Keywords

1 Introduction

How technological innovations impact and transform work has been at the very centre of the research agenda of the sociology of work for decades (see Baethge-Kinsky et al. 2018). This is especially true for studies on digitalisation (e.g., Pfeiffer and Suphan 2020). Researchers in this field developed numerous heterogeneous approaches and methods to grasp the relation of work and digital technologies, with a multitude of different findings. Despite the heterogeneity and the wide variety of approaches, two relevant findings can be obtained: First, most studies on digitalisation show that the digital transformation does not simply relate to technological innovation but rather can be understood as a complex socio-technical process (Henke et al. 2018; Pfeiffer 2018; Hirsch-Kreinsen 2020). Nevertheless, digital technologies are a relevant factor within the transformation of work, especially when socio-technical constellations at work are modified (Gläser et al. 2018; Pfeiffer 2018; Hirsch-Kreinsen 2020, 2021). In addition, several studies show that technologies affect different segments of the labour market to different degrees and that the influence of technology differs strongly depending on the socio-technical constellation, leading to variations by occupational field, organisation, and design of work processes (Orlikowski 2000; Baethge et al. 2007; Helmrich and Zika 2010). Following these findings, digital transformation should not be understood as a coherent process but rather as a multi-layered, contradictory, and unsynchronised process undergoing social preparation, technical enablement, and discursive negotiation (Henke et al. 2018).

Second, it can be said that the current debate on digitalisation is both fuzzy and fragmented. It is fragmented given the large variety of multifaceted research about a broad spectrum of different technologies—each with specific influences on work—in varying empirical fields (cf. Gläser et al. 2018; Boes et al. 2020). For example, some authors examine human-computer interaction in production logistics (Klumpp et al. 2019) and others human-robot interaction in healthcare (Buxbaum et al. 2019). Some authors use the overall debate on digitalisation to focus on computerisation and automatisation in container terminals (Gekara and Thanh Nguyen 2018), and still others are talking about big data systems in modern retail stores (Evans and Kitchin 2018). A significant number of debates evolve around imprecise buzz words related to digitalisation like industry 4.0, virtualisation, or algorithmicising (see Harteis 2018). Other discussions focus primarily on digitalisation as a context and mainly examine, for example, quantitative and structural consequences for the labour market (Eichhorst et al. 2017; Matraeva et al. 2020) or the structure and conditions of work and education (Selwyn and Facer 2014; Carls et al. 2020). Yet other authors distinguish between the concepts of ‘digitization’ and ‘digitalization’ (Legner et al. 2017). At the same time, it is fuzzy given that some authors examine the same facet of digitalisation using different terminologies, do not define their understanding of digitalisation, and/or do not distinguish their concept from other concepts of digitalisation (cf. Gläser et al. 2018; Boes et al. 2020). Together, the unspecific focus of research topics and the conceptual blurriness result in a high degree of fuzziness for the studies on digitalisation.

In addition, capturing the complex socio-technical process of digitalisation is a major methodological challenge and an objective of the Priority Programme “Digitalization of Working Worlds” (SPP 2267). The current debate on digitalisation can be differentiated in two methodological poles depending on the level of abstraction. On the one hand, several (mainly quantitative) studies use a high level of abstraction and thus focus on overarching phenomena of digitalisation (e.g. Arntz et al. 2020; Hanebrink et al. 2021), while typically neglecting company, technological, and work-process related aspects. This leads to the critique that they primarily discuss generalisations of a one-sided and hence inadequate kind (cf. Gläser et al. 2018; Henke et al. 2018; Pfeiffer 2018). On the other hand, there are various (mainly qualitative) studies with a low abstraction level (cf. Gekara and Thanh Nguyen 2018; Ruiner and Klumpp 2020), which leads to difficulties in comparing or generalising the highly contextual findings. In turn, that makes it hard to examine overarching phenomena of digitalisation.

Overall, the complexity of the digitalisation process in combination with the fuzziness and strong fragmentation of the current debate results in the absence of a sound theoretical-conceptual basis for investigations on the digitalisation of working worlds (Henke et al. 2018; Gong and Ribiere 2021). To capture the multi-layered, contradictory, and unsynchronised socio-technical process in an adequate way, the methodological concepts and tools should offer the possibility of taking both the company, technological, and work-process related aspects as well as overarching phenomena of the digital transformation into account.

To go beyond unspecific generalisations about digitalisation and to neither overestimate nor underestimate the relevance of sector-, company-, and domain-specific embeddedness, we argue in this article why it is necessary to apply mixed-methods approaches to adequately capture digitalisation. With a pragmatic approach on the use of quantitative and qualitative methods, we simultaneously apply these methods and juxtapose the results for an integrated interpretation of the state of digitalisation and permeation. The aim of this article is to offer a methodological template which may be applied for further investigations of the digitalisation of working worlds.

