Exploring individuals’ patterns of personal information management practices: factors influencing the representation, organization and credibility assessment of information

Introduction

Personal information management refers to both the practice and the study of activities that people perform to acquire, organize, maintain, retrieve, use and control the distribution of information items, including paper and digital documents, online resources and e-mail messages used to accomplish personal and work-related tasks as well as fulfil various roles such as parent, employee, or community member (Jones and Teevan, 2007). Research on personal information management has addressed various aspects of how people find, organize, share and maintain their personal information, including whether and if so, how people use file folder organization and/or assign tags and how they search for their files; the effectiveness of their file folder hierarchies; their utilization of e-mail systems for file storage and sharing; and other strategies of personal information management (Bergman, 2013; Jones, 2007).

Bergman, Beyth-Marom and Nachmias (2003) suggest that personal information management systems reflect the subjective, value-added attributes that the user gives to the data stored in the system and propose three generic principles: the subjective categorisation principle, that all information items related to the same subjective topic should be categorised together regardless of their technological format; the subjective importance principle, that the subjective importance of information should determine its degree of visual salience and accessibility; and the subjective context principle, that information should be retrieved and viewed by the user in the same context in which it was previously used.

Bergman (2013) identified and mapped five categories of variables for personal information management: organization related variables (order, redundancy and name meaning), structure variables (collection size, folder depth, folder breadth and folder size), work process variables (attendance time and modality), memory related variables (memory reliance, dominant memory) and retrieval variables (retrieval type, retrieval success, retrieval time and ubiquity).

Several studies have examined patterns of personal information management by various users, such as academics (e.g., Kwasnik, 1991); managers (e.g., Barreau, 1995), educators (e.g., Diekema and Olson, 2014) and office workers (e.g., Malone, 1983). For example, Diekema and Olsen's (2014) analysis of primary and secondary teachers' personal information management processes found that teachers tend to inherit information from their predecessors. They also observed that information technology supporting education creates inordinate information management demands as teachers are expected to have classroom Websites, maintain records and provide individualized instruction using online educational resources.

Personal information management research highlights the necessity of continued examination of different aspects of individuals' personal information management in various contexts. This study aims to extend previous research and contribute to the literature by investigating the personal information management behaviour of information science graduate students. This study is led by the following broad question: What contextual forces determine the graduate students' personal information finding and their assessment of its credibility?

In this study, Gibson's (1979) understanding of context as ecology and Suchman's (2007) conceptualisation of individuals' actions as situated in the ‘complex world of object, artefacts and other actors, located in space and time‘ (p. 177) were applied to an exploration of the ways in which various tools shape our personal information environments.

Related work

Personal information management is an inherently complex process (Jones, 2007), which involves various strategies to manage information (Fitchett, Cockburn and Gutwin, 2013). Individuals' decisions whether to organize their personal information depend on their assessment of its anticipated future use (Kwasnik, 1989). According to Jones (2007), the key personal management activities include: finding, which moves individuals from a knowledge need to information that meets that need; keeping, which moves individuals from encountered information to expected future needs for which this information might be useful; and meta-level activities, which focus on the mapping that connects information to need and on meta-level issues concerning organizing structure, strategies and supporting tools (p. 493). An individual's personal information management is based on information needs which lead to finding relevant information. Jones (2007) argues that in the context of personal information management the term finding accurately reflects the nature of meeting an information need as it implies finding information in a space that is presumably under individuals' control. Information needs can motivate a range of tasks from a simple search, for instance, of an area code to a more complex pursuit such as compiling a bibliography on a special topic of interest. The process of finding information that meets an individual's need involves searching, sorting, scanning and assessing the relevance of the found information (Jones, 2007). However, meeting the information need often entails a series of interactions or, in Bates's (1989) term, berrypicking, wherein affect and emotions are important constituents of how individuals seek and use information (Nahl, 2007).

