Data literacy – using data

As I have explored the topic for which I have yet to find a suitable descriptive phrase, I have encountered similar existing initiatives. Claiming originality should always be done with great care. One source I intend to use for several posts seems to have been the focus of school libraries who describe this topic as data literacy and see an important role for librarians relevant to this topic in guiding and supporting student activities in developing this literacy. Without a formal claim of responsibility, I am left to speculate regarding how the connection with the role of librarians explains their interest. My guess is that librarians see data literacy as involving answering certain questions based on a type of information resource (data sources) and librarians exist to help connect users with information resources. I will explore this position and provide some references as my series of learning to generate and use data continues.

One category of public data your librarian might acquaint you with would consist of datasets and databases such as what might be available from the CDC (https://www.cdc.gov/datastatistics/ – selected as an example because of our present situation) or data.gov (https://www.data.gov/). Various activities might be based on the analyses of the data provided. Before I retired, I was aware of academic colleagues who were applying advanced statistical procedures to public data to look for relationships among variables in data that had not been considered.

My example here is far less sophisticated. It is also not original as it is proposed by one agency providing such data. The data provided by the U.S. Department of Education (https://www2.ed.gov/rschstat/landing.jhtml) even proposes the use of a graphing tool students can apply to data – Kids’ Zone – Create a Graph (https://nces.ed.gov/nceskids/createagraph/).

I thought I would use graphing to explore data that interests me. The data that caught my attention was taken from a table summarizing 8th graders daily use of a computer broken down by various categories (Table 218.50). I think you can expand this image to read the two rows of data I am focusing on here. These data concern whether a given student qualifies for free/reduced lunch (a proxy for SES) and amount of use.

The data table can be downloaded as a xls file (the type of file for use in an Excel Spreadsheet or other spreadsheet program) or just read from the chart.

With a spreadsheet, I can select the cells that interest me and turn the values into a graphic representation.

With the Create a Graph program, the data would be entered directly from the keyboard to generate a similar representation.

Now we have bar graphs representing data summarizing 8th graders daily use of a computer. What can be observed? Perhaps you find the percentage of 8th graders who report no use interesting. What does this mean? What would you need to know about the methodology that would help you understand? For example, would an iPad be counted as a computer? What about the difference in family resources? The largest gaps are in the groups claiming no use and the group claiming a small amount of use? Here low income seems to matter. It does not seem to matter among the groups using a computer quite a bit. Do you wonder if this was because computers were provided by the schools and this is where most computer use occurred? Research always leads to more questions when the data are limited and the methodology does not allow an answer to questions asked after the process of data collection is complete.

I intend to write a couple more posts focused on the recommendations from the library association. I would describe my goals as very much overlapping with the way this association describes data literacy. If there is a distinction, I would describe it as my interest in data and data collection methods that are not numerical. If data literacy is limited to quantitative data, I must still find another terminology to describe this broader perspective.

Loading

Exploring energy consumption – data collection and thinking

Energy consumption is an important component of so many issues. Energy consumption is a component of global warming and awareness can contribute to more thoughtful consumption or interest in a form of energy generation less dependent on fossil fuels. Energy consumption is a component of economics although for most of us the cost of the energy we consume is not a serious concern. What energy is in its various forms is a topic we learn about as part of our education. Typically, we learn about energy in a rather abstract way based on terms and formula. This is another post on the data and methods we might use to gain some insight into energy as it relates to the multiple topics I describe here.

Back in 2015, I became interested in solar energy and purchased a simple kit to be able to store electricity generated from a solar panel. The basic items of this kit are shown here with their corresponding prices as these prices are a component of how one must think about the cost of production and the items offer a concrete way to understand what is involved in a very basic method of production.

The two significant items in this kit are the 100 Watt solar panel and the charge controller. The value of this kit was $185 at the time of purchase. This 100 watt panel was supposedly capable of storing 500 watth (watt hours) per day under ideal conditions. Think of a 60 watt light bulb consuming 60 watts per hour as a way understand how much electricity it could generate. Ideal means best alignment with full sun. The charge controller protects the battery should the input of energy come too fast and should the battery reach capacity and the input need to be terminated.

