tag:theconversation.com,2011:/global/topics/data-visualization-40900/articlesData visualization – The Conversation2021-03-31T05:27:54Ztag:theconversation.com,2011:article/1575032021-03-31T05:27:54Z2021-03-31T05:27:54ZWhere is Australia at with the COVID vaccine rollout? Our interactive shows how we compare with the rest of the world<figure><img src="https://images.theconversation.com/files/392731/original/file-20210331-15-p4l02h.png?ixlib=rb-1.1.0&rect=0%2C7%2C2396%2C1340&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">
</span> <span class="attribution"><span class="source">Shutterstock</span></span></figcaption></figure><p>As COVID vaccination programs roll out around the world, our new interactive tool allows you to see where each country is at with its immunisation program. Using data from <a href="https://ourworldindata.org/covid-vaccinations">Our World in Data</a>, you can see how each country is performing when it comes to:</p>
<ul>
<li><p>total COVID-19 vaccinations administered given so far</p></li>
<li><p>total COVID-19 vaccinations administered per 100 people.</p></li>
</ul>
<p>Due to some limitations with the <a href="https://covid.ourworldindata.org/data/owid-covid-data.csv">source data</a>, not all countries are shown. Our World in Data explains how they collect their data, which is being used by <a href="https://twitter.com/MaxCRoser/status/1369401412715679753">the WHO</a>, over <a href="https://ourworldindata.org/covid-vaccinations#source-information-country-by-country">here</a>. We’ve shown data as of March 28 below because that’s the latest data set available that’s relatively complete for a good range of countries.</p>
<p>Australia has administered about 600,000 doses at the time of writing (although, as shown on the interactive, it was 510,000 on March 28). That is a long way short of the <a href="https://www.theguardian.com/australia-news/2021/mar/31/missed-it-by-that-much-australia-falls-34m-doses-short-of-4m-vaccination-target-by-end-of-march">target</a> set by the prime minister to administer 4 million doses by the end of March.</p>
<p><iframe id="tc-infographic-575" class="tc-infographic" height="850px" src="https://cdn.theconversation.com/infographics/575/bc7dfdbf25654d74e1242aefa1e1696e266375a1/site/index.html" width="100%" style="border: none" frameborder="0"></iframe></p>
<figure>
<figcaption>Data visualisation: Kaho Cheung. Data source: Our World in Data https://ourworldindata.org. Australian data via https://covidlive.com.au/. </figcaption>
</figure>
<p>The Conversation asked Hassan Vally, an epidemiologist and infectious diseases expert, to reflect on what the data reveal at date of this article’s publication. </p>
<p>Here’s what he told us:</p>
<hr>
<p>This visualisation provides a useful and easy way to track where countries are at in their COVID vaccine rollouts. </p>
<p>We are in the midst of one of the largest logistical exercises in world history and it’s easy to drown in the sheer magnitude of the numbers.</p>
<p>The numbers are astounding even though we have only just begun. As of March 31, we have delivered about <a href="https://www.bloomberg.com/graphics/covid-vaccine-tracker-global-distribution/">574 million doses of the vaccine across 141 countries</a> across the globe, reaching 3.7% of the population.</p>
<h2>Total doses administered</h2>
<p>The United States has done extremely well in terms of the total number of doses given. It has clearly administered the most doses so far in total.</p>
<p>President Joe Biden <a href="https://www.whitehouse.gov/briefing-room/speeches-remarks/2021/03/18/remarks-by-president-biden-on-the-100-million-shot-goal/">promised</a> to deliver 100 million doses of vaccine by his 100th day in office, which was achieved by March 19, <a href="https://www.bloomberg.com/news/articles/2021-03-18/biden-says-u-s-to-meet-100-million-shot-goal-six-weeks-early">approximately six weeks ahead</a> of schedule. Regardless of how hard a marker you are, given the obstacles in the US, this is an impressive achievement.</p>
<p>China comes in second, having recently passed a significant milestone of delivering more than <a href="http://english.www.gov.cn/news/topnews/202103/28/content_WS60607883c6d0719374afba53.html">100 million doses</a> of the vaccine. The numbers of people currently being vaccinated each day are extraordinary, with goals to increase delivery to more than <a href="https://www.scmp.com/news/china/science/article/3127166/coronavirus-china-piles-pressure-get-vaccinated-country-targets">10 million doses</a> per day and immunise <a href="https://time.com/5943700/china-vaccinate-population/">40%</a> of the population by June.</p>
<p>India comes in next at around 61 million doses, which is impressive given the unique challenges it faces. </p>
<p>The United Kingdom is also doing well on total number of doses given. After these countries, you have Brazil a fair way back at about 18 million doses at the time of writing, and then a pack of countries further back.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/vital-signs-israel-shows-how-to-do-vaccinations-right-its-a-race-and-were-behind-157242">Vital Signs: Israel shows how to do vaccinations right. It's a race, and we're behind</a>
</strong>
</em>
</p>
<hr>
<h2>Total doses given per 100 people</h2>
<p>A different picture emerges when you adjust the data for the population sizes of countries by looking at the number of doses given <em>per 100 people</em>.</p>
<p>As has been well publicised, <a href="https://theconversation.com/vital-signs-israel-shows-how-to-do-vaccinations-right-its-a-race-and-were-behind-157242">Israel</a> leads the pack by a fair margin.</p>
<p>United Arab Emirates is doing really, as is Chile, the UK, Bahrain and the US.</p>
<p>Interestingly, the large countries such as China, India and the US that have delivered large total amounts of vaccine fall back from the lead when you adjust for population size.</p>
<p>Taking into account the size of countries to better assess the progress of the vaccination programs highlights the benefit of having a smaller population size and small geographical size.</p>
<h2>Wealthier countries are ahead</h2>
<p>This interactive tool also highlights that wealthier countries are generally ahead of poorer countries when it comes to the vaccination rollout. Unless addressed, that’s going to be a significant problem. Until the pandemic is stamped out everywhere, huge risks remain for all countries. </p>
<p>Global equitable access to vaccines is the right thing to do. But it’s also in the interests of rich nations, too. As the WHO says:</p>
<blockquote>
<p>With a fast-moving pandemic, no one is safe, unless everyone is safe.</p>
</blockquote>
<p>It’s fantastic to see Australia playing its role as a good regional citizen by providing <a href="https://theconversation.com/australia-wants-to-send-1-million-vaccine-doses-to-png-but-without-reliable-electricity-how-will-they-be-kept-cold-156798">vaccines and assistance to PNG</a> to help deal with their serious COVID situation. One could argue however, that we could and should be doing more.</p>
<hr>
<p>
<em>
<strong>
Read more:
<a href="https://theconversation.com/australia-wants-to-send-1-million-vaccine-doses-to-png-but-without-reliable-electricity-how-will-they-be-kept-cold-156798">Australia wants to send 1 million vaccine doses to PNG – but without reliable electricity, how will they be kept cold?</a>
</strong>
</em>
</p>
<hr>
<h2>How’s Australia doing compared with the rest of the world?</h2>
<p>At first glance, Australia looks to be tracking poorly compared to the rest of the world. </p>
<p>However, context is really important. We need to remember Australia is virtually COVID-free, making it the envy of the world. </p>
<p>We have access to two good vaccines suitable for all age groups, and the immunisation program has begun.</p>
<p>Although things have started slowly and <a href="https://theconversation.com/australias-covid-vaccine-rollout-is-well-behind-schedule-but-dont-panic-157048">we are behind</a> where we would like to be, our <a href="https://theconversation.com/as-australias-covid-vaccine-rollout-splutters-we-need-transparency-about-when-international-borders-might-reopen-157399">slow start</a> will likely ramp up significantly in the coming weeks. </p>
