Adopting situationally relevant modes of music information at different stages of information-seeking processes : A longitudinal investigation among music students

Adopting situationally relevant modes of music information at different stages of information-seeking processes A longitudinal investigation among music students Antti Mikael Rousi, Reijo Savolainen and Pertti Vakkari Faculty of Information Technology and Communication Sciences, University of Tampere, Tampere, Finland Abstract Purpose –The purpose of this paper is to elaborate the picture of situational relevance by examining how modes of music information are viewed as situationally relevant at different stages of information-seeking processes among music students. Design/methodology/approach –Empirical data of the present longitudinal study were collected in two phases by utilizing questionnaire and interview methods. Informants comprised of 14 university-level music students representing the fields of music performance, music education and music theory and composition. Modes of music information were approached through the information typology presented by Rousi, Savolainen and Vakkari. Findings –The findings indicate that not only the modes of music information were seen as situationally relevant for different reasons by the three participating music student groups when at the beginning of their tasks, but also that the perceived situational relevance of the information modes underwent changes as their tasks progressed to focus formulation and post-focus stages. Research limitations/implications –Due to the small number of participants, further research is needed to verify the results concerning the differences in information-seeking processes between diverse music student groups. Originality/value –The paper showcases that approaching music information through frameworks that classify information sources at diverse levels of abstraction enables an accurate description of information-seeking processes and illuminates context-sensitive development of situational relevance of music information of diverse modes. Keywords Music, Relevance, Music information, Information seeking, Information search process, Music students Paper type Research paper Introduction Kuhlthau’s (2004, p. 112) information search process (ISP) model suggests that information types at varying levels of abstraction may have differentiated interpretations during information-seeking process, and that these interpretations can undergo change as such processes progress. Music is a domain where information resides not merely in written language. In their works on musicology and musical semiotics, Bengtsson (1977) and Tarasti (1994, p. 4) state that musical knowing transpires through sign systems such as Journal of Documentation Vol. 75 No. 6, 2019 pp. 1230-1257 Emerald Publishing Limited 0022-0418 DOI 10.1108/JD-12-2018-0210 Received 17 December 2018 Revised 5 April 2019 Accepted 8 April 2019 The current issue and full text archive of this journal is available on Emerald Insight at: www.emeraldinsight.com/0022-0418.htm © Antti Mikael Rousi, Reijo Savolainen and Pertti Vakkari. Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial & non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode The authors would like to thank the anonymous reviewers for their constructive comments that greatly helped us to improve the manuscript. 1230 JD 75,6 music notations, aural experiences and gestural language of music making. This suggests that the interpretation and relevance of music information of varying kinds may change during an information-seeking process. When juxtaposing the dynamic nature of relevance suggested in the ISP model with the above multifaceted notion of music information of musicology and musical semiotics, an anomaly becomes apparent. Prior studies have left open the following question: which modes of music information are viewed as situationally relevant during the different stages of information-seeking processes? In general, studies on situational relevance examine the relationships between information and the user´s information problem situation (Schamber, 1994, p. 8; Saracevic, 2007, p. 1930). The present paper approaches the relevance of music information in relation to situational requirements of accomplishing a study-related information-seeking task on music. Prior to the present paper, the issue of music information objects and relevance have predominantly been discussed within the field of music information retrieval (MIR) (e.g. Casey et al., 2008; Kim, 2015). Even though there are situational factors, such as user’s location (e.g. Cheng and Shen, 2014) and time of day (e.g. Su et al., 2010), included in MIR’s approach to relevance, we lack investigations examining how music students view the situational relevance of music information of varying kinds during the different stages of performing their information-seeking tasks. Neither studies describing and modeling information seeking and needs of musicians (e.g. Brown, 2002; Kostagiolas et al., 2015; Lavranos et al., 2015, 2016) have devoted due attention to the particular nature of the modes of music information and their situational relevance. Even though there are a few studies of information seeking examining the situational relevance of music information (see Rousi et al., 2018), none of these prior studies focus on how the situational relevance of music information of diverse modes is viewed during the different stages of information-seeking tasks. To fill gaps in empirical research on context-sensitive information-seeking processes, the present paper examines how music students representing different fields and levels of university studies adopt situationally relevant modes of music information during the different stages of performing their information-seeking tasks. The term music information mode refers to a set of information sources that are seen to represent music information at a certain level of abstraction. In other words, the concept of music information mode is used to group information sources based on their method of representation, be it gestural language, non-conceptual aural experiences of music or symbolic written representations, for example. To conceptualize these modes, the present paper utilizes the typology of music information proposed by Rousi et al. (2016). Six modes were