Process

Chapter 2: Process

The project was borne of a combined interest in natural approaches to human health and wellbeing, philosophy, the natural environment, spirituality, and the perception of the human as either distinct from or connected with her environment, and the implications for this on lived human experience. Ecofeminism encompasses many of these interests, and was therefore considered to be a sound theoretical background on which to base this project. Researching correlations between human and Earth health became the focus, with the creation of a digital artefact to investigate the hypothesis:

Human disconnection from nature is a result of patriarchal “power over” structures of control and subordination which have caused parallel degradations of the natural environment and human health.

The idea behind the project was to create a digital artefact which would allow the interrogation of relevant data and therefore prove or disprove this hypothesis. The digital artefact would be a website, containing an article of data journalism using interactive (where possible) data visualisations which would be discussed and utilised to support the narrative. During the process of creating the website and the first data visualisations, it was decided that including a blog of as a part of the reflective and creative process would act as an integral part of the artefact, acting in a self-reflective way by allowing the recording of the process and any challenges. Including the blog section on the website seemed to be a more holistic approach, a reflection on a reflection, but also a means of exposing and interrogating the process itself. In the spirit of economy, this website utilises and includes all created data visualisations, even if they do not appear in the final piece of data journalism, thereby creating less scrap for the digital junk-heap, as all parts are creating the whole, and are on display. This ties with the academic themes behind the project, and as such reflects upon itself in a more holistic and interconnected sense. Many digital projects have a volume of work at the back end, that will never be seen by the end consumers of the product, so in the spirit of digital environmentalism, this website is as non-wasteful as possible.

Considering the definite need for a Content Management System (CMS), and exploring the possibilities for this, along with limitations and alternatives, and considering the implications of technical ability and a knowledge of the functionality of specific tools and how they may positively or negatively impact the project and the standard of the resulting product, WordPress via Reclaim Hosting was selected as the appropriate CMS for this project. WordPress is the market leader for website creation (“Usage Statistics and Market Share of WordPress for Websites, April 2017,” n.d.), and has been the primary digital tool for developing the digital artefact for this project, alongside data visualisation software and Microsoft Excel.

After first creating the website using WordPress.com, designing and customising the site, adding pages, content, and blog posts, it became apparent that the limitations of the free WordPress.com software would create problems for the standard of the finished digital product. After dealing with technical issues to do with embedding Tableau interactive data visualisations into blog posts, it seemed prudent to switch the website over to Reclaim Hosting, which does have a small fee attached, but for the difference in functionality provided, combined with the adaptability and freedom gained by using WordPress through Reclaim, it was a necessary decision for the quality of the finished artefact.

After consideration and exploration of possible CMS, WordPress seemed to be the most practical tool for creating and building the artefact. WordPress, Omeka, Scalar, Joomla, and Drupal were also considered as possibilities to manage the content, however, Omeka appeared to be more suited to an archival project with a volume of specific items to be presented, and as such did not fit the specific requirements for this artefact. Scalar offered an attractive interface and an easy to use experience with content linked externally rather than embedded. It would have been easy to link content internally and to present those connections with Scalar, however, for this project it was decided that the functions of Scalar weren’t quite appropriate. Joomla appeared to offer another, deeper level of functionality, but that gain in functionality needed to be balanced with ease of use and the complexity of the tool. After exploration, it became clear that Joomla offered the option to install a multitude of plugins which could drastically improve functionality, however, there could have been issues of compatibility between some of the plugins, and others were paywalled. It is this complexity which would have made the tool more difficult to use, which raised concerns in terms ability. If the extra functionality cannot be exploited to the benefit of the project, then using it would not have been a prudent choice. Overall, Joomla appeared slightly too complex for the needs of the project, considering both technical skill level and required functionality. Drupal was also considered, as another CMS choice. Drupal’s website was confusing and bulging with misleading sections which seemed irrelevant to this project and its needs. Drupal was intimidating as a CMS and seemed designed for a user with advanced technical capabilities. Building a website from scratch using HTML and CSS was also considered, but the requirements for this project in terms of aesthetics, functionality, time constraints, and technical speed and ability were important considerations which, on balance, made this choice a risk to the overall quality of the artefact.

 

After investing considerable time into creating visually impactful data visualisations in Tableau Public, Data Hero, and using the World Bank website data visualisation tool, it was too much of a sacrifice to abandon those creations and their functions due to the limitations of free WordPress. It is not possible to properly embed data visualisations as interactive into a free WordPress website, and even though the correct code was inserted into the HTML section, an inferior quality non-interactive image of the data visualisation with a ‘play’ arrow obscuring the image (because it is assumed to be video) was simply not an acceptable outcome, and did not provide the level of functionality required by this project.