<p>It is significant that we have now entered phase 1b of the rollout, which means many millions more are now eligible to get the vaccine. We have also now started onshore vaccine production, which ensures vaccine supplies into the future.</p>
<p>Yes, there have been frustrations. But unlike many places, Australia has the luxury of time to carefully and safely deliver the vaccines due to our excellent performance so far in containing the spread of COVID-19.</p><img src="https://counter.theconversation.com/content/157503/count.gif" alt="The Conversation" width="1" height="1" />
At first glance, Australia looks to be tracking poorly compared to the rest of the world. However, context is really important here.Sunanda Creagh, Senior EditorLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1413482020-07-24T12:22:49Z2020-07-24T12:22:49Z3 questions to ask yourself next time you see a graph, chart or map<figure><img src="https://images.theconversation.com/files/349233/original/file-20200723-23-1c9tv31.jpg?ixlib=rb-1.1.0&rect=167%2C5%2C3464%2C2454&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">White House Coronavirus Task Force members reference a misleading chart in a press briefing.</span> <span class="attribution"><a class="source" href="http://www.apimages.com/metadata/Index/Virus-Outbreak-Trump/f2c5f8d116a24062b563a32cea88235e/1/0">AP Photo/Alex Brandon</a></span></figcaption></figure><p>Since the days of painting on cave walls, people have been representing information through figures and images. Nowadays, data visualization experts know that <a href="https://www.forbes.com/sites/evamurray/2019/01/28/how-data-visualization-supports-communication">presenting information visually</a> <a href="https://ed.ted.com/lessons/david-mccandless-the-beauty-of-data-visualization">helps people better understand</a> <a href="https://doi.org/10.1016/j.envsoft.2010.12.006">complicated data</a>. The problem is that data visualizations can also leave you with the wrong idea – whether the images are sloppily made or intentionally misleading. </p>
<p>Take for example the bar graph presented at an <a href="https://www.c-span.org/video/?470990-1/president-trump-coronavirus-task-force-briefing">April 6 press briefing</a> by members of the White House Coronavirus Task Force. It’s titled “COVID-19 testing in the U.S.” and illustrates almost 2 million coronavirus tests completed up to that point. President Trump used the graph to support his assertion that testing was “<a href="https://www.whitehouse.gov/briefings-statements/remarks-president-trump-vice-president-pence-members-coronavirus-task-force-press-briefing-21/">going up at a rapid rate</a>.” Based on this graphic many viewers likely took away the same conclusion – but it is incorrect.</p>
<p>The graph shows the total cumulative number of tests performed over months, not the number of new tests each day.</p>
<p><iframe id="pG025" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/pG025/2/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>When you graph the number of new tests by date, you can see the number of COVID-19 tests performed between March and April did increase through time, but not rapidly. This instance is one of many when important information was not properly understood or well communicated.</p>
<p>As a <a href="https://scholar.google.com/citations?user=T7vRKkQAAAAJ&hl=en">researcher of hazard and risk communication</a>, I think a lot about how people interpret the charts, graphs <a href="https://doi.org/10.1016/j.ijdrr.2020.101487">and maps</a> they encounter daily.</p>
<p>Whether they show COVID-19 cases, global warming trends, high-risk tsunami zones, or utility usage, being able to correctly assess and interpret figures allows you to make informed decisions. Unfortunately, not all figures are created equal.</p>
<p>If you can spot a figure’s pitfalls you can avoid the bad ones. Consider the following three key questions the next time you see a graph, map or other data visual so you can confidently decide what to do with that new nugget of information.</p>
<h2>What is this figure trying to tell me?</h2>
<p>Start by reading the title, looking at the labels and checking the caption. If these are not available – be very wary. Labels will be on the horizontal and vertical axes on graphs or in a legend on maps. People often overlook them, but this information is crucial for putting everything you see in the visualization into context.</p>
<p>Look at the units of measure – are they in days or years, Celsius or Fahrenheit, counts, age, or what? Are they evenly spaced along the axis? Many of the recent COVID-19 cumulative case graphs use a logarithmic scale, where the the intervals along the vertical axis are not equally spaced. <a href="https://www.youtube.com/watch?v=O-3Mlj3MQ_Q">This creates confusion for people</a> unfamiliar with this format.</p>
<iframe width="100%" height="315" src="https://www.msnbc.com/msnbc/embedded-video/mmvo80534597724" scrolling="no" frameborder="0" allowfullscreen=""></iframe>
<figure><figcaption><span class="caption">A March 12 broadcast of ‘The Rachel Maddow Show’ included a graph with unlabeled numbers and a tricky horizontal axis.</span></figcaption></figure>
<p>For instance, a graph from “<a href="https://www.msnbc.com/rachel-maddow/watch/u-s-unprepared-for-expected-explosion-in-coronavirus-cases-80534597724?cid=sm_fb_maddow">The Rachel Maddow Show” on MSNBC</a>, showed coronavirus cases in the United States between Jan. 21 and March 11. The x-axis units on the horizontal are time (in a month-day format) and the y-axis units on the vertical are presumably cumulative case counts, though it does not specify.</p>
<p>The main issue with this graph is that the time periods between consecutive dates are uneven.</p>
<p><iframe id="yzUp1" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/yzUp1/5/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>In a revised graph, with dates properly spaced through time, and coronavirus diagnoses plotted as a line graph, you can see more clearly what <a href="https://theconversation.com/coronavirus-cases-are-growing-exponentially-heres-what-that-means-135181">exponential growth</a> in the rate of infection really looks like. It took the first 30 days to add 33 cases, but only the last four to add 584 cases.</p>
<p>What may seem like a slight difference could help people understand how quickly exponential growth can go sky high and maybe change how they perceive the importance of curbing it.</p>
<h2>How are color, shape, size and perspective used?</h2>
<p><a href="https://eos.org/features/visualizing-science-how-color-determines-what-we-see">Color plays an important role</a> in how people interpret information. Color choices can make you notice particular patterns or draw your eye to certain aspects of a graphic.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/349252/original/file-20200723-23-wgpj48.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/349252/original/file-20200723-23-wgpj48.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=196&fit=crop&dpr=1 600w, https://images.theconversation.com/files/349252/original/file-20200723-23-wgpj48.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=196&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/349252/original/file-20200723-23-wgpj48.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=196&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/349252/original/file-20200723-23-wgpj48.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=246&fit=crop&dpr=1 754w, https://images.theconversation.com/files/349252/original/file-20200723-23-wgpj48.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=246&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/349252/original/file-20200723-23-wgpj48.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=246&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Oregon landslide susceptibility.</span>
