learning analytics.


How does your research background in biology inform how you undertake inquiry in learning analytics, and how does it help you understand the role of developing technologies in the learning experience?

I think one’s disciplinary training always influences one’s approach to inquiry and problem-solving. That’s one of the things one learns when joining a research community. I was trained as a behavioral ecologist and evolutionary biologist. That involves understanding behavioral observational methods and their distinct limitations. It embraces a systems-oriented view to help understand the patterns and phenomena one is observing. There is a keen awareness that one is capturing data through a ‘lens’ that imposes what is selected and what passes through. Choosing a different lens changes what you capture, and the sensitivity of the lens is predictive of the range of and acuity of the information that is recorded. There is an acute awareness that the amount of information before you is enormous and frankly overwhelming. Humans have very restricted sensory capacities. We think, for example, that what we ‘see’ represents what is in front of us. Yet our eyes use only limited frequencies to trigger neural responses. Our brain filters further after what is ‘let in’ and discards some information, preferencing others.  This is a result of millions of years of natural selection, chance, random mutations, and who knows what else? That knowledge is something that comes with training in that discipline. In practical terms it simply means that one must be cautious about interpretation, aware of the possible influences the very methods we use impose on what we capture. And most importantly it forces a strong bias toward analysis that is grounded in a theoretical context. The intention is always two-fold: challenge a framework and at the same time use it to focus one’s limited tools on the questions that the framework enables you to ask. In the simplest case with learning analytics, many earlier studies started with clickstreams and the data collected used to infer behavior, motivation, engagement, and host of other psychological constructs. This was and is a very crude ‘lens’ onto a problem. What makes us think that the mere frequency of a click, the ability of which to even be expressed was shaped by the UI/UX software designer, gives us a reasonable path to infer motivation? It is perhaps the greatest challenge of learning analytics to address the limited availability and the limited breadth of possible information that can be collected about human behavior when the only available lens is through digital systems mediated by a screen, mouse and keyboard. We’re slowly getting beyond that in certain areas, e.g., video motion analysis, but the disciplinary training I received reminds me constantly of limitations that arise and confusion that is imposed when we conflate the digital world with the world of full sensory experience that humans inhabit. And, as I noted earlier, that world is itself constrained by our own sensory evolution. How does my own disciplinary training frame my understanding of developing technologies in learning? It presents a constant reminder of the enormity of what we don’t know about learning, and the awareness of the tiny windows developing technologies afford. It also brings wonder and appreciation for what we have accomplished with such crude—if improving—tools.  

 

We are thrilled to welcome Dr. Phillip Long as a CNDLS Senior Scholar. Previously the Chief Innovation Officer for Project 2021 & the Associate Vice Provost at the University of Texas, Austin, Long’s work in the learning sciences focuses on emerging technologies and our cognitive interactions with them. This is the second post in our series highlighting his background in Biology and how it informs his work in learning analytics.


How does your research background in biology inform how you undertake inquiry in learning analytics, and how does it help you understand the role of developing technologies in the learning experience?

I think one’s disciplinary training always influences one’s approach to inquiry and problem-solving. That’s one of the things one learns when joining a research community.

I was trained as a behavioral ecologist and evolutionary biologist. That involves understanding behavioral observational methods and their distinct limitations. It embraces a systems-oriented view to help understand the patterns and phenomena one is observing. There is a keen awareness that one is capturing data through a ‘lens’ that imposes what is selected and what passes through. Choosing a different lens changes what you capture, and the sensitivity of the lens is predictive of the range of and acuity of the information that is recorded.

There is an acute awareness that the amount of information before you is enormous and frankly overwhelming. Humans have very restricted sensory capacities. We think, for example, that what we ‘see’ represents what is in front of us. Yet our eyes use only limited frequencies to trigger neural responses. Our brain filters further after what is ‘let in’ and discards some information, preferencing others.  This is a result of millions of years of natural selection, chance, random mutations, and who knows what else?

That knowledge is something that comes with training in that discipline. In practical terms it simply means that one must be cautious about interpretation, aware of the possible influences the very methods we use impose on what we capture. And most importantly it forces a strong bias toward analysis that is grounded in a theoretical context. The intention is always two-fold: challenge a framework and at the same time use it to focus one’s limited tools on the questions that the framework enables you to ask.

In the simplest case with learning analytics, many earlier studies started with clickstreams and the data collected used to infer behavior, motivation, engagement, and host of other psychological constructs. This was and is a very crude ‘lens’ onto a problem. What makes us think that the mere frequency of a click, the ability of which to even be expressed was shaped by the UI/UX software designer, gives us a reasonable path to infer motivation?

It is perhaps the greatest challenge of learning analytics to address the limited availability and the limited breadth of possible information that can be collected about human behavior when the only available lens is through digital systems mediated by a screen, mouse and keyboard. We’re slowly getting beyond that in certain areas, e.g., video motion analysis, but the disciplinary training I received reminds me constantly of limitations that arise and confusion that is imposed when we conflate the digital world with the world of full sensory experience that humans inhabit. And, as I noted earlier, that world is itself constrained by our own sensory evolution.

How does my own disciplinary training frame my understanding of developing technologies in learning? It presents a constant reminder of the enormity of what we don’t know about learning, and the awareness of the tiny windows developing technologies afford. It also brings wonder and appreciation for what we have accomplished with such crude—if improving—tools.