A large body of research has examined organization strategies across different types of information, including paper documents (Malone, 1983), e-mail communications (Whittaker and Sidner, 1996), electronic files (Henderson and Srinivasan, 2009) and Web resources (Tauscher and Greenberg, 1997). Analysis of paper organization strategies revealed that individuals tend to organize their documents both by filing and by piling them (Whittaker and Hirschberg, 2001). Moen (2007) observed that people tend to use paper or some other durable medium for their health-related information as well, to ensure that their information is accessible to other household members. They also discovered that filers amassed more information and accessed it less frequently than pilers. Research on e-mail organization revealed a range of approaches to organizing e-mails, including saving and regularly cleaning e-mails (Bälter, 1997); folder filing (Fisher, Brush, Gleave and Smith, 2006) and archiving and filtering (Bellotti, Ducheneaut, Howard and Smith, 2003). Studies have also demonstrated use of e-mail as a prevalent method of file sharing (Whittaker, 2011). Individuals' organization of Web and electronic resources also include filing (Boardman and Sasse, 2004), bookmarking, tagging and sharing (Heckner, Heilemann and Wolff, 2009).

Rapid development of information communication technologies has not only created new opportunities to access and store a vast amount of information across various platforms and devices, but also challenged the process of managing personal information. Bellotti and colleagues' (2005) analysis of individuals' e-mail use, for example, showed that a single e-mail communication may initiate a range of tasks involving different software applications and requiring significant amounts of time to accomplish tasks. Bergman, Whittaker and Falk's (2014) large-scale investigation of differences between group and personal information management and retrieval found that subjects tend to be more successful in sharing and retrieving their files in personal information than in group information management systems. The question is how individuals' increasingly connected world affects their personal information management, in particular, finding and assessing the credibility of information accessible through various platforms, application and devices.

There are a number of factors that come into play in the selection of a credible source to satisfy an information need, including source quality, accessibility, trustworthiness, and such user-related factors as age, sex, and health status (Rowley, Johnson and Sbaffi, 2017). Research suggests that trustworthiness and expertise (such as domain expertise, information skills and source experience) (Lucassen and Schraagen, 2011) are key criteria of assessment of information credibility (Flanagin and Metzger, 2007; Lucassen and Schraagen, 2013). Metzger and Flanagin's (2013) review of research into trust formation and credibility evaluation suggests that users relate to reputation, endorsement, consistency, self-confirmation, expectancy violation and persuasive intent. Hargittai, Fullerton, Menchen-Trevino and Yates Thomas (2010) found that for undergraduates, the main factors of credibility assessment include identifiability and currency of information, availability of other sources for validation, factual or editorial nature of content, legitimacy of authorship and provision of site linkages.

In the context of Google's search engine, evaluation of information may also depend on the user's system of perceived ease of use (Johnson, Rowley and Sbaffi, 2016) and trust in search engines (Andersson, 2017). However, compared to more traditional sources, the assessment of credibility of online information can be more challenging because of a lack of structural features and expert gatekeepers to monitor the content and the overall converged nature of online information, such as news and advertisement (Flanagin and Metzger, 2000). In this regard, Rieh and Hilligoss (2007) view credibility as relative to the social context in which information seeking is pursued and credibility judgments are made. Rieh's (2014) analysis of individuals' credibility assessment as a process-oriented notion, which incorporates various information use contexts, revealed that the score-based trust perceptions provide a limited view of credibility perceptions. The author also observed the importance of topicality of information when it comes to assessment of user-generated rather than traditional media content.

There is also increasing concern over obscurity of the algorithms behind the search engines, such as Google's ranking of the search results, suggested search terms, and the consequent influence on individuals' ways of getting informed and of knowing. Huvila (2016) argues that as search engines ‘conveniently provide us with reasonably good answers, it feels counter-intuitive to be too critical of their shortcomings' (p. 576). Besides, ‘reasonable goodness of answers is determined less by a purely rational choice and meticulously analysed evidence than a personal feeling of adequateness' (p. 576).

The process of personal information organization manifests individuals' cognitive understanding of encounters with the environment and mental ability to partition space and create categories by separating things from the context in which they are embedded (Zerubavel, 1991). Rosch (1978) reasons that cognitive economy is central to the primary functionality of a category system: ‘[T]he task of category systems is to provide maximum information with the least cognitive effort' (p. 28). In this regard, Oh (2017) has identified three different types of personal digital information categorisers: (i) rigid categorisers, who employ well-developed structures and rules to organize personal information; (ii) fuzzy categorisers, who lack folder structure and rules to organize their information and (iii) flexible categorisers, who use developed folder organization but exercise flexibility in using, changing and maintaining their information organization structure as well as possessing uncategorised or loosely categorised information.

