Blurts

I have written about the use of technology in classrooms since the mid-1990s. Looking back over some of the topics that interested me, it seems that some topics that were emphasized at one time and would still be equally as valuable now seldom show up in what I would describe as the “online discussion”. Two examples that I would place in this category would be a) project-based learning and b) the classroom use of digital probes.

I have been thinking about how the observation that such trends exist might be demonstrated. I remember years ago that Google made data available about the frequency of search terms and I looked to see if I could still find such data. Google still offers what it calls Google Trends.

My exploration involved the following: a) science of reading, b) project-based learning, c) computer literacy, and d) classroom AI. A direct comparison among search goals is difficult because some issues are likely to be searched using a consistent phrase and others with a variety of terms. I think Google recognizes this challenge and differentiates search term from search topic. However, with the topics I wanted to investigate both term and topic were not always available and a consistent approach would not work. I decided that focusing on relative trends would be more meaningful than comparing relative frequency at a point in time. I have no idea if this is true or not.

This graph compares data from 2004 through the present. The blue line represents “project-based learning”, the red line “science of reading”, the yellow line “computer literacy”, and the green line “classroom AI”. Again, the trends rather than the relative frequency seem most useful because of the confounds I have already mentioned. The Science of Reading and Classroom AI (red and green) have shown recent increases in interest. Computer literacy (much more general than something such as learning to program) and project-based learning have shown declines (blue line, yellow line).

This tool would seem interesting to explore within a classroom setting. What trends might interest students?

Here is a recommendation for educators at all levels interested in a broad perspecctive on present AI applications in education. Lawrence Holt has updated his and colleagues’ existing work titled “A map of generative AI for education”. This work describes the multiple ways AI is being used by educators, learners, and administrators and provides links to a large selection of applications available to implement these applications. This is a great and complete work identifying many applications and services I had not tried. For any educator, you will learn something new from this content.

As a side note, this resource is provided through Medium. Medium is a subscription service asking that readers pay $5 a month for access to all content. Compared to approaches such as Substack that charge a similar amount for each author you follow, this is a great bargain. With Medium, authors receive compensation depending on the popularity of their contributions. The $5 is divided based on what you read with a percentage going to Medium. I do publish some of my longer articles on Medium and I have yet to generate enough revenue to cover the cost of my membership. I don’t expect to generate a lot of income. I value the principle that those who invest time producing quality content should be compensated in some way. The article I link here is a great example of why. The research that went into this work must have been extensive and assumptions that this work does not deserve compensation is absurd.

I bring up this point based on a personal observation. I am an educational technologist so I write both about technology and about education. With Medium, I find that what I write about technology receives a good deal of attention while what I have to say about education less so. My true expertise would suggest the opposite reaction should be the case. My hypothesis is that this response reflects core values and expectations.

Online content and how it is delivered seems to be at a cross road of a sort. Free services are criticized because these services collect information about readers that allows targeted ads and manipulative propaganda. This is especially the case when 3rd party cookies are allowed, These criticisms are legitimate. The alternative is really subscription-based services.

I found a notice that the National Library Board was experimenting with AI as a way to increase interest in reading. An AI activity named Playbrary was their first venture. The application uses ChatGPT to turn public-domain classics into choose-your-own adventures. You don’t need the paid version of ChatGPT for the text version and if you have the paid version of ChatGPT the adventures will include some graphics. I explored the Sherlock Holmes adventure.

It was great fun and then I hit the end and was told to read the book and that would be fun too.

I started to wonder how ChatGPT did this and just for kicks tried the following prompt.

Prompt: I want to read a choose-your-own adventure story. Include an occasional image that supports the story. Wait after each choice for my response.

It turns out ChatGPT will create choose-your-own adventure games.

I don’t know if this is a good or a bad thing. I don’t know if it is a waste of the resources ChatGPT must bring to bear to do this. I just know it works.

You can request an adventure on pretty much any subject. I never got to the end of an adventure so I wonder if ChatGPT even knows where the story is going.

Prompt: I want to read a chose-your-own adventure story set in a middle school. Include an occasional image that supports the story. Wait after each choice for my response.