Reflecting on how the level of technical ability of the user influenced technical choices for this project, and considering whether the goal should be self-challenge or a focus on working from a place of comfort is an interesting process. Frustrations that the end product would have been more engaging and multi-faceted had it been created using Joomla or Drupal due to the higher functionality were overcome by switching to WordPress via Reclaim, and combining ease of use with manipulability. However, after considering all the options and weighing the pros and cons with an awareness of user skill level, while being sensitive to the current needs of the project and how those needs could be best met by a CMS, WordPress rose to the top when considering that content management, ease of use, visual attractiveness and audience engagement were the key factors in this process.

 

During initial searches for data, it became clear that the best course of action to obtain reliable data to investigate the hypothesis would be to use the websites and databanks of large, well-respected, national, international or worldwide organisations. These data stores provide a wider scope within them for finding a range of relevant data, but their data is also compatible, comparable, and contains measures that are well-known, respected, and relevant. The data used for this project came from the OECD, the UN, and The World Bank. These websites had the facility to download large datasets in excel format, with the World Bank website being particularly helpful to the user by allowing very specific criteria to be entered into a search facility before download, saving time and space.

 

Selecting a tool to create engaging and informative data visualisations brought its own set of difficulties. Previous work with Tableau public meant that this functionally dense tool could be used without the need for learning the basics. Tableau Public provides a data cleaning facility, and produces high quality, visually impressive and interactive data visualisations which can be embedded into appropriate CMS. After two projects using this tool there are still many functions not utilised; the reality of using Tableau Public to its full capabilities went beyond the scope of this project. Data Hero also works as a user-friendly, helpful data visualisation tool, but is less functional than Tableau. There was no scope for manipulating the data once uploaded on either tool, except on very specific parts, like column titles or value definitions, so if the in-built data cleaner was unable to understand the data, it was necessary to go back, clean the data manually, and then re-upload to the tool. Data Hero is not as intelligent as Tableau Public, the tool automatically generates data visualisations based on assumptions about the data uploaded, but those assumptions are often incorrect, meaning irrelevant or confusing charts are created from small sections of data. While some visually compelling charts were created using Data Hero, the lack of an embed option is disappointing, and results in static data visualisations which can be downloaded, and a link provided, but they could not be interactive within another website, reducing the overall impact and user-experience.

 

While searching for relevant datasets, The World Bank was discovered; a free and open source worldwide data bank, providing access to a vast selection of world development data, which is quite comprehensive. This website also provides a user-friendly interface for creating data visualisations from the data provided. The tool works well with the provided data because they are already compatible (no data cleaning needed), easily searchable, filters can be added, many different datasets can be chosen at the same time and then selected specifics shown in the data visualisation window. Some visually impressive maps were created using this tool, which were eventually embedded into the artefact. There were technical issues with the website initially; there was a requirement to create an account to save or download data visualisations, but the server wasn’t working properly, and the website repeatedly crashed when trying to save the created visualisations, so some time and effort was lost. The embed codes were provided, but, like with Tableau Public embed codes, they also would not function correctly within the free WordPress software. These embedding issues were solved by transferring over to WordPress via Reclaim Hosting. The first website was created using free WordPress software, and can still be viewed at connectedenvironments.wordpress.com.  After switching to Reclaim Hosting, all the posts were migrated manually from the original site to www.interconnectedenvironments.bernadettesmart.com.continuousflux.com.

 

Deciding the measures on which to base the possible correlations between human and earth health, it was necessary to seek out data first, to find what was available, and whether it provided the scope required to interrogate the question. Many datasets are incomplete, and not all countries have data available for each year, so a balance between available data, appropriate measures, and the ability to see patterns or correlations within the data was a necessary consideration. There was a wealth of data available, too much to use everything that could be considered a good measure, so decisions had to be made. C02 appeared a good measure of environmental degradation, because it is considered a “key indicator” by the OECD (“Environment at a Glance, Statistics / Environment at a Glance / 2015,” n.d.) and because it demonstrates the level to which a country has reduced (or not) its carbon emissions. Looking at the overall C02 emissions is interesting, and shows the difference in volumes of production between different countries, but it does not break down the figures effectively. Per capita measurement is much more telling; obviously there are other factors involved, like industry, development, use of non-renewable sources of fuel, use of fossil fuels, farming etc, but overall this is good measure of the impact a country has on the environment in general.

