PAPER Inspiring Future Marine and Data Scientists Through the Lure of Ocean Tracks AUTHORS Randall E. Kochevar Education Development Center, Inc., Waltham, Massachusetts, and Hopkins Marine Station, Stanford University Ruth Krumhansl Kira Krumhansl Education Development Center, Inc., Waltham, Massachusetts Cheryl L. Peach Scripps Institution of Oceanography, University of California, San Diego Erin Bardar Josephine Louie Education Development Center, Inc., Waltham, Massachusetts Jessica Sickler Lifelong Learning Group, Columbus, Ohio Julianne Mueller-Northcott Amy Busey Silvia LaVita Jacqueline DeLisi Education Development Center, Inc., Waltham, Massachusetts ABSTRACT The Oceans of Data Institute (ODI) at the Education Development Center (EDC), Inc.; Stanford University; and the Scripps Institution of Oceanography have been collaborating, with the support of three National Science Foundation grants over the past 5 years, to bring large scientific data sets into secondary and postsecondary classrooms. These efforts have culminated in the development of a Web-based stu- dent interface to marine science data called Ocean Tracks (http://oceantracks.org), which incorporates design principles based on a broad range of research findings in fields such as cognitive science, visual design, mathematics education, and learning science. The Ocean Tracks interface was tested in high school classrooms in spring and fall of 2013 with a total of 195 high school students. These tests indicate that students appeared to find many aspects of the interface simple and intuitive to use. Teachers and students indicated that working with real data was highly engaging, pointing to the tremendous potential for “big data” to transform the way science is taught. Interest among college faculty in Ocean Tracks indicates a need in undergraduate classrooms for similar tools that allow students to interact with data. So in the fall of 2014, we began to collect baseline data on students attending undergraduate oceanography classes at the Scripps Institution of Oceanography (Scripps) and Palomar College, where we will also be developing curricula and conducting classroom tests. Preliminary results from this work are presented here. Keywords: Ocean Tracks, data literacy, big data, marine science, data science Introduction The fact that it was real animals being tracked was cool ….I learned a lot about elephant seals and biodiversity hotspots. It was interesting that so many animals go to the California coast. — Undergraduate student reflecting on the Ocean Tracks Web interface “B ig data” has become a big deal. The adjective “big” has an evolv- ing meaning but refers not just to the ever increasing volume of data available to use in scientific research but also to the dizzying velocity at which real-time data are generated, and the expanding variety of data that are generated as new sensors and technologies are de- ployed to quantify the nuances of the natural, built, and social environments (Dijcks, 2013). The use of data as a decision driver has expanded into every scientific, industry, and business enterprise (Manyika et al., 2011). Not only is big data pushing the bound- aries of scienti fic discovery, but car mechanics, nurses, real estate agents, and teachers are now all expected to use data in their professions. Analyzing data, discerning meaningful patterns, and extracting useful information have become gateway skills to full par- ticipation in the workforce and civic engagement in the 21st century, and individuals skilled in the variety of tasks and duties required to acquire, aggregate, clean, organize, and analyze these data sets are in high demand. There ’ s been widespread recog- nition of the large and growing gap be- tween the need for a workforce skilled 64 Marine Technology Society Journal
Transcript
P A P E R
Inspiring Future Marine and Data ScientistsThrough the Lure of Ocean TracksA U T H O R SRandall E. KochevarEducation Development Center,Inc., Waltham, Massachusetts,and Hopkins Marine Station,Stanford University
Ruth KrumhanslKira KrumhanslEducation Development Center,Inc., Waltham, Massachusetts
Cheryl L. PeachScripps Institution of Oceanography,University of California, San Diego
Erin BardarJosephine LouieEducation Development Center,Inc., Waltham, Massachusetts
Julianne Mueller-NorthcottAmy BuseySilvia LaVitaJacqueline DeLisiEducation Development Center,Inc., Waltham, Massachusetts
A B S T R A C TThe Oceans of Data Institute (ODI) at the Education Development Center (EDC),
