At the Nexus of Big Data and Dispute Resolution

A Case Study on Conflict-Related Search Data



Report Author




Justin R. Corbett   

Linkedin  |  Google  |  ORCID



Recommended
Citation



Corbett, J.R. (2016) At the Nexus of Big Data and Dispute Resolution: A Case Study on Conflict-Related Search Data. Mediation Theory and Practice, 1(1): 5-33. https://doi.org/10.1558/mtp.v1i1.29197


Report Extras



Journal Article: Mediation Theory and Practice

Podcast Interview: Earbud_U

Coverage: Solicitors Journal Article



Related Presentations 


[KEYNOTE]  Conflict in America: Conflict-Related Queries & the Search for Resolution

Texas Association of Mediators (TAM)  |  2015 Annual Conference  |  2015-02-20  |  San Antonio, TX



[KEYNOTE]  The Language of Conflict: Lessons from a Big Data Dive for Conflict Specialists

Conflict Resolution Minnesota (CRM)  |  2015 Annual Conference  |  2015-04-23  |  St. Paul, MN



[KEYNOTE]  In Search for Resolution: Finding Opportunity in Conflict-Related Queries

Virginia Mediation Network (VMN)  |  2015 Fall Conference  |  2015-10-17  |  Richmond, VA



[KEYNOTE]  Searching for and Selling Peace in an Adversarial World

Academy of Professional Family Mediators (APFM)  |  2015 Annual Conference  |  2015-10-23  |  Reston, VA



In Search for Resolution: Finding Opportunity in Conflict-Related Queries

New York State Dispute Resolution Association (NYSDRA)  |  2015 Annual Conference  |  2015-09-29  |  Saratoga Springs, NY



Trends in Conflict Throughout the United States as Recorded in Online Search Queries

Cyberweek  |  2014-11-04  |  Webinar




Abstract

 

This paper presents findings from an ongoing project that has analysed over 225 million annual online conflict-related search queries to better understand the scope and scale of conflict in America. The frequency, location, sentimentality, and timing of 52,203 unique conflict search terms were monitored for every US County. Analysis has revealed the public’s collective interests and conflict intensities within 225 distinct conflict contexts, many of which have cyclical trends that can inform resource development and deployment efforts. It has allowed for the creation of maps detailing where conflict strongly arrests or strangely avoids local populations. Finally, it has revealed new insights into how the public prioritises frames for their conflicts and their preferred resolutions. Creative integration of traditional and previously unaligned data, such as presented here, will change the dispute resolution field with regard to its understanding of the dispute resolution marketplace, the surfacing of opportunities to satisfy that market, and the ability to monitor the resulting impact of interventions from the personal to broad societal levels.




Dispute Resolution Big Data

 

Data informs the professional climate in which we live. Big data, with a scale only newly and still nascently capable of penetrating analyses, has the potential to inform that climate with consequence beyond current comprehension (McKinsey Global Institute 2011). Its quantified nature increasingly pervades our existence, influences our behaviour, and contours our position and possibility (Economist 2010; Pew Research Center 2012). As with the natural climate, it often does this largely unnoticed and without the critical reflection its impact deserves. Instead, our psychology and our social condition favour the anecdote of our artificially conditioned familiarity (Plous 1993). As professionals, we often measure our success and failures based on limited understandings of narrowly aligned competition and inarticulately defined local markets (Welch, Immelt, Dammerman and Wright 2001; Gaglio and Katz 2001). We design our practice in response to a perception of and preference for what occurs within our conceptual walls, neglecting the irrefutable and not inconsequential influence of the broader climate which swirls about outside; attending to our programmatic and practice thermostats without regard for the seasons or storm fronts chronicled and foretold by data.

 

This myopia to data and its influence is decreasingly sustainable. In our age of technological innovation wherein garage-based start-ups can upend billion dollar industries with near predictable regularity (Marchetti and Jazbec 2015), the field of alternative dispute resolution (ADR) in its distributed coordination and paltry resources (Corbett and Corbett 2011) is poorly protected from or positioned to respond to disruptive moonshots across our bow from savvy new entrants or non-traditional competitors alike. Indeed, it seems ADR’s only meaningful insulation from such disruption is its relative practice obscurity (McGillis 1997) and the rurality of its traditional revenue prospects (Velikonja 2009); protective shoals away from which the field desperately continues its decades-long navigation. As professional opportunists-for-hire for even the direst of contexts, the prospect of a sea changing tipping point – one which finally lets loose our alternative anchor – should be entirely thrilling. As skilled seers of realities not yet realised, it is increasingly easy to envision how the nexus of big data and dispute resolution may well be the channel through which our next major advancements originate.

 

As we fling wide the doors and windows of our practices to assess the broader climate, opportunities to incorporate existing and emerging data into our decision-making processes abound. Traditional sources, such as client evaluations, satisfaction surveys and common outcome metrics are not only far easier to gather at the programme and practitioner levels, but are now also easily anonymised and aggregated across professional contexts. A typical practitioner’s small-data spreadsheet approach to client feedback can now be compared to thousands of online reviews for practitioners across the street or around globe; providing the inquisitive insights into which skills to turn or which competitive strengths to trumpet. A court-sponsored mediation programme can now assess its nuanced standing amongst thousands of aligned initiatives in far-flung jurisdictions to identify cohorts capable of lending hard-earned and hitherto narrowly distributed smart practices that help further its local impact. An enterprising startup can quickly assemble and analyse massive sets of dispute resolution contracts and case outcomes, becoming the go-to expert on clause drafting or predictive outcome analytics.

 

New and previously unaligned datasets may hold even greater impact and intrigue still. Targeted consumer research that allows us to understand the conflict experiences and dispute resolution preferences of representative samples of any population segment can now be assembled at scale. Various wearable technologies can now track our physiological responses to conflict and its resolution; creating a literal dashboard of parties’ palpitations to augment the field’s current fascination with brain science theoretics. Data from online search activity can now reveal the existence, frequency and trends of any conceivable conflict context from around the globe, producing the most detailed, real-time map of the dispute resolution marketplace as has ever been developed. Indeed, this latter project has now been undertaken and constitutes the following case study on the application and professional impact of big data on the field of dispute resolution.

 

In truth, the following is not the first exploration of this convergence (Moses and Chan 2014; Soltas and Stephens-Davidowitz 2015). Importantly, though, neither will it be its last. Data – large and otherwise – will only increase in utility. Those who contour this arc of our collective horizon are the thought and practice leaders of our emerging future.

 

Case Study: In Search of Resolution


Project Rationale

 

The decision to explore online search data as a way to map and deepen our understanding of the dispute resolution marketplace was an acknowledgement of the near ubiquity of search behaviour within the project’s initial target market. The inclination to type or speak into a growing network of connected devices in an attempt to locate relevant information and resources has become more commonplace than most futuristic projections could have even imagined. In the US market alone, recent figures estimate over one billion online searches are performed every day (comScore 2015; Singhal 2015). Caught within this staggering frequency is a mix of motivations and topics as diverse as the seekers themselves. For a rapidly expanding majority, our digital search histories have become the most consistent, intimate (Motorola 2015), and representative reflection of our experiences and our evolving selves as we had ever dreamed for our once shiny, uncracked primary school diaries. ‘I search, therefore I am’ (Hillis, Petit and Jarrett 2013) may well be one of our time’s most resonant unspoken mantras. In their vast aggregate, however, another subtle, yet considerable consequence has emerged: ‘they search, therefore we know’. And we know much from their searches.

