Sampling and Recruiting

Planning started in November 2019, and we spent some time choosing an adequate sample and a useful diary format that reflected our needs and the diarists’ work routines. Our sampling strategy aimed at adequately representing basic patron groups in our sample. At the same time, we were going for maximum heterogeneity regarding those attributes we could not represent proportionally and those whose distribution among our readers we did not know. Accordingly, we looked for researchers at different stages of their careers, students, professional users with diverse backgrounds, and private readers, that is people who visit the library mostly for personal reasons such as exploring their family history. In addition, we considered people representing various disciplines and research subjects, language backgrounds, preferred working environments, disabilities, parenthood status, and experience in using the library. Balancing time schedule, effort, and expected saturation—the point at which additional participants would confirm existing data rather than providing new information—we aimed to recruit fifteen participants.

In addition to a call on the library’s website, blog, and social media sites (Twitter, Facebook, and Instagram), we created posters for various sites in the library and placed 100 flyers on desks in the reading room, at the entrance of the cafeteria, and at the information desks. See appendix A for recruiting materials. Using different online and offline communication channels was a good way to recruit participants, as registrations reached us during the whole enrollment period—about two months—via all those channels. Diarists later recommended toilet doors as another effective way to display information for library users.

As an incentive, we promised small surprises during journaling, a guided tour behind the scenes of the library, and a drawing for use of a carrel for one week. However, interviews with participants later showed that most did not notice these incentives but were motivated by personal interests such as the desire to participate in the improvement of library services and curiosity and interest in the diary method. Others expressed feelings of commitment and obligation to the library.

During the enrollment period we realized that we had not included private readers and patrons with disabilities in the sample. Hence, we published calls aimed specifically at these groups and succeeded in adjusting our sample. In the end, we received eighteen commitments for taking part in the diary study. Following our sample strategy and taking into account possible dropouts, we chose sixteen people to participate. Our sample largely corresponded to the library’s basic user groups, but there were fewer private readers and professional users without a university background compared to all registered users. Regarding other attributes such as language, age, library experience, disabilities, or parenthood status, we succeeded in creating a very heterogeneous sample. All in all, recruitment turned out to be much easier than expected as we had slightly underestimated our users’ willingness to participate.

Diary Format

Our decision to use the diary format followed from one premise: our choice of format should not disadvantage any participant. Hence, we would not conduct a purely digital diary study, as we did not expect everyone to come to the library with the required devices. It turned out that two of the diarists sometimes worked without mobile devices or laptops in the library. On the other hand a strict pen-and-paper approach promised to be elaborate and demanding for the research team as we would have to transcribe all data before starting the analysis. Taking into account additional factors such as budget restraints and GDPR requirements—a European Union regulation that unifies the rules governing the processing of personal data by most private and public controllers—as well as participants’ preferences, we decided on a pen-and-paper approach. For those who preferred, we offered a simple digital template as a diary structure. However, no diarist requested it. Furthermore, most diarists were happy about the pen-and-paper approach as it set journaling apart from their own research work. An entry in the diary did not interrupt their computer work because the diary always lay next to them, and they shifted to a different writing style when making an entry.

We asked the participants to keep a diary for three days, as we did not want to strain their engagement, and we expected less new information each day. We did not expect diarists to write on three consecutive days, especially if that did not correspond with the frequency of their visits; instead, we asked diarists to choose three authentic and, ideally, typical days in the library.

Every morning the library’s information desk provided the diarists with a small package in a colorful shiny envelope that contained a pen, paper for taking notes, an information sheet including our contact details, and a floor plan of the reading room to mark the chosen seat. We also included items for a little fun such as a card with fun facts about the library (e.g., highest reminder fee of all times), gummy bears, and sticky notes. Every legal pad had a handwritten welcome note on the first page, followed by observational tasks for the day and an example of our preferred diary structure. Day 1 started with the basic diary structure for participants to use throughout the study: participants should briefly describe every activity, noting its location and start and end times. We asked participants to enhance these entries on Day 2 with an evaluation of specific locations with respect to the activities they undertook there. On Day 3 participants added their own ideas concerning the library space and the library services—whatever suggestions for improvement they had in mind, we wanted to know. See appendix B for examples of these prompts.

