Milk Comes from the Store; Data Comes From the Internet

Kim Kastens
Author Profile
published Dec 11, 2009

Milk from Cow<image source>
As I mentioned in an earlier post, I've been working with a doctoral student, Sandra Swenson, who has asked 8th, 9th and 12th graders a series of open-ended questions about a widely-used geoscience data map. One of her questions is "How do you think this was made?" A substantial fraction of these kids provided answers along the lines of "it came from a computer," giving no inkling that data had been gathered from the actual Earth in order to create the data representation. I was reminded of the stereotype of the city kid who thinks that milk comes from the grocery store.

Imagine my outrage when I came across a website made by adults who should know better. The geoSPATIAL DATA ACQUISITON website provides access to bathymetry, coastline, landuse, soil type, precipitation, and other useful and interesting data types. But the so-called "acquisition" procedure is to download files from various government websites. Do the students who use this website understand as much about where these data came from as they do about where milk comes from?

GeoSpatial logo<image source>

The site includes an extended student exercise, Lost Creek: A Case Study in Data Acquisition. In this exercise, the student successively downloads elevation data, aerial photo images, soil type, landuse classifications, precipitation data, and hydrographic data, and then runs hydrologic modeling software. Although I'm always happy to see activities in which students engage substantially with geoscience data, it troubles me that at no point along the way is the student either asked or told anything about how the data were obtained from the earth.

To someone like myself, who spent many months at sea extracting data from the bottom of the ocean, it feels essential that students should understand in some deep intuitive way that geoscience data ultimately come from the Earth--not just from the Internet and not just from a government agency.

I also think they should understand something about how the data were acquired and processed, so as to cut down on the chances of misinterpreting the data. But I don't know where to draw the line: how much of the back story of geoscience data do students need to know in order to learn effectively from the data?

Milk Comes from the Store; Data Comes From the Internet --Discussion  

This is also a bugbear of mine. At least the general public 'think' they know what forensic scientists do from watching the likes of CSI. I know of very few depictions in the media of what earth scientists 'actually' do. Even earth science documentaries generally have geologists as talking heads (and occasionally turning over a rock to show a fossil - being careful to hide the museum label on the sample) rather than showing what they do. Perhaps it would be boring for the layperson but as a consequence nobody actually knows what a geologist 'does' and this is to the detriment to science as a whole.


Share edittextuser=3052 post_id=3654 initial_post_id=0 thread_id=1250

I'm loving this series, Kim. Looking forward to more metacognitive posts along these lines. Thanks!


Share edittextuser=2987 post_id=3655 initial_post_id=0 thread_id=1250

This very much reminds me of when I taught intro geology and we discussed the damages associated with earthquakes.... I had many students who were perplexed by the notion that after an earthquake, people would not have access to drinking water. Their logic: water comes from the faucet. When I explained that the pipes leading to their faucet might have been broken in the quake, they said, well, then go buy bottled water from the store. They really had no concept of how water gets from the ground to their house or to the store. And of course no real understanding of how impassable roads might be.


Share edittextuser=541 post_id=3656 initial_post_id=0 thread_id=1250

Interesting question. I wonder how important it is that they understand how the data is acquired? As a geoscientist, I collect some data on my own, but I also make use of data and interpretations that I wouldn't have been able to collect on my own. (For example, I've never done U-Pb zircon dating, but I think about data collected by other people when I think about whether my models are reasonable or not.)

On the other hand, it would be nice if students (and the general public) had some understanding of where scientific ideas come from. (And it would be nice the ideas were fairly realistic.) But how do we get there? I have a long project in which intro students collect data from a local river, but I don't know if the students extrapolate from that experience to other kinds of geologic data.


Share edittextuser=1095 post_id=3657 initial_post_id=0 thread_id=1250

Interesting blog, I'm flattered you identified it as a source of providing geoscience data since I have developed it with the help of several graduate students.
I have to say though that I do feel the outrage is a little misplaced. The site was developed specifically for civil engineers who do hydrologic modeling and without access to this kind of data the process is much more laborious - and prone to error. While making the download of data (from the "best" sites) the most obvious links you will also find further information about it's accuracy, use, sources (from original locations) if you want to.
Another example you might have used would be cars. How much of the inner-workings of a car do you know, or do you have to know in order to effectively use a car? I know if I had to know how every part worked and functioned I'd still be walking or riding a bike (a debate for a different blog :). I don't disagree that the more you know about a car the better you can use and take care of it, but there simply is not capacity to know in depth all of the science around us, and thanks to good geoscientists like you and others we can "trust" that the data have been put there with a reasonable expectations of their application - in this case for hydrologic modeling studies. We have to trust the mechanics, the doctors, the farmers/grocers in our lives at some level.
Great discussion , and one that will (and should) always be asked. Thanks!


Share edittextuser=3055 post_id=3658 initial_post_id=0 thread_id=1250

Hi Jim,

Thanks for coming over to join our conversation.

Your point is well taken. It's not just cars; there are a whole range of engineered structures and devices that I use without delving into the details of how they were designed and constructed, trusting that the appropriate experts used the appropriate quality control techniques. And the risk I take by trusting is higher when the devices in question are things like bridges and airplanes, much higher than for a mere bathymetric data set. So it seems fair and reasonable that engineers in turn should be able to rely on work done by geoscientists without delving into details of how the data were acquired and processed. Division of labor is one of humanity's most powerful strategies for problem solving, and we sure have plenty of problems that need solving, including many that need both engineering and geoscience input.