 

Ad-supported search engines, in their efforts to ensure their online advertisers are equipped to realise successful campaigns, endow advertisers with a wealth of aggregated and anonymised attributes about our unending torrent of inquiries. First, we have access to the actual wording typed or spoken by the searcher. These words and phrases, in their natural language presentation, represent the most succinct distillation of what the searchers are experiencing or seeking in their moments of inquiry. In dispute resolution terms, they often represent dense and revealing opening statements. Take, for example, the search phrase: ‘Should I leave my husband if he cheats on me’. We can contextualise this search as concerning a spousal relationship. We can identify cheating as the current central focus of the searcher’s ongoing conflict. We can appreciate the inquiry’s frame as a request for advice. We can infer there may be communication difficulties within the relationship which have resulted in an online search in replacement of – or at least as a supplement to – a direct discussion with the spouse. Finally, we can begin to identify which dispute resolution procedures may be most appropriate for addressing these presenting issues. From each unique inquiry – particularly wordy, long-tail searches – we are often able to identify and infer a revealing amount of information simply through analysing the inquirers’ chosen words.

 

Search engines offer access to even more data about these searches still. We can also know when and where these searches occurred, through what types of devices they were facilitated, and the likely responses of the searchers when viewing query results, to name just a few. In terms of search timing, it is possible to capture aggregated metadata that is parsable on an annual, monthly, weekly, daily and even hourly basis. Location data are available for over 200 countries at varying levels of granularity, which in some locations can be as specific as a few square blocks.

 

As an example of the implications of this attached metadata, at the level of this project’s current analysis – monthly queries in each of the 3,143 US counties and county-equivalent geographies – we are now able to watch online searches related to domestic violence ebb and flow as seasonal changes push the country colder and many victims into longer, in-home contact with their offenders. At the level of specificity for which search metadata are currently available, however, that same analysis can move from charting seasonal trends to witnessing storm fronts pass through a city and watching the resulting real-time implications on local domestic violence. In this example, knowing the timing, location, and intensity of these trends has obvious implications for both law enforcement and victim service providers. It also presents an opportunity for the dispute resolution field to identify when and where our partnership with aligned professions may be of greatest impact or, when appropriate, both where our direct service provision is in greatest demand and the types of contexts we may encounter during our response.

 

Many additional examples of the implications of identifying and tracking search data on various conflict contexts is easy to envision. Judicial programmes at both the individual courthouse and centrally coordinated levels could now forecast filing trends by sussing out and monitoring those terms that most directly and predictively correlate with various complaint categories. Local schools can engage outside trainers and on-site contractors at the most opportune times in anticipation of cyclical spikes in bullying and student–teacher conflicts. Community mediation programmes can use data-supported trend analysis to realign programmatic priorities and limited resources according to shifting local conflict dynamics. Individual practitioners can more effectively maximise their online and offline advertising and outreach efforts to coincide with both regular and unexpected interest peaks within their respective practice areas.

 

As our collective record of online search queries becomes an increasingly nuanced and representative chronicle of our interests and needs, the consequence of mining its stores for insights and opportunities of both professional and social gain will only increase. For the dispute resolution field specifically, search data offers a map of unrivalled detail of both our current market and our path toward greater public resonance and utilisation.

 

Methodology

 

The preparatory design of this project required contouring a conceptual parameter for the varied contexts of conflict itself, and then engaging numerous search engine tools to source data on terms meaningfully connected to those contexts. To the first, it is important to note that this project did not seek to narrowly examine the contexts of conflict for which individuals may engage a search engine in their quests for assistance. Unlike other projects of varying rigor which have assembled disparate indicators of social strife or local context-specific conflict (Institute for Economics and Peace 2015; Gardeazabal 2012; Brück, Justino, Verwimp and Avdeenko 2010; Corbett 2015), the opening frame for this project assumed little in terms of the content of searchers’ conflict-related queries or their comfort in engaging a search engine to pro- or reactively address those matters. Instead, a broad parameter of conflict was established by first coupling contemporary trends in conflict research with the author’s own familiarity with the field’s dilating dimensions, and then programmatically propagating that schema using relational algorithms provided by the chosen search engine to capture both historic and emerging trends from the public’s actual search activity.

 

To begin, a set of broad conflict categories were identified to represent the numerous overlapping spheres of one’s life in which conflict could and often does manifest. Those spheres included: Community, Education, Family, Market, Self-help, and Social.

 

The Community category is intended to capture conflicts occurring within the traditional notion of one’s local, physical neighbourhood. These include the annoyance of seemingly unending dog barking, parking on another’s property, fence line disagreements, and housing matters including foreclosures and homeowner association regulations, to name a few. The Education category captures conflicts encountered by those in regular contact with a variety of learning environments, including youth and young adults in both formal institutions and less structured educational contexts; the individuals who provide the administration and instruction within those contexts; and their proximal stakeholders (e.g. school boards and parents). Example conflicts in this category would present as student bullying, student–teacher discord, and parent–teacher disagreements. The Family category includes matters involving those with whom one has either a familial or intimate relationship. Conflicts, here, would be parent–teen strife, tension related to elder caregiving, infidelity within a marriage or other significant relationship, and inheritance disputes. The Market category casts a wide net to capture economic, interpersonal, and inter-entity conflicts diversely situated within the marketplace, including business-to-business, business-to-consumer, and workplace interactions. Conflicts found in this category include troublesome co-workers and hostile working environments, legal-focused issues of employment discrimination, and customer complaints. The Self-Help category includes introspective attempts at bettering oneself, particularly in their management of troublesome emotions and the enhancement of personal conflict engagement and negotiation skills. The Social category includes one’s friends, roommates, and acquaintances. It represents the notion of a conceptual neighbourhood similar to the Community category’s aligned, but physically bound counterpart; social communities supporting one’s hobbies, interests, or ideological persuasions, but which are not persistently physically co-located. Examples here would be arguments with a best friend and rows within a social club. Finally, a broad category termed ‘ADR’ was subsequently added to capture and monitor terms which may not intuitively fit into one of the aforementioned categories, but which would likely be of direct interest to the ADR educators, service providers, trainers, and other thought leaders for whom the benefits of this project are intended.

 

Conflict, thus broadly categorised, will span an insight-blurring mix of eclectic contexts. Within the Family category alone, one could experience and search for conflicts related to custody, divorce, elder caregiving, infidelity, inheritance, paternity, and parenting, among many others. As such, these categories were further refined using traditional field research (Jehn, Northcraft and Neale 1999) that has tended to group conflict as either task/event, relational, or structural in nature. For example, within the Family category, again, one could have task or event conflicts related to an anniversary, dating, reunion, vacation, or wedding. Relational Family conflicts could involve a bride, brother, child, ex-spouse, or grandparent. Structural Family conflicts could be framed within adoption, extended family, or marriage contexts. Spread between these three groupings, a mind-mapping exercise intended to cursorily identify the range of possible Family category inclusions resulted in 117 separate groupings, each of which represented a zone of potential friction. Similar exercises were performed on each of the remaining broad categories, producing a total of 400 zones of potential friction across all of the aforementioned conflict spheres.



Visualisation 1. Outlining Conflict. Four hundred zones of potential friction were identified within the primary categories of conflict. These zones -- identified as task/event, structural, or relational in nature -- became the seeds of algorithmic inquiry; prompting search engine reporting tools to identify related queries and aligned zones for further investigation. Examples of some of the identified prospective education conflicts zones are identified above.


Next, these potential friction points were cross-tabulated with 128 conflict terms (e.g. ‘argument’, ‘bullying’, ‘conflict’, ‘dispute’, ‘fight’, and ‘mediation’) to produce pairings following a ‘conflict term’ + ‘zone of potential friction’ format (e.g. ‘contested’ + ‘divorce’). These pairings were then fed into a search engine keyword volume reporting tool to determine which pairings had ever been searched for online. (Google was chosen as the preferred search engine from which to source these data because of its command of roughly two-thirds of the US search volume market share, as well as the relative ease it provided for extracting the data required for this project.) Of the 51,200 combinations produced through simple cross-tabulation, only 2,732 produced a historic, reportable search volume. These resulting terms were then entered into an advertiser-oriented keyword planning tool designed to algorithmically identify related search terms based on searchers’ preceding and subsequent queries, as well as other significant measures weighted by the search engine (Brin and Page 1998). This step produced a set of 115,078 unique search terms identified as being substantially related to some conflict context.