You can use power from a solar panel as it is generated, but typically the approach involves sending power from the panel to a battery and then drawing on the power from the battery as necessary. This is called a deep cycle battery and is designed to be charged and run down. There are many different ways to design batteries. A contrasting type might be designed to serve a function such as would be required in starting a car – big push, but most of the time not being drawn down. The cost of the battery was $75. If you are curious, the items described here might be the collection of resources someone would use to put together a system should they want to use a camper “off grid”.

To power most electrical items we use, a 12 volt battery is not of much value. You need to convert the DC (direct current) output from a battery into AC (alternating current). The device to do this job is call an inverter. It is possible you have one in your car you might use (as a passenger) to plug your laptop in while someone else is driving. Cost of this inverter – $28.

So In 2015 I purchased this system capable of charging a battery and then powering low wattage devices for a total of $288. It was for experimenting, but I really never had a good place to use it. I set it up and knew it worked, but that was the end of the story.

Fast forward to the lock down required by the great pandemic of 2020. I am bored and need productive things to do. I decide to make use of my solar system and explore related issues.

We have this gazebo in our backyard. It is a beautiful little building that is pretty much too hot to use in the summer because of all the windows and too cold to use in Minnesota winters. We are presently in early Spring and I decided given my circumstances of limited movement I would use the Gazebo as a reading/writing shack. I can usually get by without heat. I am running a space heater today because we just received about 6 inches of snow, but this is atypical. Why not explore the utility of the solar panel system?

I needed one more tool to do the type of exploring that I thought might be useful. I purchased from Amazon (delivery right to your door even during the lock down) a usage monitor. This device is plugged into an outlet (from my DC/AC converter) and it then measures the power consumed when something is plugged into it. All kinds of data – volts, amps, watts, time power is being used, kwh (consumption in kilowatts).

Who buys these things? I am not certain. I think the idea is you use this device to try to understand what different devices might be contributing to your electricity bill and perhaps trying to decide if you should invest in a more efficient version of something you already use if it is very inefficient. I have a completely different agenda for this device. What a great learning through authentic investigation opportunity. There are plenty of things you can measure to understand power consumption and really appreciate things like the formula – watts = amps x volts.

So, for example, if I can actually measure the three variables in this formula independently, I should be able to see if the formula really works. This device should let me do this. I am using my MacBook pro connected to my solar system through my power consumption meter. The read outs are as follows:

.28 amps

35 watts

124 volts

The values jump around constantly so there is a slight delay as I switch the monitoring device from variable to variable. You can take any two variables and plug them in to the formula watts=amps x volts to predict the third variable and you should see that things come out pretty close.

What I was really interested in was how much did I save running my solar system while I worked in the gazebo. I can plug multiple devices into my solar system, but the monitoring device only allows me to measure the power used by one at a time. I ran multiple devices through the system during the week – mostly my laptop, but also my iPad and my Echo (for music).

My idea was to use the rate provided by more power company to see how much money I had saved.

So, the following two values were generated by my monitoring device.

So, in the past week, I spent 24 hours plugged into the solar system in my gazebo. Curiously, the hours came out to exactly a day. I used .41 kilowatts of electricity during that time. .41 x .088 equals about 4 cents. Given my up front cost was $288, four cents does not seem like much.

I found a site that allows you to estimate the cost to use different devices and a laptop was among the devices included. This site used 60 watts as the power consumption of a laptop (I measured my laptop at 35 watts an hour or so ago). My laptop was not doing anything other than powering the screen at the time. I have not compared the load when say I was watching on online video or something that would place a load on the machine. If I use the value from the site instead of my actual total and assume that the ipad and Echo would draw the same as my laptop, I now estimate my cost savings would have been about 16 cents. This result seems very different and I am going to have to collect some more data and do a better job of controlling the variable I ignored (the load required for the different devices) to see if I can come up with something closer to what the online site suggested.