identified: music making as the first mode of enactive representations; music listening as the second mode of enactive representations; iconic representations of music; technological models of music as the first mode of symbolic representations; and ideological models of music as the second mode of symbolic representations and other symbolic information. The present paper utilizes a longitudinal research design where the situational relevance of the modes of music information are qualitatively analyzed at two phases during the information-seeking tasks performed by the music students. Furthermore, the progress of the participants’information-seeking task processes was examined from the viewpoint of thoughts, actions and emotions by using a modified process survey instrument of the ISP model (see Kuhlthau, 2004, p. 60). The empirical material was collected in 2013–2015 by using survey and interviews methods. The participants comprised of 14 music students representing music performance, music education and music theory and composition. All the participants were Finnish music students of Sibelius Academy, Helsinki University of the Arts. The rest of the paper is structured as follows. First, to provide a background, the literature review introduces the typology of the modes of music information used in the present study; this section also describes diverse approaches to situational relevance and 1231 Music information characterizes the main features of studies of relevance of music information objects and Kuhlthau’s ISP model. Then, the research questions and the empirical research design are specified. The main part of the paper focuses on the report of the empirical findings. The concluding section discusses the findings and reflects their significance. Literature review Modes of music information The typology of music information used in the present study was created by integrating Tarasti’s (1994) music semiotic ideas with Bruner’s (1966) approach to modes of knowledge representations introduced in Bruner’s book, Toward a Theory of Instruction. According to Bruner, any domain of knowledge and every single problem within that domain can be presented to the learner through using the following modes of representation: Enactive mode of representation refers to sequences of activities for creating desired results. Iconic mode of representation refers to presenting a concept through a graph without exhaustively defining it. Most abstract of the modes is the symbolic mode where through a system that defines rules of expression, a set of arguments is created for describing a concept (Bruner, 1966, pp. 44–45). The proposed typology further defined the content of Bruner’s modes within the domain of music by positioning sign systems relevant to music presented by Tarasti (1994) in Theory of Musical Semiotics into them, which led into altogether six following modes of information (Rousi et al., 2016): (1) Music making as the first mode of enactive representations refers to information resided in different sequences of actions that produce sounds for musical purposes. This action may appear in varied forms, such as playing the violin, singing or creating electronic music with a computer. (2) Receiving music as the second mode of enactive representations refers to receiving musical performances, while being present at a performance or through a recording, without the control over the performance’s sonic results. Audio recordings played using either home audio devices or mobile online music streaming services function as examples of sources incorporating this second mode of music information. (3) Iconic representations of music refer to the graphic illustrations presenting musicrelated information. Different notations, such as modern staff notations, function as an example of sources incorporating this mode of music information. (4) Technological models of music as the first mode of symbolic representations refer to examining structures of music, i.e. tonal organization of harmony and counterpoint and sonic formulae of orchestration and interpretation, for example, through symbolic means such as written language. Information sources that incorporate the first mode of symbolic representations are, for example, textbooks of harmony and counterpoint. (5) Ideological models of music as the second mode of symbolic representations address music at a symbolic level but not directly the qualities that transpire in audible music. In this mode, conceptual symbolism regarding music are negotiated with other fields and their conceptual symbolic representations. For example, it is possible to produce narratives on concepts such as “Western classical music”and “history”or “music”and “aesthetics”without reference to actual phenomena present in sonic reality of music. Examples of information sources that can be more geared towards the second symbolic mode include monographs on philosophy of music and some texts concerning the history of music, such as some biographies of composers. 1232 JD 75,6 (6) As music students might also have other than music-related information needs, such as information needs about academic writing, a sixth information category titled other symbolic information was added into examination. This category was defined to include all symbolic, i.e. conceptual, information sources from other than musicrelated disciplines, such as conventions of scientific writing. The typology’s purpose is not to exhaustively classify phenomena as belonging to one mode alone. The position and needs of the information seeker play a crucial role in providing the angle through which the different modes of music information transpire. For example, the technological models of harmony and counterpoint are closely weaved with the history of Western music, which is evident in concepts such as “Palestrina style counterpoint”(see also Tarasti, 1994, p. 17). Thus, the same information sources can very well include many of the modes of music information. Situational relevance Whereas systems or algorithms create relevance by retrieving documents based on the query, people derive relevance from obtained information objects (Saracevic, 2007, p. 1919). Situational relevance refers to examining relationships between information and the user’s information problem situation (Schamber, 1994, p. 8; Saracevic, 2007, p. 1930). Relevance inferences