Tag Archives: text analytics

Instagram is changing the way I see

I recently joined Instagram (I’m late, I know).

I joined because my daughter wanted to, because her friends had, to see what it was all about. She is artistic, and we like to talk about things like color combinations and camera angles, so Instagram is a good fit for us. But it’s quickly changing the way I understand photography. I’ve always been able to set up a good shot, and I’ve always had an eye for color. But I’ve never seriously followed up on any of it. It didn’t take long on Instagram to learn that an eye for framing and color is not enough to make for anything more than accidental great shots. The great shots that I see are the ones that pick deeper patterns or unexpected contrasts out of seemingly ordinary surroundings. They don’t simply capture beauty, they capture an unexpected natural order or a surprising contrast, or they tell a story. They make you gasp or they make you wonder. They share a vision, a moment, an insight. They’re like the beginning paragraph of a novel or the sketch outline of a poem. Realizing that, I have learned that capturing the obvious beauty around me is not enough. To find the good shots, I’ll need to leave my comfort zone, to feel or notice differently, to wonder what or who belongs in a space and what or who doesn’t, and why any of it would capture anyone’s interest. It’s not enough to see a door. I have to wonder what’s behind it. To my surprise, Instagram has taught me how to think like a writer again, how to find hidden narratives, how to feel contrast again.

Sure this makes for a pretty picture. But what is unexpected about it? Who belongs in this space? Who doesn't? What would catch your eye?

Sure this makes for a pretty picture. But what is unexpected about it? Who belongs in this space? Who doesn’t? What would catch your eye?

This kind of change has a great value, of course, for a social media researcher. The kinds of connections that people forge on social media, the different ways in which people use platforms and the ways in which platforms shape the way we interact with the world around us, both virtual and real, are vitally important elements in the research process. In order to create valid, useful research in social media, the methods and thinking of the researcher have to follow closely with the methods and thinking of the users. If your sensemaking process imitates the sensemaking process of the users, you know that you’re working in the right direction, but if you ignore the behaviors and goals of the users, you have likely missed the point altogether. (For example, if you think of Twitter hashtags simply as an organizational scheme, you’ve missed the strategic, ironic, insightful and often humorous ways in which people use hashtags. Or if you think that hashtags naturally fall into specific patterns, you’re missing their dialogic nature.)

My current research involves the cycle between social media and journalism, and it runs across platforms. I am asking questions like ‘what gets picked up by reporters and why?’ and ‘what is designed for reporters to pick up?’ And some of these questions lead me to examine the differences between funny memes that circulate like wildfire through Twitter leading to trends and a wider stage and the more indepth conversation on public facebook pages, which cannot trend as easily and is far less punchy and digestible. What role does each play in the political process and in constituting news?

Of course, my current research asks more questions than these, but it’s currently under construction. I’d rather not invite you into the workzone until some of the pulp and debris have been swept aside…

Is there Interdisciplinary hope for Social Media Research?

I’ve been trying to wrap my head around social media research for a couple of years now. I don’t think it would be as hard to understand from any one academic or professional perspective, but, from an interdisciplinary standpoint, the variety of perspectives and the disconnects between them are stunning.

In the academic realm:

There is the computer science approach to social media research. From this standpoint, we see the fleshing out of machine learning algorithms in a stunning horserace of code development across a few programming languages. This is the most likely to be opaque, proprietary knowledge.

There is the NLP or linguistic approach, which overlaps to some degree with the cs approach, although it is often more closely tied to grammatical rules. In this case, we see grammatical parsers, dictionary development, and api’s or shared programming modules, such as NLTK or GATE. Linguistics is divided as a discipline, and many of these divisions have filtered into NLP.

Both the NLP and CS approaches can be fleshed out, trained, or used on just about any data set.

There are the discourse approaches. Discourse is an area of linguistics concerned with meaning above the level of the sentence. This type of research can follow more of a strict Conversation Analysis approach or a kind of Netnography approach. This school of thought is more concerned with context as a determiner or shaper of meaning than the two approaches above.