I get more concerned, though, when earth science students don't know or care where the data came from and how it was acquired and processed. I think that there is at least a danger of mis-mediated knowledge of the earth if they learn from data they don't really understand. In my opinion, this is a very under-researched question in science education research: what is the interplay between understanding where data came from and understanding the Earth phenomenon of which the data are our best available representation?

I wonder if in part we are dealing with a vocabulary disconnect? The term "data acquisition" in the disciplines I am familiar with is unambiguously associated with the concept of: "... the process of sampling of real world physical conditions and conversion of the resulting samples into digital numeric values that can be manipulated by a computer." I took that definition from but it aligns with several other definitions on the web, as well as the way the term is used by the scientists and technologists I know in and around geosciences. Engaging with the real world is an essential component of that definition, making what the scholar of science studies Bruno Latour would call a "first inscription" or what the anthropologist Chuck Goodwin would refer to as the moment when Nature is turned into culture (when something non-human-made is turned into something human-made).

I get the impression that in your field the term is closer in meaning to the common English usage: "the act of gaining possession of something," so one would "gain possession of," for example NOAA bathymetric data sets, by downloading them following the directions on your website.


Share edittextuser=75 post_id=3659 initial_post_id=0 thread_id=1250

This reminds me of the spectacle of a hearing of the US Senate Commerce Committee in which a distinguished senator (who will here go unnamed) who was calling into question the need for federal support of the National Weather Service. His question (paraphrased) to the Director of the National Weather Service was: Why should we spend federal dollars to support the National Weather Service when people can just go to the Weather Channel to get the day's forecast? The curt response from the Director was (to the effect): Sir, the Weather Channel gets their data from the National Weather Service.

In a module on Geochemical Instrumentation and Analytical Techniques, we hope to help students become "critical consumers and producers of data." See this module at: We provide essential information about the underlying theory of the technique (e.g. properties of X-rays), instrumentation components, sample selection and preparation, standardization and related procedures, data acquisition, data rendering and presentation, examples of appropriate applications, and a review of limitations of the technique, and references and links for future exploration. Our hope is that students will be able to crictially analyze data presented in a departmental seminar or journal article, understand the context of how the data are presented and used, and perhaps even ask an insightful next question.

Data, and data products (e.g. maps, graphs, images) must be presented in the full context of how they were acuqired, what questions were being asked, who collected the data and for what purpose, and in light of the assumptions and procedures that were used in acquiring/developeing the data/products. Without this full context, "data" are merely numbers.


Share edittextuser=7 post_id=3660 initial_post_id=0 thread_id=1250

David Mogk's last paragraph in this blog section is exactly what science education programs should help their teachers to understand. Just to be clear, I am restating it here:

Data, and data products (e.g. maps, graphs, images) must be presented in the full context of how they were acquired, what questions were being asked, who collected the data and for what purpose, and in light of the assumptions and procedures that were used in acquiring/developing the data/products. Without this full context, "data" are merely numbers.

My research (referred to earlier in this blog) has shown that students most often described data about the Earth as being acquired from computers and satellites (including satellite pictures) with only a small percent actually understanding that the data was collected by instruments specific to the data type. While it is true that a lot of data about the Earth is gathered from satellites, it's my concern that students state "satellite" as a means for gathering data as a general term because satellites are out in space and can take "pictures" of the Earth. It doesn't seem to matter if the representation is not a photographic picture or what the data in the key represents; many students surveyed in my research still mentioned satellite and computers as being the method of data acquisition.

It's not that I am suggesting that students should know about all of the details of satellite instrumentation and other sophisticated instrumentation (e.g. U-Pb zircon dating or the inner workings of cars) but that they should be introduced to a variety of instrumentation used to gather data and for what purpose and by whom. As David also pointed out, this should be discussed in light of the assumptions and procedures that scientists make when developing the data and/or products of data.

Teacher and student understanding of these processes and instrumentation is one aspect of taking a closer look at the process of science. I personally appreciate how NASA Earth Observatory quite often describes the image, the data, and how the data was acquired. (e.g. and The GeoMapApp website does this too, but there are many images out there on the Internet that do not and I think the image and its purpose lose credibility when they are lacking this information. Images just become, "wow, that's a really cool picture."


Share edittextuser=3517 post_id=6864 initial_post_id=0 thread_id=1250

It strikes me that one of the contentions in this discussion is that we can't all understand how everything works- but isn't the point here that when teaching students about earth science, they should know how Earth Science works? In other words, teaching mechanics how to fix cars would not work without teaching them how the car works. Along the same vein, teaching students how earth scientists work means teaching them how the data were collected (among other things). I guess one could argue that not every student in an earth science class is going to actually become an earth scientist, but how can we adequately give them that chance without making these connections? I suppose the main difference is that probably most students in classes for car mechanics actually intend to become practicing mechanics someday... this may boil down to the difference between a specific trade school and a liberal arts approach. But, it seems like even things like "how a car works" must include such basics as "you need to put gas in" (at least today, until it requires plugging in or something else!) which is a corollary to "how a geoscientist works" includes "how you collect and analyze data."


Share edittextuser=2126 post_id=13978 initial_post_id=0 thread_id=1250

My brother-in-law, Dan Chayes, has pointed out to me the source of the "milk comes from the grocery store" meme. It comes from Aldo Leopold, Sand County Almanac, in which he wrote: "There are two spiritual dangers in not owning a farm. One is the danger of supposing that breakfast comes from the grocery, and the other that heat comes from the furnace." This quote and others are at:


Share edittextuser=75 post_id=18401 initial_post_id=0 thread_id=1250

Join the Discussion

Log in to reply

« My Catastrophe is Bigger than Your Catastrophe       Where do Data-Driven Visualizations Come From? »