depend not only on the features of the information in document, but also on user’s context, user’s previous knowledge and specific qualities sought (Barry, 1998, p. 1302). The situational relevance approach differs from topical relevance approach where the focus is on the relationship between the topic expressed in a search query and topic covered by resulting information objects (Saracevic, 2007, p. 1929; 1931). Concepts such as success dimensions, criteria categories and factors influencing selection of information objects, e.g. documents, have been used to describe the outcomes of previous research examining situational relevance (Schamber, 1994, pp. 24-25). In the present paper, however, the construct of situational relevance type was preferred. This is due to the fact that such types are approached as situational factors that make the seeking of diverse modes of music information pertinent to the performance of ongoing study tasks among students working on their theses and dissertations. Music information objects and relevance The issue of music information objects and relevance has predominantly been discussed within the field of MIR. MIR studies have examined relevance of music information objects through context-based approaches matching string search queries with textual metadata representations and through content-based approaches matching music similarity of audio data examining rhythm and melodies (e.g. Downie, 2003; Downie, 2004; Casey et al., 2008; Kim, 2015). In MIR, the approaches to relevance are often more system than user oriented. MIR’s conceptualization of music information provide useful summaries for MIR system design (e.g. Downie, 2003, pp. 293-301) but may not be sufficient from the viewpoint of a person engaged in music information seeking to whom gestural language of music making (Godøy and Jensenius, 2009) and diverse literature (Brown, 2002, p. 82) may also be relevant. Currently, there are also users’situational or contextual factors incorporated into MIR techniques. For example, context-aware music recommender systems utilize the following contextual data when suggesting content. User’s mood or emotion (Kim et al., 2010); daily activities, such as working, sleeping and running (Wang et al., 2012); user’s location (Cheng and Shen, 2014); and time of day (Su et al., 2010). However, these factors are general in their nature, and do not tell why, for example, music students see the different modes of music information situationally relevant in different stages of their information-seeking tasks. 1233 Music information Studies describing and modeling information seeking and needs of musicians and music scholars confirm that both groups have information needs spanning to diverse information sources representing music information at varying levels of abstraction (Brown, 2002; Hunter, 2006; Liew and Siong, 2006; Lai and Chan, 2010; Matson and Shelley, 2013; Dougan, 2015; Kostagiolas et al., 2015; Lavranos et al., 2015, 2016). For example, music performances, recordings, notations and music-related literature are identified as information need types, albeit amongst many, by previous studies (e.g. Kostagiolas et al., 2015, p. 7). Brown’s (2002, p. 82; 86) findings suggest that music scholars utilize, for example, audio recordings and music notations while conducting research. In general, the above studies have not devoted due attention to the particular nature of the modes of music information and the ways in which music students evaluate their situational relevance. Currently, there are some studies of music-related information seeking and information behavior that have approached music information’s situational relevance through the many layers of music information (Rousi et al., 2018; see also Pohjannoro and Rousi, 2018). However, prior to the present paper, there are no longitudinal studies examining which modes of music information are viewed as situationally relevant during the different stages of information-seeking processes? ISP model As stated in the introduction, Kuhlthau’s (2004) ISP model enables the process approach by providing a framework for the analysis of the information-seeking tasks of the music students. In a series of empirical studies, Kuhlthau (2004) demonstrated that information-seeking tasks consist of several stages. The ISP model proposes that information-seeking tasks typically consist of six consecutive stages: initiation, selection, exploration, formulation, collection and presentation (Kuhlthau, 2004, p. 44-50). These stages are significant as person’s thoughts, actions and emotions typically undergo change as one’s information-seeking task progresses from stage to another. According to the model, the formulation is a pivotal stage during which a more focused perspective to the task is gathered. After focus formulation, person’s sense of direction and confidence increases and actions change from exploring relevant information to seeking and documenting pertinent information (Kuhlthau, 2004). The ISP model has gained a prominent role among theories of information seeking and has affected theory building in Information science (Vakkari, 2001). It has previously provided a process framework for studies examining the changing relevance criteria of textual documents during information-seeking tasks (Bateman, 1998; Vakkari and Hakala, 2000). Even though music-related information seeking has been described and modeled in previous studies (Lavranos et al., 2016), which situationally relevant modes of music information are adopted by music students during different stages of their information-seeking tasks has not been previously examined through the ISP model or any other process-oriented framework of information seeking. Methods Research question The goal of the present paper is to elaborate the picture of situational relevance by examining how modes of music information are viewed as situationally relevant at different stages of information-seeking processes among music students representing different fields of university studies. To this end, the present paper addresses the following research question: RQ1. Which situationally relevant modes of music information are adopted during the information-seeking processes of music students representing the fields of music performance, music education and music theory and composition? 