Update to AI-Supported Social Annotation with Glasp

Social annotation is a digital and collaborative practice in which multiple users interact with text or video through comments, highlights, and discussions directly linked to specific parts of the source. This practice extends the traditional act of reading and watching into a participatory activity, allowing individuals to engage with both the text and each other in educational ways.

For learners functioning within a formal educational setting or an informal setting, social annotation can benefit learners in multiple ways. It can transform reading from a solitary to a communal act, encouraging students to engage more deeply with texts. Students can pose questions, share interpretations, and challenge each other’s views directly on the digital document. This interaction not only enhances comprehension and critical thinking but also builds a sense of community among learners. Potentially, educators can also participate guiding discussions or reacting to student comments.

Beyond the classroom, social annotation is used in research and professional fields to streamline collaborations. Researchers and professionals use annotation tools to review literature, draft reports, and provide feedback. This collaborative approach can accelerate project timelines and improve the quality of work by incorporating multiple expertises and viewpoints efficiently.

I have written previously about social annotation as a subcategory of my interest in technology tools that allow layering and even earlier in the description of specific annotation tools such as Hypothesis. As now seems the case with many digital topics, social annotation eventually was expanded to incorporate AI. This post updates my description of the capabilities of the AI capabilities of Glasp. Glasp is a free tool used to annotate web pages, link comments to videos, and import annotations from Kindle books. It functions as a browser extension when layering comments and highlights on web pages and videos. The accumulated body of additions is available through a website which is where the AI capability is applied as a mechanism for interacting with the collected content and for connecting with other Glasp users. 

The following content is divided into two sections. The first section focuses on the AI capabilities applied to personally collected content and the content collected by others. The second section explains how to locate the content of others who have used Glasp to collect content designated as public. This second section describes capabilities I have personally found very useful. As a retired individual, I no longer have access to colleagues I might interact with frequently. Collaborative tools are only useful when collaborators are available and developing connections can be a challenge.

Interacting with stored annotations using AI

The following image displays the personal browser view from the Glasp site. The middle column consists of thumbnails representing multiple web pages that have been annotated and the right-hand column the highlighted material (no notes were added to the source I used for this example) from the selected source. The red box was added to this image to bring your attention to the “Ask digital clone” button. This image is what you would see when connecting to my site to interact with my content. The button would read “Ask your clone” if I was connecting to my own account to interact with my content. Here is a link you can use to interact with my content. After you have read just a bit further, return and use this link to duplicate my example and then try a few requests of your own. 

The next image displays what happens when the “Ask digital clone” button is selected. You should see a familiar AI interface with a text box at the bottom (red box) for initiating an interaction. I know the type of content I have read so I have generated a prompt I know should be relevant to the content I have annotated. 

The prompt will generate a response if relevant information is available. However, here is what I find most useful. The response will be associated with a way to identify sources (see red box). Typically, I am most interested in reviewing original material from which I can then write something myself.

The link to relevant highlights should produce something that looks like the following.

Locating content saved by others

Glasp offers a capability that addresses the issue I identified earlier. How do you locate others to follow?

The drop-down menu under your image in the upper right-hand corner of the browser display should contain an option “Find like-minded people”. This option will attempt to identify others with some overlap in interests based on the type of content you have annotated. So, you must start by building at least a preliminary collection of annotated sites yourself. If you have no content, there is nothing available to use as the basis for a match.

Glasp should then generate something like the following. You can click on someone from this display to query their existing public material. If you then want to follow that individual, their site should contain a “Follow” button.

Summary

I hope this is enough to get you started. You can use the link to my account to explore. It seems unlikely to me that Glasp will always be free. They must have development and infrastructure costs. For now, the company has offered an interesting approach that has grown in capability during the time I have used it.

Loading

Tutoring and the Big Picture

The period since the end of the COVID lockdown has been difficult for K-12 schools. Students have struggled to attain traditional rates of academic progress. Student absenteeism has reached alarming levels. Where we live some districts have struggled to find teachers to fill vacant positions and other districts are now cutting staff because of massive budget overruns. Educators feel a lack of support from politicians and groups of parents complaining about various issues and sometimes pulling their children out of one school to enroll them in another. 