So, this is real research and it is messy. This is why research invests so much time in replication and careful methodology. Perhaps my case for educational applications would have been more persuasive had I not shared my actual data and the confusion these data have generated. Still, I decided this would be the wrong message. I conducted research and published scientific papers for a living (having nothing to do with electricity or solar energy) and I decided no, the messiness of science is part of the experience. I am trying to make the case for learning about data collection, methodology, and analysis as important educational goals. The messiness and the confusion caused by different ways of operationalizing questions and interpreting data are part of the process. I decided this is real science and this is what students need to experience in appreciating the process.

Educators, the power consumption meter was $26. What real questions about power and power consumption could your students ask and answer with this device.

I have taken considerable liberty in describing the mathematics, physics, and engineering involved in collecting, storing, and using energy captured from the sun. The simple system I have described could be constructed using a wide variety of alternatives for each of the devices and these devices interact in complex ways to influence efficiency and longevity. In addition, systems that use multiple panels, multiple batteries, and connecting solar systems to the existing electrical grid add complexity that could be considered. A reasonable source should you be interested in these issues – Build your own low-budget solar system.

Loading

I see data everywhere

First, data is plural [when it refers to information and not used as an example of a word]. It is the version of datum used to indicate more than one datum. It is difficult to keep that straight as it is so often used incorrectly that many have given up asking for a correction. Data as a singular noun even sounds right. I try to be careful, but no one corrects me any more. Try submitting a research manuscript though claiming “the data is ….” and you will be notified.

So, this my first effort to bring the attention of educators and those who try to focus what educators do on the importance of student experience developing expertise in data collection (methodology) and interpretation. My interest is in the absence of such experiences and the personal belief that the development of data-related collection and analysis skills is more important than other semi-important initiatives such as the focus on coding and computational thinking. I have yet to come up with catchy one or two-word description of the capabilities and dispositions I am focusing on. Data literacy is close and would work, but is already taken and it tends to be defined in a way that is too narrow. The existing use of data literacy is too focused on numerical data and other important categories of data (by my way of thinking) exist,

The collection of data can lead to thinking which I think is key to learning. Thinking does not necessarily have to follow data access, but thinking as an educational task is a matter of taking the opportunities that are there.

I have a record of a very high proportion of the songs I have listened to since 2006. The total number of songs at the start of the time I was writing this was 280,063, but it will be different by the time I am finished because I almost always have music playing in the background when I work. [Remember this comment because it is a clue to something I will ask of you later.] The number 280,063 is potentially interesting because it may indicate something about me relative to you. I may be a bigger music fan than you and maybe not. The number alone hides other interesting data. 280,063 is a statistic that summarizes something about the data. The data in the rawest form consist of a list of song/artist titles. A very long list. Each item on this list is connected with a time and date (when does he and did he listen) and has been categorized by the company responsible for the service by genre (has this old guy’s tastes changed over the years). BTW – the service I use is called Last.FM and the process of sending a signal from the multiple digital devices I use to listen to music to the giant database maintained by the service is called scrobbling. I can’t scrobble when I listen to the radio in my truck or when some member of my family uses the Amazon Echo. But when I am listening to music I select, I am nearly always using a phone, tablet, ipod, ipad, or computer. For those who have now returned to listening to music recorded on records, I get you, but aside from the hiss I think your claim to authenticity is bogus. Authenticity changes with the times. You aren’t listening to piano rolls are you? I admit the general digital-oriented basis for my existence is also somewhat uncommon, but I am what I am.

Just for kicks, I want to show you two charts that summarize these data and because of the trends that are evident ask for an explanation. I will refer to these summary representations as graph A and graph B. I think they show the same thing even though they seem to be moving in opposite directions. The first is the number of songs heard by year and the second the number of songs heard in recent months.