Inc.; Stanford University; and the Scripps Institution of Oceanography have beencollaborating, with the support of three National Science Foundation grants overthe past 5 years, to bring large scientific data sets into secondary and postsecondaryclassrooms. These efforts have culminated in the development of a Web-based stu-dent interface to marine science data called Ocean Tracks (http://oceantracks.org),which incorporates design principles based on a broad range of research findings infields such as cognitive science, visual design, mathematics education, and learningscience. The Ocean Tracks interface was tested in high school classrooms in springand fall of 2013 with a total of 195 high school students. These tests indicate thatstudents appeared to find many aspects of the interface simple and intuitive to use.Teachers and students indicated that working with real data was highly engaging,pointing to the tremendous potential for “big data” to transform the way scienceis taught. Interest among college faculty in Ocean Tracks indicates a need inundergraduate classrooms for similar tools that allow students to interact withdata. So in the fall of 2014, we began to collect baseline data on studentsattending undergraduate oceanography classes at the Scripps Institution ofOceanography (Scripps) and Palomar College, where we will also be developingcurricula and conducting classroom tests. Preliminary results from this work arepresented here.Keywords: Ocean Tracks, data literacy, big data, marine science, data science
Introduction
The fact that it was real animalsbeing tracked was cool… . Ilearned a lot about elephantseals and biodiversity hotspots.It was interesting that so manyanimals go to the California coast.—Undergraduate s tudentreflecting on the Ocean TracksWeb interface
“Big data” has become a bigdeal. The adjective “big” has an evolv-ing meaning but refers not just to theever increasing volume of data availableto use in scientific research but also tothe dizzying velocity at which real-timedata are generated, and the expandingvariety of data that are generated asnew sensors and technologies are de-ployed to quantify the nuances of thenatural, built, and social environments(Dijcks, 2013). The use of data as adecision driver has expanded intoevery scientific, industry, and businessenterprise (Manyika et al., 2011). Notonly is big data pushing the bound-
aries of scientific discovery, but carmechanics, nurses, real estate agents,and teachers are now all expected touse data in their professions. Analyzingdata, discerning meaningful patterns,and extracting useful informationhave become gateway skills to full par-ticipation in the workforce and civicengagement in the 21st century, andindividuals skilled in the variety oftasks and duties required to acquire,aggregate, clean, organize, and analyzethese data sets are in high demand.
There’s been widespread recog-nition of the large and growing gap be-tween the need for a workforce skilled
64 Marine Technology Society Journal
in working with big data and the sup-ply of qualified graduates (Manyikaet al., 2011; Shaw, 2014). Wheredo students acquire these skills? Moststudents enrolled in traditional sci-ence laboratories often work withsmall data sets that they collectedthemselves. Although work withsmall data sets teaches important foun-dational skills, students are not ade-quately engaged in developing a newand broader set of skills necessary toextract meaning from large, complexdata sets. The nature of the new dataanalytical skills that students needwas clarified by an occupational profileof the “big-data-enabled specialist,”which was recently generated and vali-dated by more than 150 experts froma broad variety of industry sectors(Oceans of Data Institute [ODI],2014a). Among the array of skillshighlighted, the profile emphasizesthe need for graduates who are ableto define problems and articulate ques-tions that can be addressed by large,complex data sets. It also revealed aneed for scientists and technologistswho are able to develop appropriateanalytical methods and tools that re-flect a deep knowledge of a variety ofdata sources (beyond what they havecollected themselves). The studystrongly endorsed the critical need for“soft skills” such as analytical thinking,critical thinking, and problem solvingand asserted that a successful bigdata-enabled specialist should be aseeker of patterns, open-minded, andcurious (ODI, 2014b).
Although the definition of big datais not well established in the publicsphere, those concerned with educat-ing K-16 students define big data asdata sets that are Complex (i.e., theyinclude multiple parameters or datatypes), Large (i.e., they include moredata than are appropriate to answer
any particular question), Interactivelyaccessed (i.e., the user has choices, typ-ically online, about what data to exam-ine and how to visualize and analyzethose data), and Professionally collected(i.e., they go beyond what students orordinary citizens are able to collectthemselves). To work effectively withthese so-called CLIP data sets, studentsmust have the ability to go beyondroutine analyses. For example, stu-dents who are skilled in CLIP dataset analyses must have the abilityto select appropriate professionallycollected data to investigate a ques-tion, create a variety of unique datavisualizations appropriate to answera question, relate multiple data pa-rameters to each other, and use mul-tiple lines of evidence to support ahypothesis.