I certainly have no answer to this collection of challenges as it seems a proposed solution to one problem only leads to complaints due to consequences from that recommendation for a different issue. I look to educational research for potential solutions and there are certainly strategies with proven positive consequences in specific areas, but this is as far as it seems practical to go.

Tutoring has been one of the recent areas I have been investigating. My specific interest has been in the potential of AI as a substitute for human tutors, but the ample research on the effectiveness of human tutors has also been a fruitful ground for considering what proven experiences and roles AI might be applied to approximate. I have continued to review what we know about the effectiveness of human tutors and what it is tutors provide that students do not receive from their traditional classroom experiences.

For those wanting to review this research themselves, I recommend a fairly recent working paper by Nickow and colleagues (2020). This paper is both a general descriptive review of how tutoring is practiced and a meta-analysis of a large number of quality studies (96) of the effectiveness of this addition to K12 instruction. The meta-analysis considers only the impact of adult tutors, but the descriptive review does include some comments on both peer tutoring and technology-based tutoring. The technology comments were written before recent advances in AI. My summary of this paper follows.

Key Findings

I will begin with the overall conclusion. While there are more and less effective implementations and situations, adult tutoring consistently results in positive outcomes. The authors concluded that tutoring was responsible for substantial and consistent positive effects on learning outcomes, with an overall effect size of 0.37 standard deviations. The results are the strongest for teacher-led and paraprofessional tutoring programs, with diminished but still positive effects for nonprofessional and parent-led tutoring. Notably, tutoring is most effective in the earlier grades and shows comparable impacts across both reading and mathematics interventions, although reading tends to benefit more in earlier grades and math in later ones. A large proportion of the studies and tutoring activity appears focused in the first couple of years of school and in the subject area of reading. There are also far more studies of tutoring in the early grades than in secondary school. Only 7% of the studies included students above 6th grade.

Categories of tutors

The study by Nickow, Oreopoulos, and Quan (2020) identifies four categories of tutors:

  1. Teachers: These are certified professionals who have undergone formal training in education. They are typically employed by schools and likely have a deeper understanding of the curriculum, teaching strategies, and student assessment.
  2. Paraprofessionals: These individuals are not certified teachers, but they often work in educational settings. Paraprofessionals may include teaching assistants, educational aides, or other support staff in schools. They often work under the supervision of certified teachers.
  3. Nonprofessionals: This category typically includes volunteers who may not have formal training in education. They could be community members, older students, or others who offer tutoring services. Their level of expertise can vary widely.
  4. Parents: Parents often play a tutoring role, especially in the early years of a child’s education. Parental involvement in a child’s education can have a significant impact, but the level of expertise and the ability to provide effective tutoring can vary greatly among parents.

The study found that paraprofessionals are the largest provider of tutoring services, followed by nonprofessionals. The effectiveness of tutoring was found to be higher when provided by teachers and paraprofessionals compared to nonprofessionals and parents. Tutors who have teaching certificates and classroom experience were rare, but most effective in the research studies. As mentioned in the introduction to this post, the question of how schools must apply their budgets appears highly influential in whether known sources of impact can be applied.

Mechanisms of Tutoring Effectiveness

This paper reviewed mechanisms potentially responsible for the effectiveness of tutoring. While these hypothesized mechanisms were not evaluated in the meta-analysis, I find such lists useful in considering other approaches such as AI. Which of these mechanisms could AI tutoring address and which would not be possible?