1234 JD 75,6 Methodology Participants. All of the 14 participants of the present study were students in Sibelius Academy, Helsinki University of Arts in Finland. The participants of the present paper represented three different music student groups of Sibelius Academy. Six of the participants were Doctor of Music (DocMus) students focusing on music performance. Their dissertation projects included bothaseriesoffiveconcertsandawrittenpart. During the present study, the participating DocMus students had written and submitted their final dissertation proposal for examination. In their proposals, both the concert programs and the topic of the written dissertation were defined. Five of the participants were students of music performance specializing in music education. Their theses were written and no music performances were included in their grading. One of the students who focused on music education was a DocMus student, i.e. PhD level student, and four were working on their Master’s theses. Lastly, three of the participants were students of music theory and composition who were also working on their Master’s theses. Also, the theses of the students of music theory and composition were written and focused on music analyses; no music performances were included in their grading. Before the first data collection phase of the longitudinal study, both groups of Master-level students had begun their Master’s thesis. These three main student groups were selected as the participants of the present study due to that they were seen to represent three different task types common to music students: preparing music performances, writing academic texts on music education and conducting music theoretical analyses. By including these three participant groups, the present study seeks to elaborate the picture on how music students representing different fields view the situational relevance of music information modes during the different stages of performing their information-seeking tasks. The participating DocMus students were recruited from an information retrieval course arranged for the DocMus students by Sibelius Academy, Helsinki University of Arts Library. The participating Master-level students of music education and music theory and composition were recruited from their respective Master’s thesis seminars. All 14 music students were awarded a sum of €40 for their participation in the longitudinal study of the present paper. Even though the amount of participants per student group is small, they represent a fairly large sample of all students annually participating in both the DocMus programs and the respective Master-level seminars. Within the period of 2010–2018,onaverage,six Doctors of Music focusing on music performance have graduated per year from the Sibelius Academy, Helsinki University of Arts. As the recruitment of DocMus students (n¼6) spanned from the semesters of 2013–2014 and 2014–2015, it may be estimated that the three participants per semester constitute a significant proportion of all students starting in the respective program on the given year. The DocMus student focusing on music education (n¼1) was from the researcher program of the doctoral school of Sibelius Academy, Helsinki University of Arts. Respectively, on average, four Doctors of Music have graduated within the period of 2010–2018 from the researcher program. The Master’s level students of music education (n¼4) were recruited from two Master’s level seminars in which there were circa ten active participants in total. The students of music theory and composition (n¼3) were recruited from a Master’s thesis seminar with circa five active participants. With both Master’s level students focusing on music education and music theory and composition, the data collection was limited to the semester of 2013–2014. This was due to that several students continued in the seminar also in the following semester, which enabled the recruiting of new participants by continuing data collection in semester 2014–2015. 1235 Music information Data collection methods The data collection methods of the present study included survey questionnaire and interview methods. To examine how the music students progressed in their informationseeking tasks, the present investigation utilized survey questionnaires. In the initial section of the questionnaire, the participants were asked to assess whether they thought their tasks had progressed. During both data collection phases, the participants chose the current stage of their information-seeking task from the three stages identified by Vakkari (2001, p. 47): start of the thesis and identifying its general topic (pre-focus); examining information concerning a general topic and formulating a specific focus ( focus formulation); and gathering information pertinent to a focused topic or finalizing the thesis (post-focus). Whereas the rest of the questionnaire used in the present study drew on the process survey instrument developed by Kuhlthau (2004, p. 60), Vakkari’s (2001, p. 47) operationalization of the stages of information-seeking tasks was made at a higher level of generality so that it would suit to the examination of tasks incorporating music performances. To further support the participants’self-assessment of the progress of task performance, as well as the changes in thoughts, actions and emotions, the process survey instrument developed by Kuhlthau (2004, p. 60) was modified for the purposes of the present study as follows. The questions examining the information seeker’s current emotions and thoughts were translated into Finnish and used as such in the data collection. The section examining current activities was also modified to include music-related activities. Music listening, performing music and conducting music analyses were added as options among the original actions listed in Kuhlthau’s (2004, p. 60) process survey instrument. To include the process aspect in the music-related activities, two options of each of these activities of playing music, listening to music and conducting music analyses were given in the questionnaire. One of the options reflected a more focused approach to the activity, such as “I conduct detailed music analyses related to my topic”while another option exemplified a less focused approach, such as “I conduct preliminary music analyses related to my topic.”Both the thoughts, actions and emotions in the