Barsalou (1987) demonstrated that individuals construct cognitive representations for achieving goals, wherein an individual may have multiple goals associated with a category. Barsalou (1987) contends that how things are categorised will vary across individuals because they use different features to represent the same category in different contexts at different times. Barsalou (1985) argues that categories have graded internal structure or a ‘continuum of category representativeness, beginning with the most representative members of a category and continuing through its atypical members to those nonmembers least similar to category members' (pp. 629-630). Furthermore, the graded structure is an important factor in predicting many categorisation tasks: how long it takes to categorise an entity as a member of category, the frequency with which people generate members of categories and how easily people learn categories and make category decisions (Barsalou, 1987, p. 103). Barsalou (1985) also argues that graded structures within categories do not remain stable across situations but can shift substantially with changes in context, including linguistic contexts.

However, there has been limited research investigating how individuals find and assess personal information. This study aims to extend research on personal information management by considering contextual factors affecting the graduate students' criteria for finding and evaluating credible sources of information to fulfil their personal information needs.

Theoretical framework

Barreau (1995) observes that personal information management depends on a context in which a ‘person works, including nature of the task, the subject matter, the available tools, the intended audience for the work and the intended use for the information and factors which are in play at the time' (p. 329). The term context has been conceptualised in various ways, from a basic understanding of it as a concrete situation or environment (e.g., Barreau, 1995) to a more nuanced reference to an individual's physical and/or conceptual engagement with an entity of interest.

Hilligoss and Rieh's (2008) framework of information credibility assessment considers context as a main factor influencing individuals' assessment of credibility of information in terms of constructs, heuristics and interaction. The authors argue that people's perceptions of context influence credibility judgments at the interaction level in fundamental ways that go beyond information source, content and visual appearance. Rieh (2014) characterises context in relation to users' goals and intentions, topicality of information and information activities. Gwizdka and Chignell (2007) argue that the ‘combination of the four factors – people, tasks, tools and context – creates a huge diversity in PIM behaviour, with considerable scope for individual differences to express themselves in how work gets performed in various circumstances' (p. 209).

This research adapts Gibson's (1979) understanding of context as ecology to establish the way in which various tools shape our personal information environments. Gibson argues that people, like all animals, always perceive objects with which they interact within their environments in terms of their affordances:

The perceiving of an affordance is not a process of perceiving a value-free physical object to which meaning is somehow added in a way that no one has been able to agree upon; it is a process of perceiving a value-rich ecological object. (p. 60)

Gibson's conceptualisation of affordances has been adopted across various domains of human-computer interaction to explain embedded and emerged affordances of objects (Leonardi, 2011). As information communication technologies have become pervasive in people's daily lives, the theory of affordances has been increasingly utilized to explain the relational nature of actions that emerge from interaction with objects among people and technologies (Faraj and Azad, 2012). Suchman (2007) suggests the dialogical relationship of ‘knowledge and action to the particular circumstances in which knowing and acting invariably occur' (pp. 178-179). For Suchman, individuals' actions are situated in the ‘complex world of object, artefacts and other actors, located in space and time' (p. 177), wherein the complex world is an ‘essential resource that makes knowledge possible and gives action its sense' (p. 177). For Suchman, configured materiality including dialogical and reciprocal transformative power, is inherently discursive and co-constituted with configuring agencies, which, in turn, introduces a very different understanding of the human-machine interface as a ‘category of contingently enacted cuts occurring always within sociomaterial practices, that affect "person" and "machines" as distinct entities and that in turn enable particular forms of subject-object intra-actions' (p. 268). According to Suchman, the perceived ‘singularity of the interface explodes into a multiplicity of more and less closely aligned, dynamically configured moments of encounter within socio-material configurations, objectified as persons and machines' (p. 268).

Applying Gibson's (1979) formulation of affordances and Suchman's (2007) conceptualisation of situated action, this study is an attempt to shed light on individuals' personal information management, not in relation to the context of a particular tool, service or device, but rather as an ecosystem wherein individuals organically engage with a range of individuals, tools, services and devices to make sense of their daily information interactions. Individuals' cognitive, affective, sociocultural and sociopolitical states, coupled with system affordances, constitute the contextual forces that determine how individuals navigate various interactions with human and non-human agents.

Research methods

In this study, criterion sampling (Patton, 2002) was used to collect data from graduate students who were willing to share insights into their personal information management. Criterion sampling involves reviewing and studying a body of cases that meet some predetermined criterion of importance (Patton, 2002). A predetermined criterion of this study was inclusion of information science graduate students who were open to sharing details of their personal information interaction and management.