The paper discusses several mechanisms through which tutoring enhances student learning:

  1. Increased Instruction Time: Tutoring provides additional instructional time, crucial for students who are behind.
  2. Customization of Learning: The one-on-one or small group settings of tutoring allow for tailored instruction that addresses the individual needs of students, a significant advantage over standard classroom settings where diverse learning needs can be challenging to meet. The authors propose that tutoring might be thought of as an extreme manipulation of class size. I know that studies of class size have produced inconsistent results extremely small class sizes are not investigated in such research.
  3. Enhanced Engagement and Feedback: Tutoring facilitates a more interactive learning environment where students receive immediate feedback, enhancing learning efficiency and correction of misunderstandings in real time.
  4. Relationship Building: The mentorship that develops between tutors and students can motivate students, fostering a positive attitude toward learning and educational engagement.

Program Characteristics

The document evaluates the variation in the effectiveness of tutoring programs based on several characteristics:

  1. Tutor Type: Teacher and paraprofessional tutors are found to be more effective than nonprofessionals and parents, suggesting that the training and expertise of the tutor play a critical role in the success of tutoring interventions.
  2. Intervention Context: Tutoring programs integrated within the school day are generally more effective than those conducted after school, possibly due to higher levels of student engagement and less fatigue.
  3. Grade Level: There is a trend of diminishing returns as students age, with the most substantial effects seen in the earliest grades. This finding emphasizes the importance of early intervention.

Summary

This research and analysis paper might be summarized by proposing that schools cannot go wrong with tutoring. The paper suggests several policy implications based on its findings. Given the strong evidence supporting the effectiveness of tutoring, especially when conducted by trained professionals, the authors advocate for increased investment in such programs. They suggest that education policymakers and practitioners should consider tutoring as a key strategy for improving student outcomes, particularly in early grades and for subjects like reading and mathematics

Reference

Nickow, A., Oreopoulos, P., & Quan, V. (2020). The impressive effects of tutoring on prek-12 learning: A systematic review and meta-analysis of the experimental evidence. https://www.nber.org/papers/w27476.

Loading

GPS and GIS: Interpreting Data in Relationship to Place

I have long been interested in cell phone photography as a way students could learn about GPS and GIS. The capabilities of our phones in this regard are probably now taken for granted and consequently ignored as a learning opportunity. If anything, the capability of connecting the phone to a location marked in captured images may be considered a security risk rather than a capability to be applied when useful.

Global Positioning Systems (GPS) and Geographic Information Systems (GIS) allow the investigator to search for interpretations of data related to place. GPS uses the signals from multiple satellites to allow an individual with a GPS device (hardware) to determine the location of the GPS device (place) in terms of precise latitude, longitude, and altitude. Put another way, the device allows you to determine exactly where you are standing on the earth. Typically, your position can be determined with greater than 10-foot accuracy. You may be familiar with GPS navigation because you have GPS hardware installed in your car or perhaps your phone. These devices know where you are and can establish a route between where you are and where you would like to go. The most basic function of a GPS is to locate the device in three-dimensional space (longitude, latitude, and altitude), but most map location (show location on a map), navigate, and store data (e.g., the coordinates of designated locations). Smartphones do so using true GPS (using the signals from satellites) but may also determine the location of the phone by calculating the phone’s location to multiple cell phone towers. The triangulation process with the cell towers is similar to that dependent on satellites but less accurate.

GIS is a software tool that allows the user to see the relationship between “layers” of information. In most cases, one layer is a map. Other layers could be the presence or quantity of an amazingly diverse set of things— e.g., voters with different political preferences, cases of a specific disease, fast food stores, a growth of leafy spurge, or nitrate concentration in water. The general idea is to expose patterns in the data that allow the researcher to speculate about possible explanations. 

With the addition of specialized software, some GPS devices and smartphones provide similar capabilities. The device knows your location and can identify restaurants, gas stations, or entertainment options nearby. The field of location-based information is expanding at a rapid pace. One approach involves providing useful, location-specific information. For example, how close am I to the nearest gas station? A second approach allows the user to offer his or her location as information. Being able to locate a phone can be useful if the phone is lost and some applications allow parents to locate their children using the location of the phone the child is carrying. Sometimes, individuals want to make their present location available in case others on a designated list of “friends” may be in the vicinity providing the opportunity for a face-to-face meeting. Obviously, there can be significant privacy concerns related to sharing your location.