original process survey instrument (Kuhlthau, 2004, p. 60), and the added music-related activities were listed in the questionnaire. The participants were asked to choose all that apply during both data collection stages. The questionnaire used is included as Appendix 1 (translated from Finnish to English by the first author). The third main part of the questionnaire focused on the situational relevance of the six information modes presented in the literature review (Rousi et al., 2016). Each information mode had its own section in the questionnaire. Within these sections, the participants were first asked to indicate on a four-point scale (not at all important, not that important, important and very important) how pertinent each of the modes was for their tasks of either a dissertation or a Master’s thesis in music and then to briefly write down the reasons behind these assessments. The information mode-specific sections of the questionnaire also included definitions of the information modes and examples of information sources of individual modes were provided (see Appendix 1). The use of questionnaires was complemented with interview methods. A thematic interview guide was developed to ask more detailed questions regarding the participants’ thesis tasks and to allow the participants to elaborate answers given in the questionnaires. Overall, the interview guide consisted of two main sections: general questions regarding the topic and progress of the participants’thesis tasks and sections examining the situational relevance of the six information modes. Within the music information mode-specific sections, the participants were able to explain in depth why a specific information mode was or was not important for their information-seeking tasks and whether they saw that its situational relevance had changed during the information-seeking process. The interview data consist of circa 630 min of audio-taped records. 1236 JD 75,6 Data collection process The empirical data were collected in two phases. First, during autumn semesters of 2013 or 2014, the 14 participants were asked to fill in the survey consisting of the sections described previously. All of the 14 students also participated in the latter phase of data collection, which took place during the following year’s spring semester. During the latter phase, the participants were first asked fill in again the same set of survey questions regarding their thoughts, actions and feelings after which their answers were further elaborated in an interview. Second, the participants were asked to the fill in the sections concerning the importance individual modes of music information as in the previous phase; each of these were complemented with interview questions. Once the individual music information mode sections were covered during the second phase of data collection, the previous answers given in the first phase survey were shown to the participants; thereafter, any differences identified in mode-specific situational relevance assessments were discussed. The average time period between the first and second data collection phases was circa 135 days. The shortest period was circa 100 days and longest circa 160 days. Table I summarizes the data collection process. Research ethics The informed consent form was given for the scrutiny of the participants at both data collection stages. The form offers information about the voluntary nature of the study, the participants to right withdraw from the study at any given stage and the anonymized reporting of the results and data. Agreeing to participate in the interviews was interpreted as the informant’s consent to participate in the study. If a participant returned the questionnaire during phase 1, but did not participate in the second phase of data collection, their data were discarded and thus not used in the present study. Direct identifiers were collected only as part of the interview data, i.e. the voice of the participant. However, given that the overall music student population is small, the indirect identifiers included in the data become increasingly efficient in identifying the participants. Therefore, the questionnaire data, interview audio recordings and the transcript data will be deleted after they are no longer used for scientific purposes. The participants’anonymity is also protected when reporting quotes from the transcribed data in the findings section. Italicized notes are used to replace a specific section of the quote with a more general expression in cases in which the specific part contained an indirect identifier that would compromise the anonymity of the participant. Data collection phase 1 Data collection phase 2 Questionnaire 1 Questionnaire 2 Interview guide Progress in the task performance Participants’self-assessments regarding the progress of their information-seeking tasks Participants’self-assessments regarding the progress of their information-seeking tasks Questions regarding the task of the participants, e.g. What is the topic of your thesis? Do you have a focus in your task? What are you planning to do next? What kind of information are you looking for right now? Current thoughts, actions and feelings Current thoughts, actions and feelings Situational relevance of the six information modes Situational relevance of the information modes The perceived importance of the mode Description of why the mode was considered as important or not important Situational relevance of the information modes The perceived importance of the mode Description of why the mode was considered as important or not important Questions regarding the changing situational relevance of the information modes Has the reason why the mode is important changed? Comparison of mode-specific answers of questionnaires 1 and 2 Table I. Summary of data collection process 1237 Music information Data analysis and validity issues To examine the progress occurring in the tasks of the participating music students, their self-assessments regarding the progress of their information-seeking tasks were first summarized. To achieve this, it was examined how the participants were distributed regarding the self-reported stages of the task performance process per phase of data collection. Second, the participants’answers to the modified process survey questionnaire were summarized so that the percentual distributions of participants