These bar graphs use one form of data to demonstrate something interesting about my life [see the clue I warned you about in one of the first paragraphs]. I obviously know things about my life you don’t and what I know brings meaning to these representations.

Some things I know: I work a lot and when I work I most frequently am reading, writing, coding, etc. using a computer. I retired a few years ago [graph A]. All of our lives have changed recently and we are spending most of our time now isolated in our homes [graph B] finding things to do that are somewhat defined by this reality.

Some reference points. In 2010, I listened to 32,728 songs. In Mar 2020, I listened to 2072 songs and 1226 songs through the first 8 days of April. The average length of a song is between 3-5 minutes, but varies significantly with genre. I suppose this fact is somehow useful in interpretation, but even 3 vs 5 is a large difference when you multiple the limits by 320,000 or so.

I wonder if these data say anything about my psychological state. I listen to a lot of blues, but blues for me is a happy genre. I listen to classical when I am in a somber mood.

Lifetime top artist – Miles Davis – 3919 listens

List 365 days – Bob Marley – 116 plays

Loading

Educational goals shape classroom practice

When we first began writing about the role technology could play in education, we focused on the idea that technology provided tools for thinking and learning. This idea came from the work of David Jonnasen and his book MindTools. We built on his observation that technology tools many folks used to accomplish tasks in their everyday lives (writing to communicate, spreadsheets to organize and ask questions of numerical data, etc.) could be used in classrooms not just with the goal of learning how to use the tools, but to develop an understanding of the concepts and skills important to content areas by applying the technology tools to the processing of content area information. Our first textbook used the umbrella of technology integration to describe this general approach, expanded the list of technology tools that could be useful, and created a theoretical and research-based framework based on authentic tasks, constructivism, and contextualized learning to explain and justify such tasks. 

I don’t know if blog posts are the place for exploring what I consider core concepts that drive curriculum development and classroom practice, but I want to try out some big ideas that I think have value and are not being considered by enough educators. 

Let me start with a comparison. Coding/programming has received a great deal of recent attention as identifying concrete tasks that can be used in classrooms with proposed benefits that are either not presently achieved or that can generalize in ways that make a focus on such tasks more efficient than existing classroom activities. Based on these arguments educators and administrators have been lobbied to turn over classroom time to coding/programming experiences. The benefits of this reallocation could result in a) the development of an important vocational skill (coding), b) deeper insight into how technology applications actually work (technology literacy), and c) the development of computational thinking. As you move through this list, I would argue that the proven benefits become more difficult to demonstrate. There is a difference between speculating about benefits and demonstrating these benefits exist.

OK, so coding for all has caught people’s attention and this attention has resulted in real change in how students spend time. There are secondary issues that must follow if educators continue to implement classroom experiences based on a focus on this category of classroom activities. For example, at some point, it makes sense that educators providing such experiences actually be prepared to do so and their competency in this area be evaluated as a component of their preparation to teach. We are yet to reach the point at which this is happening at a meaningful level.

The issue I am trying to address concerns how a certain type of classroom experience gains leverage in determining how classroom time is spent. Is it based on the identification of skills that are to result from education and a consideration of what learning experience best develop these skills? I don’t have the sense that computational thinking was identified as something all students need and an effort was then followed to determine how best to develop such skills. The emergence of computational thinking seemed more a justification that followed something already gaining momentum to add more pressure to the cause.

I understand the public’s interest in technology and the mystery in many people’s minds as to how the magic of technology is accomplished. OK – this fascination in some areas could result in a preparatory course at the secondary level and the opportunity to major in a discipline in college. How did we get from such possibilities to the argument that all students need to develop computational thinking? As might now be obvious, I don’t think computational thinking is an actual thing and the concept is just an area-specific version of problem-solving. Why stop with coding? One could easily argue that compositional thinking as exhibited by those heavily involved as writers and promoted by educators as writing across the curriculum should be recognized for its proven general benefits, but is commonly ignored because committing more time to writing activities is less trendy. The generalizability of higher-order cognitive skills has long been controversial. We call this transfer and achieving transfer from a specific application of a cognitive skill to a more general application of this skill is tricky and difficult to demonstrate. It simply does not happen automatically with the amount of experience schools can devote to coding and the existing practices of teaching coding are not designed to transfer whatever computational thinking is expected to emphasize.