Visualisation 2. Conflict Cross-Tabulation. To ensure the search engine reporting tools were limiting their results to conflict-relevant queries, each of the 400 zones of potential conflict were serially paired with 128 conflict terms, producing over 50,000 combinations. Each of those combinations were evaluated for possible search histories, resulting in over 2,700 unique terms with reportable histories. A separate tool was then used to identify search terms substantially related to that active subset. In total over 115,000 terms were algorithmically identified and then assessed for their conflict relevance.  


Unfortunately, upon review, many terms within this expanded list were insufficiently related to any particular conflict context as to justify their inclusion in the project’s final analysis. For example, the search term ‘online dictionary’ was identified by the search engine keyword planning tool as being substantially related to conflict, likely on the grounds that a statistically significant number of individuals who did search for a conflict-related term subsequently searched for ‘online dictionary’ in their attempt to better understand some of the obtuse nomenclature encountered on websites suggested within their initial search results.

 

To correct for this statistical blindness, the full list of over 115,000 suggested terms were manually reviewed by the researcher for the likelihood of a relationship to any conflict context conceivably connected to the identified categories or any other area for which dispute resolution services could be reasonably rendered. Dismantling and codifying this colossus of conflict resulted in a manually verified list of 52,203 unique conflict-related search terms for which historical search volume and related metadata were available. When accounting for all search providers and device platforms, these conflict-related search terms are projected to represent over 225 million annual queries; a deafening and unending cacophony of ‘HELP!’ occurring more than seven times a second (every second!), and quickening still even as you read.



Visualisation 3. U.S. Rate of Online Conflict Queries. Across all conflict categories, nearly a quarter billion online conflict-related queries occur annually; more than seven queries every second. Each query is a call for help, a search for assistance, and an opportunity to help another in need.








Visualisation 4. U.S. Rate of Online Conflict Queries by Category. 1. Community Category (roughly 22 searches per minute, nationwide); 2. Education Category (≈ 2 searches / minute); 3. Family Category (≈ 94 searches / minute); 4. Market Category (≈ 92 searches / minute); 5. Self-Help Category (≈ 20 searches / minute); 6. Social Category (≈ 34 searches / minute).



Visualisation 5. U.S. Rate of Online Family Conflict Queries by State. Each state has its own frequency of online conflict queries informed by its population, its policies toward such conflicts, and the presence of psycho-social dynamics that either encourage or discourage their propagation. Here, a real-time representation of family-oriented online conflict queries is presented based on the identified state-level frequencies.


These verified terms and their corresponding metadata and individual volumes were then coded into specific conflict contexts intended to reflect the diversity and nuance of situations possible within the aforementioned broad categories. The Market category, for example, included numerous distinct contexts, including many workplace-related Equal Employment Opportunity Commission (EEOC)/Equal Opportunities Commission (EOC) conflicts for which search terms were labelled as belonging to age, disability\medical, gender, parenting\pregnancy status, race, religion, or sexual orientation discrimination contexts.

 

In total, 225 distinct conflict contexts were identified and populated using the final set of unique terms. The creation of these distinct conflict contexts was influenced by classifications of conflict commonly seen throughout the legal and dispute resolution fields. It incorporated law enforcement incident reporting categorisations, case filing schemas employed by state Supreme Courts, and practice specialties as recorded by local bar associations and numerous ADR rosters. Each context monitors the reported activity of at least 50 and as many as 4,290 unique conflict-related search terms. Each term was assigned to a maximum of two specific conflict contexts. For example, within the broad Family category, the term ‘how to get a divorce’ was coded as a ‘Divorce’ context, while the term ‘how to get a divorce with a mediator’ was coded into both of the ‘Divorce’ and ‘Family ADR’ contexts. Collectively, these contexts represent the aggregated perturbation and pain of our population. From a professional perspective, they also represent the existing and emerging dispute resolution practice areas for which search volume – at least – suggests enormous opportunity.



Visualisation 6. Data Classifications. The amassed trove of conflict-related online query data can be numerously and insightfully parsed. Several of the top-level classifications include data categories/tags, contexts, unique terms, and annual query counts. For each of the over 52,000 terms, data on their monthly, County-level activity was captured and analysed for this project.  


With the data thus categorised, contextualised, and coupled with its related search volumes and metadata, it was then possible to begin a granular and groundbreaking analysis of conflict in America. In performing this analysis, certain data artefacts and limitations are acknowledged and represent opportunities for future refinements to the project’s evolving design, including:

 

  1. The review of only Google search data;
  2. The inclusion of only those terms which met the researcher-imposed 100 searches per year threshold;
  3. A further researcher-imposed limitation which resulted in the analysis of only English language search terms;
  4. The possible existence of repeat queries from a single individual – a possibility which resulted in the analysis focusing on a metric identified as local ‘conflict intensity’ rather than on a more traditional per capita basis; and
  5. A general acknowledgement that the continued existence of a digital divide will result in certain segments of the population not having their concerns chronicled in an analysis which exclusively relies on access to online technologies.

 

These acknowledgements notwithstanding, the project presents an unrivalled opportunity to begin harvesting insights from the fertile nexus of big data and dispute resolution.

 

Findings and Discussion

 

Conflict is a pernicious, profound and pervasive condition. Our online activities as we search our way through one of its iterations and into the next are responsively prolific. In chronicling this search activity and coupling it with its revealing metadata, we can construct the most detailed picture of conflict as has ever been compiled. This project analysed the frequency, location, sentimentality, and timing of 52,203 unique conflict-related search terms over time within each of the 3,143 US counties and county-equivalent geographies. These terms were manually coded into 225 distinct conflict contexts, allowing the aggregation of aligned terms without obfuscating contextually nuanced data insights from the smoothing effects of unrelated term inclusions.

 

The resulting dataset is capable of providing a chronological review of the collective interests and conflict intensities of hundreds of distinct conflict contexts, many of which have cyclical trends and/or predictive correlations that can now be used to inform future resource development and deployment efforts. It is capable of mapping where these conflicts occur; generating heatmaps of the physical locations where conflict strongly arrests or strangely avoids a local population. Through its unmoderated natural language reporting, these data are suited to sentiment analysis – allowing us to score each term’s positive or negative polarity, thereby furthering our understanding of how the public prioritises frames for their various conflicts. Finally, given the scale of available data, we can analyse any of these variables individually (e.g. the number of infidelity-related searches in the US) or in combination with any mix of others (e.g. chronological fluctuations of positively-framed infidelity-related queries from Harrison County, Iowa) without relinquishing the possibility of revealing meaningful and statistically significant insights.

 

Category, Context, and Term Prevalence

Coding each unique search term into a hierarchical classification involving one parent category and up to two distinct conflict contexts makes it possible to measure the diversity and combined volume of related search terms in a way which exposes the overall prevalence of various types of conflict. By assigning each of the 52,203 unique terms and their over 225 million annual collective searches into the 225 distinct conflict contexts we can contour the scale of and construct possible solutions to the rough edges of our communal, interpersonal, and economic existence.

 

Examining the sheer volume of searches performed within each category, Family-related queries were the most prevalent, representing 33 per cent of all US conflict-related search queries. Market-related conflict terms were nearly as prominently represented at 32 per cent of all queries. Each of the remaining broad categories commanded substantially smaller portions of the public’s collective search interests – and likely everyday conflict experiences – represented in descending order: Social 12 per cent, ADR 8 per cent, Community 8 per cent, Self-Help 7 per cent, and Education 1 per cent.