expressing different thoughts, actions and emotions were examined per phase of data collection. Due to the small number of the participants, no correlations or statistical tests will be conducted. However, descriptive statistics are calculated with regard to the stage of the participants’tasks and thoughts, actions and emotions at both data collection phases to support the qualitative analysis that result in student group summaries. The findings regarding these are presented using descriptive statistics so that they highlight the changes occurred in the participants’ thoughts, actions and emotions during the longitudinal study. With the questionnaire and interview data examining the perceived importance of the music information modes and their situational relevance, the analysis was conducted as follows. First, the questionnaire data were summarized by calculating information modespecific averages of perceived importance. Second, the interview data were transcribed and scrutinized through qualitative content analysis. More specifically, the open-ended answers explaining the perceived importance assessment of an information mode were interpreted to indicate its types of situational relevance. To this end, an attempt was made to identify reasons by which the participants considered a music information mode important to task performance; such reasons were used as a point of departure to identify the situational relevance types reviewed in the findings section below. To detect the changes occurring in the situational relevance types, the constant comparative method (Tesch, 1990, p. 96; Silverman, 2005) was used to discern conceptual similarities between the open-ended answers obtained during both data collection phases. If a conceptual dissimilarity was observed in the open-ended answers, the full interview data were further scrutinized for more evidence of changing situational relevance of an information mode. Even though the empirical data are already six years old, it was seen as valid for the particular needs of the present investigation. It examines which situationally relevant modes of music information are adopted during preparing a thesis focusing music performance, music education and music theoretical analyses. Analysis is not conducted on the level of individual information sources per se, but on the more general level of music information modes, each categorizing music information at a different level of abstraction. As the processes of giving concerts in the Western art music tradition, writing academic texts on music education and conducting music theoretical analyses are grounded in their cultural traditions (see e.g. Kuhn, 1996; Trehub et al., 2015), their core information elements do not change so rapidly than the quickly developing field of online music information sources and music streaming services, for example. To improve the validity of this study, the following section incorporates quotes taken from the interview data. The quotes were translated from Finnish into English by the first author. Care was put into preserving the quotes as close as possible to verbatim form during the translation process. Findings The findings section is organized as follows. First, the summary of participants’ self-assessments regarding the progress of their information-seeking tasks is presented. Second, descriptive statistics regarding the changes of the participants’thoughts, actions and emotions between data collection phases are presented to demonstrate whether they –in line with Kuhlthau’s (2004) ISP model –indicated progress in the task performance. 1238 JD 75,6 relevance type of the information modes could change significantly after and during the task-stage of focus formulation. This further supports the finding of prior research that it is useful to approach the concept of relevance as a context-sensitive and process-oriented user construct (Bateman, 1998; Vakkari and Hakala, 2000). Previous research on the changing relevance judgments of textual documents has shown that once information seeker’s task progresses, the notion of relevance also becomes more focused. This is due to that the information seeker has more readiness to identify pertinent texts in the later stages of the task, even if with text documents these changes could be subtle (Vakkari and Hakala, 2000, p. 559). The findings of the present paper regarding students focusing on music education suggest a similar dynamic with regard to task progress and relevance of textual documents; during and after the focus formulation, this participant group was able to narrow its focus to the symbolic or textual mode which was viewed as most relevant and in general the changes in the situational relevance of the important information modes were subtle. Within the tasks of DocMus students, the enactive and iconic modes were of most importance during the pre-focus stage topic exploration. During and after focus formulation, the increased ability to discern pertinent information led to the perceiving of the symbolic information modes as important and relevant due to new diverse reasons. This seems to suggest that tasks focusing on enactive outputs such as music performances are more expansive what comes to using the different modes of music information at different stages of tasks. This appears as somewhat contradictory to the music education students working to create a written thesis because they strived to narrow their focus into the most relevant symbolic information mode. Some of the students of music performance focusing on music education found the enactive modes relevant to the extent that their situational relevance type underwent changes during the period of examination. Thus, the enactive modes may not only be relevant for music students writing written theses, but they also might incorporate several situational relevance types during a task aimed at a written thesis. The situational relevance types of the modes of music information among students of music education and music theory, as well as students of music composition will be examined in detail in forthcoming publications (for DocMus students, see Rousi et al., 2018). The present paper utilized Kuhlthau’s (2004) process survey instrument in order to identify the progress of the task performance. The modified process survey instrument also helped to identify participants whose