| Mapping the Field: Signposting |
| As I research, I think “some history of AI would be
good - but not too much. Maybe the first Lit Review
section?” I also notice scholars talk mainly about learning, teaching and administration. And here’s a gap - people always neglect administration! |
And I notice a lot of literature speculates on the future of ed. So
maybe a conclusion that reviews future
schooling? |
| A History of AI in Education AI and Learning AI and Teaching AI and Administering Pedagogy Research Gap: Impact of AI on Higher Ed Admin Schools of the Future |
| A History of AI in Education In the 1950s… By the 1980s… Then in the 2010s… AI and Learning Tailored Learning… …But also Plagiarism AI and Teaching Productivity Technology Fatigue (“This wasn’t the job I signed up for!”) |
4. AI and Administering Pedagogy Lowered cost of Education Delivery Surveillance Schools? 5. Impact of AI on Higher Ed Admin a. Automating procedures (communications, enrolments etc). b. Greater efficiency, less control 6. Schools of the Future An AI on every device Ubiquitous, Life-long Learning The End of Vocational Training? |
| The integration of AI technologies in educational administration has
shown promising potential for reducing the cost of education delivery.
Several studies have explored this aspect, with varying conclusions.
Smith et al. (2022) conducted a comprehensive analysis of 50 higher
education institutions that implemented AI-driven administrative
systems, finding an average cost reduction of 15% in operational
expenses over a three-year period. This aligns with earlier predictions
by Johnson (2019), who argued that AI could significantly streamline
administrative processes, particularly in areas such as enrollment
management and resource allocation. However, Patel and Lee (2023)
caution against overestimating these cost savings, highlighting that
initial implementation costs and ongoing maintenance of AI systems can
be substantial. Their case study of a large public university revealed
that while AI reduced certain administrative costs, these savings were
partially offset by new expenditures in IT infrastructure and staff
training. Despite these challenges, the consensus among researchers like
Zhang et al. (2024) is that AI’s potential to lower education delivery
costs is significant, particularly when implemented strategically and
with a long-term perspective. They emphasize that the true value of AI
in reducing educational costs may lie not just in direct savings, but in
its ability to enhance efficiency and free up resources for core
educational activities. |
| An example: [New text] Contrary to what scholars have argued in relation to teaching and learning, the literature on higher education administration is generally positive. [Old text] The integration of AI technologies in educational administration… (blah blah) …not just in direct savings, but in its ability to enhance efficiency and free up resources for core educational activities. [New text] Such positive interpretations largely ignore potentially destructive effects of AI on professional staff morale, and how this in turn impacts teaching and learning outcomes. My research explores whether these effects exist, and thus addresses a gap. As I show in the next section, this gap helps to develop more realistic projections of how schools of the future may be designed to address both challenges and opportunities of AI. |
| An example: [New text] Contrary to what scholars have argued in relation to teaching and learning, the literature on higher education administration is generally positive. [Old text] The integration of AI technologies in educational administration… (blah blah) …not just in direct savings, but in its ability to enhance efficiency and free up resources for core educational activities. [New text] Such positive interpretations largely ignore potentially destructive effects of AI on professional staff morale, and how this in turn impacts teaching and learning outcomes. My research explores whether these effects exist, and thus addresses a gap. As I show in the next section, this gap helps to develop more realistic projections of how schools of the future may be designed to address both challenges and opportunities of AI. Nod back** to *****previous***** sections Point *****forward***** to next section**… |
| Famous AI paper: Attention is All You Need (Vaswani et
al. 2017). Kickstarting Transformer-based models, generative AI, GPT
etc But also true of course for human readers – including professors Part of the craft of academic writing is reminding the reader: Where you’ve been Where you’re going If a literature review is a map, constellation, wall etc, and you are showing the reader where the gap is You also want to show your control of the literature – “here’s the gap – and I know its a gap, because of what I showed you here, here and here” Compared to your impression as an academic writer, you can almost never bore your reader with signposts, stage directions etc. (to a point) |
| Chapter Intro: include a para that says words to the effect: “In the
last chapter, I argued / showed / demonstrated X. In
this chapter, I will do Y. That
involves covering A,
B and C”. (X = past tense / Y = future
tense) Chapter Conclusion: include a para that says words to the effect: “In this chapter, I argued / showed / demonstrated Y. That involved talking about A, B and C. In the next chapter, I will be looking at Z”. (Y = past tense / Z = future tense) Of course you rarely do things so formulaically. This is one area where Gen AI can help - “rephrase this for me: …” Lack of signposting is typically for early drafts, and is the reason for many comments during review |