The increasing availability of scien-tific CLIP data sets poses a tremendousopportunity to develop these skills.More broadly, carefully constructeduse of authentic scientific data in theclassroom enables students to engagein learning activities that are moredeeply inquiry based. It offers the po-tential for higher development ofproblem-solving skills, addressesmore complex concepts, and offersgreater relevance to students’ lives thantraditional learning activities (Hotaling,2005; Parsons, 2006; Simmons et al.,2008). Learning research has garneredconsiderable evidence that knowledgeacquired by students via the simple“taking in” or memorization of infor-mation is fragile and can be superficial.To build a more robust and enduringunderstanding of content, students inscience classrooms need to activelythink with new information, con-necting and applying concepts as theyconstruct scientific explanations forobserved phenomena (National Re-search Council [NRC], 2000, 2012a).
The NRC framework, which formsthe basis for the recently released NextGeneration Science Standards (NGSS),asserts that participation in the prac-tices of science “makes students’ knowl-edge more meaningful and embeds itmore deeply into their world view”(NGSS Lead States, 2013). At the un-dergraduate level, research suggeststhat inquiry-based activities are under-utilized in undergraduate classroomsand that faculty still largely rely ona lecture format in their courses. Inrecognition of this, the President’sCouncil of Advisors on Science, Tech-nology, Engineering, and Mathemat-ics (PCAST) “advocate and providesupport for replacing standard labo-ratory courses with discovery-basedresearch” at the undergraduate level(PCAST, 2012). The incorporationof student investigations with an arrayof large scientific data sets, now avail-able online, provides a significant op-portunity to do so.
However, scientific data portalsand analytical tools are generally de-signed for experts, which poses a sig-nificant barrier to their use in theclassroom. Recognizing this, the Na-tional Science Foundation (NSF)funded the Oceans of Data project togather knowledge relevant to the de-sign of student-friendly interfaces toscientific data from a broad range offields such as cognitive science, visualdesign, mathematics education, andlearning science. The resulting guide-lines, published in Visualizing Oceansof Data: Educational Interface Design(Krumhansl et al., 2014), were thenapplied to the design and developmentof a student interface to marine sciencedata called Ocean Tracks. This articledescribes what we are learning fromclassroom testing of Ocean Tracksat the high school and (now) under-graduate levels, focused on developing
July/August 2015 Volume 49 Number 4 65
students’ skills in working with CLIPdata.
The OceanTracks Interface
Funded by theNSF in 2012,OceanTracks: Investigating Marine Migra-tions in a Changing Ocean—a col-laboration between the EducationDevelopment Center (EDC), StanfordUniversity, and technology partnerEarthNC—provides customized ac-cess to data collected by the Taggingof Pacific Predators program, alongwith physical oceanographic data setsfrom the National Oceanic and Atmo-spheric Administration (NOAA) andNASA (http://oceantracks.org). Theability to visualize the movements offree-ranging marine species has onlycome about in the past decade due torecent developments in electronictagging technology (see sidebar).These tags have revealed fascinatingpatterns in habitat usage by theselarge marine animals and have raisedcritical questions about their futurepersistence in ocean ecosystems(Block et al., 2011; Hazen et al., 2012;Maxwell et al., 2013). TheOcean Tracksmap interface and accompanyinganalysis tools enable students to be atthe frontier of this field of knowledgeand to engage in classroom learningactivities that address the same kindsof questions being considered bywildlife researchers, such as, “How doyou design conservation strategies forhighly mobile species whose habitatusage is projected to change as theclimate changes in the future?”
The core of theOcean Tracks inter-face is a familiar, Google Maps–basedview of the earth, centered on theNorth Pacific Ocean (Figure 1). Whenthe interface is launched, a red line(the track of a northern elephant seal)
appears on the map, marking the seal’spath northwest from the Californiacoast to Alaska’s Aleutian Islands,where it loops around before swim-ming south and returning to thebeaches of California. The map is sur-rounded on both sides by pull-outtabs labeled Tracks, Tools, Overlays,Login, Measurements, and Library.Using these tabs, users are able toadd or remove tracks from the map,dynamically graph data associatedwith a specific track (e.g., daily maxi-mum diving depth and daily averagespeed), or add more data layers (e.g.,sea surface temperature and chloro-phyll a). Users can log in to annotateand save map views as well as recorddata in a data entry table structuredspecifically for the data types in the in-terface. The Library contains custommultimedia content about the animals,tools, and technologies and providesaccess to video tutorials that guideusers through the use of each interfacefunction.