For these approaches, the dataset cannot just come from anywhere. The analyst should understand where the data came from.

One could divide these traditions by programming skills, but there are enough of us who do work on both sides that the distinction is superficial. Although, generally speaker, the deeper one’s programming or qualitative skills, the less likely one is to cross over to the other side.

There is also a growing tradition of data science, which is primarily quantitative. Although I have some statistical background and work with quantitative data sets every day, I don’t have a good understanding of data science as a discipline. I assume that the growing field of data visualization would fall into this camp.

In the professional realm:

There are many companies in horseraces to develop the best systems first. These companies use catchphrases like “big data” and “social media firehose” and often focus on sentiment analysis or topic analysis (usually topics are gleaned through keywords). These companies primarily market to the advertising industry and market researchers, often with inflated claims of accuracy, which are possible because of the opacity of their methods.

There is the realm of market research, which is quickly becoming dependent on fast, widely available knowledge. This knowledge is usually gleaned through companies involved in the horserace, without much awareness of the methodology. There is an increasing need for companies to be aware of their brand’s mentions and interactions online, in real time, and as they collect this information it is easy, convenient and cost effective to collect more information in the process, such as sentiment analyses and topic analyses. This field has created an astronomically high demand for big data analysis.

There is the traditional field of survey research. This field is methodical and error focused. Knowledge is created empirically and evaluated critically. Every aspect of the survey process is highly researched and understood in great depth, so new methods are greeted with a natural skepticism. Although they have traditionally been the anchors of good professional research methods and the leaders in the research field, survey researchers are largely outside of the big data rush. Survey researchers tend to value accuracy over timeliness, so the big, fast world of big data, with its dubious ability to create representative samples, hold little allure or relevance.

The wider picture

In the wider picture, we have discussions of access and use. We see a growing proportion of the population coming online on an ever greater variety of devices. On the surface, the digital divide is fast shrinking (albeit still significant). Some of the digital access debate has been expanded into an understanding of differential use- essentially that different people do different activities while online. I want to take this debate further by focusing on discursive access or the digital representation of language ideologies.

The problem

The problem with such a wide spread of methods, needs, focuses and analytic traditions is that there isn’t enough crossover. It is very difficult to find work that spreads across these domains. The audiences are different, the needs are different, the abilities are different, and the professional visions are dramatically different across traditions. Although many people are speaking, it seems like people are largely speaking within silos or echo chambers, and knowledge simply isn’t trickling across borders.

This problem has rapidly grown because the underlying professional industries have quickly calcified. Sentiment analysis is not the revolutionary answer to the text analysis problem, but it is good enough for now, and it is skyrocketing in use. Academia is moving too slow for the demands of industry and not addressing the needs of industry, so other analytic techniques are not being adopted.

Social media analysis would best be accomplished by a team of people, each with different training. But it is not developing that way. And that, I believe, is a big (and fast growing) problem.

Storytelling about correlation and causation

Many researchers have great war stories to tell about the perilous waters between correlation and causation. Here is my personal favorite:

In the late 90’s, I was working with neurosurgery patients in a medical psychology clinic in a hospital. We gave each of the patients a battery of cognitive tests before their surgery and then administered the same battery 6 months after the surgery. Our goal was to check for cognitive changes that may have resulted from the surgery. One researcher from outside the clinic focused on our strongest finding: a significant reduction of anxiety from pre-op to post-op. She hypothesized that this dramatic finding was evidence that the neural basis for anxiety was affected by the surgery. Had she only taken a minute to explain her  hypothesis in plain terms to a layperson, especially one that could imagine the anxiety a patient could potentially experience hours before brain surgery, she surely would have withdrawn her request for our data and slipped quietly out of our clinic.

“Correlation does not imply causation” is a research catchphrase that is drilled into practitioners from internhood and intro classes onward. It is particularly true when working with language, because all linguistic behavior is highly patterned behavior. Researchers from many other disciplines would kill to have chi square tests as strong as linguists’ chi squares. In fact, linguists have to reach deeper into their statistical toolkits, because the significance levels alone can be misleading or inadequate.