Participants were thirty-seven graduate students in the college of communication and information: 28 females, 9 males; ages 21-53, with an average of 29 years. All participants of the study were information science graduate students who were taking one of the programme’s required courses.

Naturalistic inquiry was implemented as a holistic, contextual and discovery-oriented approach to minimize researcher intervention in the study setting and places (Guba, 1978) and focus on the natural behaviour occurred (Naumer and Fisher, 2007). For a period of two weeks participants were asked to keep a daily log of their personal information interactions and record the four most representative actions of their choice.

After completion of this phase, participants were asked to complete a questionnaire eliciting details of their approaches to finding and keeping information and their meta-level activities, such as organizational strategies and information credibility assessment in relation to their information needs. Subjects were asked to reference their log to reflect on the emerged patterns of their personal information management in the questionnaire. In total, the sample of this study comprised 2,072 logged actions and thirty-seven questionnaires.

Constructivist grounded theory (Charmaz, 2014) was used to analyse participants’ logs and follow-up responses. The constructivist approach treats research as a construction but acknowledges that research occurs under particular conditions, of which one may not be aware and which may not be of one’s choosing (Charmaz, 2014, p. 13).

The kinds of information participants logged in their diaries was looked at first, their questionnaire responses analysed in relation to reported actions to understand how they found, interacted with, evaluated and kept or discarded information. Beginning with inductive logic, the researcher and a research associate engaged in close reading and systematic comparison, by which means they generated successively more abstract concepts through an inductive process of comparing data, categories and concepts (Charmaz, 2014).

To assess inter-rater agreement, 40% of the responses were coded by the author and a research assistant. Inter-rater agreement was satisfactory for all content analytic variables: patterns of file and e-mail organization 85%, finding information 94%, and assessing credibility 84%. All coding disagreements were resolved through discussion and unified coding was applied to all responses by one researcher (see Table 1).

Results

Patterns of file and e-mail organization

Participants of this study were found to utilize a range of strategies to satisfy their immediate as well as long-term information needs, from simple mental notes to remember certain action to complex strategies to manage information to ensure its future discoverability. Several participants noted the use of a hierarchical folder system to organize their personal files (ID1; ID2; ID11; ID12; ID18). Affordances of systems and accessibility of storage seemed to be underlying motivations for keeping information. As one of the participants noted,

I probably save too much information in (very well described) files, but do not see that as a problem yet. Storage is cheap and it is relatively easy to maintain (ID1).

I tend to have more rigorous filing system of my physical personal financial and medical records, such as specifically labelled folders. But I end up with a lot of piles of documents at work (ID37).

Several participants (ID2; ID26; ID37) preferred to navigate (Whittaker, 2011) their file system and/or utilize both navigation and search, using such cloud-based platforms as Google drive (ID22). Nonetheless, they employed various strategies for facilitating retrieval of resources in the future using either affordances of platforms or their own cognitive capabilities. For example, ID22 used the star feature on frequently used Google documents for easy navigation. Others (ID10; ID12) kept required information open on the Internet browser:

I tend to leave the browser tabs open as long as I need rather than archiving. But it helps to keep my attention on things I’m working on (ID12).

Another participant mentioned a tendency to reinforce memorization of online resources or placing notes on screen rather than saving online resources (ID1). Several participants mentioned bookmarking online resources for easy retrieval in the future (ID1; ID3; ID4; ID26).

In line with previous studies exploring individuals’ e-mail organization patterns (Whittaker 2011), participants were engaged in activities such as allocating attention for handling the e-mail messages, deciding actions in response to incoming e-mail communications, managing tasks and organizing e-mail communications and folders (ID19; ID37). Participants were engaged in folder organization of their e-mail communications (ID9), filter settingand managing separate e-mail accounts for various purposes, such as work and shopping (ID4). In this regard, one of the participants noted use of folders and tags to organize e-mails:

My primary e-mail account includes both important and unrelated e-mails which I struggle to sort and organize, I try to use both folders and tags to organize my e-mails. My folders are task based wherein tags are topical (ID4).

Several participants used their mailbox as ‘an active workspace‘ (Whittaker, 2011, p. 19) by keeping items that needed their attention in the mailbox before organizing their e-mails in folders (ID11; ID14; ID18; ID20; ID37). One mentioned using such e-mail application features as flagging or marking e-mail as unread or as a reminder to handle action requiring e-mails (ID22).