Beyond providing access to datathat have typically been out of reachbut may be of great interest to stu-dents, the Ocean Tracks interface wasdesigned to optimize students’ oppor-tunities to focus their cognitive re-sources on viewing and comparingdata to test hypotheses, while mini-mizing the time spent on download-ing, filtering, and creating displays(Krumhansl et al., 2014). By automat-ing many of the complex and time-consuming processes necessary tomake customized data representations,students are able to delve deeper intomaking meaning and building moresophisticated understandings of thedata.
Using the Ocean Tracks interactivemap and data analysis tools, studentsare able to explore and quantify pat-terns in the migratory tracks of sharks,
The Tagging of Pacific Pred-ators (TOPP) program began in2000 as part of the InternationalCensus of Marine Life. Usingelectronic tags, a multidisciplin-ary team of 75 scientists fromfive nations deployed morethan 4,300 tags on 23 differentspecies of apex predators in theNorth Pacific Ocean, includingwhales, seals, sharks, tunas, sea-birds, sea turtles, and evensquid. By the year 2010, morethan 365,000 days of trackingdata had been collected, repre-senting the largest data set ofits kind. The electronic tagsused in the TOPP program in-cluded tiny archival tags that re-cord information on light (usedto calculate position), tempera-ture, and depth; pop-up satellitearchival tags that record similardata but release automaticallyfrom the animals at a presettime and telemeter their datavia satellite back to the labo-ratory; and satellite tags thatallow transmission of positionand oceanographic data in real-time. By the end of the TOPPprogram, these technologieshad reached a sufficient level ofreliability and precision suchthat they are now routinelyfeeding oceanographic datainto NOAA’s Integrated OceanObservation System (IOOS),which is used to predict oceanweather around the world.
66 Marine Technology Society Journal
elephant seals, and other marine ani-mals in the northern Pacific Oceanby taking measurements, such asspeed and diving depth, to support hy-potheses about marine animal be-havior (Figure 2). The interface thensupports students in relating these be-haviors to fluctuations and trends inphysical oceanographic variables,such as sea surface temperature andocean currents (Figure 3).
These interface features allow stu-dents to engage with the data in inves-tigations that mirror those currentlybeing conducted by scientists tounderstand the broad-scale effects ofchanges in climate and other humanactivities on top predators in ocean
ecosystems (Block et al., 2011; Hazenet al., 2012; Maxwell et al., 2013). Byinvolving students with these authen-tic data sets, Ocean Tracks supportsthe teaching of content related to ma-rine ecology, including productivity,ocean circulation and physics, andglobal climate change.
High SchoolClassroom Testing
Iterative development of alpha andbeta versions of the interactive Webinterface was informed by two roundsof classroom pilot testing of OceanTracks with high school students andteachers in spring and fall of 2013
(Table 1). Drawing on multiple datasources (Table 2), we examined the us-ability of the interface in classroom set-tings and the range of data skills thatstudents were able to demonstratewhen investigating Ocean Tracksdata. The classroom tests also providedopportunities to gather informationabout the types of supports needed(such as curriculum, teacher guides,and video tutorials, all available athttp://oceantracks.org) to developstudents’ abilities to use CLIP datasets to learn about the natural world.In addition, our research examinedhow teachers implemented draftOcean Tracks curriculum modules,the challenges and successes they
FIGURE 1
Ocean Tracks utilizes the Google Maps interface, which we found most students to have already used extensively—alleviating the need for instructorsto teach users routine functions like scrolling and zooming in and out. It also features tabbed pull-out menus that give users access to more sophis-ticated tools, which interact with the map display.
July/August 2015 Volume 49 Number 4 67
perceived during implementation, andthe ways in whichOcean Tracks can beused to meet curricular goals andteaching standards.