Data Collection

We conducted the survey January to March 2020. This included an introductory interview with every participant, three days of journaling, and a debriefing interview with every diarist. The initial semi-structured interviews gave us insight into individual routines and preferences when using the library and when doing intellectual work in general. This included the kind of work participants wanted to accomplish with the support of the library (e.g., finishing a PhD thesis, writing a novel, etc.) and the reasons they preferred to do that in the State Library instead of other places, such as other libraries, a coworking space, home, or their offices. We asked participants to describe the course of an ordinary day in the library, their preferred spots in the library, and how often and at which time of day they usually spent time in the library. In addition to providing us with valuable context, these interviews let us explain the study’s aim and design, which stimulated interest and commitment among the participants. The interviews created an atmosphere of trust between us and the diarists, so they felt comfortable leaving their journals with us.

Overall, the diary data proved to be very useful for the analysis, and most participants felt at ease with the journaling approach. Obviously, they were familiar with the use of diaries as tools for reflecting and documenting, as Hyers (2018) described. However, for the analysis it would have been useful if all diarists had stuck more closely to the structure we had proposed. We learned that the mixture of a semi-structured diary is fine as long as diarists maintain the structure in a strict way—an aspect we did not point out strongly enough during the initial interviews.

Concerning the survey period, three days were fine but two days might have been enough as well. Some participants mentioned afterwards that they somewhat fell into journaling and spent a little too much time with their diaries because they liked to reflect. Others pointed out that days in the library are much alike and that they could not add a lot of information on Day 3. This aligns with our experience that the density of information reduced on Day 3. Two days of journaling might produce the same new value and reduce time for data cleanup considerably. Another possibility to reduce that time could be a limit of ten diarists instead of fifteen, if the sample still shows sufficient heterogeneity.

The semi-structured final interview after the journaling phase left room for queries from both sides, facilitated joint reflection on the diaries, and ensured methodological feedback for the project team. Two interviewers conducted all interviews, one responsible for interviewing and the other one in charge of documentation. The interviews were especially helpful for more deeply understanding the participants and their library use, but they were also time consuming, which, of course, we could have anticipated. Fortunately, a few close colleagues-in-training stepped in and were of crucial help. Lesson learned. While we still would opt for two interviewers instead of one with additional recording, it makes sense to spread out the start dates of the diary phase for the individual participants over several weeks.

Data Wrangling

Diaries are incredibly rich sources of data—that is what makes a diary study so insightful. However, data does not talk. To answer our research questions we had to find a reasonable way of selecting and organizing the data to extract useful information from the diaries. This was one of the most time-consuming steps in the entire study. But once we had selected the relevant data, categorized it, and put it in a well-structured Excel spreadsheet, we had the answers to many of our research questions, such as duration of stay, number and location of breaks, and time spent on library services, at our fingertips. Literature on diary studies has rarely covered this aspect in detail. In our experience this is the most important and the most difficult step, so here is a detailed description of how we proceeded.

Since one aim of the study was to learn more about how much time people spent on specific activities (e.g., breaks) and at specific places (e.g., reading room or cafeteria), we had to exclude the diaries that did not have sufficient timestamps explaining the activities. Most participants included starting times for each activity that we could use to calculate their duration. In addition, we could sometimes deduce this information from existing data. This step allowed us to see how much time the diarists spent at each location. However, some diaries were of a very literary style or just did not include enough timestamps. We searched these eight diaries for mentions of attractive or unattractive spaces, descriptions of important library functions or dysfunctions, and the like.

In the end, we needed to prepare thirty-seven diaries for analysis. First, we determined which activities would constitute the smallest units of our data set. Our spreadsheet contained one line for every single activity described in one of the thirty-seven diaries. For example, we treated a working session by one participant on Day 2 from 10 a.m. to 11:30 a.m. as one activity; we treated another working session by the same participant later in the afternoon as another activity. To allow for a more aggregated view of the data, we then developed a coding scheme that merged identical activities with different participant descriptions. In the end, we worked with fourteen different activities such as arriving, concentrated work, and interaction with other patrons. Next, we developed other coding schemes for locations and topics to which participants referred positively or negatively when they evaluated what they experienced in the library (library staff, lockers, behavior of other users, food, etc.).

Our aim was to describe every activity as comprehensively as possible, so we added a column for each of these aspects: date, day of the week, start time, end time, calculated duration, location as described in the diary, location code, evaluation as described in the diary, topic code, and ideas or suggestions as described in the diary. To find patterns tied to individual demographics of our participants—patron group, age, library experience, disability, international background and many more—we added columns for this information, which we had documented in the interview protocols.