A great example of student use of GPS, GIS, and the Internet is the GLOBE program (http://www.globe.gov/). GLOBE is an international program led by a collaborative group of U.S. federal agencies (NOAA, NSF, NASA, EPA). Over 140 colleges and 10,000 schools from over 90 countries are also involved. GLOBE involves students in authentic projects led by scientists in the areas of air quality, land cover, water quality, soil characteristics, and atmospheric sciences. 

In the GLOBE program, students in classes taught by specially trained teachers work with scientists to collect data according to precisely defined protocols. The advantage to the scientists is the massive and distributed data collection system made available by the Internet. Data gathered from precise locations (identified with GPS) can be integrated (with GIS) on an international scale. Students have access to educational materials and learn by contributing to authentic projects.

The GLOBE projects are presented in ways that have local relevance and have been matched to K–12 standards. While the topics of study most closely address standards in the areas of math and science, the international scope of the project also involves students with world geography, diverse cultures, and several languages (the project home page is available in seven languages). The data are also available online, and groups of educators are encouraged to propose and pursue related projects.

Readily available software and hardware also allow educators to design projects that are not dependent on formal, large-scale programs. We all have become much more familiar with GPS devices and many of us own navigation or phone devices that could be used in educational projects. Digital cameras tag images with GPS coordinates. Once we have a way of determining location, we might then consider what data we can match to location. Fancy equipment is not always necessary. Sometimes the data are what we see. Do you know what a Dutch Elm tree looks like? Have any Dutch Elms in your community survived? Where are they located? There are also many easy ways to use location to attach data, in this case photos, to maps. For example, Google Photos offers some amazing capabilities. If you store cell phone pictures in Google Photos, try searching for a location (e.g., Chicago). Google Photos knows where some things are located (e.g., the Bean), but will also return photos based on the embedded EXIF data that includes GPS information.  

Probes you may already own, your phone and data collection

Your cell phone has an interesting feature. It can store the exact location from which each picture was taken and other information with the same file containing the image. These data are stored as part of EXIF (exchangeable image file format). You may know that some images are accompanied by information such as the camera used to take the picture, aperture, shutter speed, etc. This is EXIF data. The longitude and latitude (I can never remember which is which) can also be stored as EXIF data.

I have a record of my first experience exploring these capabilities. I was using a camera with GPS capabilities rather than a phone, but it is the personal insights about capabilities that resulted that are relevant here. In was 2009 and my wife and I were in Washington, DC, for a conference. We spent some time visiting the local landmarks and I took the following picture. As you can see, we were standing at the fence that surrounds the White House and I took the following photo. I think tourists would no longer be allowed in this location, but that was a different time. 

I used the EXIF data to add the photo to Google Maps. In the following image, you can see the image and the mapped location in street view. At first the map information confused me – no white house. Then I realized, we are standing on Pennsylvania Ave. on the other side of the fence shooting through the trees to frame the picture. We were the pin looking through the fence, over the flower bed, toward the White House. I have often said I have a different understanding of technology because I have always been a heavy tech user and experienced life as technological capabilities were added. I was there before and after and thus have a sense of how things changed. When technology capabilities are already there you often learn to use them without a need to understand what is happening and without a sense of amazement that can motivate you to seek understanding and creative applications. 

Mapping photo collections with Google services

The following is a tutorial. The specific example that is the basis for the tutorial is not intended to be relevant to classroom use, but the example is authentic and the processes described should transfer. Now retired, we were wintering in Kauai when I decided to write this post. I do a lot of reading and writing in coffee shops and I had decided to begin collecting photos of the various shops I frequented. Others would probably focus on tourist attractions, but coffee shops were the feature that attracted me.

Mapping photos in Google is best understood as using two interrelated Google services – Google Photos and Google MyMaps. If you are cost conscience and are not interested in advanced features or storing a lot of images, you can get away with the free levels of Google tools. The two-stage process involves first storing and isolating the images you want to map (Google Photos) and then importing this collection to be layered on a Google map (MyMaps).