Some of the participants noted they keep their e-mails in active mailbox and, then, regularly perform spring cleaning (ID9; ID11; ID23; ID24). Whereas some of the participants (ID11; ID18) noted extensive e-mail organization strategies and regular e-mail cleaning (ID20) to make sure that they would be able to retrieve potentially needed e-mail communication in future. In line with the previous studies (Barreau and Nardi, 1995; Whittaker and Hirschberg, 2001), participants (ID19; ID37) also noted the challenge of handling their volume of e-mails and uncertainty as to whether communications saved for future use will actually be used:

I repeatedly transfer e-mails from my inbox to other folders ‘for future use.’ It is often unlikely that I may retrieve these e-mails ever again. But if I needed to, how would I go about fetching them? (ID31).

Search is considered as the key capability that makes personal information management useful and it comes as no surprise that participants relied on both folder navigation and search when it came to find needed communications (ID22; ID31). As one noted,

often, I will know an e-mail is in a certain folder and then use the search feature within that folder (ID31).

Participants also used e-mail to send useful resources to themselves (ID3) or as reminders:

I failed to internalize how much I use my e-mail to communicate with myself, e.g., sending Web links to peruse later, or sending more structured reminders telling me to, say, recharge my Fitbit (ID34).

Finding information and assessing credibility

Some participants (ID34; ID36) reported preferring to re-find information rather than preserve files:

I tend to rely on a quick Google search or streaming via YouTube or SoundCloud more than archiving (ID9);

I just look up information as many times as I need to I do not really put much effort to organize my information (ID10).

One participant referred to the facility in finding needed resources as a generational feature:

As digital native I’m able to process information faster and we live in age when much more accessible that it ever has before (ID10).

Some reported relying on their memories of action items:

I simply keep information on what I have to do in my mind and took action based on the information stored there (ID5).

Participants’ information finding activities can be characterised as situational depending on their information needs as well as affective states (Dervin, 1998). Several participants recognized the habitual nature of their personal information seeking and organization and their preferences for certain sources (ID1; ID2; ID19; ID30; ID32):

I am a creature of habit and my information interactions are very predictable. Although I will occasionally do something out of the ordinary, I generally keep to daily or at least weekly routines (ID1).

This personal information management pattern resembles Chatman’s (1991, 1992) conceptualisation of everyday information seeking as a small world in which people are mainly interested in information ‘that is perceived as useful, that which has firm footing in everyday reality and responds to some practical concern’ (Chatman, 1991, p. 447).

For most of the participants, information seeking was collaborative and social due to reliance on system affordances such as Pinterest, YouTube and Google search engine suggested search terms (or autocomplete). (ID8; ID19; ID20; ID22; ID23; ID26; ID27; ID28; ID29; ID30; ID31; ID32; ID35). In line with previous studies, a number of participants noted reliance on the affordances of social network sites to keep, find and share their personal information, such as Instagram photos (ID1; ID2; ID3; ID12; ID10). Suggested terms were considered useful in refining and correcting spelling of search terms (ID30; ID32), as well as a source of credible information as it was assumed that the terms were based on queries by users who had experienced similar information needs (ID22), which also connected users effectively with those with similar information needs:

Google suggested terms help to associate with those who have similar queries, to learn about trending searches, and to connect with the world, even if I do not and will never know who those people are… (ID35).

However, suggested terms were also considered distracting as they tended to sway attention away from an original search intention (ID15).

Researchers have found that people are likely to engage in online health information seeking to verify and/or challenge information provided by medical professionals, especially when they experience incongruences and/or uncertainties (Lin, Zhang, Song and Omori, 2016). Choudhury, Morris and White (2014) found that users frequently not only utilize search engines to learn more about serious or highly stigmatized conditions, but also share the sensitive health information on Twitter. Participants of this study have also emphasized value of Google suggested terms in seeking the health-related information (ID1; ID7; ID12; ID14; ID15; ID17; ID18; ID35; ID36). As one participant noted:

Medical terms are so difficult to spell, suggested terms are extremely helpful (ID32).

Another appreciated the superior phrasing of experts’ suggested medical terms and questions:

I struggle to define medical search terms, and those who are more informed about these issues than I am are more likely to have phrased my question in a better, more Google-friendly format (ID35).