At the end of the day our spreadsheet contained 884 lines—one for every activity documented—and thirty-five columns containing additional information on every activity. This was the basis of our analysis. See appendix C for excerpts from our spreadsheets.

Visit Structure

An average visit to the State Library takes about seven hours, and readers spent two-thirds of this time with focused work on their individual project. However, that does not imply that this kind of intensive work passes without interruption. Other activities more or less regularly disrupt the concentrated focus, resulting in roughly three working episodes. Interrupting activities include about 30 minutes that go into organizational matters like answering e-mails, making phone calls, or structuring the day ahead. This kind of organizational work is less intensive but still closely connected to the individual projects or to other forms of necessary wage labor. In addition, the participants spent about 10 minutes with tasks that are part of using the library, like extending the lending period of books or managing their library account. Last but not least, readers take breaks. Of course, this fact in itself was no surprise, but we were astonished to realize that breaks did not take longer than 60 minutes total and that these 60 minutes were segmented into three smaller breaks. We had expected longer lunch breaks and—to be honest—more and longer breaks altogether.

Work Environment

It was striking that almost every diarist highlighted their pivotal reason for coming to the Haus Potsdamer Straße lies in its architecture and atmosphere. Other libraries in Berlin cannot compete with the spacious layout of the reading room, the natural light, the plants that form smaller seating areas, and the vast views throughout the building. What is maybe even more compelling is the incredible atmosphere of concentration, which results from the State Library’s status as a library used for research—a lot of reading and writing compared to less learning or group work. Concerning the working spaces themselves, the diaries described a mostly satisfactory situation. Some of the few suggestions included more storage possibilities close to patrons’ seats and reading lights that offer more options for individual adjustments.

In addition, the diaries provided a detailed model of the perfect working space in the library, which is characterized by a balanced relationship between wide space on the one hand and intimacy on the other. It is located in a quiet spot with few passers-by and half-height room dividers or shelves that block the desk from passing routes behind one’s seat. However, it allows for views through the bright reading room and to the outside—without any dazzling light. It is aligned in such a way that there is no seat opposite and so the distance to the neighboring table is large enough to ensure sufficient privacy. Given that ideal, it became clear that the existing high desks in the reading room are not in the right spot and need improvement.

Besides remarks about the temperature in the reading room and hints to worn furniture, two findings stuck out. First, most disturbances the diaries documented were due to the behavior of other library users. Second, there was no suitable library space for one activity that occurred on a regular basis: phone calls. This presents the downside to the wide and open architecture of the library. Wherever one decides to go for that one important phone call during the day, or the urgent call from a child’s daycare, other people could overhear the call. Readers feel disturbed by others, and they have to deal with an absence of privacy. More than once the diarists told the story of listening to job interviews while using the bathroom because that seemed to have been the most private location the caller could find. Since then, the library has provided call boxes that users can easily reach from everywhere inside the library.

Break Environment

Taking breaks during a seven-hour library day is crucial, and most of the documented breaks occurred either in the cafeteria, which is not open to the public, or in the spacious and public lobby of the library. The cafeteria is only accessible from the reading room, and as there is a strictly enforced no-food policy, it is not possible to bring homemade food inside the cafeteria. Hence, readers can only eat homemade food in the library’s lobby. More than half of the diarists brought their own food to the library, and they criticized the lack of suitable furniture in the lobby, such as tables and chairs for dining.

Concerning the cafeteria, the diaries showed clearly that the cafeteria’s offerings and prices do not meet the patrons’ expectations. That includes the costs for basic drinks like water that diarists perceive as quite high. This results in the common use of taps in the bathrooms to fill water bottles halfway through the day. It is no surprise that diarists described these refills as unsanitary and a little undignified. Until now, the library did not realize the importance of a more sanitary way to refill one’s drinking bottle during the day, and we definitely need to take care of that fundamental problem.

In addition, the diaries mentioned a specific kind of break we had already heard about via the graffiti wall months earlier: the power nap. It seems like the need for a short nap is as common with library patrons as with library staff—regardless of age or user group. Who is not familiar with that hour after lunch when time goes so slowly and energy levels are so low? However, we still need to figure out how urgent the need for such a designated break room is and which setup would be comfortable, practical, and truly preferred.