<span class="attribution"><span class="source">Oregon Department of Geology and Mineral Industries</span></span>
</figcaption>
</figure>
<p>Consider two maps depicting landslide susceptibility, which are exactly the same except for reversed color schemes. Your eye may be be drawn to darker shades, intuitively seeing those areas as at higher risk. After looking at the legend, which color order do you think best represents the information? By paying attention to <a href="https://www.khanacademy.org/humanities/hass-storytelling/storytelling-pixar-in-a-box/ah-piab-visual-language/v/color-visual">how color is used</a>, you can better understand how it influences what stands out to you and what you perceive.</p>
<p>Shape, size and orientation of features can also influence <a href="https://doi.org/10.1111/j.1756-8765.2011.01150.x">how you interpret a figure</a>. </p>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="confusing pie chart of employment data" src="https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=608&fit=crop&dpr=1 600w, https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=608&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=608&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=764&fit=crop&dpr=1 754w, https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=764&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/348978/original/file-20200722-32-o99maq.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=764&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">What industries employ Coloradans?</span>
<span class="attribution"><a class="source" href="https://dossier.ink-live.com/html5/reader/production/default.aspx?pubname=&edid=5f3a495a-fdef-463f-b826-6b92609f04c5">Hemispheres</a></span>
</figcaption>
</figure>
<p>Pie charts, like this one showing employment breakdown for a region, are notoriously difficult to parse. Notice how hard it is to pull out which employment category is highest or how they rank. The pie chart’s wedges are not organized by size, there are too many categories (11!), the 3D perspective distorts the wedge sizes, and some wedges are separate from others making size comparisons almost impossible.</p>
<p><iframe id="yCDTo" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/yCDTo/2/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>A bar chart is a better option for an informative display and helps show which industries people are employed in.</p>
<h2>Where do the data come from?</h2>
<figure class="align-right zoomable">
<a href="https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="screen shot of Twitter poll about Trump's performance" src="https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=361&fit=crop&dpr=1 600w, https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=361&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=361&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=453&fit=crop&dpr=1 754w, https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=453&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/345601/original/file-20200703-33935-elrvg1.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=453&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Survey posted on ‘Lou Dobbs Tonight,’ requesting viewers vote on Twitter about Trump’s performance.</span>
<span class="attribution"><a class="source" href="https://www.mediaite.com/tv/lou-dobbs-invites-viewers-to-vote-on-trumps-coronavirus-leadership-superb-great-or-very-good/">Fox Business Network</a></span>
</figcaption>
</figure>
<p>The source of data matters in terms of quality and reliability. This is especially true for partisan or politicized data. If the data are collected from a group that isn’t a good approximation of the population as a whole, then it may be biased.</p>
<p>For example, on March 18, Fox Business Network host Lou Dobbs polled his audience with the question “How would you grade President Trump’s leadership in the nation’s fight against the Wuhan Virus?” </p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"1240421216692961284"}"></div></p>
<p>Imagine if only Republicans were asked this question and how the results would compare if only Democrats were asked. In this case, respondents were part of a self-selecting group who already chose to watch Dobbs’ show. The poll can only tell you about that group’s opinions, not people in the U.S. generally, for instance.</p>
<p>[<em>Get facts about coronavirus and the latest research.</em> <a href="https://theconversation.com/us/newsletters/the-daily-3?utm_source=TCUS&utm_medium=inline-link&utm_campaign=newsletter-text&utm_content=coronavirus-facts">Sign up for The Conversation’s newsletter.</a>]</p>
<p>Then consider that Dobbs provided only positive responses in his multiple choice options – “superb, great or very good” – and it is clear that this data has a bias.</p>
<p>Spotting bias and improper data collection methods allows you to decide which information is trustworthy. </p>
<h2>Think through what you see</h2>
<p>During this pandemic, information is emerging hour by hour. Media consumers are inundated with facts, charts, graphs and maps every day. If you can take a moment to ask yourself a few questions about what you see in these data visualizations, you may walk away with a completely different conclusion than you might have had at first glance.</p><img src="https://counter.theconversation.com/content/141348/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Carson MacPherson-Krutsky receives funding from The National Science Foundation. She is the co-owner of HazardReady, LLC. </span></em></p>Visualizations can help you understand data better – but they can also confuse or mislead. Here, some tips on what to watch out for.Carson MacPherson-Krutsky, PhD Candidate in Geosciences, Boise State UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/1089842019-01-02T23:35:09Z2019-01-02T23:35:09ZQuantifying the Holocaust: Measuring murder rates during the Nazi genocide<figure><img src="https://images.theconversation.com/files/252031/original/file-20181227-47298-16ylpjh.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">A concentration camp in Poland.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/concentration-camp-poland-1155091006?src=hledUMN2FMmlzs7ySp-ARQ-1-26">AkzuzkA/shutterstock.com</a></span></figcaption></figure><p>Even though the Holocaust is one of the best documented genocides in a historical sense, there is surprisingly little <a href="https://encyclopedia.ushmm.org/content/en/article/documenting-numbers-of-victims-of-the-holocaust-and-nazi-persecution">quantitative data</a>
<a href="https://www.sapiens.org/archaeology/the-darkest-truths/">available</a>, even on major critical events.</p>
<p>What’s more, <a href="https://encyclopedia.ushmm.org/content/en/article/introduction-to-the-holocaust">this history is often</a> told in figures too large to comprehend on the human scale. Large numbers – like the infamous 6 million people murdered – obscure the significance of key operations that shaped this genocide, leaving instead just a vague characterization of a massively devastating event. </p>
<p>In a digital age, mathematics, data science and visualization can help make sense of these events for new generations. By examining a rare and neglected dataset of human train deportations from the period, <a href="http://advances.sciencemag.org/content/5/1/eaau7292">my study</a>, published on Jan. 2, begins to uncover the true scale of slaughter. </p>
<h2>Operation Reinhard</h2>
<p>My work investigates a period in 1942, referred to as <a href="https://en.wikipedia.org/wiki/Operation_Reinhard">Operation Reinhard</a>, when the Nazis efficiently shuttled about 1.7 million victims, often whole Jewish communities, across the European railway network in train carriages to Treblinka, Belzec and Sobibor. Almost all of those who arrived at these death camps were murdered, usually within hours, in the gas chambers. Because the Nazis destroyed nearly all records of the massacre, it is important to try to uncover what actually happened at the time.</p>