Visualisation 7. Prevalence of Conflict Queries by Category. If the nation's entire conflict-related search history was aggregated and broadly categorized, family and market-related conflicts would dominate nearly two-thirds of our online search for conflict assistance. From most to least prevalent, the categories are: Family (33%), Market (32%), Social (12%), ADR (8%), Community (8%), Self-Help (7%), and Education (1%).     


Combined, Family and Market-related terms account for two-thirds of the conflicts for which the US sought assistance. Anecdotally, this mirrors and supports the heightened focus afforded these categories within the current dispute resolution profession. In their scale, however, their combined 152 million annual searches suggest extensive opportunity exists to further attend to these conflict spheres occupying the central attention of so many.


Otherwise noteworthy is the apparent near exclusion of Education-related conflict queries within the broader dataset. This may be a product of two factors. First, bullying queries, which were substantial within the broader dataset, were coded into contexts under the Social category unless the term contained an explicit or inferred connection with an educational setting. As such, many school-related bullying queries may have been categorised as social in nature rather than educational. Second, it may be that as digital medium choices proliferate and preferences evolve, the majority of those individuals currently in an educational setting (i.e. school-aged youth and young adults) are decreasingly relying upon a relatively impersonal search engine when seeking assistance for their conflicts – turning instead to socially-aware or anonymity-focused mobile apps to broadcast their discontent.


Turning to specific contexts, we can further refine our understanding of the nuance that encompasses and escalates each conflict category. Peering into the Family category, we can explore the placement of each of the 16,285 unique Family-specific terms and their respective annual volumes. In doing so, we see divorce is the single most searched context, including 4,290 unique terms that collectively generate 16,217,200 annual queries – the equivalent of one divorce-related search every two seconds. Abuse, harassment, intimidation, and violence terms are the second most frequently searched (15,584,990 per year), followed in order by custody, guardianship, and visitation (11,790,830 per year); infidelity (11,033,460 per year); family legal information (10,014,730 per year); counselling and therapy (3,208,810 per year); inheritance and wills (2,994,760 per year); dating (1,581,660 per year); separation (1,338,730 per year); and parenting (1,222,500 per year), to round out the top 10 most frequent Family conflict contexts by annual volume.



Visualisation 8. Family Conflict Context Tree Map. More than 225 distinct conflict contexts were identified and populated as part of this project. Each of the more than 16,000 unique family-related conflict queries were manually assigned into up to two distinct conflict contexts. The above tree map identifies the prominence of the top 12 family contexts within the broad Family category. 


The top contexts in each of the broad categories represent areas of considerable activity within the dispute resolution field’s many corners. Terms within these and other contexts are almost always inquiries to which the dispute resolution profession is sufficiently – if not uniquely – positioned to respond. As such, it bears repeating that given the scale of annual searches catalogued, even the least populated context identified through this research often contained enough local volume (i.e. public interest) to substantially support the exploration – if not outright establishment – of specialty practices in many locations throughout the analysed geographies. Potential client interest and conflict intensities abound. The work to us now, as discussed more thoroughly later, is in crafting messaging and services responsive to those specific interests.

 

A third level of analysis was undertaken to examine the individual terms themselves, unpackaged from their respective contexts and categories. Here, we identified the specific terms of greatest interest to the broadest number of searchers. Unsurprisingly, these terms represented extremely short-tail laconic queries, some of which were likely applicable to a large number of potential contexts. ‘Bullying’, for example, was the most searched term within the entire dataset; matching – and likely circularly feeding into – its persistent prominence in media and so much of the public’s collective mindset. ‘Unemployment’, ‘domestic violence’, ‘divorce’, and ‘child abuse’ constituted the remaining top five most frequently searched terms throughout the US. By category, at least one ADR-related term was included in the top 10 most frequently searched terms in every US state except Missouri, which had an ADR term place at number 12. The highest rated Community term ‘dog barking’ ranked no higher than eighth place for any state. As a testament to the earlier notation on the dearth of Education-related terms in the overall collection, those terms were entirely absent within the top 25 most frequently searched terms in every state. Family and Market terms were consistently represented throughout the top 25 placements, capturing as many as nine Family rankings in Connecticut and 12 Market rankings in both the District of Columbia and New York. Self-help terms narrowly escaped Education’s rankings by placing in the top 25 in only eight states, most commonly with the term ‘communication skills’. Finally, ‘bullying’ commanded the leaderboard of most frequently searched terms for the Social category, ranking first in all but six states.

 

Mapping Conflict

Whether it be the precise GPS coordinates from a mobile or wearable device, or the triangulated position of a computer’s IP address, affixing offline, physical locations to our online activities is our digital world’s new norm. In online advertising, location plays an increasingly (Lawson 2015) critical role in the micro-moments that dot consumers’ journey and ultimately determine how that journey ends (Google 2015b):

 

  • ‘With traffic, I can just make it to the grocery before closing.’
  • ‘Free expedited shipping is available for my location.’
  • ‘Help is literally just around the corner.’

As location matters, it’s made available as part of the metadata packaged alongside our amalgamated search histories. In this way, interpersonal and inter-entity conflicts of any stripe can be mapped in the same familiar way we plot voting blocs and with near the same specificity we have come to expect from streaming weather radar. Near-future newscasters could realistically and confidently intro hyper-local coverage with such examples as:

 

  • ‘We’re seeing an earlier than expected start to dog barking disturbances in Cambridge’s lower east side this evening.’
  • ‘Forecasts are calling for a 25% increase in student–teacher tensions tomorrow morning as KS1 testing gets under way.’

For conflict, location matters. Knowing the conflict iterations and intensities for a particular region can influence how local service providers assign and apply their relevant assets. Having a data-informed sense of the sort and size of a local market can – and should – influence a professional’s decision as to whether and how a local practice can reasonably thrive. And in a more provocative sign of the times and further research to come, location is an increasingly recognised and influential vector in charting everything from the spread of seemingly benign rudeness (Foulk, Woolum and Erez 2015) to violence-imbued forms of conflict (Skogan, Hartnett, Bump and Dubois 2009) akin to metastasising contagions, and subject to the similar constructive effects from targeted proactive and restorative interventions.

 

This project mapped each of the 225 monitored conflict contexts to every US state and the District of Columbia, as well as to each of the 3,143 county and county-equivalent boundaries within the US. In doing so, a literal map of conflict was charted. Choropleths that color-coded geographic regions by overall category or context search volumes predictably displayed conflict condensed within population centres. California, Texas, New York, Florida, and Illinois, by sheer volume of searches, are homes to the most number of US-based conflict searches. State-level analyses of more granular county regions – as is also anticipated from forthcoming zip code-level analyses has – found similar favouring toward intrastate population centres.



Visualisation 9. State Rankings by Volume and Conflict Intensity. Examining the location of conflict queries at the state level presents two starkly different Americas. By sheer volume of online queries, the most populous states produce the highest rates of conflict-related queries. When adjusted for population, however, less-populous states produce higher conflict-intensity scores. 


To further explore geography’s relationship with conflict, per capita analyses were performed within each of the monitored regions to craft a more representative image of where conflict correlatively congregates. As a single individual may repeatedly search the same term and will often search various terms within the same context, the resulting measure may best be described as a location’s ‘conflict intensity’. Here, individuals’ repeated term or intra-context queries are understood to represent the increasing intensity with which they – and the counters, counsellors, and collateral others around them – are experiencing the related conflict. Higher rates of conflict-related searches per capita suggest a higher perceived intensity of that conflict by the local populace performing those searches.