People who use language but don’t study linguistics usually aren’t aware of the degree of patterning that underlies the communication process. Language learning has statistical underpinnings, and language use has statistical underpinnings. It is because of this patterning that linguistic machine learning is possible. But, linguistic patterning is a double edged sword- potentially helpful in programming and harmful in analysis. Correlations abound, and they’re mostly real correlations, although, statistically speaking, some will be products of peculiarities in a dataset. But outside of any context or theory, these findings are meaningless. They don’t speak to the underlying relationship between the variables in any way.

A word of caution to researchers whose work centers around the discovery of correlations. Be careful with your findings. You may have found evidence that shows that a correlation may exist. But that is all you have found. Take your next steps carefully. First, step back and think about your work in layman’s terms. What did you find, and is that really anything meaningful? If your findings still show some prospects, double down further and dig deeper. Try to get some better idea of what is happening. Get some context.

Because a correlation alone is no gold nugget. You may think you’ve found some fashion, but your emperor could very well still be naked.

Time for some Research Zen

As the new semester kicks into gear and work deadlines loom, I find myself ready for a moment of research zen.

2012-12-16 14.18.00

Let’s take a minute to stand in a stream and think about the water. Feel the flow of the water over your feet and by your calves. Feel the pull of constant motion. Feel yourself sink against the current, rooting deeper to keep steady. Breathe the clean outdoor air. Observe the clouds and watch the way the sky reflects in the water in the stream. The stream is not constant. The water passing now is not the water that passed when you started, and the water that passes when you leave will be still different. And yet we call this a stream.

As I observe sources of social media, thinking about sampling, I’m faced with some of the same questions that the stream gives rise to. Although I would define my sources consistently from day to day, their content shifts constantly. The stream is not constant, but rather constantly forming and reforming at my feet.

For a moment, I saw the tide of social media start to turn in favor of taxi drivers. In that moment, I felt both a strong sense of relief from the negativity and a need to revisit my research methods. Today I see that the stream has again turned against the drivers. I could ignore the momentary shift, or I could use this as a moment to again revisit the wisdom of sampling.

If I sample the river at a given point, what should I collect and what does it represent? How, when the water is constantly moving around me, can I represent what I observe within a sample? Could my sampling ever represent a single point in the stream, the stream as a whole, or streams in general? Or will it always be moments in the life of a stream?

In the words of Henry Miller, “The world is not to be put in order. The world is in order. It is for us to put ourselves in unison with this order.” In order to understand this stream, I need to understand what lies beneath it, what gives it its shape and flow, and how it works within its ecosystem.

The ecosystem of public opinion around the taxi system in DC is not one that can be understood purely online. When I see the reflection of clouds on the stream, I need to find the sky. When I see phrases repeated over and over, I need to understand where they come from and how they came to be repeated. In the words of Blaise Pascal “contradiction is not a sign of falsity, nor the lack of contradiction a sign of truth.” No elements in this ecosystem exist independent of context. Each element has its base.

Good research involves a good deal of reflection. It involves digging in against currents and close observation. It involves finding a moment of stillness in the flow of the stream.

Breathe in. Observe carefully. Breathe out. Repeat, continue, focus, research.

Turns out Ethnography happens one slice at a time

Some of you may have noticed that I promised to report some research and then didn’t.

Last semester, for my Ethnography of Communication class, I did an Ethnography of DC taxi drivers. The theme of the Ethnography was “the voice of the drivers.” It was multilayered, and it involved data from a great variety of sources. I had hoped to share my final paper for the class here, but that won’t work for three reasons.

1.) The nature of Ethnography. Ethnography involves collecting a great deal of data and then choosing what to report, in what way, and in what context. The goal of the final paper was to reflect on the methodology. This was an important exercise, but I really wanted to share more of my findings and less of my methodology here.