Jason Howell and Jeff Jarvis interview Sal Kahn about the Kahn Academy’s AI tool. The interview includes descriptions of how Kahnmigo is designed to interact with learners, the capabilities it provides educators, and some of the topics covered in Kahn’s future book about AI in education.

Here is an example describing data collection using probeware. We are particularly fond of this example because we conducted the data collection in our backyard.

Using a data logger to measure stress

An interesting thing sometimes happens when powerful tools are readily available. Authentic tasks, the tasks we described in the first chapter as mimicking the activities of those we would describe as practitioners, seem to present themselves. Sometimes the tools provide the means to answer some of the questions that occur to all of us. Here is a personal example.

We know a health and physical education teacher who engages her students in projects that require the collection and analysis of data. Some of her projects use a device that records heart rate data. The device, which consists of a strap worn around the chest and a data recorder worn like a wristwatch, allows complete freedom of activity. The data recorder holds more than eight hours of information and is easily connected to a computer for data transfer. She described a project that involved the parents of her middle-school students. She asks for three parent volunteers with three different occupations who are willing to wear the heart monitor during their workday. When they can remember, the volunteers are asked to push a button on the recorder when they change activities and to keep notes identifying the different activities. The data logger marks the time when these transitions occur. Students bring the heart monitor back to school, offload the data onto a computer, create a chart from the data, and enter labels on the chart corresponding to the activities that the parents report. Students then compare the heart rate patterns of each subject and discuss which activities seemed to create the greatest stress.

We borrowed the heart rate monitor and tried to generate our own authentic investigation. Here is what we came up with. We own a hot tub (called a spa in some locations) and, in reading the operating instructions, noted some health warnings. Individuals with certain medical conditions are cautioned against using the hot tub. Remembering the description of the use of the heart monitor as a measure of stress, we wondered if sitting in 104°F water could be demonstrated to stress the circulatory system. The design of our experiment was simple: Establish a 20-minute baseline, sit in the hot tub for 20 minutes, and conclude with a 20-minute cooling-off period.

The data generated clearly demonstrate the stress heat imposes on the body. The data are graphed in the image that appears below. Note the two marks on the x-axis. These marks indicate the transitions between stages of our experiment. The middle segment of the graph, the time during which Mark was in the hot water, demonstrates an elevated and accelerating heart rate. Note irregularities in the data appearing near the transitions between stages of the experiment. During these transitions, Mark was in his swimming suit working to remove and then reattach the hot tub cover with an air temperature of 20°F.

The computer can also store the raw heart rate data as a text file consisting of the numerical values recorded by the data logger. Once this has been done, the text file can be opened with a spreadsheet, and the data can be manipulated and analyzed in various ways. For example, we selected data values for the first and second segments that did not include the irregularities associated with removing the hot tub cover. The spreadsheet average function was then applied to these two sets of numbers. The average heart rate before entering the hot water was 74 beats per minute, and the average heart rate after entering the hot water was 89 beats per minute. It appears that our hypothesis concerning the stress experienced because of heat has been supported. Many related questions might follow: Why does the heart work harder when the body temperature is raised? Would you expect heart rate to be affected in the same way on a very warm day? Who should be cautioned against spending time in a hot tub?

Information about Data Loggers, Calculator-Based Laboratories (CBLs), Microcomputer-Based Laboratories (MBLs) and probes that attach to the CBLs and MBLs can be located on the Internet. Some of these sites also sell curriculum materials related to these products. Some sources follow.

HOBO data loggers are available from Onset Computer Corporation. (http://www.onsetcomp.com).

The Polar Heart Rate Monitor is available from Heartmind Heart Rate Monitors. Software associated with this product is available for both the Macintosh and Windows operating systems. (http://www.polarusa.com/).

Vernier sells sensors, kits of sensors suited to particular math and science courses, and lab manuals describing a variety of experiments that can be performed with the sensors. The Vernier site also identifies more companies that can use these probes and this information may be useful in extending the small list we can provide here (http://www.vernier.com).

I have been exploring the use of ChatGPT as a prewriting activity. I have been asking it to generate an outline on a topic which works great. I had read that ChatGPT could also create MindMaps. It can, but the entries are gibberish. I had read that ChatGPT has been having a problem with the spelling in graphics and this certainly seems to be the case.