Suggested terms were also generated from the previous individuals’ searches and helped to effectively connect with those who faced the similar issues:

Google suggested terms help to associate with those who have similar queries, to learn about trending searches, and to connect with the world, even I do not and will never know who those people are… (ID35).

Participants paid attention to health-related information they collected and monitored through wearable technologies, such as FitBit (ID32; ID33; ID34; ID35). However, some expressed uncertainty concerning how their data was being collected and used and whether they were sufficiently attentive to what information they were accumulating:

There are many critical details of my life that FitBit now has access to and it would make sense for me to determine what kind of data is being captured by my input and how it is being used. I will actively try and be more discerning about the kinds of information that will actually help me in the future versus just saving a document or e-mail because I can. (ID33).

All the participants in this study applied various techniques to assess the credibility of information with which they interacted. Reliability of the content was considered a primary criterion, including source (ID2; ID8), authority of the author (ID21; ID23) and date (ID2) as well as design and format of the resources (ID23). However, credibility assessment was considered highly perceptual and situational (Gunther, 1992); for instance, one participant (ID28) pointed out that when shopping on a new online store,

I evaluate images… if they [items] all look similar in pose or quality, I would trust the site. But if there are a variety of images, such as with and without models, I would be less likely to trust (ID28).

Assessment of credibility seemed to be significantly influenced by opinions of participants’ immediate social groups as well as general public or the crowd (ID6; ID16; ID24; ID 26; ID28). For example, one noted:

I have a basis for trusting my family members as they have been a reliable source of information in the past, so I always trust the information they provide (ID5).

Another participant noted:

As I have limited time, I usually rely on sources that were deemed credible by my social circle (ID11).

Reliance on a wider domain of users seemed to reflect the social nature of credibility assessment:

When I get information that I am unfamiliar with I search to see what people have to say. If it is credible you can see from the reviews and the lack of complaints (ID20).

Another participant considered an amateur forum as a credible source:

to see if anyone else had experienced the same situation. It was one of the first searches to appear that clearly addressed and identified the problem I experienced (ID10).

A number of participants emphasized that they considered information credible because they over time had developed trust in the individual(s) and/or businesses providing the information (ID1; ID6; ID8):

I am a long-time subscriber to [Website]… and I trust their information. If they get facts wrong, they usually address it and welcome readers to write and express their opinions (ID1).

Another participant noted the importance of credibility to the source:

Any information obtained from a well-known source, such as Amazon or a school, I consider credible because if it is not true that would damage their reputation (ID21).

Most of the participants mentioned they trust faculty and those who were parents expressed their reliance on information provided by school teachers:

I also inherently trust the credibility of the information presented by the faculty because I recognize them as authorities (ID30).

They have also expressed strong reliance on particular platforms and/or devices,

I trust my phone’s GPS because I have used in multiple occasions and arrived at the places I needed with little fault (ID3).

While some participants regarded user-generated reviews and ratings provided by publicly known parties as more trustworthy than personal sites (for example: ‘I would not trust a personal blog or a personal site to provide me with the correct information’ (ID 34)), others took mixed approaches to comparisons of user-generated versus authoritative sources of information:

My highest confidence was with content created by myself, people I know, or people I could see creating the content (for example, YouTube videos). I also had a high level of confidence in content created by trusted names I had previously used, including … Google Maps, the Mayo Clinic and Goodreads (ID25).

Some also expressed being cautious with information available on social media (ID19), such as YouTube:

when I search for a YouTube video, I’m predisposed to research the results (ID9).

Participants also emphasized the emotional and cognitive aspects of assessing the credibility of information:

evaluation of presented materials are based on my own intuition. I think the credibility is something that is almost subliminal and not really deliberate unless it is a new source of information (ID19).

Information was also considered credible based on a previous positive experience as well (ID5; ID13):

I skim each link before I find one that I know from previous experience is most likely correct (ID13).

Discussion

This study was an attempt to understand individuals’ personal information management as an ecosystem wherein they organically engage with other agents, tools and services to make sense of their daily information interactions. Application of Gibson’s (1979) conceptualisation of affordances and Suchman’s (2007) understanding of situated action revealed that individuals’ personal information management involves a constellation of techniques, processes and practices that are situated in their sociocultural and sociotechnical environments (see Table 2).