Prevalence of overweight amongst national populations seemed an appropriate measure of population health, because it is not only possibly indicative of a divorce from natural health and nutritious foods, but also has wide-reaching health implications (Kopelman, 2007). When compared to data, there seemed to be a correlation between overweight people and C02 emissions, suggesting a connection between the health of a national population, and that nation’s treatment of the environment. Interestingly, when compared to data for non-communicable (more often lifestyle related) diseases, while countries with improved sanitation and water source and higher healthcare spending may have less deaths from communicable diseases, the deaths from non-communicable diseases (heart disease, diabetes, etc) are considerably higher. While it could be argued that these numbers are due to healthcare spending, the evidence suggests this is not the case.

Mortality rate; there are so many factors which contribute to this data, meaning this is not a straightforward measure of health. Some countries had very high mortality rates of over 600 per 1000 adults (“DataBank | The World Bank,” n.d.), so while at first mortality rate seemed to be a good measure of human health, on reflection, these deaths could be due to war, disease, famine, murder, and many other factors which could contribute to the figures, so it was not acceptable to assume the high numbers correlated with poor human health, although these numbers do reflect a poor quality of life (Smart, n.d.).

Considering the usage of fossil fuels compared to alternative and nuclear fuel usage as a measure; while these are certainly good measures of governmental environmental attitudes within a country, the data is weighted so that nuclear and alternative energy are inextricable, which presents ethical issues for this project. Nuclear energy cannot be considered in the interests of the health of any living creature or any part of nature; this means of energy production creates waste which is impossible to safely dispose of, and the consequences of any nuclear accidents are vast. Use of fossil fuels was an appropriate measure of environmental degradation, due to the disconnection from environmental awareness it represents, in terms of pollution, C02 emissions, and the use of finite natural resources.

Freshwater withdrawals were relevant particularly due to the Murrumbidgee River Basin study (Kandasamy et al., 2014), which exemplifies the issues with prioritising economic goals over environmental concerns. Within this data (“DataBank | The World Bank,” n.d.), Bahrain was shown to withdraw such an excess of water compared to the available natural resources of the country, that the data visualisations produced were not easy to read, which necessitated a rethink of the methods used to display this data to clarify the situation among the top users of freshwater resources.

Forestry usage; while initial thoughts were that this would be a good measure of environmental degradation, data was not available on replanting, or the impact of this practice on C02 emissions. The resulting data visualisations did not show a tangible measure of environmental degradation for the purposes needed, so this was rejected as a measure while accepting that felling trees for purely economic gains is clearly an anti-environmentalist practice.

Deaths from communicable diseases as a measure of human health; this connects with data on access to an improved water source and improved sanitation, while it may have been illuminating to use this data, as deaths from communicable diseases correlate historically with improvements in sanitation (Alemu, 2017). Also considering the implications of investment in low-efficacy for health outcomes (Westerhaus MD, n.d.) biomedical healthcare in countries with lower prevalence of improved sanitation would have been relevant, however, it seemed beyond the scope of this project to investigate this comprehensively.

The creation of data visualisations with a measure of human health compared to a measure of environmental degradation on one chart proved more challenging than expected. Different measures, and vastly different numbers within the measures made this awkward, and the resulting data visualisations did not have the visual impact or clear correlations that were needed to conclusively show patterns. The tools (or possibly the density of the data) were somewhat resistant to this, although Data Hero did allow the creation of a few passable charts with comparisons, there is still more work to be done on creating data visualisations with the visual impact required; clear correlations shown on charts seemed illusive. Tableau Public has the capacity to create dashboards and impressive visualisations, but they seemed consistently, frustratingly, just out of reach. After creating a few very good-looking data visualisations, it became clear that they weren’t really showing any correlations or comparisons as effectively as had been hoped, because there was only one dataset on each visualisation. The best method of direct comparison to view possible patterns within the data seemed to be to embed two versions of the same map visualisation one above the other into the blog section of the website, this allowed the direct comparison of data on the same maps, but was not a dynamic and impactful comparison which showed patterns clearly.

In creating the data journalism piece which would form a large part of the artefact, the most relevant data visualisations were selected and a framing, explanatory narrative written around them. The goal was to create an insightful analysis of the data visualisations through the narrative and thereby prove or disprove the hypothesis. The scope of the data was larger than expected, resulting in a piece that did not have the impact that had been hoped for. There is room for further study, given time, and the blog section could certainly be continued to further investigate more fully whether human health and the health of the planet correlate in a conclusive sense.