I have written online content since 2002. My first blog started then. I don’t really remember my motivation for making the initial effort, but I suspect it had something to do with exploring what at the time was a new way of using technology and the opportunity to share personal ideas. Because my interests focused mainly on how K12 classroom teachers could use technology and I had a relatively popular textbook on that subject, I wrote to explore ideas in the same domain and to share these ideas with educators whether they used our textbook or not. I didn’t think about my online writing as a source of revenue. It was college profs who made the decision to assign the textbook. There were no ads on the original blog and I hosted the blog on a server I ran myself. 

As blogs became more popular, some of my ideas changed. I had an idea for drastically reducing textbook prices and keeping textbooks more current by shortening textbooks and supplementing the material in textbooks with free online content. I began to explore the idea by creating a separate online site linked directly to our existing textbook as an experiment to see what type of use the online material would generate. Because there might then appear to be a financial benefit, I decided it was time to purchase server access outside of the university so there would be no questions about the inappropriate use of university resources. I was never able to convince our publisher to implement my plan for a $29 textbook, but the idea was planted in my mind and I continued to spend time working on online resources. 

Alternate ways of supplying content to students had become a personal interest. I never got into the OER approach. Knowing the amount of time it took to research and write educational materials and being aware that universities really don’t reward such efforts in comparison to research publications and grant submissions, I included the issue of compensation in my explorations. Google ads were relatively easy to add to my content and I gave it a try. Just to put things into a financial perspective, my monthly payments for server space and the fees I pay to own the names for my accounts (Learningaloud and Curmudgeon Speaks) total approximately $200 a year. I could probably lower this amount if I did not want as much control as I have over the types of things I can do on the server, but $200 a year is a great investment for a hobby. As far as ad revenue goes, my total take since starting the ad-enabled versions of my blogs (maybe since 2010) has been less than what I pay for one year of server use. What people may not know about ad revenue is that it is not the appearance of ads that generates revenue, but clicks on these ads. Thinking about how the ad revenue works I admit to being puzzled. I can’t remember ever having purposefully clicked on an ad. I must display a hundred ads a day just being online a lot, but I don’t produce clicks. Others must operate differentially for anyone to make the effort to write for ad revenue.

What I think is often not understood is that the textbook company and authors make zero money after the first sale. I remember our income as 12% of the price to bookstores (the wholesale price). The revision schedule for the book was three years (six semesters). We made most of the total income from the book in the second and third semesters. After that, the used book market took over. The “Technology for Teachers” course has a far smaller potential audience than courses required for large numbers of students (Psychology, Math, Chemistry) and the price for such books is significantly lower, but the competition is also less and there is a demand for frequent revisions because of changes in the field. We were able to generate five editions so we did pretty well. 

The textbook business is tough. There are many complaints about the cost of textbooks and there is no doubt the cost of new books is high. However, there are signifiant costs and risks to the companies and the investments they must make. I have some sympathy for the companies knowing the pressure they are under and the number of companies that have given up and sold their holdings so that the number of actual publishers has declined drastically. Our book was part of the holdings of three different companies in fifteen years. 

With all of this as background, I remain interested in where the support for educational content is headed. I am still exploring in my retirement and I have the time and body of content I can use to test different compensation models. My more recent explorations have included Kindle, Substack, and Medium. Kindle is an obvious alternative to commercial publishing companies. Authors self-publish at a much lower price (say $10-15 instead of $100). Authors are pretty much on their own in generating interest in their books. Substack and Medium are probably not best understood as book sources. Both provide an audience that authors try to attract. The typical approach is to offer some content for free in hopes of attracting readers who will then pay a subscription fee. Authors typically continue with some free content and some paywalled in hopes of finding a balance that will grow their subscribers. Substack and Medium use different subscription models. With Substack, a reader subscribes to the work of individual authors paying maybe $50 or so to each author. With Medium, readers subscribe to the service ($60 a year) and then read all they want from any author. Medium then has a system to divide up the subscription fees based on how much attention is devoted to the content of a given author. It seems to me that Medium offers the best value. I do subscribe to some Substack content, but I keep thinking I could purchase three Kindle books for each Substack subscription and I would get more out of these three books. 