2.) The particular aspect of my findings that I most want to share here has to do with online discourse. Specifically, I want to examine the lack of representation of the drivers perspective online. There are quite a few different ways to accomplish this. I have tried to do it a number of ways, using different slices of data and using different analytic strategies. But I haven’t decided which is the best set of data or method of analysis. But I am a very lucky researcher. Next week I’m headed to a workshop at Radbound University in Nijmegen, Netherlands. The workshop is on the Microanalysis of Online Discourse. I am eager to bring my data and methodological questions and to recieve insight from such an amazing array of researchers. I am also very eager to see what they bring!

Much of the discussion in the analysis of online discourse either excludes the issue of representation altogether or focuses on it entirely. Social media is often hailed as the great democratizer of communication. Internet access was long seen as the biggest obstacle to this new democracy . From this starting point, much of the research has evolved to consider more of nuances of differential use, including the complicated nature of internet access as well as behavior and goals of internet users. This part of my findings is an example of differential use and of different styles of participation. Working with this data has changed the way I see social media and the way I understand the democratization of news.

3.) Scope. The other major reason why I haven’t shared my findings is because of the sheer scope of this project. I was fortunate enough to only have taken one class last semester, which left me the freedom to work much harder on it. Also, as a working/student mom, I chose a project that involved my whole family in an auto-ethnographic way, so much of my work brought me closer to my family, rather than farther apart (spending time away from family to study is one of the hardest parts of working student motherhood!)

I have amassed quite a bit of data at this point, and I plan to write a few different papers using it.

Stay tuned, because I will release slices of it. But have some patience, because each slice will only be released in its own good time.

 

At this point, I feel the need to reference the Hutzler Banana Slicer

Turns out, Ethnography is more like this:

 

than like this:

What do all of these polling strategies add up to?

Yesterday was a big first for research methodologists across many disciplines. For some of the newer methods, it was the first election that they could be applied to in real time. For some of the older methods, this election was the first to bring competing methodologies, and not just methodological critiques.

Real time sentiment analysis from sites like this summarized Twitter’s take on the election. This paper sought to predict electoral turnout using google searches. InsideFacebook attempted to use Facebook data to track voting. And those are just a few of a rapid proliferation of data sources, analytic strategies and visualizations.

One could ask, who are the winners? Some (including me) were quick to declare a victory for the well honed craft of traditional pollsters, who showed that they were able to repeat their studies with little noise, and that their results were predictive of a wider real world phenomena. Some could call a victory for the emerging field of Data Science. Obama’s Chief Data Scientist is already beginning to be recognized. Comparisons of analytic strategies will spring up all over the place in the coming weeks. The election provided a rare opportunity where so many strategies and so many people were working in one topical area. The comparisons will tell us a lot about where we are in the data horse race.

In fact, most of these methods were successful predictors in spite of their complicated underpinnings. The google searches took into account searches for variations of “vote,” which worked as a kind of reliable predictor but belied the complicated web of naturalistic search terms (which I alluded to in an earlier post about the natural development of hashtags, as explained by Rami Khater of Al Jezeera’s The Stream, a social network generated newscast). I was a real-world example of this methodological complication. Before I went to vote, I googled “sample ballot.” Similar intent, but I wouldn’t have been caught in the analyst’s net.

If you look deeper at the Sentiment Analysis tools that allow you to view the specific tweets that comprise their categorizations, you will quickly see that, although the overall trends were in fact predictive of the election results, the data coding was messy, because language is messy.

And the victorious predictive ability of traditional polling methods belies the complicated nature of interviewing as a data collection technique. Survey methodologists work hard to standardize research interviews in order to maximize the reliability of the interviews. Sometimes these interviews are standardized to the point of recording. Sometimes the interviews are so scripted that interviewers are not allowed to clarify questions, only to repeat them. Critiques of this kind of standardization are common in survey methodology, most notably from Nora Cate Schaeffer, who has raised many important considerations within the survey methodology community while still strongly supporting the importance of interviewing as a methodological tool. My reading assignment for my ethnography class this week is a chapter by Charles Briggs from 1986 (Briggs – Learning how to ask) that proves that many of the new methodological critiques are in fact old methodological critiques. But the critiques are rarely heeded, because they are difficult to apply.