All this aside, I think I can make a better argument for a completely different area of skills, knowledge, and dispositions I think have greater generality. Part of the challenge I see in my effort is the generation of a cutesy name. I can’t come up with something for what I have in mind that is the equivalent of computational thinking. Perhaps I will eventually make an effort, but I will first offer the background for this proposal.

What I have in mind can be identified at the level of application in such claims as data-driven or empirically supported. As consumers, clients, patients, etc., we want to believe that the value, accuracy, etc. of the experiences, information, etc. we are provided have been established. How such goals are met depends on the discipline, but all disciplines have processes based in the collection of data through processes I like to describe as methods. We are probably most familiar with the approach we associate with “the sciences” and that abstract model we call the scientific method. Still, historians have data and apply processes and dispositions to these data to generate accounts of history that they argue are accurate and unbiased. Economists and political scientists do the same. Marketers attempt to identify public opinions in order to make decisions on product preferences and persuasive approaches. Data collection, analysis, and practices based on the consequences of collection and analysis are everywhere.

At best, we give lip service to the scientific method in K-12 science instruction. Students typically follow carefully scripted labs and have no experience designing experiments (methods) to test ideas (hypotheses) through the collection and analysis of data (statistics). 

Like computational thinking, I see the benefits on multiple levels. Methods and analytical frameworks vary with domain. There is more. With experience, you learn not just the methods of a given domain, but also the limits of these methods. An example familiar to many is the use of a conclusion generated from an experiment that does not involve manipulation in arguing a causal relationship. This problem is commonly described as “correlation is not equal to causation”. Another example is that I would describe as anecdotal reasoning. For example, a politician tells the story of an individual in justifying some policy. The example may be accurate but the conclusion that this example is the most common result of similar circumstances is unproven.  The skills involved in the specification of a problem and the establishment of a methodology to collect data relevant to this problem and then the evaluation of what such data indicate are everywhere. Where and when do we acquire such skills and in how many areas are such skills developed? We should develop such skills even when we do not apply them directly as we need to evaluate the cases based in data that others offer us to influence our behavior.

This post is getting very lengthy and I am guessing readers want me to get past my rationale to something that is actionable. I have some very concrete suggestions for classrooms in mind. I propose that educators take a look at the content they teach and think about the skills they are trying to develop and begin to ask the following questions:

  1. What data are relevant to the topics I teach? What data can my students collect that have some bearing on these topics?
  2. After relevant examples of data have been identified, the thinking then should move to the methods that can be used to collect such data and what about these methods should be considered that could limit accurate conclusions being reached.
  3. Finally, educators should consider the dispositions that must accompany data collection and analysis. What attitudes encourage critical thinking and openness to the examination of personal beliefs that data collection might challenge?

Once you understand these goals, data sources, data collection and analysis methods, and desirable dispositions are everywhere. You collect data when you go to the library or search online to answer a question. What search tactics increase the likelihood useful information will emerge from your search? What commitments to thinking about this information will allow you to discount misinformation and use valid information to challenge existing beliefs? You collect data when you create a questionnaire to identify the choice a class prefers among multiple options. How can the questions in the questionnaire be written to avoid bias? The examples of  opportunities are limitless, authentic, contextualized, and focused on skills necessary throughout life. This is a mindset that is consistent with existing curriculum goals, but emphasizes skills and methods of developing these skills that are seldom emphasized.

With this position statement as a starting point, I intend over the next several weeks to identify some examples of methodology, data collection, and analysis that encourage others to imagine their own possibilities. 

More related content to follow.

Loading