The interface and supportingmate-rials were revised in between the firstand second pilot tests in accordancewith needs that arose from analyses ofPilot 1 data. During both pilot tests,we collected data in the form of stu-dent surveys, student focus groupinterviews, student screencast record-ings, student work, classroom observa-tions, and teacher logs. Because mostof the collected data were qualitative,the project team developed codingschemes for each data source and sum-marized findings across these sourcesrelated to interface usability, types
of data skills displayed, and areasof student engagement. Primary find-ings from Pilot 2 data are discussedbelow.
FindingsData from 22 screencast recordings
collected and analyzed from the secondpilot test indicate that students ap-peared to find many aspects of theinterface simple and intuitive to use.Students were observed performingwith ease each of the following majorinterface functions: turning on animaltracks and finding data for specifictrack points, activating tracks andtailoring intervals along time-seriesgraphs for specific track or oceano-
graphic variables, selecting map over-lays corresponding with individualanimal tracks and track segments, se-lecting multiple animal tracks andusing the hotspot tool to reveal specieshotspots, annotating maps with mapmarkers, navigating to and within theOcean Tracks library to find relevantinformation, and zooming and pan-ning around the primary map view.Data from classroom observationswere consistent with these findings:Observers noted that most studentsfound it easy to navigate the majorfunctions of the interface and to accessthe animal tracking and oceanographicdata available.
We also found that Ocean Tracksprovided students with opportunities
FIGURE 2
The Ocean Tracks interface allows users to take measurements of animal tracks using a customized set of tools. A “time slider” is used to selectspecific track points, shown in bold red, and visualize a plot of depth, speed, and curviness of the track over the selected interval.
68 Marine Technology Society Journal
to demonstrate important data skills.In an analysis of 23 student work as-signments that were randomly sam-pled from the 92 that were collected,the vast majority of assignments(87%) contained a data visualization
such as a map, table, or graph that stu-dents had generated to address a ques-tion posed in the Ocean Tracks draftcurriculum (Figure 4). In at least 70%of sampled work assignments, studentsdescribed relationships between two or
more data variables and presented aninterpretation of a data pattern usingbackground information drawn fromthe Ocean Tracks library or anothersource. In over 60% of sampled workassignments, students produced a
TABLE 1
Ocean Tracks classroom tests: dates and sample characteristics.
Date Duration Teachers (n)a Students (n) Course Titles Grade Levelsb Settingsc
Pilot 1 Spring 2013 5–6 weeks 3 61 Marine biology, biology,urban ecology
9–12 Suburban/town,urban
Pilot 2 Fall 2013 4–6 weeks 4 134 Same as above Same as above Same as aboveaThe three teachers who participated in Pilot 1 also participated in Pilot 2. An additional marine biology teacher from one of the Pilot 1 schools joined in Pilot 2. Duringboth pilots, one marine biology teacher participated with two classrooms of students.bThe biology class targeted students in Grade 10. The marine biology classes contained students in Grades 11–12, and the urban ecology class included students inGrades 9–12.cThe biology and marine biology classes were conducted in schools in suburban/town settings. The urban ecology class was in a school in an urban setting.
FIGURE 3
The Ocean Tracks interface also allows users to display data overlays showing sea surface chlorophyll (shown) and temperature. Track points arecolored yellow along the regions of the track that correspond to the date range of the overlay to aid users in linking monthly conditions represented bythe data overlay to track points occurring during the same period.
July/August 2015 Volume 49 Number 4 69
claim based on data measures or visu-alizations and presented reasoning forhow data measures supported thatclaim. The following quote is a highschool student’s response to a questionposed in theOcean Tracks draft curric-ulum about where the prey of elephantseals may be located, illustrating thetypes of quantitative and evidence-based reasoning that the program canencourage and support.