Creating an album in Google Photos

The first step involves the creation of an album to isolate a subset of photos. In the following image, you should find a column of icons on the left-hand border. The icon within the red box is used to create an album.

This icon should then display the Create album button at the top of the display.

Name the new album

Now, return to photos and for each photo you want to map, use the drop down menu to add that photo to the appropriate album.

Continue until you have identified all of the images you want to map.

MyMaps

Google MyMaps (https://www.google.com/maps/d/) provides a means to create multiple personal maps by layering content (e.g., images) on top of the basic Google map. Using the link provided here, open your Google account and identify what you want to label your new personal map. 

If you are adding images with GPS data, the process will automatically locate the images you provide to the appropriate location. It makes sense to me, to begin by moving the location that I am interested in to the screen. In this case, I am adding images to the island of Kauai.

The following image is a panel you will see in several of the images that follow. The first use of this panel is to enter a name for the map I am creating. 

The text box to enter the name is revealed by selecting the original value (Untitled map) and this will open a box to add the name you intend.

The next step is add the layer that will contain the photos on top of this map. 

The approach I am taken is to add all of the images once and this is accomplished by referencing the Google Photos album that already exists. I select “coffee shops” from my albums.

MyMaps does not assume I intend to use all of the images in the designated album so I must now select the images I want to add to the map. When finished, select “Insert”. 

This generates the finished product.

MyMaps allows you a way to share your map with others. Try this link to see the map I have just created. Selecting one of the thumbnails appearing on the map should open a larger view. Give it a try. Without the password, your access should be “read only”.

Summary

Google Photos and Google MyMaps allow students to explore GPS and GIS. Images taken with smartphones can be added to a Google map allowing authentic GIS projects. Care should be taken to understand how to turn GPS locations associated with photos on and off. 

Loading

Improving peer editing

Clearly, the teacher is likely to be the most important source of guidance in developing the processes necessary for effective writing. However, peers are also an important resource. The generation of individual feedback is time-consuming and providing such feedback to a class of students multiple times would be extremely demanding. Our comments here reflect the proposal of writing experts who argue that peers as a group can likely respond more quickly than the teacher who would be working alone and the comments of peers can augment the feedback provided by the teacher. There is one more argument for involving peers (Bruning, Schraw & Norby, 2011, p. 300). Better writers at all ages appear to be better editors with poor writers often seeing little of substance that can be changed in their original first drafts. The learning of editing skills to provide a useful service to peers develops the same skills that can be applied to the student’s own work.

Peer editing has gained increased attention among researchers with the research offering greater insight as more specific issues are investigated. For example, Wu and Shunn (2021) note that most previous research has focused on the value of peer feedback in improving the document for which feedback was provided. This is a different issue than whether giving and receiving feedback results in improved performance on future writing tasks. In their research, which involved secondary students enrolled in an AP composition course, across multiple writing tasks, the researchers investigated both the impact of peer editing on the present and a future writing task. The study supported the positive impact of giving and receiving feedback on both document quality and future performance. 

The proposed benefits of peer editing include (Pritchard & Honeycutt, 2007);

1 A non-threatening audience,

2 Increased opportunities to write and revise, and

3 Immediate feedback – to write and revise a lot, the teacher cannot do it all.

Note the use of the qualifier “proposed”. You might argue that some students can be quite insensitive in interacting with peers and so wonder about proposing peers offer a “non-threatening audience”. Proposed here implies that skills can be developed and offer advantages when a reasonable level of competence is present.

One should not assume that effective peer editing is a simple manner of having students exchange papers and offer the comments that come to mind. Without guidance and experience, student comments may be socially awkward and focused on the most shallow of writing skills (e.g., spelling errors). 

Peers, by definition, are at approximately the same level of proficiency as those they are attempting to assist. In addition, they often lack the social skills and sensitivity to have their suggestions interpreted as helpful rather than mean. However, given some preparation, spending time responding to the written products of peers can be helpful to the peer writer and a way to develop the writing skills of the writer. (Pritchard & Honeycutt, 2007)

Here is a brief summary of a series of activities proposed by Simmons (2003) as a process for developing the skills of peer editors. 