<p>My study looks at the “kill rate,” or murders per day. This reveals a sudden massive slaughter after Hitler “ordered all action speeded up,” <a href="https://www.popularmechanics.com/military/a25725350/nazi-infrastructure-holocaust-lewi-stone/">as one SS officer put it</a>, on July 23, 1942. Approximately 1.5 million Jews were murdered in only 100 days, including in shootings outside the death camps, with nearly 500,000 victims killed each month during August, September and October. That’s approximately 15,000 murders every day. </p>
<p>The slaughter then soon terminated, as there were hardly any Jews remaining in the area to kill.</p>
<p><iframe id="tedcz" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/tedcz/2/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<p>The full scope of this genocidal slaughter appears to be undocumented in history. Available information before this study was mostly reconstructed indirectly, partially conjectured, and usually given on an annual timescale, rather than daily or monthly. That meant completely missing the three-month slaughter.</p>
<p>My analysis was based on carefully compiled train records presented <a href="https://www.nytimes.com/1987/06/28/books/the-impossible-and-unspeakable.html">in a 1987 book</a> by Holocaust historian <a href="https://en.wikipedia.org/wiki/Yitzhak_Arad">Yitzhak Arad</a>. Arad documents approximately 500 transportations from some 400 different Polish Jewish communities, recording for individual days the location, number of victims of each transportation and final death camp destination. </p>
<p>My analysis required carefully sorting and working with the dataset, as well as including other surviving data. In addition, I generated a spatio-temporal map and film of the data. These visualizations plotted the 400 communities on a map of Poland and indicated the time sequence of all deportations to the death camps over the whole year 1942.</p>
<p>While Operation Reinhard is considered the largest single murder campaign of the Holocaust, the extraordinary speed at which it operated to obliterate the Jewish people has been poorly estimated in the past and almost completely unknown to the general public. This massacre of unparalleled scale took place in just three short months, and was only captured through analysis of Arad’s dataset.</p>
<p>This minimal time indicates the enormous coordination involved by a state machinery responsive to the Fuhrer’s murderous will to eradicate a people. The train records show how zones were emptied of Jewish communities one by one in an organized manner and how intense kill rates were achieved in targeted areas that only slowed as victims ran out. My plots of the data highlight the pace and frenzy of this mass murder. </p>
<p><iframe id="6NwF0" class="tc-infographic-datawrapper" src="https://datawrapper.dwcdn.net/6NwF0/1/" height="400px" width="100%" style="border: none" frameborder="0"></iframe></p>
<h2>Measuring genocide</h2>
<p>Despite more than 70 years of research into the Holocaust, this appears to be the first attempt to graph aggregated data of the genocide, chronologically and spatially. My data-driven approach captures Operation Reinhard in a different perspective to the volumes of historical reports.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/WvqTgRWXaDY?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
</figure>
<p>Genocide scholars often compare rates of recent genocides to the rate at which the Nazi Holocaust occurred, treating the latter as a kind of benchmark for genocide severity. As such, currently many <a href="https://www.vqronline.org/genocide-rwanda">social</a> <a href="https://www.theatlantic.com/magazine/archive/2001/09/bystanders-to-genocide/304571/">scientists</a> maintain that the Rwandan genocide was the most “intense genocide” of the 20th century, with a sustained period of murders occurring at a rate three to five times more rapid than the Holocaust. </p>
<p>However, my work shows that while the Rwanda massacre killings were 8,000 victims per day for a 100-day period, the Holocaust was nearly double this rate during a similar 100-day period in Operation Reinhard.</p>
<p>That suggests that Holocaust kill rate has been underestimated on an order of six to 10 times. In my view, these sorts of comparisons have limited usefulness, and clearly diminish the Holocaust’s historical standing. </p>
<p>The Holocaust stands out as a demonstration of how the efficient machinery of government was turned on people in an unparalleled way. It transcended in its ruthlessness and systemic efficiency. This is the key lesson of the Holocaust that I believe must not be forgotten.</p><img src="https://counter.theconversation.com/content/108984/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Lewi Stone receives funding from Australian Research Council grant DP150102472</span></em></p>Even though the Holocaust is one of the best documented genocides in history, there’s limited quantitative data available. A new study investigates the number of deaths per day.Lewi Stone, Professor of Biomathematics at Tel Aviv University and, RMIT UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/882032017-12-05T04:06:56Z2017-12-05T04:06:56ZTurning hurricanes into music: Can listening to storms help us understand them better?<figure><img src="https://images.theconversation.com/files/197002/original/file-20171129-12069-kj3zgg.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Hurricane Maria, September 2017.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/hurricane-maria-makes-landfall-puerto-rica-718981030?src=UnFs_IIyu5aiiwbe6g7UHw-1-6">lavizzara/shutterstock.com</a></span></figcaption></figure><p>During the 2017 hurricane season, major storms in the North Atlantic devastated communities in and around Houston, Florida, Puerto Rico and the wider Caribbean. </p>
<p>The destruction shows how important it is to understand and communicate the serious threats that these storms pose. Scientists have made great strides in forecasting many aspects of storms, but if the people at risk don’t understand the danger they’re in, then the impact is lost. </p>
<p>We are colleagues from different areas of the Penn State campus: One of us is a professor of meteorology, and the other a professor of music technology. Since 2014, we have been working together to sonify the dynamics of tropical storms. In other words, we turn environmental data into music.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/JKqaqndHu04?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">Hurricane Sandy, sonified.</span></figcaption>
</figure>
<p>By sonifying <a href="https://www.youtube.com/watch?v=gpr0jreqVbU&feature=youtu.be">satellite videos</a> like those often seen in weather reports,
we hope that people will better understand how these extreme storms evolve. </p>
<h2>Data into sound</h2>
<p>Most of us are familiar with data visualization: charts, graphs, maps and animations that represent complex series of numbers. Sonification is an emerging field that creates graphs with sound. </p>
<p>As a simple example, a sonified graph might consist of a rising and falling melody, instead of a rising and falling line on a page.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/59AwYIP6q_E?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">A simple example of sonification.</span></figcaption>
</figure>
<p>Sonification offers a few benefits over traditional data visualization. One is accessibility: People with visual or cognitive disabilities may be <a href="https://www.ted.com/talks/wanda_diaz_merced_how_a_blind_astronomer_found_a_way_to_hear_the_stars">better able to engage with sound-based media</a>.</p>
<p>Sonification is also good for discovery. Our eyes are good at detecting <a href="http://www.ed.gov.nl.ca/edu/k12/curriculum/guides/art/art1201/sectn1.pdf">static properties</a>, like color, size and texture. But our ears are better at sensing <a href="http://deutsch.ucsd.edu/psychology/pages.php?i=601">properties that change and fluctuate</a>. Qualities such as pitch or rhythm may change very subtly, but still be sensed quite easily. The ears are also better than the eyes at following multiple patterns simultaneously, which is what we do when we appreciate the interlocking parts in a complex piece of music.</p>