There is a major difference in sealbehavior between the monthsof July and August. In July, itappears that the seal was passingthrough a rather nonproductivepart of the ocean, due to a curvi-ness factor of 1.35, and an aver-age speed of 2.4 km/h and anaverage depth of −525.6 m. Thefact that the curviness factor is solow and the speed so high showsthat the seals were travelingstraight through that part of theocean, like a tourist on a highwaytraveling between monuments
and attractions…. In August,however, the seals’ curviness factorrose drastically to 2.7 while speeddropped to 1.23 km/hr…the av-erage depth was also greater thanthe average depth in July indicat-ing that the seal was diving moreoften to greater depths, most likelyas a result of the presence of morefood. Based on this data, it is evi-dent that the location of the seal’sprey is predominantly locatedin the region of seal tracks duringthe month of August. Data fromthe chlorophyll distribution mapfurther supports this claim be-cause chlorophyll concentrationwas 2 mg/m3 during the monthof August, whereas during Julythis concentration was around0.3 mg/m3, showing that the sec-tion of the ocean in the August lo-cation was much more productive.
In another exercise, students wereasked to identify biological hotspotsin the Pacific Ocean and to examine
the data to generate ideas about theoceanographic processes that createdthese hotspots. (Hotspots are areas ona map that have a particularly hightrack-point density.) To identify bio-logical hotspots, students use a cus-tomized tool in the Ocean Tracksinterface that identifies areas of thehighest track-point density for thetracks displayed on the map. Studentsthen use map overlays to identify link-ages between habitat usage and ocean-ographic parameters (Figure 5).
We found evidence that OceanTracks is an inherently interestinglearning environment by nature ofthe data to which students are given ac-cess. Of the 117 responses received tothe open-ended survey question,“What did you like most aboutOcean Tracks?,” 87 students (74%)said that the animals and their migra-tion tracks were the most interestingaspects of the Ocean Tracks interface.Teachers identified that students wereengaged by the fact that they wereanalyzing “real” data.
Student focus group interviewsa 3 (8–13 students each) 4 (9–17 students each)
Student screencast recordingsb 15 22
Student work assignmentsc – 92 collected23 analyzed
Classroom observationsd 30 19
Teacher logse 29 28aStudent surveys were collected, and focus group interviews were conducted at the end of each pilot test.bStudent screencast recordings were collected primarily during the initial weeks of each pilot test. Each recording captured the screen activities and vocalizations ofindividual or paired students as they navigated theOcean Tracks interface and conducted a data investigation during a class period. During Pilot 2, one ormore recordingswere collected for seven individual students and two paired students over a period of 4 weeks.cDuring Pilot 2, 92 samples of student work were collected, and 23 were randomly selected for analysis. We sampled approximately 25% of all work assignmentssubmitted by three teachers; for a fourth teacher, we sampled three of eight (or 38%) of submitted assignments.dClassroom observations were conducted by multiple project team members and produced written field notes following a common observation protocol. Each separateset of field notes was counted as an observation.eTeachers were asked to reflect on the pilot activities by completing an initial log, weekly interim logs during Ocean Tracks implementation, and a final, summative log.
70 Marine Technology Society Journal
And it was real. That made adifference. The video we had thekids watch that showed howthe scientists tag the animalskind of brought a little bit morefor them that it wasn’t just a com-puter game. That really madea difference. (teacher interviewresponse)
One aspect of the interface thatproved to be particularly interestingto students was the human impactsoverlay (Halpern et al., 2008), whichvisually depicts an index calculatedfrom data on a variety of human ac-
tivities that impact the marine envi-ronment, including fishing, climatechange, and pollution. Thinkingabout how the habitats of these marineanimals are impacted by certainhuman activities allowed students tomake connections to their own lives,bringing real-world relevancy to theirwork with Ocean Tracks.
I liked the human impact tabbecause it was interesting to seewhere in the ocean was impactedthe most by us humans becausethe ocean, especially the PacificOcean, is very large and vast. I
would be interest[ed] to know howhas the human impact changedfrom 10–20 years ago to now?And how has that affected the ani-mals? (student survey response)
While students demonstrated suc-cessful use of many interface featuresand tools, our analyses also identifiedchallenges that students faced whileworking withOcean Tracks. For exam-ple, we documented in our screencastrecordings 63 instances, lasting fromseveral seconds to a few minutes,when students displayed confusionover use of an interface feature or
FIGURE 4
Percentage of student work assignments with evidence of specific data skills, Ocean Tracks Pilot 2 (total N = 23).