1) Teacher models editing. If possible, offer a document you (teacher) have written. Think aloud about what improvements might be made. Make the revisions and then compare the original and the revised documents.

2) Model how feedback should be communicated. Model praise. Model questioning – what was the author trying to do? Model how to offer suggestions.

3) Use peer pairs to initiate peer feedback experience.

4) Use examples from student work and student feedback with the class.

Those who have studied the development of peer editing skills want it to be understood that this is far from a one or two-lesson process. Often early efforts are a struggle. Student editors develop skills gradually and typically begin with superficial recommendations (spelling, grammar), unmerited praise (not to be confused with encouragement), or insensitive criticism. Regarding teacher expectations, it makes sense that the priorities of review applied to the work of others would be similar to changes, if any, developing writers would make with experience in their own work. Attention paid to the metacognitive processes of considering the audience and communication effectiveness of a document as a whole is more abstract than the recognition of grammatical rule violations. Hence, purposeful demonstration, discussion, and practice are important in developing editing skills whether applied to a document developed by the student or a document developed by a peer.

 Peer comments should include and begin with positive comments. What did you like? The targeted writing skills will change as the goals of writing change either with experience or purpose.

A computer and classroom whiteboard or projector combination is a great way for the teacher to model and provide examples. Writing tools that save comments and recommendations and writing tools that allow a comparison of drafts offer the teacher an external representation of author or peer editor thinking and provide the teacher something tangible to address. What challenges were recognized and what changes were actually implemented? We provide some examples of such capabilities in our online resources.

One interesting model for upper-elementary developed by Sarah Dennis-Shaw appears on the ReadWriteThink site. This model suggests that students offer peers comments, suggestions, and corrections.

Compliments

e.g., My favorite part was ___ because ___

Suggestions

e.g., If you gave more details, I would be certain I can understand what you mean.

Corrections

e.g., I found this misspelled word – mispell

It is worth the effort to review Dennis-Shaw lessons no matter what grade level you work at as the online resources are quite specific in outlining the steps in the instructional process and also provide sample instructional materials. For example, what might a writing sample used in the training phase look like? We also recommend that you do an Internet search for rubrics or checklists that might be suited to your own instructional circumstances (e.g., Simon Williams rubric)

References

Bruning, R.H., Schraw, G.J., Norby, M.M. (2011). Cognitive psychology and instruction (5th ed). Boston: Pearson.

Pritchard, R. J., & Honeycutt, R. L. (2007). Best practices in implementing a process approach to teaching writing. Best practices in writing instruction, 28-49.

Simmons, J. (2003). Responders are taught, not born. Journal of Adolescent and Adult Literacy, 46(8), 684-693.

Wu, Y., & Schunn, C. D. (2021). The Effects of Providing and Receiving Peer Feedback on Writing Performance and Learning of Secondary School Students. American Educational Research Journal, 58(3), 492-526.

Loading

Improve AI Generated Multiple-Choice Questions

In a recent episode of the EdTech Situation Room, host Jason Neiffer made a very brief observation that educators could improve the effectiveness of AI-generated multiple-choice questions by adding a list of rules the AI tool should apply when writing the questions. This made sense to me. I understood the issue that probably led to this recommendation. I have written multiple times that students and educators can use AI services to generate questions of all types. In my own experience doing this, I found too many of the questions used structures I did not like and I found myself continually requesting rewrites excluding a type of question I found annoying. For example, questions that involved a response such as “all of the above” or a question stem asking for a response that was “not correct”. Taking a preemptive approach made some sense and set me on the exploration of how this idea might be implemented. Neiffer proposed an approach that involved making use of an online source for how to write quality questions. I found it more effective to maybe review such sources, but to put together my own list of explicit rules.