I am now trying to repurpose one of my textbooks for Medium. It is hard to say how this will go. The service is really not designed for books although I have published a book as a series of chapters. It also does not seem that K12 educators are not that well represented among Medium readers as the content tagged K12 seems limited. I like the idea of micropayments so I am giving Medium a try just to see what level of interest I can generate. 

I welcome your interest. You can view the free Medium content without paying a subscription and I have some content available that is not behind a paywall. 

The type of writing I do requires that I identify topics of interest and then review the research relevant to that topic to combine with my knowledge to write content relevant to educational issues. My career has allowed me to witness the growth in tools for accomplishing this task. In the old days, the task was heavily based on using the reference sections of key publications to identify important older publications that should be reviewed. It was far easier to look back than forward. This changed with online services such as Research Rabbit, Google Scholar and similar services. These services provide the reference list from a given article, but also more recent articles that cite the targeted article.

Litmaps provides an interesting way to examine this same data. You begin the process with a seed map in which you identify an influential article and the service returns a visual display identifying cited articles and articles citing the seed. From there, it is possible to review the abstracts of articles from the map and explore connections among articles. The system saves searches so you can return to the information you have identified without having to find a way to export the information as you work with a topic over time.

The following is the display you first encounter with the option to propose a search seed.

One of the topics that interests me is the pro and con debates over whether reading and note taking should be conducted using a digital device or paper. One of the articles often cited when proposing that note-taking is best done with pen and paper was published in 2018. The title “The pen is mightier than the keyboard” conveys the orientation of the paper and sounds interesting when referencing this research area in the popular press. It is the type of source that makes a good starting point for Litmaps. In the following image, the seed article appears as a blue dot and links indicate earlier and later connections with this study,

Selecting one of the linked studies provides an update to the display. You see the abstract and other connections to the linked study.

Litmap is free for limited use and there is a lower rate for educational use. The pricing options are shown below.

The Associated Press has put together a great resource on conspiracy theories. The resource covers examples and analysis for researchers. This would seem a great starting point for educators. The resources strives to be politically neutral and includes a historical perspective.

Perhaps you saw parts of the grilling of some tech executives as part of committee hearings related to Big Tech and the Online Child Sexual Exploitation Act. The committee had arranged to have parents of children who had committed suicide in the audience and kept asking the tech executives to apologize and address these parents. It was difficult and I suppose made for great political television. I admit to having some sympathy for the executives put in this situation.

My reaction to these issues is to ask myself what I think would be necessary to make the situation better. I see the present situation as the equivalent of the politicians yelling at the big tech people claiming this is your fault so fix it. I keep thinking if in their place I would respond by asking just what is it you would propose we should change.

With this situation and the anti Section 230 politicians, the tech companies that essentially provide an opportunity for bad actors to act are the target. The problem with tech more generally is the same tools for accomplishing great things can be used by bad actors to do terrible things. There is no simple way to go after the bad actors so politicians go after the tech companies. It seems to me one solution would be to require the users of online technologies to establish their identities. I understand this would be challenging, but we seem to have figured it out when it comes to driver’s licenses and passports. Without anonymity, it would seem possible to go after bad actors. I understand there would be an issue here with equity and some users have very legitimate reasons for being anonymous, but this is a choice that could be made.

My point is that this is concrete and politicians could make this a requirement. Why not? Again, based on this example, Section 230 argues against such requirements because such expectations could probably be addressed by big tech, but would not be possible to implement by any of the rest of us with an online presence (even a blog) that could allow public comment. Competition with big tech would be eliminated.

I do think it essential to address big tech, but I have a much wider set of concerns. The limited options are the type of issue I think politicians should address as a root issue. Capitalism assumes things improve when individuals can move to an alternative providing the circumstances they desire. There are few options and the lack of interoperability between services limits choice.

If you are interested in this issue, I would recommend a recent post from danah boyd (I forget why she does not use capital letters). boyd has researched and written about adolescents and technology for years. Her analysis takes a broader view and while recognizing a role for online technology proposes that adolescents and the rest of us live in a messed up and toxic culture such that addressing tech alone will result in little change.