 

On this variable, the influence of population density gives way to far more nuanced local dynamics. Counties with heavier distributions of school-aged youth present greater conflict intensity for school-related bullying, while those with more senior populations are proportionally more intense for conflicts such as grandparent rights. Areas which newly introduce or concentrate focus on select legal or policy changes affect aligned conflict intensities, as was visible within parenting-related conflict queries in Nevada as that state raised awareness of its mandatory Seminar for Separating Parents (Supreme Court of Nevada 2014). Other, more intimate motivating factors eluded easy explanation from within this expansive, yet siloed dataset, such as the twofold intensity of infidelity-related terms throughout Washington state compared to nearly every other US state. What is promising, however, is that our sometimes tenuous or tortured understanding of these contexts will only improve as aligned datasets allow us to discern with greater acuity the culprits and contributors to these and other conflicts.



Visualisation 10. New York Conflict Intensity Profiles by County. By examining the County-level data, both an aggregated and numerous category-specific conflict intensity scores could be produced. Here, adjusted conflict intensity scores (ranging from 0 to 100) were assigned to each of New York's 62 Counties. This and even more granular analyses can help inform conflict-assistive practitioners in their outreach and awareness raising efforts.


Mapping location-influenced conflict intensities can not only pinpoint local novelties of passing or pressing import, it can also inform the administration of hierarchical systems such as states’ judicial branches, community mediation networks, and regional associations of dispute resolution practitioners. Detailed statewide analyses of category-level conflict intensities have now been completed for several states, including the calculation of county-level intensity scores. The assigned scores and related choropleth visualisations synthesise millions of data points and allow systems administrators to quickly identify areas where additional family court ADR messaging may net the greatest return; where new or expanded community mediation programming would likely receive the warmest reception; or where an association’s membership or training regimen may be at risk of over- or under-saturating local market demand. As local groups have been introduced to the availability of and opportunity represented within these new tools, early collaborations have already resulted in several of these very outcomes, with many others likely to develop as analyses and dissemination progress.

 

Forecasting Trends

Knowing where a plea for help originates is interesting. Knowing also when that plea occurs makes it actionable. This project tracked the monthly search volumes for each of the 52,203 unique terms in each of the 3,195 monitored geographies. Taken together, this conflict-focused location and time metadata can deepen our knowledge of which conflicts have historically weighed upon a region, equip us with the ability to responsively schedule our ongoing outreach efforts, and enhance our potential to predictively forecast conflict intensities and the mass damages and market demands they produce.

 

Even at a monthly-level analysis, search activity represents a far more representative and timely indicator of when conflict has, is, and probably will occur than the dispute resolution field has previously relied. Community mediation programmes, for example, often inform their assessments of local demand and rationalisations for future funding on some of conflict’s latest-stage lagging indicators of distress: police calls for service (Charkoudian 2005) and court filings (Hedeen and Coy 2000). Search, on the other hand, is an earlier-stage coincident indicator capable of offering near real-time insights into unfolding conflict trends. Indeed the potential of search analysis to disrupt traditional metrics is no more evident than Google and others’ ability to best the Centers for Disease Control in identifying emerging flu trends, sexually transmitted infection outbreaks, and a growing range of other infectious diseases of substantial consequence to local public health (Ginsberg, Mohebbi, Patel, Brammer, Smolinski and Brilliant 2009; Google 2015a; Yang, Santillana and Kou 2015; Jaklevic 2015). Achieving similar timeliness in identifying emerging conflict intensities holds considerable potential for the dispute resolution field as it seeks to achieve greater resonance at earlier stages of the conflict lifecycle.

 

Of greater value still, is the possibility of identifying predictable trends within various conflict contexts that enhance our ability to peer around tomorrow’s opaque corners to optimise today’s stance. Several such trends have already been identified through this analysis, with many others likely on the horizon. Monitored conflict terms within the Dating and Relationships context, for example, have been revealed to follow cyclical volume trends. For this context, search interest spikes at the start of the year (colloquially hypothesised as an en masse deluge of New Year’s resolutions to finally divest themselves of troubled relationships), remains high through April (i.e. relationship spring cleaning), declines precipitously until it reaches its lowest interest level in August (i.e. summer love), experiences a mini-spike through October (i.e. a possible fling cleaning), and declines again through the end of the year (i.e. the goodwill of sparing holiday heartache) before restarting its cyclical January spike. Conflict-related searches assigned to the ‘grandparent rights’ context has its own cycle with a slow-growth spike from March to June (i.e. strife over setting the summer visitation schedule) which then resets and repeats from July to October (i.e. planning the forthcoming holiday schedule). Student–teacher conflicts reach their highest intensities alongside local standardised testing schedules. Self-help terms for negotiation skills-building by those interested in securing a better position or salary at work spike just after the tax filing season. These are just a few of the statistically significant and sustained cycles identified within this extensive dataset.



Visualisation 11. Intensity of Conflict-Related Search Activity by Month. Many of the analysed conflict contexts were found to have predictable volume patterns, with search intensities responding to everything from seasonal changes, to school schedules, tax deadlines, and many other offline factors.




Visualisation 12. Annual Dating and Relationship-Related Conflict Search Trend. Here, the seasonality of love is charted through the highs and lows of dating and relationship conflict queries. From 'summer love's' query recession to New Year's resolutions and relationship 'spring cleaning' conflict peaks, understanding the relationship conflict cycle through online search data can help maximize message resonance by ensuring practitioners' and the public's outreach activities are smartly sequenced.


Beyond prediction through recurrent cycles, bleeding edge conflict research is beginning to validate the prospect of big data’s seemingly precognitive abilities in even anomalous or otherwise arrhythmic contexts, as well. A provocative and statistically rigorous study originating from Harvard portends an upending of a long-ascribed theory on hyper-local criminality (O’Brien and Sampson 2015). Even better than the broken windows theory (Kelling and Wilson 1982), the research provokes, it appears private conflict may offer a far more predictive indicator of future crime. By analysing data from over one million 911 calls and over 200,000 requests for non-emergency services, private conflict (e.g. relationship disturbances, parking on another’s property, etc.) – the sort very familiar to dispute resolution practitioners – was identified as most strongly predictive of future crime. Big data analysis unlocked this insight which may well influence future policing and preventative efforts around the globe.

 

The apparent corollary between the call data used in that study and select terms monitored in the ongoing search study is undeniably exciting for what future analyses may reveal. Wherein the Harvard study examined fewer than 750,000 requests from select Boston neighbourhoods for each of the two studied years, search behaviour chronicles no fewer than 225,000,000 comparable annual requests from device-parsable locations throughout the country. While the signal-to-noise ratio will require ongoing calibration, ecometric validation appears all but certain for the intuition that perceives the predictive quality of searches for late night dog barking on early morning calls to animal control services, for local divorce attorneys on future filing rates, and for workplace discrimination coping techniques on EEOC/EOC claims. By monitoring these and other contexts, one could predictively position their services in front of key decision makers mere moments before private searches escalate into public action. More arousing, still, is the prospect of emerging technologies that leverage big, disparate data to autonomously monitor an inhuman number of intuition immune variables capable of highlighting truths in conflict dynamics faster than our cognitive and carbon-limited abilities can appreciate.

 

Moving from late-stage lagging indicators to earlier-stage coincidence indicators – and even to the select, tantalisingly prophetic pre-conflict leading indicators – through the creative integration of big data analytics represents a tenfold improvement in our ability to understand and effectively manoeuvre our profession within the expansive and dynamic dispute resolution marketplace. As with location, timing matters in conflict ... and the time for more thoroughly integrating data into our understanding of its nature is decidedly at hand.