My approach used a prompt that looked something like this:

Write 10 multiple-choice questions based on the following content. The correct answer should appear in parentheses following each question. Apply the following rules when generating these questions. 

There should be 4 answer options.

“All of the Above” should not be an answer option.

“None of the Above” should not be an answer option.

All answer options should be plausible.

Order of answer options should be logical or random.

Question should not ask which answer is not correct. 

Answer options should not be longer than the question.

I would alter the first couple of sentences of this prompt if I was asking the AI service to use its own information base or I wanted to include a content source that should be the focus of the questions. If I was asking for questions generated based on the large language content alone, I would include a comment about the level of the students who would be answering the questions (e.g., high school students). For example, questions about mitosis and meiosis without this addition would include concepts I did not think most high school sophomores would have covered. When providing the AI service the content to be covered, I did not use this addition.

Questions based on a chapter

I have been evaluating the potential of an AI service to function as a tutor by interacting with a chapter of content. My wife and I have written a college textbook so I have authentic content to work with. The chapter is close to 10,000 words in length. In this case, I loaded this content and the prompt into ChatPDF, NotebookLM and ChatGPT. I pay $20 a month for ChatGPT and the free versions of the other two services. All proved to be effective.

ChatPDF

NotebookLM

With NotebookLM, you are allowed to upload multiple files that a prompt uses as a focus for the chat. For some reason rather than including my entire prompt, I had better results (suggested by the service) when I included the rules I wanted the system to apply as a second source rather than as part of the prompt.

ChatGPT

The process works a little differently with ChatGPT. I first copied the text from the pdf and pasted this content into the prompt window. I then scrolled to the beginning of this content and added my prompt. I could then ask the service to produce multiple question samples by asking for another 10 or 20 questions. I found some interesting outcomes when asking for multiple samples of questions. Even the format of the output sometimes changed (see the position of the answer in the following two examples).

**4. According to Clinton (2019), what is a potential impact of reading from a screen on metacognition?**

(A) Increased understanding

(B) Enhanced critical thinking

(C) Overconfidence and less effort

(D) Improved retention

(**C**)

**7. Which skill is considered a “higher order thinking skill”?**

(A) Word identification

(B) Critical thinking (**Correct**)

(C) Fact memorization

(D) Basic calculation

From sample to sample, some of the rules I asked ChatGPT to use were ignored. This slippage seemed unlikely in the initial response to the prompt.

What is an important consideration when designing project-based learning activities?**

(A) The amount of time available to students

(B) The availability of resources

(C) The level of student autonomy

(D) All of the above

(**D**)

Summary

The quality of multiple-choice questions generated using AI tools can be improved by adding rules for the AI service to follow as part of the prompt to generate questions. I would recommend that educators wanting to use the approach I describe here generate their own list of rules depending on their preferences. The questions used on an examination should always be selected for appropriateness, but the AI-based approach is a great way to easily generate a large number of questions to serve as a pool from which an examination can be assembled. Multiple choice exams should include a range of question types and it may be more efficient to write application questions because an educator would be in the best position to understand the background of students and determine what extension beyond the content in the source material would be appropriate.

Loading

AI Tutoring Update

This is an erratum in case my previous posts have misled anyone. I looked up the word erratum just to make certain I was using the word in the correct way. I have written several posts about AI tutoring and in these posts, I made reference to the effectiveness of human tutoring. I tend to provide citations when research articles are the basis for what I say and I know I have cited several sources for comments I made about the potential of AI tutors. I have not claimed that AI tutoring is the equal of human tutoring, but suggested that it was better than no tutoring at all, and in so doing I have claimed that human tutoring was of great value, but just too expensive for wide application. My concern is that I have proposed that the effectiveness of human tutoring was greater than it has been actually shown to be.