Other Benefits

Next to providing books, journals, and other items and apart from the working spaces in the reading room, the library fulfils a second role that is at least as important: structuring. Visiting the library helps to separate personal life from work. Opening hours support temporal structuring which may help to create a kind of everyday office routine. As the success of research work can be uncertain, aspects such as motivation, discipline, and self-control are necessary. The library fosters all these aspects through codified and non-codified behavioral expectations. Due to those expectations, some of the diarists refer to the library as “a factory” in which they voluntarily submit to visible and invisible rules helping them to “function.” In the end, there is a very German Haus und Benutzungsordnung (usage regulations) in effect which clearly regulates what people may take inside the reading room and how to behave in it.

Also, the presence and behavior of other concentrating visitors had a favorable effect on the diarists’ own work behavior and the library atmosphere. The debriefing interviews showed that users disapproved of reading room activities not consistent with the notion of “functioning” in the “knowledge factory,” such as watching videos on YouTube, engaging in social media, and in some cases even learning. To put it more positively, working next to other people who are engaged in comparable tasks helps one stay focused and motivated. It also supports acknowledging one’s own work as relevant. Even and especially if the diarists only knew their neighbors by sight and not personally, they felt socially embedded and could focus more easily.

The State Library seems to add another helpful element to the mix: it is not part of any university or research institute and as such, the library itself facilitates their work but is not expecting any results from its patrons. The State Library is a neutral space for every researcher in town, and its goal is nothing other than to help and to enable.

Choosing the Method

Deciding whether a diary method is suitable means reflecting on the kind of data you want to gather. If you are interested in qualitative information about everyday incidents, personal experiences, spatial or temporal details, the course of specific actions, or introspection, then a diary study might be a good fit. For example, Murphy (2016) did one to learn about book lending routines of students. One of the strong methodological arguments for diary studies is their ability to produce valid and reliable data by substantially reducing recall bias (Alaszewski 2006; Bartlett & Milligan, 2015; Bolger et al., 2003; Kenten, 2010). As events and impressions are documented in their natural, spontaneous context, diary studies have been praised for capturing “life as it is lived“ (Bolger et al., 2003). In our case we found that especially true when we dealt with the number and duration of specific activities (e.g., working, having a break, interacting) and events such as disturbances. Had participants reported these activities in an interview or questionnaire hours or days after they happened, recall bias would have impaired data quality considerably.

However, diaries are situated somewhere “between the spontaneity of reportage and the reflectiveness of crafted text” (Langford & West, 1999, p. 8). One should be aware of the fact that participants write their diaries with the researchers in mind. Triangulating findings from diary research with other methods, especially interviews, can mitigate self-selection or self-reflection biases (Mittelmeier et al., 2021). So if you want to explore some kind of phenomenon in depth—including context as well as feelings, perceptions, or thoughts of the participants—you still might want to go with a diary approach. If the survey should take place over an extended period of time, a diary is probably more suitable than methods like shadowing because it reduces the time necessary for data collection (Heath, 2016).

But while the diarists will do data collection for you, structuring and analyzing the data will more than counteract this benefit. You will also need time for recruiting users and for briefing and debriefing sessions, which makes a diary study a time-intensive method. The more structured your diaries are the less messy data analysis will be. If you plan on using rather unstructured diaries you should be ready to spend the largest share of your research time on data analysis and interpretation. A diary study is tied to participants’ expressive abilities and hence limited to those who are familiar with the concept of diaries and versed in such self-reflection (Hyers, 2018; Lallemand & Koenig, 2017), which makes it a method applicable for many but not all populations.

Determining the Format

In most cases, a diary method concerning library visits will build upon solicited diaries. You will commission diaries for the purpose of your study, which puts you into the position of structuring the diaries depending on your needs. Usually, you will sample participants who have exactly the knowledge you are looking for, and they will write diaries about a specific task or situation that they will not only record but also reflect on. The better you know what kind of information you need, the more precise your choice of diary can be. Therefore, the first step might be the choice between a pen-and-paper approach versus a digital one. Hyers (2018) described the pros and cons of each option. If you have access to the necessary funds, you could think about using diary software, which will enable the participants to note their observations in some kind of app via their smartphones or other device. Often it is also possible to include photography or voice messages. Bartlett and Milligan (2015) discussed the influences of technology on diary methods. On a smaller budget, you could decide to go with a messenger service, email, or text documents.