I am currently working on a project that demonstrates some of the problems with standardizing interviews. I am revising a script we used to call a representative sample of U.S. high schools. The script was last used four years ago in a highly successful effort that led to an admirable 98% response rate. But to my surprise, when I went to pull up the old script I found instead a system of scripts. What was an online and phone survey had spawned fax and e-mail versions. What was intended to be a survey of principals now had a set of potential respondents from the schools, each with their own strengths and weaknesses. Answers to common questions from school staff were loosely scripted on an addendum to the original script. A set of tips for phonecallers included points such as “make sure to catch the name of the person who transfers you, so that you can specifically say that Ms X from the office suggested I talk to you” and “If you get transferred to the teacher, make sure you are not talking to the whole class over the loudspeaker.”

Heidi Hamilton, chair of the Georgetown Linguistics department, often refers to conversation as “climbing a tree that climbs back.” In fact, we often talk about meaning as mutually constituted between all of the participants in a conversation. The conversation itself cannot be taken outside of the context in which it lives. The many documents I found from the phonecallers show just how relevant these observations can be in an applied research environment.

The big question that arises from all of this is one of a practical strategy. In particular, I had to figure out how to best address the interview campaign that we had actually run when preparing to rerun the campaign we had intended to run. My solution was to integrate the feedback from the phonecallers and loosen up the script. But I suspect that this tactic will work differently with different phonecallers. I’ve certainly worked with a variety of phonecallers, from those that preferred a script to those that preferred to talk off the cuff. Which makes the best phonecaller? Neither. Both. The ideal phonecaller works with the situation that is presented to them nimbly and professionally while collecting complete and relevant data from the most reliable source. As much of the time as possible.

At this point, I’ve come pretty far afield of my original point, which is that all of these competing predictive strategies have complicated underpinnings.

And what of that?

I believe that the best research is conscious of its strengths and weaknesses and not afraid to work with other strategies in order to generate the most comprehensive picture. As we see comparisons and horse races develop between analytic strategies, I think the best analyses we’ll see will be the ones that fit the results of each of the strategies together, simultaneously developing a fuller breakdown of the election and a fuller picture of our new research environment.

“Not everything that can be counted counts”

“Not everything that counts can be counted, and not everything that can be counted counts” – sign in Einstein’s Princeton office

This quote is from one of my favorite survey reminder postcards of all time, along with an image from from the Emilio Segre visual archives. The postcard layout was an easy and pleasant decision made in association with a straightforward survey we have conducted for nearly a quarter century. …If only social media analysis could be so easy, pleasant or straightforward!

I am in the process of conducting an ethnography of DC taxi drivers. I was motivated to do this study because of the persistent disconnect between the experiences and reports of the taxi drivers and riders I hear from regularly and the snarky (I know this term does not seem technical, but it is absolutely data motivated!) riders who dominate participatory media sources online. My goal at this point of the project is to chase down the disconnect in media participation and see how it maps to policy deliberations and offline experiences. This week I decided to explore ways of quantifying the disconnect.

Inspired by this article in jedem (the eJournal of eDemocracy and Open Government), I decided to start my search using framework based in Social Network Analysis (SNA), in order to use elements of connectedness, authority and relevance as a base. Fortunately, SNA frameworks are widely available to analysts on a budget in the form of web search engines! I went through the first 22 search results for a particular area of interest to my study: the mandatory GPS policy. Of these 22 sites, only 11 had active web 2.0 components. Across all of these sites, there were just two comments from drivers. Three of the sites that didn’t have any comments from drivers did have one post each that sympathized with or defended DC taxi drivers. The remaining three sites had no responses from taxi drivers and no sympathetic responses in defense of the drivers. Barring a couple of comments that were difficult to divine, the rest of the comments were negative comments about DC taxi drivers or the DC taxi industry. This matched my expectations, and, predictably, didn’t match any of my interviews or offline investigations.

