SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. PRELIMINARY RESULTS
  6. CONCLUSION
  7. Acknowledgements
  8. REFERENCES

Serendipity has received much attention from library and information science, psychology, and computer science. Yet not much is known about serendipity in the context of everyday information behavior. In general, a key challenge in the study of serendipity is obtaining accounts of serendipitous experiences that provide insight into the phenomenon. The exploratory research reported here approaches this problem by examining naturally occurring descriptions of serendipity as found on blogs. The paper shows how these data can be collected, stored, and analyzed. We also discuss strengths of the proposed approach in comparison to the use of descriptions elicited in controlled settings for the purposes of research. Through a grounded theory approach, we develop a model of serendipity that can inform the design of information systems. The paper contributes to the LIS field by discussing an alternative data collection method for serendipity research, outlining a tentative conceptual model of serendipity, and showing the utility of this model for the analysis of everyday accounts of serendipity found on blogs.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. PRELIMINARY RESULTS
  6. CONCLUSION
  7. Acknowledgements
  8. REFERENCES

Accidental encountering is an integral part of everyday information behavior (Erdelez, 2004). Some researchers, however, have expressed concern that opportunity for this type of encounter might be reduced because of technological facilitation of information behavior (Thom-Santelli, 2007). In response, work in information systems has proposed the development of interfaces that support or enable serendipity (Toms and McKay-Peet, 2009; Thom-Santelli, 2007). The effective integration of serendipity in information technology requires an understanding of how people experience serendipity in everyday environments.

Although there is a long history of interest in serendipity, little work has addressed this specific question. Psychological researchers tend to focus on serendipitous scientific discovery rather than on everyday serendipity (Rosenman, 1988). Much of the work in LIS has focused on serendipity in the academic context (Foster and Ford, 2003). Other work has examined descriptions of serendipitous information behavior (e.g., Erdelez 2004). Less attention has been paid, however, to the context in which serendipity occurs: what conditions promote a serendipitous encounter, what types of information are gained in these encounters, and what makes an encounter serendipitous.

Much of the past research on everyday serendipity has relied on elicited descriptions. As Erdelez (2004) notes it is difficult to evoke complete information encounter episodes using this method. We propose to address this difficulty by analyzing non-elicited, natural descriptions of serendipity collected from online blogs using keyword searches. We use an iterative grounded theory approach (Strauss, 1987) to analyze these postings with respect to potentially relevant facets identified on the basis of previous research (e.g., Erdelez, 2004; Foster and Ford 2000).

Our paper has the following three goals: 1) to test the effectiveness of an alternative data collection method for serendipity research; 2) to propose a preliminary conceptual model that outlines the facets of serendipity; and 3) to better understand serendipity in the context of everyday information behavior.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. PRELIMINARY RESULTS
  6. CONCLUSION
  7. Acknowledgements
  8. REFERENCES

We conducted a variety of keyword searches relevant to serendipity on Google Blog between April and August 2010. The search queries used terms occurring in the serendipity literature or in natural language descriptions. Forty-seven variations of natural queries were used. We included lexical and syntactical variations as well as Boolean operators and wildcards (for examples see Table 1). The queries were expanded and reworded based on query results obtained in our blog searches. The keyword searches returned a large number of reports of potential serendipitous encounters. These were reviewed to identify rich descriptions of serendipitous accounts, resulting in 94 accounts that constituted our dataset.

Table 1. Sample Search Queries Used to Retrieve Serendipity Incident Reports from Blogs
1“wasn't actively looking OR searching for” “when * found”
2“found OR discovered ** by accident OR serendipitously OR by chance”
3“had a eureka moment”
4“light light bulb went off” “connection”
5“put the two together”
6“that made me reali[s/z]e”
7“that's when I made the connection”
8“out of the blue I found *
9“and then it hit me”
10“I had an aha moment” “connection”

PRELIMINARY RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. PRELIMINARY RESULTS
  6. CONCLUSION
  7. Acknowledgements
  8. REFERENCES

Using a grounded theory approach, we analyzed the descriptions in the resulting dataset based on facets identified from the prior literature. Based on an iterative process, we identified the following relevant facets: 1) prior concern or information need; 2) activity that is reported at the time of the encounter, possibly interrupted by the serendipitous event; 3) the information or object that is serendipitously encountered; 4) an element of surprise – what is surprising about the information, object, or the encounter itself; 5) finally, we examine the relevance of the encountered object or information to the prior need and discuss how serendipitous information can extend prior thinking or alter perspective. Our next step in the data analysis is to identify whether or not these facets are reflected in naturally occurring accounts of serendipity, and, if so, what values they take in these descriptions. Finally, we wish to identify the natural co-occurrence of these values in blogger descriptions, with the goal of developing a typology of serendipitous encounters.

The following description, typical of the data collected, was the result of search query 1 in Table 1. This account demonstrates a prior need (materials for jewelry-making), an activity at the time (shopping for other items), discovery of a relevant item (the materials that are required), an element of surprise (it was not known that the items were available at that location), and relevance to the prior need (satisfies a prior need for an object):

I wasn't actively looking for sales when I found my findings. I stopped at on a whim. Both times we were on a family outing, when I spotted a sign and asked to stop. I had been on a quest for the right size and color of dominoes. I was still in thinking, “I'll give it another try”. At one particular estate, I was getting ready to pay for my broken vintage jewelry treasures when the cashier asked me if I make jewelry. “YES”, I said and elaborated, “Not only that, but I have friends who also make jewelry and other things out of bits and pieces of pretty vintage things.” The cashier proceeded to pull out a box full of findings. She offered them to me at a price I couldn't refuse.

CONCLUSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. PRELIMINARY RESULTS
  6. CONCLUSION
  7. Acknowledgements
  8. REFERENCES

Our search for incident reports showed that bloggers reflect upon their experiences and circumstances of serendipitous encounters, allowing researchers to construct a rich data set on serendipity from productions in social media environments. These data have several strengths in comparison to data obtained in controlled research settings: 1) they are freely and publicly available online, 2) created by bloggers independently of the study, and 3) are written by self-motivated writers for an unknown audience.

Analysis of these descriptions allows us to identify critical contextual facets associated with serendipity. Ultimately, the results of this research will inform the development of information interfaces that support serendipitous discovery.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. PRELIMINARY RESULTS
  6. CONCLUSION
  7. Acknowledgements
  8. REFERENCES

We would like to thank Jeremy Clark for data collection and management. This study was funded by Social Sciences and Humanities Research Council of Canada Grant R3603A07 awarded to A. Quan-Haase.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. PRELIMINARY RESULTS
  6. CONCLUSION
  7. Acknowledgements
  8. REFERENCES
  • Erdelez, S. (2004). Investigation of Information encountering in the controlled research environment. Information Processing and Management, 40, 10131024.
  • Foster, A., & Ford, N. (2003). Serendipity and information seeking: an empirical study. Journal of Documentation, 59 (3), 321340.
  • Rosenman, M.F. (1988). Serendipity and scientific discovery. Journal of Creative Behavior, 22 (2), 1328.
  • Strauss, A. (1987). Qualitative Analysis for Social Scientists. New York: Cambridge University Press.
  • Thom-Santelli, J. (2007). Mobile social software: Facilitating serendipity or encouraging homogeneity? Pervasive Computing, 6 (3), 4651.
  • Toms, E., & McCay-Peet, L. (2009). Chance encounters in the digital library. In M.Agosti, J.Borbinha, S.Kapidakis, C.Papatheodorou & G.Tsakonas (Eds.), Research and Advanced Technology for Digital Libraries (pp. 192202). Berlin: Springer-Verlag.