July/August 2015 Volume 49 Number 4 71
tool. In 23 of those instances (37%),students found a specific track func-tion unclear, such as identifying trackstart and end dates and matchingtracks on the track selector to thosedisplayed on the map. In nine in-stances (14%), slow screen responsive-ness due to data-intensive loadingtimes led to student confusion aboutwhether the interface had registeredtheir mouse commands. We also ob-served six instances when overlappingwindows on the screen made it diffi-cult for students to see the data visual-izations on the Ocean Tracks map orother underlying data views. Datafrom classroom observations indicatethat some students struggled withunderstanding the relationship be-tween different animal tracks in time,reflecting the complexity of the tem-poral and spatial nature of the data athand. Based on analyses of studentwork assignments and as shown in Fig-
ure 4, most students also had troublewith the quantitative aspects of theiranalyses. They typically presentedqualitative descriptions rather thanquantitative comparisons of data mea-sures. Future interface and curriculumdevelopment efforts will focus on ad-dressing these challenges.
Ocean Tracks:College Edition
AlthoughOcean Tracks was initiallydesigned and tested for high schoolclassrooms, the NSF’s recent fundingof Ocean Tracks: College Edition(OTCE) shows recognition that thedata interface and learning resourcewe have developed have clear potentialfor widespread use in undergraduatesettings. Education research has dem-onstrated that undergraduate studentsare not developing robust skills in basicscientific and engineering practices
(NRC, 2012b), skills that are key toinnovation, discovery, and problemsolving. Teaching students to use, an-alyze, and understand data to explorescientific questions is an essential com-ponent of basic undergraduate scienceeducation (Manduca & Mogk, 2002),and yet for many students whoseundergraduate science courses consistprimarily of lectures, reading, andnote taking, their success in the courseis assessed using standardized tests thatemphasize memorization rather thananalytical thinking. These approachesare far less effective at motivating stu-dents and at developing skills in thepractices of science and engineering(PCAST, 2012). Recently, however,undergraduate institutions have begunto adopt approaches to teaching thatare grounded in research on learning(NRC, 2015), and there is burgeoninginterest in curricula and tools that sup-port more inquiry-based approaches.
FIGURE 5
An example of student work showing how the student was able to use the hotspot tool to identify a habitat hotspot for the bluefin tuna along the coastof California and Baja California (left). The student then used an overlay of chlorophyll (right) to investigate the underlying oceanographic process thatcreated this hotspot. The quote demonstrates how the student was able to integrate their conceptual understanding of upwelling with their interpre-tation of the data.
The chlorophyll levels in this area where the hotspot is are very high…which makes it a very attractive spot for these animals.This hotspot is pretty much right on and right next to the continental shelf, which is a place in the ocean where large amounts ofupwelling occur. Also the temperature by the coast is leaning towards the colder side. It stays around 12–16 degrees Celsius,which means since it's colder water there is more upwelling. (high school student)
72 Marine Technology Society Journal
Emerging interest in college-levelcurricula and tools that help teachdata skills, promising results from thehigh school classroom testing, andstrong indications of interest fromundergraduate faculty led to an NSFaward to EDC and the Scripps Institu-tion of Oceanography in the fall of2014 to develop and test learningmodules to support the use of theOcean Tracks interactive map anddata analysis tools in college class-rooms. The 3-year study involves test-ing in both community college anduniversity settings and consists of base-line data collection and needs assess-ment, curriculum development andresearch, and a pilot study. Researchefforts will support curriculum devel-opment and investigate how studentsfrom diverse institutions engage withonline and face-to-face versions of thecurriculum materials as well as howthese materials improve their under-standing of and competence in theuse of scientific practices, knowledgeof core content, and interest in pursu-ing careers in science.
Two distinct college-level courseswill serve as the test-beds for curri-culum development for the OTCEproject. The first is an online intro-ductory oceanography course taughtat Palomar Community College inSan Marcos, California. The secondis an upper division undergraduatecourse taught at University of Cali-fornia San Diego’s Scripps Institutionof Oceanography, titled “CaliforniaCoastal Oceanography.” Throughan iterative curriculum design andresearch process, we will work collabo-ratively with our partner faculty mem-bers to explore questions including“What supports may be needed toincorporate programs such as OceanTracks into undergraduate sciencecourses?” and “To what extent do
course experiences working withCLIP marine data through OceanTracks improve undergraduate stu-dents’ demonstration of scientificpractices with data and interest inscientific careers, particularly in themarine sciences?”