The question at this point is one of denominator.

The easiest denominator to use, and, in fact, the least complicated was the number of sites. Using this denominator, only one quarter of the sites had any representation from a DC taxi driver. This is significant, given that the discussions were about aspects of their livelihood, and the drivers will be the most closely affected by the regulatory changes. This is a good, solid statistic from which to investigate the influence of web 2.0 on local policy enactment. However, it doesn’t begin to show the lack of representation the way that a denominator such as number of posts, number of posters, or number of opinions would have. But each one of these alternative denominators has its own set of headaches. Does it matter if one poster expresses an opinion once and another expresses another, slightly different opinion more than once? If everyone agrees, what should the denominator be? What about responses that contain links that are now defunct or insider references that aren’t meaningful to me? Should I consider measures of social capital, endorsements, social connectedness, or the backgrounds of individual posters?

The simplest figure also doesn’t show one of the most striking aspects of this finding; the relative markedness of these posts. In the context of predominantly short, snarky and clever responses, one of the comments began with a formal “Dear DC city councilmembers and intelligent  taxpayers,” and the other spread over three dense, winding posts in large paragraph form.

This brings up an important aspect of social media; that of social action. If every comment is a social action with social intentions, what are the intentions of the posters and how can these be identified? I don’t believe that the majority of posts left were intended as a voice in local politics, but the comments from the drivers clearly were. The majority of posts represent attempts to warrant social capital using humor, not attempts to have a voice in local politics. And they repeated phrases that are often repeated in web 2.0 discussions about the DC taxi situation, but rarely repeated elsewhere. This observation, of course, is pretty meaningless without being anchored to the data itself, both quantitatively and qualitatively. And it makes for some interesting ‘next steps’ in a project that is certainly not short of ‘next steps.’

The main point I want to make here is about the nature of variables in social media research. Compared to a survey, where you ask a question, determined in advance, and have a set of answers to work with in your analysis, you are free to choose your own variables for your analysis. Each choice brings with it a set of constraints and advantages, and some fit your data better than others. But the path to analysis can be a more difficult path to take, and more justification about the choices you make is important. To augment this, a quantitative analysis, which can sometimes have very arbitrary or less clear choices included in it, is best supplemented with a qualitative analysis that delves into the answers themselves and why they fit the coding structure you have imposed.

In all of this, I have quite a bit of work out ahead of me.

I think I’m using “big data” incorrectly

I think I’m using the term “big data” incorrectly. When I talk about big data, I’m referring to the massive amount of freely available information that researchers can collect from the internet. My expectation is that the researchers must choose which firehose best fits their research goals, collect and store the data, and groom it to the point of usability before using it to answer targeted questions or examining it for answers in need of a question.

The first element of this that makes it “big data” to me, is that the data is freely available and not subject to any privacy violations. It can be difficult to collect and store, because of its sheer size, but it is not password protected. For this reason, I would not consider Facebook to be a source for “big data.” I believe that the overwhelming majority of Facebook users impose some privacy controls, and the resulting, freely available information cannot be assigned any kind of validity. There are plenty of measures of inclusion for online research, and ignorance about privacy rules or shear exhibitionism are not a target qualities by any of these standards.

The second crucial element to my definition of “big data” is structure. My expectation is that it is in any researchers interest to understand the genesis and structure of their data as much as possible, both for the sake of grooming, and for the sake of assigning some sense of validity to their findings. Targeted information will be layed out and signaled very differently in different online environments, and the researcher must work to develop both working delimiters to find probable working targets and a sense of context for the data.

The third crucial element is representativeness. What do these findings represent? Under what conditions? “Big data” has a wide array of answers to these questions. First, it is crucial to note that it is not representative of the general population. It represents only the networked members of a population who were actively engaging with an online interface within the captured window of time in a way that left a trace or produced data. Because of this, we look at individual people by their networks, and not by their representativeness. Who did they influence, and to what degree could they influence those people? And we look at other units of analysis, such as the website that the people were contributing on, the connectedness of that website, and the words themselves, and their degree of influence, both directly an indirectly.