Thus, this study observed that individuals utilise a range of strategies to satisfy their immediate as well as continuing information needs, from simple use of the ‘star’ feature on frequently used Google documents for easy navigation, to complex strategies of file organization and storage to ensure future discoverability of the information. In line with Hilligoss and Rieh’s (2008) study, this paper suggests that credibility assessment goes far beyond the individual level of interaction and involves a number of contextual factors that affect assessment, including cognitive, emotional and social aspects. Consistent with previous studies (Marchi, 2012), this study found that graduate students embrace both strong and weak social ties to make sense of their daily information. For example, one of the study participants rely on information provided by the family members:

I have a basis for trusting my family members as they have been a reliable source of information in the past, so I always trust the information they provide (ID5).

Another participant expressed reliance on a wider domain of users or crowd:

When I get information that I am unfamiliar with I search to see what people have to say. If it is credible you can see from the reviews and the lack of complaints (ID20).

Observed patterns of strong reliance on crowd opinion seems to suggest that individuals tend to make decisions informed by biased perspectives such as reviews found to be positively weighted (Hajibayova, 2019) and product reviews are often misleading (Askay, 2014). As found in previous studies (Waddell, 2017), we observed that individuals’ reliance on crowd opinion often shapes their credibility judgment as well as their perceptions of the importance of information with which they interact. The observed lack of a structured approach to assessment of credibility of personal information raises some concern about potential impact on graduate students’ worldviews and decision-making processes. The increasing general perception that internet platforms as curators of internet knowledge, Google search engine (Liu, Zamir, Li and Hastings, 2018), are sources of knowing and curators of internet knowledge calls for more transparency of the manoeuvrings behind the ranking of relevant resources and provision of search terms. There should also be more transparency of user-generated reviews across various platforms, sources and relevance rankings of the reviews.

Conclusions

In an increasingly complex world where individuals encounter a vast amount of information daily, understanding people’s personal information management across various contexts, platforms and devices is crucial for designing effective information systems that would potentially improve individuals’ productivity as well as overall quality of life.

Investigation of a relatively diverse population of social science graduate students shed light on the complex and nuanced nature of personal information management, wherein various contextual factors influence graduate students’ decisions to find, keep and/or re-find the information. Application of Gibson’s (1979) formulation of affordances and Suchman’s (2007) understanding of situated action revealed that individuals’ personal information management are situated in their sociocultural and sociotechnical environments. This study observed the situational and affective nature of graduate students’ personal information management strategies, which are often interwoven with system affordances. The participants’ expressed reliance on their social groups as well as on the crowd in finding credible sources of information implies the complexity of the sense of trust, whether trust in those with whom they have strong ties, e.g., family and friends, or trust in those whom they perceive as having shared experiences. The observed reliance on those whom they presumably had shared experiences suggest the blurred perceptions of the crowd-generated opinions of utilized platforms as human-generated versus the precision of the algorithmically generated recommendations. For instance, one of the study participants appreciated the superior phrasing of crowds’ suggested medical terms and questions:

I struggle to define medical search terms and those who are more informed about these issues than I am are more likely to have phrased my question in a better, more Google-friendly format (ID35).

This blurred understanding of platform-recommended terms, reviews and other related information and their potential impact on individuals’ everyday lives call for more investigation of transparency of platforms to ensure ethical and accountable information services.

This study is subject of several limitations. In particular, the daily log technique and follow-up questionnaire provided in-depth understanding of the situated nature of individuals’ daily information interactions and management; however, such records are potentially subject to participants’ bias and management of impressions by selecting certain characteristics and minimizing others depending on the audience for whom they are performing (Goffman, 1959). Future research may study more diverse population and include various methods, such as interviews, to better understand contextual factors underlying individuals’ personal information interactions and management strategies.

Acknowledgements

Heartfelt thanks go to Dr. Pugh for editing the multiple iterations of this work, the editor and reviewers for the constructive suggestions that helped to improve this research.

About the author

Lala Hajibayova is Assistant Professor, School of Information, Kent State University, PO Box 5190, Kent, OH 44242-0001, USA. She received her PhD in information science from Indiana University. She investigates how individuals' contextualized experiences of engaging and representing artefacts facilitate cultural and scientific productions as well as design of human-centred systems that embrace a multiplicity of views and comply with ethical norms. She can be contacted at lhajibay@kent.edu

How to cite this paper