Phase 1 of the OTCE project con-sists of collecting baseline data andconducting a needs assessment tobetter understand students’ prior ex-perience with data and the ways inwhich oceanography and marine biol-ogy faculty use CLIP data sets in theircourses. Preliminary findings fromstudent surveys follow.
Student SurveyPreliminary Results
As part of the OTCE Phase 1 needsassessment study, we issued a survey(see http://tinyurl.com/ot-ce-survey)to 101 students taking an introductoryoceanography course at PalomarCollege and 181 students taking anintroductory course at Scripps on theEarth’s water to gauge their■ beliefs about the importance of
being able to make sense of data;■ prior experience and confidence
levels working with data, data visu-alizations, and data analysis tools;
■ interests and motivations related tothe oceans and marine life; and
■ academic and career interests (andtheir connections to data).The survey response rates for stu-
dents at Palomar and Scripps were73% and 77%, respectively. Resultsfrom this survey are being used to in-form curriculum development, help-ing us to identify data analysis skillsfor which students might need themost structured supports, and contentareas that will be most engaging andmeaningful for students.
Just under two thirds (62%) oftotal survey respondents from both in-
stitutions (n = 214) were female, andone third (33%) said that they knowthey are interested in pursuing a careerin science. To explain why they weretaking a course on the oceans, vast ma-jorities of survey respondents agreed orstrongly agreed that they have a generalcuriosity about the oceans (89%), theyfeel that ocean health and conservationare important to them (81%), and theyare concerned about the effects of cli-mate change on the oceans (80%).Similarly, high proportions of surveyrespondents also agreed or stronglyagreed that they believe it is importantto know how to make sense of data toget a good job (86%), to be successfulin whatever career they choose (86%),and to be a more effective and in-formed citizen (91%).
ConclusionsNew technologies deployed in the
oceans are generating unprecedentedquantities of data, illuminating pre-viously unknown environments, andenabling the exploration of wholenew types of questions. This same rev-olution is happening beyond therealms of marine science in diverse ap-plications ranging from law enforce-ment to agriculture. To realize thefull potential of big data to answerquestions, improve processes, andsolve problems, our schools should bepreparing students with a new set ofworkforce skills. Furthermore, beyondtomorrow’s workforce, it is imperativethat all citizens are able to make in-formed, evidence-based decisions in abig data world.
The Ocean Tracks project was de-veloped to find ways to bring the po-tential benefits of new tools of oceanexploration to a much broader audi-ence. Preliminary results from ourstudies with high school students
July/August 2015 Volume 49 Number 4 73
suggest that this approach showspromise in promoting student skillsin scientific practices with CLIP dataand interest in data and/or marine sci-ence, and our early work with under-graduate college students suggests anequally high level of interest and re-ceptiveness to this approach. Lookingforward, the goal is to determine howto build on this early success and usethe Ocean Tracks platform, as well asthe principles on which it was devel-oped, to help improve the use of datain science education more broadly.
Our experiences over the past5 years on four NSF-funded projectshave convinced us that there is tremen-dous untapped potential in the largescientific data sets that are increasinglyavailable to anyone with an Internetconnection. We have also learnedthat there are many challenges asso-ciated with making these data truly ac-cessible to novice users. We need tocontinue to find ways to make a varietyof data accessible via novice-friendlyinterfaces and data analysis tools. Weneed to better understand the skillsnecessary to prepare students for abig data world, and we need to developeffective ways to teach these skills overthe course of a student’s schooling.Weneed to develop new types of trainingprograms for teachers. Tackling thesechallenges via Ocean Tracks has re-quired the active, day-to-day engage-ment of a highly multidisciplinaryteam that includes high school andundergraduate educators, curriculumdevelopers, education researchers,marine biologists, and technology de-velopers. Just as our collaboration hasengaged those with diverse expertiseand perspectives, we believe raisinga generation equipped to unlock thepotential of big data will require thesustained and coordinated effortsof those with expertise in education,
science, and technology, compellingnew types of conversations and col-laborations that cross disciplinaryboundaries.
Corresponding Author:Randall E. KochevarEducation Development Center, Inc.43 Foundry AvenueWaltham, MA 02453Email: [email protected]