The reason I am bothering to write this post is that I have recently read several posts proposing that the public (i.e., pretty much anyone who does not follow the ongoing research on tutoring) has an inflated understanding of the impact of human tutoring (Education Next, Hippel). These authors propose that too many remember Bloom’s premise of a two-sigma challenge and fail to adjust Bloom’s proposal that tutoring has this high level of impact on student learning to what the empirical studies actually demonstrate. Of greater concern according to these writers is that nonreseachers including educational practitioners, but also those donating heavily to new efforts in education continue to proclaim tutoring has this potential. Included in this collection of wealthy investors and influencers would be folks like Sal Kahn and Bill Gates. I assume they might also include me in this group while I obviously have little impact in compared to those with big names. To be clear, the interest of Kahn, Gates, and me is really in AI rather than human tutoring, but we have made reference to Bloom’s optimistic comments. We have not claimed that AI tutoring was as good as human tutors, but by referencing Bloom’s claims we may have led to false expectations. 

When I encountered these concerns, I turned to my own notes from the research studies I had read to determine if I was aware that Bloom’s claims were likely overly optimistic. It turns out that I had read clear indications identifying what the recent posters were concerned about. For example, I highlighted the following in a review by Kulik and Fletcher (2016). 

“Bloom’s two sigma claim is that adding undergraduate tutors to a mastery program can raise test scores an additional 0.8 standard deviations, yielding a total improvement of 2.0 standard deviations.”

My exposure to Bloom’s comments on tutoring originally had nothing to do with technology or AI tutoring. I was interested in mastery learning as a way to adjust for differences in the rate of student learning. The connection with tutoring at the time Bloom offered his two-sigma challenge was that mastery methods offered a way to approach the benefits of the one-to-one attention and personalization provided by a human tutor. Some of my comments on mastery instruction and the potential of technology for making such tactics practical are among my earlier posts to this site. Part of Bloom’s claim being misapplied is based on his combination of personalized instruction via mastery tactics with tutoring. He was also focused on college-aged students in the data he cited. My perspective reading the original paper many years ago was not “see how great tutoring is”. It was more tutoring on top of classroom instruction is about is good as it is going to get and mastery learning offers a practical tactic that is a reasonable alternative.

As a rejoinder to update what I may have claimed, here are some additional findings from the Kulik and Fletcher meta-analysis (intelligent software tutoring).

The studies reviewed by these authors show lower benefits for tutoring when outcomes are measured on standardized rather than local tests, sample size is large, participants are at lower grade levels, the subject taught is math, a multiple-choice test is used to measure outcomes, and Cognitive Tutor is the ITS used in the evaluation.

However, on a more optimistic note, the meta-analysis conducted by these scholars found that in 50 evaluations intelligent tutoring systems led to an improvement in test scores of 0.66 standard deviations over conventional levels. 

The two sources urging a less optimistic perspective point to a National Board of Educators Research study (Nickow and Colleagues, 2020) indicating that human tutoring for K-12 learners was approximately .35 sigma. This is valuable, but not close to the 2.0 level.

Summary

I have offered this update to clarify what might be interpreted based on my previous posts, but also to provide some other citations for those who now feel the need to read more original literature. I have no idea whether Kahn, Gates, etc. have read the research that would likely indicate their interest in AI tutoring and mastery learning was overly ambitious. Just to be clear I had originally interpreted the interest of what the tech-types were promoting as mastery learning (personalization) which was later morphed into a combination with AI tutoring. This combination was what Bloom was actually evaluating. The impact of a two-sigma claim when translated into what such an improvement would actually mean in terms of rate of learning or change in a metric such as assigned grade seems improbable. Two standard deviations would move an average student (50 percentile) to the 98th percentile. This only happens in Garrison Keeler’s Lake Wobegon. 

References:

Bloom, B. S. (1984). The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring. Educational researcher, 13(6), 4-16.

Kulik, J. A., & Fletcher, J. D. (2016). Effectiveness of intelligent tutoring systems: a meta-analytic review. Review of educational research, 86(1), 42-78.

Nickow, A., Oreopoulos, P., & Quan, V. (2020). The impressive effects of tutoring on prek-12 learning: A systematic review and meta-analysis of the experimental evidence. https://www.nber.org/papers/w27476

Loading