 

Search Sentiment

Knowing where and when individuals seek assistance for their conflicts is demonstrably valuable in situating and scheduling dispute resolution services. Actually hearing – or reading – their specific pleas elevates our understanding of their contexts to a whole new level. Dispute resolution specialists are skilled in receiving and honouring conflict narratives, while preparing their pivot toward the positive. We can hear an opening statement, assess the depth of destructiveness or despair within, and ready the space and skills necessary for the ensuing dialogue. Across service types, practice frameworks, and conflict contexts, we attune ourselves to clients’ tone. Through the natural language presentation of search data, we can attend the same at scale; augmenting through automation the forensic attention conversation analysis scholars are giving to intake and session transcripts (Stokoe 2013).

 

A given practitioner’s ability to accurately ascribe a particular sentiment to a given conflict-related search term will understandably vary alongside its length and her skill. Where our ability lumbers and capacity levels, the ability of big data and computational capacity of technology prevail. As part of this analysis of conflict-related search terms, each term was subjected to an algorithmic assessment of its sentiment polarity using a specialty service designed to quickly surface insights from large qualitative datasets (Lexalytics 2015). From this assessment, it is possible to quantify interesting characteristics of the terms and their parent contexts and categories. Does the public frame conflict as constructive opportunities or something less? Do these frames vary between contexts or categories and, if so, to what degree?

 

Sentiment polarity for this analysis scored terms on a scale from –1 to +1, with purely neutral sentiment situated at zero. Notably negatively framed search terms included such examples as:

 

  • marriage in crisis’;
  • laws about child abuse’;
  • ‘hostile work environment retaliation’;
  • ‘revenge on horrible neighbors’.

Opportunity, here, seems distant and bleak. Conversely, particularly positively framed queries included:


  • ‘smart divorce’;
  • fathers equal rights’;
  • ways to improve productivity in the workplace’;
  • fence agreements between neighbors’.

 

Opportunity, here, is readily at hand. Each of the over 52,000 unique terms were scored along these lines, revealing in the aggregate a data-supported understanding of both how the public conceptualises conflict and where our greatest work in encouraging its reframing resides.



Visualisation 13. Conflict Query Word Cloud. In analysing the sentiment of online conflict-related search queries, the public's preferred frame for conflict is resoundingly clear, as the prevalence of negative (orange) words and phrases far exceeds the use of neutral or positive (grey) terms within the evaluated corpus. 


At the broad category level, conflict terms were roundly negative. The Community category was least negative with an average sentiment polarity of –0.07, followed in order of descending destructiveness by Self-Help (–0.17), Market (–0.19), Family (–0.22), Education (–0.30), ADR (–0.30), and Social (–0.38). Among the terms that earned a non-neutral polarity assignment, 89 per cent of the public’s conflict terms received negative scores, while only 11 per cent received positive frames. To the public at large, conflict is not the opportunity we seek for them.



Visualisation 14. Sentiment of Conflict-Related Searches, Overall and by Category Averages. Overall only 6% of conflict terms were positively framed. By conflict category, the most negatively framed conflict queries on average were for social conflicts. The least negative conflict category -- though still not positive, on average -- was community-related conflicts.


Drilling specifically into the Family category, more diverse ratings emerge. Unsurprisingly, the ‘abuse, harassment, intimidation, and violence’ context exudes negativity with an average score of –0.60. Others, in ascending order, include ‘infidelity’ (–0.43), ‘sibling conflict’ (–0.33), ‘parenting’ (–0.15), ‘in-laws’ (–0.12), ‘ADR-involved custody, guardianship, and visitation’ (–0.03), and ‘grandparent rights’ (–0.02), to name a few. Surprisingly, several Family-specific conflict contexts averaged positive scores, including ‘divorce’ (+0.01), ‘inheritance’ (+0.04), ‘separation’ (+0.05), and ‘elder caregiving’ (+0.12). These latter hopeful standouts may well represent a populace increasingly motivated to shun the entrenched negativism traditional within those contexts and – risking gratuitous doe-eyed optimism – the budding affect of the dispute resolution field’s hard-won campaign for hearts and minds. Though it lay beyond the current research scope, methodically ascribing the roots of this shift will be important as the campaign to repackage conflict as opportunity continues.

 

In the interim, however, this initial assessment presents its own opportunity to reflect upon how resonant the dispute resolution message is with those who would benefit from its intervention. Save select conflict contexts, the public roundly frames conflict negatively; occurrences to rue and rail against, rather than opportunities to relish and rise above. Yet, the public’s broad negativity toward conflict may be more prominent than even this sentiment scoring has highlighted. While this system typically assigned neutral scores to first-person pronouns – words that were heavily represented in the search corpus – research suggests these words are even better markers of depression and other troubled physiological conditions than overtly negative emotional words (Chung and Pennebaker 2007). With all this negativity, does the traditional dispute resolution narrative connect with or sail wide of the public’s interests as it seeks assistance?

 

To explore this, descriptive texts from dispute resolution practitioners’ and programmes’ websites were extracted and analysed using the same sentiment polarity scoring methodology. This was done for service providers in several geographies, as well as specialty practice areas to identify which factors, if any, may influence presentation. In Minnesota, for example, the descriptions of mediation, its appropriate contextual applications, and benefits from 21 variably situated service providers were analysed. Both community mediation (n = 8) and private practitioner (n = 10) websites were positively framed, earning average scores of +0.14 and +0.15, respectively. An ADR-focused higher education website – perhaps as a testament to its dual role of promoting both the profession and its services – netted a +0.21. Interestingly, the two primary government agencies tapped to further dispute resolution’s interests statewide had websites which saddled the neutral fulcrum with one scoring a +0.08 and the other (Minnesota Judicial Branch website) a decidedly negative –0.12. In Minnesota, as with many other studied geographies, the supply of dispute resolution service providers largely speaks a different language than the local market demands.



Visualisation 15. Text Sentiment on Minnesota ADR Service Providers' Websites, Overall and by Provider Category Averages. By analysing the descriptive text on a range of Minnesota ADR service provider websites, only 5% of sites presented a negative sentiment, 71% presented a neutral, informative sentiment, and 24% were overtly positive in their descriptions of ADR. Some government ADR websites were found to use negatively-framed language, while higher education programs were the most positive.  


Turning to a specialty practice area, the descriptive texts from 200 members of the Academy of Professional Family Mediators were also examined. Serving the spread of family contexts monitored through this research, the success of these members’ websites in selling their medium of professionalised peace in our adversarial world determines in no small part the sustainability and even survivability of their respective practices. Perhaps reflecting the field’s collective optimism, these practitioners’ sites were largely positive, aligning along a traditional bell-curve distribution centreed on a score of +0.16. The spread, however, highlights a broad divergence between practitioners whose professional objectives are otherwise closely mirrored. On the Pollyannaish pole, one member’s site earned a score of +0.48 by describing mediation using terms such as: agreements, benefits, caring environment, collaborative, support, and voluntary. Acknowledging the dark in conflict, another member’s website scored a –0.33 by describing the same process with terms such as: adversarial, bitterness, confront, hostility, resentment, and uncomfortable. These are but two of many practitioners whose trade and targets are aligned, but who have taken decidedly divergent approaches to advertising its application.



Visualisation 16. Relationship between Family Conflict-Related Search Sentiment and Family Mediator Website Sentiment. By analysing the descriptive texts of 200 family mediator websites for practices located across the US, a distribution of negative-to-positive site sentiment was identified, with most sites using nominally positive frames. In contrast, among the top 10 most frequently queried family-related conflict contexts, only three averaged positive framing.


These concurrent divergences – both our intra-profession messaging styles and the interplay between the public’s largely conflict-focused narrative and our predominantly resolution-oriented narrative – raise interesting questions from the practical to the profound. Namely, 'Which messaging approach is best at attracting future clients?', and 'Is the profession’s preoccupation with fundamentally recapitulating conflict as opportunity a Sisyphean sideshow that ultimately slows its own advancement?'

