Winne discusses self-regulated learning “traces”: ambient, or accretion, data that provides insight into a learner’s focus, and motivational, cognitive, and emotional states. Traces can be found in any learning medium: paper-based assignments, learning management systems (e.g., Canvas), software, etc. In addition to specific traces listed in the chapter, what other evidence might be important in increasing a learner’s awareness of his learning habits and supporting self-regulated learning?

In scratching together my notes while reading this selection on Self-Regulated Learning (SRL), I marked the section from page 79 of the publication (page 80 of the PDF) where it states that “SRL is risky because it may have productive or counterproductive results” (Winne) as I found that to be the part that has always been in the back of my mind when I set out to “teach myself” something. Namely, what if I am learning this the wrong way, from the wrong people, or from the wrong starting point?

For example, a physical manifestation of this is when watching a workout video for some new exercise I’ve never done before and I try to either watch my body position and motion in a reflection, but what if I miss something or don’t know that I need to watch for something (like keeping my elbows in or keeping my knee straight) and I end up hurting myself because I didn’t know what I didn’t know (or didn’t see what I couldn’t see).

I would like to see a learning analytics trace report that the learner could review after a study session that grades the quality and reliability of the reference material they used as identified by the traces for fitness of purpose. For example, some would say that newspapers are generally a reliable source of information, yet the Chicago Daily Tribune is famous for its 4 November 1948 misprint of “DEWEY DEFEATS TRUMAN” illustrated below with President-elect Truman holding the paper up for the world to get the last laugh (Rollins, 2019).

As we discuss and explore AI-powered learning analytics, the need for data cleaning (Tableau, n.d.) is even greater if we truly want to be able to trust the results and guidance of the AI we are using (Vorvoreanu & Walker, 2022).

While determining the quality of the references used in SRL is complicated enough (RUSA/RSS Evaluation of Reference and User Services Committee, 2007), creating the AI to guide and assist in advancing the SRL practitioner’s skills there needs to be a set of universally accepted (and adjustable) guidelines to ensure that any prejudice or bias an SRL practitioner has doesn’t detrimentally impact the source of the learning. AI will not make the “good instructional content” acquisition problem go away, but it can help it to if programmed correctly. It is a problem much like the RISC vs. CISC problem in microprocessor design (Sari, 2023) which lets us just move the complexity of the problem to be solved from hardware to software so we can more easily tweak and adjust as needed.

With an AI-driven qualitative analysis overlaid with the quantitative data feedback from the traces, the SRL practitioner can be informed of the validity of their SRL practices and if they are not picking fair, balanced, factually accurate, and reliable sources of information, they would be informed so they could adjust accordingly. Of course, for the AI to reliably do that, the AI itself must be pure of heart (Kalluri, 2020).

References

Kalluri, P. (2020). Don’t ask if AI is good or fair, ask how it shifts power. Nature, 583, 169. https://doi.org/10.1038/d41586-020-02003-2

Rollins, B. (2019, January 8). File:Dewey Defeats Truman.jpg. Wikimedia Commons: https://commons.wikimedia.org/wiki/File:Dewey_Defeats_Truman.jpg

RUSA/RSS Evaluation of Reference and User Services Committee. (2007, December 11). Measuring and Assessing Reference Services and Resources: A Guide. American Library Association: https://www.ala.org/rusa/sections/rss/rsssection/rsscomm/evaluationofref/measrefguide

Sari, S. (2023, May 13). RISC vs. CISC. Baeldung.com: https://www.baeldung.com/cs/risc-vs-cisc

Tableau. (n.d.). Guide To Data Cleaning: Definition, Benefits, Components, And How To Clean Your Data. Retrieved September 10, 2023, from Tableau.com: https://www.tableau.com/learn/articles/what-is-data-cleaning

Vorvoreanu, M., & Walker, K. (2022, February 1). Advancing AI trustworthiness: Updates on responsible AI research. microsoft.com: https://www.microsoft.com/en-us/research/blog/advancing-ai-trustworthiness-updates-on-responsible-ai-research/

Winne, P. H. (n.d.). Learning Analytics for Self-Regulated Learning. In C. Lang, G. Siemens, A. F. Wise, D. Gašević, & A. Merceron (Eds.), Handbook of Learning Analytics (Second ed., pp. 78-85). Society for Learning Analytics Research (SOLAR). https://doi.org/10.18608/hla22.008