<p>Sound is also processed <a href="https://www.bloomsbury.com/us/the-universal-sense-9781608198849/">more quickly</a> and <a href="https://www.penguinrandomhouse.com/books/298964/this-is-your-brain-on-music-by-daniel-j-levitin/9780452288522/">more viscerally</a> than visuals. That’s why we involuntarily tap our feet and sing along to a favorite song. </p>
<h2>Turning storms into songs</h2>
<p>A hurricane lifetime can last anywhere from a day to a few weeks. Agencies such as the U.S. National Oceanic and Atmospheric Administration continuously measure all sorts of features of a storm. </p>
<p>We distilled the changing characteristics of a hurricane into four features measured every six hours: air pressure, latitude, longitude and asymmetry, a measure of the pattern of the winds blowing around the storm’s center.</p>
<p>To create the sonifications, we export these data into the music synthesis program <a href="http://supercollider.github.io/">SuperCollider</a>. Here, numerical values can be scaled and transposed as necessary so that, for example, a storm lasting several days can be played over just a few minutes or seconds. </p>
<p>Each type of data is then treated like a part in a musical score. Data are used to “play” synthesized instruments that have been created to make sounds suggestive of a storm and to blend well together. </p>
<p>In our recordings, air pressure is conveyed by a swirling, windy sound reflecting pressure changes. More intense hurricanes have lower values of air pressure at sea level. The winds near the ground are also stronger in intense storms.</p>
<p>As pressure lowers, the speed of the swirling in our sonic recordings increases, the volume increases and the windy sound becomes brighter.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/RiLgaQRXqSs?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">This demonstration (not based on actual data) gives the sound that would result from pressure values decreasing and then increasing again.</span></figcaption>
</figure>
<p>The longitude of the storm center is reflected in stereo pan, the position of a sound source between the left and right speaker channels.</p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/dZt0jbKN-4A?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">The demonstration (not based on actual data) plays longitude positions moving from west to east (left to right). (This is best heard over stereo headphones.)</span></figcaption>
</figure>
<p>Latitude is reflected in the pitch of the swirling sound, as well as in a higher, pulsing sound. As a storm moves away from the equator toward one of the poles, the pitch drops to reflect the drop in temperatures outside the tropics. </p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/zL2Ku1xe9k8?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">This is a demonstration (not based on actual data) of latitudes tracking away from the equator and then back toward it. Although there are a very few exceptions, storms typically do not move back toward the equator.</span></figcaption>
</figure>
<p>A more circular storm is typically more intense. Symmetry values are reflected in the brightness of a low, underlying sound. When the storm has an oblong or oval shape, the sound is brighter. </p>
<figure>
<iframe width="440" height="260" src="https://www.youtube.com/embed/Y6S2QmlokH4?wmode=transparent&start=0" frameborder="0" allowfullscreen=""></iframe>
<figcaption><span class="caption">This demonstration plays values that outline the life cycle of a storm, evolving from an oval shape to becoming more circular, then returning to an oval shape. This progression reflects what would happen when a weak storm forms, becomes stronger, then dies.</span></figcaption>
</figure>
<h2>Using sound</h2>
<p>So far, we have sonified 11 storms, as well as mapped <a href="https://www.youtube.com/watch?v=TKTLE1rRUDA&feature=youtu.be">global storm activity from the year 2005</a>.</p>
<p>Storm sonifications could potentially benefit those who are tracking storm systems or updating the public about weather activity. Sonifications could be played over the radio, for example. They might also be useful for people who have limited phone bandwidth and are better able to receive audio content than video content.</p>
<p>Even for experts in meteorology, it can be easier to get a sense of interrelated storm dynamics by hearing them as simultaneous musical parts than by relying on graphics alone. For example, while a storm’s shape is typically tied to air pressure, there are times when storms change shape without changing in air pressure. While this difference can be difficult to see in a visual graph, it’s easily heard in the sonified data.</p>
<p>Our goal is to introduce sonifications of all kinds of graphs into science classes, particularly those with younger students. Sonification is becoming an <a href="http://digitalcommons.unl.edu/psychfacpub/444/">acknowledged research method</a>, and <a href="http://sonification.de/handbook/">several studies</a> have proven it effective at communicating complex data. But its uptake has been slow. </p>
<p>Nationwide, scientists, teachers and school administrators are recognizing the importance of <a href="https://www.usnews.com/news/stem-solutions/articles/2014/02/13/gaining-steam-teaching-science-though-art">the arts</a>, <a href="https://www.huffingtonpost.com/nicholas-ferroni/music-in-the-classroom_b_2072777.html">including sound and music</a>, when teaching science and mathematics. If a generation of students grows up experiencing science through more of their senses – sight, hearing and touch – then they may find the sciences more inviting and less intimidating.</p><img src="https://counter.theconversation.com/content/88203/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Mark Ballora receives funding from NAKFI (National Academies Keck Futures Initiative). Two seed grants for sonification work do be done in the area of oceanography. Not directly connected with the work described here, </span></em></p><p class="fine-print"><em><span>Jenni Evans receives funding from the National Science Foundation. She is the president-elect of the American Meteorological Society.</span></em></p>A meteorologist and a music technologist team up to turn the data from tropical storms into musical graphs.Mark Ballora, Professor of Music Technology, Penn StateJenni Evans, Professor of Meteorology, Penn StateLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/811592017-07-20T06:18:25Z2017-07-20T06:18:25ZHow open data can help the Global South, from disaster relief to voter turnout<figure><img src="https://images.theconversation.com/files/178913/original/file-20170719-14920-ceu2p1.png?ixlib=rb-1.1.0&rect=0%2C0%2C1441%2C1003&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">Display of Colombia's main export countries on the "Globe of Economic Complexity" application provided by The Center for International Development (CID), Harvard University
</span> <span class="attribution"><a class="source" href="http://globe.cid.harvard.edu/?mode=gridSphere&id=CO#">CID, Harvard University</a>, <a class="license" href="http://creativecommons.org/licenses/by-sa/4.0/">CC BY-SA</a></span></figcaption></figure><p>The modern era is marked by growing faith in the power of data. “Big data”, “open data”, and “evidence-based decision-making” have become buzzwords, <a href="http://www.undatarevolution.org/">touted</a> as solutions to the world’s most complex and persistent problems, from <a href="https://ssir.org/articles/entry/data_disrupts_corruption">corruption</a> and <a href="https://www.tableau.com/about/blog/2017/3/fighting-famine-mobile-data-67499">famine</a> to the <a href="http://www.computerworld.com/article/3027117/big-data/big-datas-big-role-in-humanitarian-aid.html">refugee crisis</a>.</p>