 

Practically, we probably benefit from the thousand definitional blossoms offered up on our variably targeted – ideally data-informed – websites. Regardless how narrowly sliced, the dispute resolution market is undoubtedly large enough to contain those who resonate with pitches toward both pain and promise. Marketing (Forbes 2015) and public persuasion (Kim 2015) research spheres undoubtedly have much to suggest along these lines. Still, as mediators deftly neutralise without needlessly neutering parties’ statements intra-session, there is surely similar value in acknowledging within our outreach the harm and hardship reflected in the public’s introductory invocations. Sites that exclusively serve syrupy, tone-deaf doctrine to prospective clients’ desperation are an effortless click away from a more receptive competitor.

 

On an even more technical level, the practical implications of the profession’s majority disinterest in embracing the public’s conflict-focused narrative may leave us largely absent in the responses search engines serve to those seeking assistance. If the valley between the actual words used and sentiment conveyed by searchers and service providers is so uncanny, the algorithms designed to couple the two will simply miss the connection or devalue it into a blue-link wasteland devoid of serendipity. Alas, given the far-flung poles between many searchers’ chosen phraseologies and the current text populating practitioners’ websites, to the cold calculators which construct our online connections, we are already ships – fleets, really – in a very dark night.

 

Current and Compelling Research Extensions

 

Where this research informs, it also inspires further inquiry. We now know the content and character of conflict’s cacophonous calls for assistance, but the discoveries of which terms most consistently convert mere calls into clients await. What revelations of conflict’s vibrancy may yet be discovered with ever more granular or multivariate analyses? How might market contours redraw with the incorporation of even broader or less frequently used search terms? Devoid professional allegiance and ambition, what might the brute and binary brilliance behind machine learning make of these data?

 

How this project will continue to unfold, what reception it awaits, and what further inquiry it might inspire is at this early stage uncertain. Still, next steps are decidedly in order. New geographic and practice area analyses are underway, coordinated online public awareness campaigns informed by research findings are scheduled, and practitioner-level website assessments are beginning to help bend the arc of our collective resolution-oriented narrative in ways which help achieve greater market resonance and practice profitability. Additional developments which continue to leverage evolving technology and data science resources present particularly enticing prospects, as well.

 

One such development is a computer kaizen approach to automating future analysis and hastening the discovery of new insights (Varian 2013). In this scenario, the over fifty thousand manually categorised terms would serve as an expansive training set from which to refine a recurrent neural network – the sort recently open-sourced (Google 2015c) – capable of automatically identifying, ingesting, and categorising future term additions. By pivoting from a primarily manual processing to a machine learning approach, future analyses could incorporate a set of unique terms many multiples of its current size and representing an even more complete diversity of terms used without concern for the artificially elevated volume threshold necessary during the initial manual processes. Additional automations of the data collection (i.e. through scheduled API calls) and analysis (i.e. through trigger sensitive templates) could be combined to produce a continuously updating dashboard of every imaginable conflict context in nearly any conceivable geographic boundary with both rich research and predictive applications. In this scenario that relies exclusively on currently available technology, these data contain the prospect of illuminating the context of where conflict has occurred, is occurring, and will probably occur, allowing the dispute resolution and aligned professions to design and deploy their services accordingly.

 

Widening the Lens

 

Cataloguing the entirety of interpersonal and inter-entity conflict throughout the country as chronicled in search entries is a notable development for many aspects of the dispute resolution field. This research is revealing new insights into the facets and fullness of the marketplace, the framing practices and preferences of the populace, and the vast opportunity that remains as we seek to maximise the utility and prominence of our profession. In doing so, it serves as a powerful example of the influence big data will increasingly command from practice and thought leaders throughout our field. In truth, though, it is but a segment in the obtuse relief propagating from the emerging nexus of big data and dispute resolution; an intriguing harbinger of the data and disruptions yet to come.

 

The frequency and consequence of future case studies in dispute resolution data, whether chronicled in journals or in market upheavals, will surely increase as tradition rests and ingenuity innovates. As practitioners and programme leaders vested in the furtherance of the dispute resolution field, it is dependent upon us to encourage and ensure such innovation occurs, and that it does so at a pace well-beyond our ongoing rickety rush to recognise technology’s potential (Rule 2015). We hasten this development through simple curiosity and structured analysis alike – demanding data-informed decisions of ourselves and of the cottage services supporting our work.

 

Data empowers. It qualifies us to more confidently speak to the scale and scope of dispute resolution’s significance; allows for the sharing of more resonant narratives with those we seek to serve and sway; helps sate stakeholders’ increasing appetite for relevant, actionable insights; enhances our own contextual awareness; and imbues us with greater agency over our profession’s malleable future. In collectively balancing the Zen of occupying mediation’s unfolding moments with dissociatively analysing the alloy of conflict’s consequence (i.e. attending to the qualitative and quantitative impact of our services and their stimulants), we elevate our understanding of the personal and public hardships we seek to moderate. By thus embracing this duality and the opportunities streaming forth from the nexus of big data and dispute resolution, the field matures into a profession better equipped to fulfil its promise in both precise and profound measure.

 

References

 

Brin, S. and Page, L. (1998) The anatomy of a large-scale hypertextual web search engine. Computer Networks and ISDN Systems 30: 107–17. http://dx.doi.org/10.1016/S0169-7552(98)00110-X

 

Brück, T., Justino, P., Verwimp, P. and Avdeenko, A. (2010) Identifying Conflict and Violence in Micro-Level Surveys. Discussion Paper No. 5067. Bonn: Institute for the Study of Labor. Retrieved on 10 October 2015 from http://ftp.iza.org/dp5067.pdf.

 

Charkoudian, L. (2005) A quantitative analysis of the effectiveness of community mediation in decreasing repeat police calls for service. Conflict Resolution Quarterly 23(1): 87–98. http://dx.doi.org/10.1002/crq.126

 

Chung, C. and Pennebaker, J. (2007) The psychological functions of function words. In K. Fiedler (ed.) Social Communication 343–59. New York: Psychology Press.

 

comScore(2015)comScore releases September2015US desktop search engine rankings. Retrieved on 10 October 2015 from www.comscore.com/Insights/Market-Rankings/comScore-Releases-September-2015-US-Desktop-Search-Engine-Rankings

 

Corbett, W. E. H. and Corbett, J. R. (2011). Community mediation in economic crisis: the reemergence of precarious sustainability. Nevada Law Journal 11(2): 469.

 

Corbett, X. (2015) First-person observations of youth-oriented conflict during play. Retrieved on 25 September 2015 from www.advancingdr.org/x#TOC-Play.

 

Economist (2010) Data, data everywhere: a special report on managing information. The Economist (25 February). Retrieved on 10 October 2015 from www.economist.com/node/15557443?story_id=15557443.

 

Forbes, D. (2015) The Science of Why: Decoding Human Motivation and Transforming Marketing Strategy. New York: Palgrave Macmillan. http://dx.doi.org/10.1057/9781137502049

 

Foulk, T., Woolum, A. and Erez, A. (2015). Catching rudeness is like catching a cold: the contagion effects of low-intensity negative behaviors. Journal of Applied Psychology 101(1): 50–67. http://dx.doi.org/10.1037/apl0000037

 

Gaglio, C. M. and Katz, J. A. (2001) The psychological basis of opportunity identification: entrepreneurial alertness. Small Business Economics 16: 95–111. http://dx.doi.org/10.1023/A:1011132102464

 

Gardeazabal, J. (2012) Methods for measuring aggregate costs of conflict. In M. R. Garfinkel and S. Skaperdas (eds) The Oxford Handbook of the Economics of Peace and Conflict 227–51. Oxford: Oxford University Press. http://dx.doi.org/10.1093/oxfordhb/9780195392777.013.0011

 

Ginsberg, J., Mohebbi, M. H., Patel, R. S., Brammer, L, Smolinski, M. S. and Brilliant, L. (2009) Detecting influenza epidemics using search engine query data. Nature 457: 1012–14. http://dx.doi.org/10.1038/nature07634

 

Google (2015a) Google flu trends and Google dengue trends. Retrieved on 10 October 2015 from www.google.org/flutrends/about.