Given those elements of understanding, we are able to provide a framework from which the analysis of the data itself is meaningful and useful.

I’m aware that my definition is not the generally accepted definition. But for the time being I will continue to use it for two reasons:

1. Because I haven’t seen any other terms that better fit
2. Because I think that it is critically important that any talk about data use is tied to measures that encourage the researcher to think about the meaning and value of their data

It’s my hope that this is a continuing discussion. In the meantime, I will trudge on in idealistic ignorance.

Repeating language: what do we repeat, and what does it signal?

Yesterday I attended a talk by Jon Kleinberg entitled “Status, Power & Incentives in Social Media” in Honor of the UMD Human-Computer Interaction Lab’s 30th Anniversary.

 

This talk was dense and full of methods that are unfamiliar to me. He first discussed logical representations of human relationships, including orientations of sentiment and status, and then he ventured into discursive evidence of these relationships. Finally, he introduced formulas for influence in social media and talked about ways to manipulate the formulas by incentivizing desired behavior and deincentivizing less desired behavior.

 

In Linguistics, we talk a lot about linguistic accommodation. In any communicative event, it is normal for participant’s speech patterns to converge in some ways. This can be through repetition of words or grammatical structures. Kleinberg presented research about the social meaning of linguistic accommodation, showing that participants with less power tend to accommodate participants with more power more than participants with more power accommodate participants with less power. This idea of quantifying social influence is a very powerful notion in online research, where social influence is a more practical and useful research goal than general representativeness.

 

I wonder what strategies we use, consciously and unconsciously, when we accommodate other speakers. I wonder whether different forms of repetition have different underlying social meanings.

 

At the end of the talk, there was some discussion about both the constitution of iconic speech (unmarked words assembled in marked ways) and the meaning of norm flouting.

 

These are very promising avenues for online text research, and it is exciting to see them play out.

Getting to know your data

On Friday, I had the honor of participating in a microanalysis video discussion group with Fred Erickson. As he was introducing the process to the new attendees, he said something that really caught my attention. He said that videos and field notes are not data until someone decides to use them for research.

As someone with a background in survey research, the question of ‘what is data?’ was never really on my radar before graduate school. Although it’s always been good practice to know where your data comes from and what it represents in order to glean any kind of validity from your work, data was unquestioningly that which you see in a spreadsheet or delimited file, with cases going down and variables going across. If information could be formed like this, it was data. If not, it would need some manipulation. I remember discussing this with Anna Trester a couple of years ago. She found it hard to understand this limited framework, because, for her, the world was a potential data source. I’ve learned more about her perspective in the last couple of years, working with elements that I never before would have characterized as data, including pictures, websites, video footage of interactions, and now fieldwork as a participant observer.

Dr Erickson’s observation speaks to some frustration I’ve had lately, trying to understand the nature of “big data” sets. I’ve seen quite a bit of people looking for data, any data, to analyze. I could see the usefulness of this for corpus linguists, who use large bodies of textual data to study language use. A corpus linguist is able to use large bodies of text to see how we use words, which is a systematically patterned phenomena that goes much deeper than a dictionary definition could. I could also see the usefulness of large datasets in training programs to recognize genre, a really critical element in automated text analysis.

But beyond that, it is deeply important to understand the situated nature of language. People don’t produce text for the sake of producing text. Each textual element represents an intentioned social action on the part of the writer, and social goals are accomplished differently in different settings. In order for studies of textual data to produce valid conclusions with social commentary, contextual elements are extremely important.

Which leads me to ask if these agnostic datasets are being used solely as academic exercises by programmers and corpus linguists or if our hunger for data has led us to take any large body of information and declare it to be useful data from which to excise valid conclusions? Worse, are people using cookie cutter programs to investigate agnostic data sets like this without considering the wider validity?

I urge anyone looking to create insight from textual data to carefully get to know their data.