<p>While perhaps most pronounced in higher income countries, this trend is now emerging globally. In Africa, Latin America, Asia and beyond, hopes are high that access to data can help developing economies by increasing transparency, fostering <a href="http://africaopendata.net/">sustainable development</a>, building climate resiliency and the like.</p>
<p>This is an exciting prospect, but can opening up data actually make a difference in people’s lives? </p>
<h2>Getting data-driven about data</h2>
<p><a href="thegovlab.org">The GovLab</a> at <a href="http://engineering.nyu.edu/">New York University</a> spent the last year trying to answer that question. </p>
<p>In partnership with the <a href="https://www.usaid.gov/GlobalDevLab">U.S. Agency for International Development (USAID)</a>, the non-profit <a href="https://www.fhi360.org/projects/mobile-solutions-technical-assistance-and-research-mstar">FHI 360</a> and the <a href="http://webfoundation.org/">World Wide Web Foundation</a>, we scoured the evidence about what roles open data, particularly government data, can play in developing countries.</p>
<p>The results of our 12 in-depth case studies from around the world are now out. The report <a href="http://odimpact.org/">Open Data in Developing Economies: Toward Building an Evidence Base on What Works and How</a> offers a hard look at the results of open data projects from the developing world. </p>
<p>Our conclusion: the enthusiasm is justified – as long as it’s tempered with a good measure of realism, too. Here are our six major takeaways:</p>
<p><strong>1. We need a framework</strong> - Overall, there is still little evidence to substantiate the enthusiastic claims that open data can foment sustainable development and transform governance. That’s not surprising given the early stage of most open data initiatives.</p>
<p>It may be early for impact evaluation, but it’s not too soon to develop a model that will eventually allow us to assess the impact of opening up data over time. </p>
<p>To that end, the GovLab has created an evidence-based framework that aims to better capture the role of open data in developing countries. The Open Data Logic Framework below focuses on various points in the open data value cycle, from data supply to demand, use and impact.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=538&fit=crop&dpr=1 600w, https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=538&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=538&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=676&fit=crop&dpr=1 754w, https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=676&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/178875/original/file-20170719-26705-196u0lq.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=676&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">Logic model of open data.</span>
<span class="attribution"><span class="source">The GovLab</span></span>
</figcaption>
</figure>
<p><strong>2. Open data has real promise</strong> - Based on this framework and the underlying evidence that fed into it, we can guardedly conclude that open data does in fact spur development – but only under certain conditions and within the right supporting ecosystem. </p>
<p>One well-known success took place after <a href="http://odimpact.org/case-nepal-earthquake-recovery.html">Nepal’s 2015 earthquake</a> when open data helped NGOs map important landmarks such as health facilities and road networks, among other uses. </p>
<p>And in Colombia, the International Centre for Tropical Agriculture launched <a href="http://odimpact.org/case-aclimate-colombia.html">Aclímate Colombia</a>, a tool that gives smallholder farmers data-driven insight into planting strategies that makes them more resilient to climate change. </p>
<p><div data-react-class="InstagramEmbed" data-react-props="{"url":"https://www.instagram.com/p/BIpa7QjAgIu","accessToken":"127105130696839|b4b75090c9688d81dfd245afe6052f20"}"></div></p>
<p>Beyond these genuinely transformative experiences, we found several examples that ran into challenges. </p>
<p>A pair of education-information dashboards in <a href="http://odimpact.org/case-tanzanias-open-education-dashboards.html">Tanzania</a>, for example, were launched with good intentions (to improve student test scores by empowering families with information on school quality). But lacking long-term strategies to scale and sustain their use and impact, these efforts soon fizzled out. </p>
<p><strong>3. Open data can improve people’s lives</strong> Examining projects in a number of sectors critical to development, including health, humanitarian aid, agriculture, poverty alleviation, energy and education, we found four main ways that data can have an impact.</p>
<p>Open data can improve governance, as it did in Burundi when the country <a href="http://odimpact.org/case-burundi-open-rbf.html">made public</a> its results-based financing system. By linking development aid to pre-determined target results, this information increased transparency and accountability. </p>
<p><div data-react-class="Tweet" data-react-props="{"tweetId":"862616345141059584"}"></div></p>
<p>Data can also empower citizens by enabling more informed decision-making. For example, by providing information on voter registration centres, <a href="http://odimpact.org/case-kenya-improving-voter-turnout.html">Kenya’s GotToVote!</a> system increased voter awareness – and, consequently, turnout.</p>
<p>By enabling economic growth and innovation, data also has the power to create opportunities. In Ghana, the <a href="http://odimpact.org/case-ghanas-esoko.html">Esoko platform</a> is helping smallholder farmers maximise the value of their crops by providing useful information about increasingly complex global food chains.</p>
<p>Finally, data can assist governments, NGOs, and citizens in solving major problems. <a href="http://odimpact.org/case-paraguays-dengue-prediction.html">Dengue has been endemic since 2009 in Paraguay</a>. Recently, open data helped researchers develop a new tool for predicting outbreaks of the disease.</p>
<p><strong>4. Data can be an asset in development</strong> While these impacts are apparent in both developed and developing countries, we believe that open data can have a particularly powerful role in developing economies. </p>
<p>Where data is scarce, as it often is in poorer countries, open data can lead to an inherently more equitable and democratic distribution of information and knowledge. This, in turn, may activate a wider range of expertise to address complex problems; it’s what we in the field call “open innovation”. </p>
<p>This quality can allow resource-starved developing economies to access and leverage the best minds around. </p>
<p>And because trust in government is <a href="http://ida.worldbank.org/theme/governance-and-institutions">quite low</a> in many developing economies, the transparency bred of releasing data can have after-effects that go well beyond the immediate impact of the data itself.</p>
<figure class="align-center zoomable">
<a href="https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=45&auto=format&w=1000&fit=clip"><img alt="" src="https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=538&fit=crop&dpr=1 600w, https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=538&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=538&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=676&fit=crop&dpr=1 754w, https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=676&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/178873/original/file-20170719-11699-74exh4.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=676&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px"></a>
<figcaption>
<span class="caption">The unique features of open data.</span>
<span class="attribution"><span class="source">The GovLab</span>, <span class="license">Author provided</span></span>
</figcaption>
</figure>