 

Google (2015b) Micro-Moments: Your Guide to Winning the Shift to Mobile. Mountain View, CA: Google. Retrieved on 10 October 2015 from https://think.storage.googleapis.com/images/micromoments-guide-to-winning-shift-to-mobile-download.pdf.

 

Google (2015c) TensorFlow: smarter machine learning, for everyone. Retrieved on 9 November 2015 from https://googleblog.blogspot.com/2015/11/tensorflow-smarter-machine-learning-for.html.

 

Hedeen, T. and Coy, P. G. (2000) Community mediation and the court system: the ties that bind. Mediation Quarterly 17(4): 351–67. http://dx.doi.org/10.1002/crq.3890170407


Hillis, K., Petit, M. and Jarrett, K. (2013) Google and the Culture of Search. New York: Routledge.

 

Institute for Economics and Peace (2015) Global Peace Index 2015: Measuring Peace, Its Causes and Its Economic Value. Sydney: Institute for Economics and Peace. Retrieved on 10 October 2015 from www.visionofhumanity.org/sites/default/files/Global%20Peace%20Index%20Report%202015_0.pdf.

 

Jaklevic, M. C. (2015) Where are STDs rampant? Google wants to help researchers find out.Kaiser Health News (10 December).Retrieved on 10 December 2015 from http://khn.org/news/where-are-stds-rampant-google-wants-to-help-researchers-find-out.

 

Jehn,K.A.,Northcraft,G.B.and Neale,M.A.(1999)Why differences make a difference: a field study of diversity, conflict, and performance in workgroups. Administrative Science Quarterly 44(4): 741–63. http://dx.doi.org/10.2307/2667054

 

Kelling, G. L. and Wilson, J. Q. (1982) Broken windows: the police and neighborhood safety. The Atlantic (March). Retrieved on 10 October 2015 from www.theatlantic.com/magazine/archive/1982/03/broken-windows/304465.

 

Kim, K. E. (2015) Framing as a strategic persuasive message tactic. In D. Holtzhausen and A. Zerfass (eds) The Routledge Handbook of Strategic Communication 285–302. New York: Routledge.

 

Lawson, M. (2015) I-want-to-go moments: from search to store. Retrieved on 10 October 2015 from https://www.thinkwithgoogle.com/articles/i-want-to-go-micro-moments.html.

 

Lexalytics (2015) Sentiment Extraction: Measuring the Emotional Tone of Content. Boston, MA: Lexalytics. Retrieved on 10 October 2015 from www.lexalytics.com/content/whitepapers/Lexalytics-WP-Sentiment-Extraction.pdf.

 

Marchetti,H.and Jazbec,A.(2015) Getting stabbed by a unicorn.Hélice 4(2).Retrieved on 10 October 2015 from www.triplehelixassociation.org/helice/volume-4-2015/helice-issue-2-4/triple-helix-scientific-news/getting-stabbed-by-a-unicorn.

 

McGillis, D. (1997) Community Mediation Programs: Developments and Challenges. Washington, DC: National Institute of Justice.

 

McKinseyGlobalInstitute(2011)Big data: the next frontier for innovation, competition, and productivity. Retrieved on 10 October 2015 from https://goo.gl/HNwiJC.

 

Moses, L. B. and Chan, J. (2014) Using big data for legal and law enforcement decisions: testing new tools. UNSW Law Journal 37(2): 643–78.

 

Motorola (2015) 2015 Motorola global smartphone relationship survey. Retrieved on 10 October 2015 from http://motorola-blog.blogspot.com/2015/07/2015-motorola-global-smartphone.html.

 

O’Brien, D. T. and Sampson, R. J. (2015) Public and private spheres of neighbor- hood disorder: assessing pathways to violence using large-scale digital records. Journal of Research in Crime and Delinquency 52(4): 486–510. http://dx.doi.org/10.1177/0022427815577835

 

Pew Research Center (2012) The future of big data. Retrieved on 10 October 2015 from www.pewinternet.org/2012/07/20/main-findings-influence-of-big-data-in-2020.

 

Plous, S. (1993) The Psychology of Judgment and Decision Making. New York: McGraw-Hill Book Company.

 

Rule, C. (2015) Technology and the future of dispute resolution. Dispute Resolution Magazine 21(2). Retrieved on 10 October 2015 from www.americanbar.org/publications/dispute_resolution_magazine/2015/winter/technology-and-the-future-of-dispute-resolution.html.

 

Singhal, A. (2015) Interview statements made at code/mobile. Re/code (11 October). Retrieved on 11 October 2015 from http://recode.net/2015/10/11/googles-search-boss-talks-surviving-and-thriving-in-an-app-world-full-video.

 

Skogan, W. G., Hartnett, S. M., Bump, N. and Dubois, J. (2009) Evaluation of CeaseFire-Chicago. Chicago, IL: CeaseFire. Retrieved on 10 October 2015 from www.skogan.org/files/Evaluation_of_CeaseFire-Chicago_Main_Report.03-2009.pdf.

 

Soltas, E. and Stephens-Davidowitz, S. (2015) The rise of hate speech. New York Times (12 December). Retrieved on 12 December 2015 from www.nytimes.com/2015/12/13/opinion/sunday/the-rise-of-hate-search.html.

 

Stokoe, E. (2013) Overcoming barriers to mediation in intake calls to services: research-based strategies for mediators. Negotiation Journal 29(3): 289–314. http://dx.doi.org/10.1111/nejo.12026

 

Supreme Court of Nevada (2014) Rules of practice for the Eighth Judicial District Court of the State of Nevada, rule 5.07: seminar for separating parents. Retrieved on 10 October 2015 from www.leg.state.nv.us/courtrules/EighthDCR.html.

 

Varian, H. R. (2013) Beyond big data. Paper presented at the NABE Annual Meeting, 10 September, San Francisco, CA. Retrieved on 10 October 2015 from http://people.ischool.berkeley.edu/~hal/Papers/2013/BeyondBigDataPaperFINAL.pdf.

 

Velikonja, U. (2009) Making peace and making money: economic analysis of the market for mediators in private practice. Albany Law Review 72: 257–91.

 

Welch, Jr., J. F., Immelt, J. R., Dammerman, D. D. and Wright, R. C. (2001) To our customers, share owners and employees. In General Electric, GE Annual Report 2000 1–7. Fairfield, CT: General Electric Company. Retrieved on 10 October 2015 from www.ge.com/annual00/download/images/GEannual00.pdf.

 

Yang, S., Santillana, M. and Kou, S. C. (2015) Accurate estimation of influenza epidemics using Google search data via ARGO. Proceedings of the National Academy of Sciences 112(47): 14,473–8. http://dx.doi.org/10.1073/pnas

 

Acknowledgements

 

Outsized gratitude is extended to his conspirator Wendy E. H. Corbett and to X. Corbett for their playful inspiration and unending support during the pursuit of this professional passion project.

 

About the Author

 

Justin R. Corbett, chief project officer of Advancing Dispute Resolution, is particularly interested in the convergence of data, technology, and the dispute resolution field. His current work leverages the aforementioned research to inform a coordinated, multi-charitable organisation million dollar public education campaign designed to raise awareness of local dispute resolution services. Prior to his current work, he led a local community mediation centre, the National Association for Community Mediation (NAFCM, US), and the Association for Conflict Resolution (ACR, International).