In our case, we found that a low-threshold access for all participants, institutional requirements, data protection, and limited funds combined with the hope of providing the participants with a warm and playful atmosphere indicated pen and paper rather than a more technological solution. In addition, it turned out that even diarists who, at first, were hesitant toward handwriting enjoyed it. Pen and paper contributed to an experience of self-reflection because it set itself apart from computer-based research activities. However, it is crucial to keep in mind that transcribing handwritten diaries will require a lot of time later on.

After you settle on a basic format, you need to decide how much structuring is necessary to answer your research question (Bartlett & Milligan, 2015). If you aim for detailed records that allow you to gather numerical data, then you will want to opt for a more structured diary. Jamieson (2016) described their diaries as a “log sheet with entry slots for time of day and the activity being undertaken” (p. 175).

On the other hand, if you prefer very free and open reflections you might want to give the diarists a little more space and let them develop their own structure, one that might change from entry to entry or from day to day. For our purposes, a modified version of a classical time-space diary format (Ellegård, 2019) was a good decision. We did not provide diarists with a log sheet but with a general schema for the entries, and we asked them to maintain that schema throughout their diaries. However, that did not happen in every diary, so we had to calculate missing values and exclude the diaries that did not provide the information we needed. Hence, we recommend a more rigid diary structure, which should still leave enough room for personal insights and the diarists’ priorities.

Organizing and Analyzing the Data

Organizing and analyzing the data will be time consuming because diaries tend to produce a lot of data. We did not transcribe word by word, but we extracted decisive numerical and locational data, and we paraphrased and partially transcribed activities, ratings, and ideas. We organized the first few diaries in very close exchange, and discussions sometimes took quite some time until we could figure out helpful guidelines. It became obvious that the researchers themselves are important participants in this kind of study. Even if interaction between researcher and diarist during the diary phase is not that close, researchers will be “shaping the conclusions” (Hyers, 2018, p. 79) from the moment data editing starts. Thus, you should reflect on your choices, discuss them whenever possible, and note them for transparency of the following report.

After editing the data, the first step toward analysis was coding the entries. To facilitate the content analysis, we worked with a thematic coding system that emerged from the research questions and from a cursory reading of the diaries. We followed a “constant comparative approach” (Bartlett & Milligan, 2015, p. 43) during which we discussed most codes and refined them until we achieved clearly structured data. Eventually, the coding scheme provided us with both detailed insight into our data and a good overview of what topics we needed to think about. In our experience, it is crucial to take enough time for editing and coding the data, even if that might take weeks of stamina. Regular debates within the team proved to be very fruitful. We’d always advise on a study team rather than a single researcher.

Reporting the Results

Of course, the whole point of such a study is not only to work toward a better understanding between library staff and readers, but to achieve tangible steps that improve the library in a user-orientated way (Goodman, 2011). In a conservative institution like the Berlin State Library, a report is the indispensable first step toward that goal. As we intended to spread our findings and the supporting quotes as widely as possible in the library, we tried to create a readable report that still met methodological requirements. Hyers (2018) gave an overview of different styles of reports.

We wrote a standardized section concerning research questions, sample, method, etc. followed by a narrative description of a visit to the library including all relevant findings. Charts and diagrams presenting quantitative information, such as the duration and timing of the documented library visits, the distribution of areas used for recreation episodes, and the average duration of all activity types, supported the findings. Ellegård (2019) had manifold suggestions for visualizing quantitative data from diary studies. We combined that kind of synoptic meta-diary with side notes that allow easy navigation through the report.

Having satisfied the formal requirements, we translated the findings into visual presentations and designed a poster, which we presented in the reading room. We also wrote short summaries for the library’s blog. See appendix D for our reports. We presented the study to our colleagues on several occasions, which helped to install the project as a valuable source of information for strategic decisions. For instance, the current pandemic forced the library to regulate access to the reading room in order to comply with hygiene regulations. As the library wants to allow as many readers as possible in the reading room while offering adequate timeslots for concentrated work, the diary study helped us determine the ideal duration of bookable timeslots.

However, tangible improvements do not seem to result from any kind of report automatically. Library staff tends to be overwhelmed with day-to-day tasks, so it is hard to really engage with changes or take charge of needed improvements and push them forward. That is why we decided to offer workshops to stakeholders in the library to develop action plans and put them into place. We are curious to find out whether we will be able to encourage our colleagues sufficiently to really get things done. Let’s hope so.