tasks did not progress during the circa 130 days period of examination. As only half of the participants made significant progress in their task, the period of examination between the phases of data collection could have been longer. This was especially evident with DocMus students’extensive tasks of dissertations Important information modes Receiving music as the second mode of enactive representations Music making as the first mode of enactive representations Technological models of music as the first mode of symbolic representations Iconic representations of music Changes in situational relevance Pre-focus Focus formulation Post-focus Study of the topic and analyzed pieces at more general level through listening of audio recordings (2nd enactive) The relevance of audio recordings may diminish after the analyses become more detailed Once the music analyses get more detailed, own playing (1 st enactive) becomes more relevant from the viewpoint of analyzes. Own playing is used to bring forth details of the pieces to be used in the analyses Note: n¼3 Table V. Important information modes and changes in situational relevance among students of music theory and composition 1245 Music information on music. However, with this period of examination, the participants seemed to be able to recall their thoughts when answering the questionnaire in the first data collection phase which helped in identifying changes in the situational relevance of the information modes. Another alternative approach in methodology could have been not to lengthen the time period between data collection phases but to insert a third data collection phase after circa four months from the second one. This would have increased the share of music students making significant progress in their tasks while participating in this longitudinal study and thus likely would have provided more evidence regarding changes in situational relevance of the information modes. However, the methodology currently employed was nevertheless seen to produce results of interest in a topic where previous studies are scarce. Theoretical and practical implications The present paper was the first step towards examining how the situational relevance of music information of varying kinds is viewed during the different stages of informationseeking task performance. The findings demonstrate that not only the modes of music information were seen as situationally relevant for different reasons by the three music student groups when they started their tasks. Moreover, the perceived situational relevance of the information modes underwent changes as their tasks progressed to focus formulation and post-focus stages. The findings also showcase that approaching music information through frameworks that classify information sources at diverse levels of abstraction enables an accurate description of information-seeking processes and illuminates context-sensitive development of situational relevance of diverse music information of diverse modes. With regard to implications on MIR design, it is noteworthy how much the nature of the tasks, i.e. focus on music performance, writing academic texts on music education or conducting music analyses affected the ways in which the participants viewed the situational relevance of the music information modes while performing information-seeking tasks. Conducting similar context-sensitive longitudinal studies examining the relevance of music information could deepen our understanding on how the situational relevance of music information varies in different types of tasks. This could, in turn, influence the MIR system design to better acknowledge the nature and the stage of a music-related task when recommending relevant materials. Limitations Due to the small number of the study participants, the findings describing informationseeking processes of students of music performance, music education and music theory and composition are not generalizable to all students representing these fields. For the same reason, it is possible that individual factors such as the student’s information-seeking styles and habits have affected the empirical findings. Therefore, further research is needed to verify the results concerning the differences in information-seeking processes of the participating music student groups. Further research The situational relevance types of music information modes for students of music education and music theory and composition should be examined in more depth. More research is also needed to verify the present paper’s findings regarding the differences in relevance dynamics of tasks aimed at music performances and of those aimed at written theses. Furthermore, similar research should be conducted to information-seeking tasks related to other artistic disciplines, such as other performing arts or visual arts. The previous approaches could prove to be useful in understanding how the situational relevance of various information modes change during information-seeking processes. 1246 JD 75,6 Conclusion This paper presents an account of how situational relevance of diverse music information modes changes in the ISPs of music students representing the fields music performance, music education and music theory and composition. The findings demonstrate that not only the modes of music information were seen as situationally relevant for different reasons by the three participating music student groups at the initial stage of task performance, but also that the situational relevance of the information modes was perceived differently between the music student groups as their tasks progressed to focus formulation and post-focus stages. The findings highlight the potential of context-sensitive approaches to music information seeking and the importance of conducting longitudinal studies elaborating how situational relevance may undergo changes during information-seeking processes. References Barry, C. (1998), “Document representations and clues to document relevance”,Journal of the American Society for Information Science, Vol. 49 No. 14, pp. 1293-1303. Bateman, J. (1998), “Changes in relevance criteria: a longitudinal study”, in Preston, C. (Ed), Proceedings of the 61st American Society for Information Science Annual Meeting, Vol. 35, Information Today, Medford, NJ, pp. 23-32. Bengtsson, I. (1977), Musikvetenskap: en översikt [Music Science: An Overview], Esselte Studium, Stockholm. Brown, C. (2002), “Straddling the humanities and social sciences: the research process of music scholars”,Library & Information Science Research, Vol. 24 No. 1, pp. 73-94. Bruner, J. (1966), Toward a Theory of Instruction, Harvard University Press, Cambridge. Casey, M., Veltkamp, R., Goto, M., Leman, M., Rhodes, C. and Slaney, M. (2008), “Content-based music information retrieval: current directions and future challenges”,Proceedings of the IEEE, Vol. 96 No. 4, pp. 668-696. Cheng, J.Z. and Shen, J. (2014), “Just-for-me: an adaptive personalization system for location-aware social music recommendation”,Proceedings of International Conference on Multimedia Retrieval, ICMR2014, pp. 1-8. Dougan, K. (2015), “Finding the right notes: an observational study of score and recording seeking behaviors of music students”,The Journal of Academic Librarianship, Vol. 41 No. 1, pp. 61-67. Downie, S. (2003), “Music information retrieval”,Annual Review of Information Science and Technology, Vol. 37, pp. 295-340. Downie, S. (2004), “The scientific evaluation of music information retrieval systems: foundations and future”,Computer Music Journal, Vol. 28 No. 2, pp. 12-23. Godøy, R. and Jensenius, A. (2009), “Body movement in music information retrieval”,Proceedings of the 10 th International Music Information Retrieval Conference, ISMIR2009, pp. 45-50. Hunter, B. (2006), “A new breed of musicians: the information-seeking needs and behaviors of composers of electroacoustic music”,Music Reference Services Quarterly, Vol. 10 No. 1, pp. 1-15. Kim, S.-M. (2015), “Towards organizing and retrieving classical music based on functional requirements for bibliographic records (FRBR)”, dissertation, University of Pittsburgh, Pittsburgh, PA, available at: https://pdfs.semanticscholar.org/ba2e/56b9b6d0300199d0f690 7e50619dc94d1f2a.pdf (accessed July 10, 2019). Kim, T.E., Schmidt, E.M., Migneco, R., Morton, B.G., Richardson, P., Scott, J., Speck, J.A. and Turnbull, D. (2010), “Music emotion recognition: a state of the art review”,Proceedings of the 11 th International Society for Music Information Retrieval Conference, ISMIR2010, pp. 255-266. Kostagiolas, P.A., Lavranos, C., Korfiatis, N., Papadatos, J. and Papavlasopoulos, S. (2015), “Music, musicians and information seeking behaviour”,Journal of Documentation, Vol. 71 No. 1, pp. 3-24. 1247 Music information Kuhlthau, C. (2004), Seeking Meaning: A Process Approach to Library and Information Services, 2nd ed., Libraries Unlimited, Westport, NY. Kuhn, T.S. (1996), The Structure of Scientific Revolutions, 3rd ed., University of Chicago Press, Chicago, IL. Lai, K. and Chan, K. (2010), “Do you know your music users’needs? A library user survey that helps enhance a user-centered music collection”,The Journal of Academic Librarianship,Vol.36No.1, pp. 63-69. Lavranos, C., Kostagiolas, P. and Martzoukou, K. (2016), “Theoretical and applied issues on the impact of information on musical creativity: an information seeking behavior perspective”,in Kostagiolas, P., Martzoukou, K. and Lavranos, C. (Eds), Trends in Music Information Seeking, Behavior, and Retrieval for Creativity, IGI Global, Hershey, PA, pp. 1-16. Lavranos, C., Kostagiolas, P., Korfiatis, N. and Papadatos, J. (2016), “Information seeking for musical creativity: a systematic literature review”,Journal of the Association for Information Science and Technology, Vol. 67 No. 9, pp. 2105-2117. Lavranos, C., Kostagiolas, P.A., Martzoukou, K. and Papadatos, J. (2015), “Music information seeking behaviour as motivator for musical creativity: conceptual analysis and literature review”, Journal of Documentation, Vol. 71 No. 5, pp. 1070-1093. Liew, C. and Siong, S. (2006), “Beyond the notes: a qualitative study of the information-seeking behavior of ethnomusicologists”,The Journal of Academic Librarianship, Vol. 32 No. 1, pp. 60-68. Matson, J. and Shelley, A. (2013), “In search of music: a study of music consumption and search behaviors in undergraduate students”,Music Reference Services Quarterly, Vol. 16 No. 4, pp. 218-231. Pohjannoro, U. and Rousi, A.M. (2018), “The modes of music information in a compositional process. A case study”,Journal of Documentation., Vol. 74 No. 5, pp. 987-1007. Rousi, A.M., Savolainen, R. and Vakkari, P. (2016), “A typology of music information for studies on information seeking”,Journal of Documentation, Vol. 72 No. 2, pp. 265-276. Rousi, A.M., Savolainen, R., Harviainen, M. and Vakkari, P. (2018), “Situational relevance of music information modes: an empirical investigation among Doctor of Music students”,Journal of Documentation, Vol. 74 No. 5, pp. 1008-1024. Saracevic, T. (2007), “Relevance: a review of the literature and a framework for thinking on the notion in information science. Part II: nature and manifestations of relevance”,Journal of the American Society for Information Science, Vol. 58 No. 13, pp. 1915-1933. Schamber, L. (1994), “Relevance and information behaviour”,Annual Review of Information Science and Technology, Vol. 29, pp. 3-48. Silverman, D. (2005), Doing Qualitative Research, 2nd ed., Sage Publications, London. Su, J.-H., Yeh, H.-H., Yu, P.S. and Tseng, V.S. (2010), “Music recommendation using content and context information mining”,IEEE Intelligent Systems, Vol. 25 No. 1, pp. 16-26. Tarasti, E. (1994), A Theory of Musical Semiotics, Indiana University Press, Bloomington, IN. Tesch, R. (1990), Qualitative Research: Analysis Types and Software, Falmer Press, London. Trehub, S.E., Becker, J. and Morley, I. (2015), “Cross-cultural perspectives on music and musicality”, Philosophical Transactions of the Royal Society B, Vol. 370 No. 1664, pp. 1-9. Vakkari, P. (2001), “A theory of the task-based information retrieval process: a summary and a generalisation of a longitudinal study”,Journal of Documentation, Vol. 57 No. 1, pp. 44-60. Vakkari, P. and Hakala, N. (2000), “Changes in relevance criteria and problem stages in task performance”,Journal of Documentation, Vol. 56 No. 5, pp. 540-562. Wang, X., Rosenblum, D. and Wang, Y. (2012), “Context-aware mobile music recommendation for daily activities”,Proceedings of the 20 th ACM International Conference on Multimedia, MM2012, pp. 99-108. 1248 JD 75,6 Appendix 1. Questionnaire used in the present study 1249 Music information 1250 JD 75,6 1251 Music information 1252 JD 75,6 1253 Music information 1254 JD 75,6