<p><strong>5. The ingredients matter</strong> To better understand why some open data projects fail while others succeed, we created a “<a href="http://odimpact.org/periodic-table.html">periodic table” of open data</a> (below), which includes 27 enabling factors divided into five broad categories.</p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/178868/original/file-20170719-13593-17hxuph.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/178868/original/file-20170719-13593-17hxuph.png?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=538&fit=crop&dpr=1 600w, https://images.theconversation.com/files/178868/original/file-20170719-13593-17hxuph.png?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=538&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/178868/original/file-20170719-13593-17hxuph.png?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=538&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/178868/original/file-20170719-13593-17hxuph.png?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=676&fit=crop&dpr=1 754w, https://images.theconversation.com/files/178868/original/file-20170719-13593-17hxuph.png?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=676&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/178868/original/file-20170719-13593-17hxuph.png?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=676&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Periodic Table of Open Data.</span>
<span class="attribution"><span class="source">The GovLab</span></span>
</figcaption>
</figure>
<p>For those who can truly unlock the potential of open data – from practitioners to policy-makers – this table could serve as a checklist of the factors and ingredients to be considered and addressed. Researchers assessing the impact of an open data project can also use it to determine what variables made a difference.</p>
<p><strong>6. We can plan for impact</strong> Our report ends by identifying how development organisations can catalyse the release and use of open data to make a difference on the ground. </p>
<p>Recommendations include:</p>
<p>· Define the problem, understand the user, and be aware of local conditions;</p>
<p>· Focus on readiness, responsiveness and change management;</p>
<p>· Nurture an open data ecosystem through collaboration and partnerships;</p>
<p>· Have a risk mitigation strategy;</p>
<p>· Secure resources and focus on sustainability; and</p>
<p>· Build a strong evidence base and support more research.</p>
<h2>Next steps</h2>
<p>In short, while it may still be too early to fully capture open data’s as-of-yet muted impact on developing economies, there are certainly reasons for optimism. </p>
<p>Much like <a href="http://thegovlab.org/the-govlab-selected-readings-on-blockchain-technology-and-its-potential-for-transforming-governance/">blockchain</a>, <a href="https://theconversation.com/how-drones-can-improve-healthcare-delivery-in-developing-countries-49917">drones</a> and other much-hyped technical advances, it’s time to start substantiating the excitement over open data with real, hard evidence.</p>
<p>The next step is to get systematic, using the kind of analytical framework we present here to gain comparative and actionable insight into if, when and how open data works. Only by getting data-driven about open data can we help it live up to its potential.</p>
<p><em>This article was co-authored by <a href="http://www.thegovlab.org/team.html">Andrew Young</a>, Knowledge Director at the GovLab.</em></p><img src="https://counter.theconversation.com/content/81159/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>The GovLab received funding from the U.S. Agency for International Development and FHI 360 to complete the report featured in this article.</span></em></p>Can open data change the world? We looked beyond the hype to find out.Stefaan G. Verhulst, Co-Founder and Chief Research and Development Officer of the Governance Laboratory, New York UniversityLicensed as Creative Commons – attribution, no derivatives.tag:theconversation.com,2011:article/785702017-07-20T01:50:11Z2017-07-20T01:50:11ZThe Library of Congress opened its catalogs to the world. Here’s why it matters<figure><img src="https://images.theconversation.com/files/176956/original/file-20170705-28939-r7aodn.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=496&fit=clip" /><figcaption><span class="caption">The Library of Congress is in Washington, D.C.</span> <span class="attribution"><a class="source" href="https://www.shutterstock.com/image-photo/washington-dc-usa-august-5-2016-570867745?src=lGpFCvv5qT3HOH2r3LGHFA-1-44">Valerii Iavtushenko/Shutterstock.com</a></span></figcaption></figure><p>Imagine you wanted to find books or journal articles on a particular subject. Or find manuscripts by a particular author. Or locate serials, music or maps. You would use a library catalog that includes facts – like title, author, publication date, subject headings and genre. </p>
<p>That information and more is stored in the treasure trove of library catalogs. </p>
<p>It is hard to overstate how important this library catalog information is, particularly as the amount of information expands every day. With this information, scholars and librarians are able to find things in a predictable way. That’s because of the descriptive facts presented in a systematic way in catalog records.</p>
<p>But what if you could also experiment with the data in those records to explore other kinds of research questions – like trends in subject matter, semantics in titles or patterns in the geographic source of works on a given topic?</p>
<p>Now it is possible. The Library of Congress has made <a href="https://web.archive.org/web/20170525222956/https://www.loc.gov/item/prn-17-068/">25 million digital catalog records</a> available for anyone to use at no charge. The free data set includes records from 1968 to 2014.</p>
<p>This is the largest release of digital catalog records in history. These records are part of a data ecosystem that crosses decades and parallels the evolution of information technology. </p>
<p>In my <a href="http://quod.lib.umich.edu/c/crmstoolkit">research about copyright and library collections</a>, I rely on these kinds of records for information that can help determine the copyright status of works. The data in these records already are embodied in library catalogs. What’s new is the free accessibility of this organized data set for new kinds of inquiry.</p>
<p>The decision reflects a fresh attitude toward shared data by the Library of Congress. It is a symbolic and practical manifestation of the library’s leadership aligned with its mission of public service.</p>
<h2>Some history</h2>
<p>To understand the implications of this news, it helps to know a bit about the <a href="https://americanlibrariesmagazine.org/2016/01/04/cataloging-evolves/">history of library catalog records</a>. </p>
<p>Today, search engines let us easily find books we want to borrow from libraries or purchase from any number of sources. Not long ago, this would have seemed magical. Search engines use data about books – like the title, author, publisher, publication date and subject matter – to identify particular books. That descriptive information was gathered over the years in library catalog records by librarians. </p>
<figure class="align-center ">
<img alt="" src="https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&fit=clip" srcset="https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=400&fit=crop&dpr=1 600w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=400&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=400&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=502&fit=crop&dpr=1 754w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=502&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/176954/original/file-20170705-3057-15zfueh.jpg?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=502&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Card catalog at the Library of Congress.</span>
<span class="attribution"><a class="source" href="https://www.flickr.com/photos/rrenomeron/8493938344/">Rich Renomeron/flickr</a>, <a class="license" href="http://creativecommons.org/licenses/by-nc-nd/4.0/">CC BY-NC-ND</a></span>
</figcaption>
</figure>
<p>The library’s action sheds light on this unseen but critical network. This infrastructure is invisible to most of us as we use libraries, buy books or use search engines. </p>
<p>For many, the idea of a library catalog conjures up the image of card catalogs. The descriptions contained in catalog records are “metadata” – information about information. Early catalog records date back to 1791, just after the French Revolution. The revolutionary government used playing cards to document property seized from the church. The idea was to make a <a href="https://archive.is/20121215093454/http://gslis.simmons.edu/wikis/LIS415OL_History_Encyclopedia/Origins_of_the_Card_Catalog">national bibliography of library holdings</a> confiscated during the Revolution. </p>
<p>For many years, library collections were organized individually. As the number of books and libraries grew, the increased complexity demanded a more consistent approach. For example, when the <a href="https://www.loc.gov/loc/lcib/0806/jefferson.html">Library of Congress purchased Thomas Jefferson’s personal library</a> in 1815, it arranged its collections around Jefferson’s personal system organized around the themes of memory, reason and imagination. (Jefferson based this on <a href="http://www.loc.gov/exhibits/thomas-jeffersons-library/">Francis Bacon’s own model</a>.) The library sought to arrange its collections on that model into the 19th century. </p>
<figure class="align-left ">
<img alt="" src="https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=237&fit=clip" srcset="https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=600&h=450&fit=crop&dpr=1 600w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=30&auto=format&w=600&h=450&fit=crop&dpr=2 1200w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=15&auto=format&w=600&h=450&fit=crop&dpr=3 1800w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=45&auto=format&w=754&h=566&fit=crop&dpr=1 754w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=30&auto=format&w=754&h=566&fit=crop&dpr=2 1508w, https://images.theconversation.com/files/175715/original/file-20170626-3062-53ilmt.JPG?ixlib=rb-1.1.0&q=15&auto=format&w=754&h=566&fit=crop&dpr=3 2262w" sizes="(min-width: 1466px) 754px, (max-width: 599px) 100vw, (min-width: 600px) 600px, 237px">
<figcaption>
<span class="caption">Books on my shelf, marked with KF and HB. The K indicates that the book relates to law, the H that it relates to social science. The second letter indicates a subcategory.</span>
<span class="attribution"><span class="source">Melissa Levine</span>, <a class="license" href="http://creativecommons.org/licenses/by/4.0/">CC BY</a></span>
</figcaption>
</figure>
<p>As the number of books and libraries grew, a more systematic approach was needed. The Dewey Decimal System appeared in 1876 to tackle this challenge. It combined consistent numbers (“classes”) with particular topics. Each class can be further divided for more specific descriptions. </p>
<p>In the 1890s, the library developed the <a href="https://www.loc.gov/catdir/cpso/lcc.html">Library of Congress Classification System</a>. It is still used today to predictably manage millions of items in libraries worldwide.</p>
<h2>Catalogs, cards and computers</h2>
<p>By the 1960s, systematic descriptions made the transition from analog cards to online catalog systems a natural step. <a href="https://www.loc.gov/marc/umb/um01to06.html">Machine-Readable-Cataloging (or MARC) records</a> were developed to electronically read and interpret the data in bibliographic cataloging records. The structured categorization coincided naturally with the use of computers. </p>
<p>Now, <a href="http://lj.libraryjournal.com/2002/10/ljarchives/marc-must-die/#_">MARC records</a> too are on the way out, making room for <a href="https://www.loc.gov/bibframe/faqs/">more modern and flexible standards</a>. </p>
<p>The Library of Congress remains a primary – but not the only – source for catalog records. Individual libraries produce catalog records that are compiled and circulated through organizations like <a href="https://www.oclc.org/en/home.html">OCLC</a>. OCLC connects libraries around the globe and offers an online catalog. <a href="https://www.oclc.org/en/worldcat.html">WorldCat</a> coordinates catalog records from many libraries into a cohesive online resource. Groups like these charge libraries through membership fees for access to the compiled data. Libraries, though, typically do not charge for the catalog records they produce, instead working cooperatively through organizations like OCLC. This may evolve as more <a href="https://everybodyslibraries.com/2016/09/12/the-value-of-catalogs-in-the-linked-data-era-two-recent-talks/">shared effort and crowdsourced resources can be combined </a>with the library’s data in ways that improve search and inquiry. Examples include <a href="http://www.share-research.org/">SHARE</a> and <a href="https://www.wikipedia.org/">Wikipedia</a>. </p>
<h2>One month later</h2>
<p>In the short time since the Library of Congress’ data release, we see inklings of what may come. At a <a href="http://blogs.loc.gov/thesignal/2017/06/hack-to-learn-at-the-library-of-congress/?loclr=eadpb">Hack-to-Learn event</a> in May, researchers showed off early experiments with the data, including a <a href="https://medium.com/@thisismattmiller/library-of-congress-lists-57ddd177f1e2?loclr=blogsig">zoomable list of nine million unique titles</a> and <a href="http://www.vam.ac.uk/blog/digital-media/visualising-library-catalogues">a natural language interface with the data</a>. </p>
<p>For my part, I am considering how to use the library’s data to learn more about the history of publishing. For example, it might be possible to see if there are trends in dates of publication, locations of publishers and patterns in subject matter. It would be fruitful to correlate copyright information data retained by the U.S. Copyright Office to see if one could associate particular works with their copyright information like registration, renewal and ownership changes. However, those records remain in formats that remain difficult to search or manipulate. The <a href="https://www.copyright.gov/digitization/records.html">records prior to 1978</a> are not yet available online at all from the U.S. Copyright Office. </p>
<p><a href="https://www.lib.umich.edu/users/joque">Colleagues</a> at the University of Michigan Library are studying the recently released records as a way to practice map-making and explore geographic patterns with visualizations based on the data. They are thinking about gleaning locations from subject metadata and then mapping how those locations shift through time. </p>
<p>There’s a growing expectation that this kind of data should be freely available. This is evidenced by the expanding number of open data initiatives, from institutional repositories such as <a href="https://deepblue.lib.umich.edu/data/">Deep Blue Data</a> here at the University of Michigan Library to the U.S. government’s <a href="https://www.data.gov/open-gov/">data.gov</a>. The U.K.’s <a href="http://www.universitiesuk.ac.uk/policy-and-analysis/research-policy/open-science/Pages/open-research-data-task-force.aspx">Open Research Data Task Force</a> just released a <a href="http://www.universitiesuk.ac.uk/policy-and-analysis/research-policy/open-science/Documents/ORDTF%20report%20nr%201%20final%2030%2006%202017.pdf">report</a> discussing technical, infrastructure, policy and cultural matters to be addressed to support open data.</p>
<p>The Library of Congress’ action demonstrates an overarching shift in use of technology to meet historical research missions and advance beyond. Because the data are freely available, anyone can experiment with them.</p><img src="https://counter.theconversation.com/content/78570/count.gif" alt="The Conversation" width="1" height="1" />
<p class="fine-print"><em><span>Melissa Levine has received funding from the Institute for Museum and Library Services.</span></em></p>Catalog data are a library’s most important map to knowledge. What does it mean that
the Library of Congress just released 25 million records to the public?Melissa Levine, Lead Copyright Officer, Librarian, University of MichiganLicensed as Creative Commons – attribution, no derivatives.