Following the ideas of the Priority Programme (SPP 2267) we understand digitalisation as a systemic transformation that is shaped by three motion dynamics: permeation, perpetuation, and making available (Henke et al. 2018). To develop our line of argument, the article restricts itself to the heuristic of permeation, which is understood as an information technology permeation of, for example, workers, work activities, work processes, value chains, companies, industries, and a wide variety of economic and social spheres (Henke et al. 2018).

We assume that, as one motion dynamic of digitalisation, permeation must be socially prepared, technologically enabled, and discursively negotiated. Likewise, we assume that permeation cannot be understood as a coherent process, but rather as a multidimensional process characterised by (at least) three dimensions—a quantitative dimension of permeation, a qualitative dimension of permeation, and a subjective dimension of permeation. These three dimensions can reinforce, inhibit, or contradict each other (see Fig. 1). The quantitative dimension of permeation refers to aspects such as the number of technologies, which technologies are available, or how frequently these technologies are used in work processes. The qualitative dimension of permeation refers to aspects such as the intensity of the influence of technologies on work activities, workarounds, or the impact of technology use on work processes. The subjective dimension refers to aspects such as the perceived relevance of technologies in everyday work or the perceptual permeation of work tasks.

Fig. 1
figure 1

Conceptualisation of the multidimensional process of permeation

Full size image

In this article, we discuss concepts of permeation using various methodological approaches. From there, we develop our own methodological concept to capture permeation. We use the empirical example of the healthcare sector—specifically hospitals—in Germany to illustrate our mixed-methods approach by analysing what is understood as permeation in this sector and how permeation can be comprehensively captured. Last, we discuss the extent to which the proposed study design facilitates both an explorative perspective and in-depth analysis of specific issues (low abstraction level) and the possibility to generalise results (high abstraction level) across sectors, domains, and occupations. We also elaborate on the extent to which the proposed approach promotes the depth and granularity of the data while also ensuring the broader contextualisation of the results. In future, this approach can be used as a methodological template for further investigations of the digitalisation of working worlds.

2 Measuring Permeation in Quantitative Studies

In the sociology of work and education, the concept of permeation is often used as a synonym for digitalisation. Digitalisation is commonly interpreted as a permeation of society and the working world with digital technologies, and, hence, includes all aspects of life (Hauff and Reller 2020). Since we regard digitalisation as a socio-technological process that encompasses aspects from technological and non-technological as well as economic, social, and labour policy matters (Hirsch-Kreinsen 2020), permeation is a multidimensional concept. And indeed, in quantitative sociology a universal approach to combine these multiple dimensions and measure digitalisation is still under development (Gong and Ribiere 2021). In the past few years, some scholars created surveys for the sole purpose of measuring digitalisation, while other (large-scale) surveys developed and implemented questions on digitalisation. Some researchers then used these questions either for methodological discussions or to conduct quantitative studies to assess the impact of digitalisation on various topics such as labour market, workplace, or education.

What all of these have in common is that they use digitalisation as a synonym for permeation and only regard the technological part of the socio-technological transformation process. The means to assess digitalisation in these surveys and studies can be categorised into two strands: (I) quantitative operationalisations via the usage or availability of different devices, technologies, or skills and (II) subjective operationalisations. These operationalisations are extremely heterogenous. They vastly differ and range over several topics, further highlighting the fuzziness and fragmented debate on digitalisation as the many different examples in Table 1 demonstrate.

Table 1 Examples of operationalisation used in quantitative surveys
Full size table

3 Investigating Permeation in Qualitative Studies

Compared to quantitative approaches, qualitative approaches have an inherently lower degree of abstraction (Lamnek 1988), allowing for explorative and non-standardised methodological measures (Sturm 2007). Qualitative measures are especially advantageous when the research object is complex, multi-layered, and when there is no generally shared body of knowledge, as these measures can emphasise the variety and uniqueness of the particular research object (e.g. Sturm 2007). This is particularly important when analysing permeation, as we assume that technologies affect different segments of the labour market to different degrees and that the influence of technology differs strongly depending on the socio-technical constellation, leading to variations by occupational field, organisation, and design of work processes (Orlikowski 2000; Baethge et al. 2007; Helmrich and Zika 2010). This, in return, requires differentiated sector-, domain- and occupation-specific analyses, especially to better understand the incoherent and multi-layered process of permeation.

Admittedly, qualitative studies have difficulties in generalising findings. One reason for this is that there is no consensus on the definition on permeation, so that qualitative studies may examine the same facet of digitalisation but still use different terminology (Gläser et al. 2018; Boes et al. 2020). Another reason that findings are difficult to generalise is the wide range of technologies and digitalisation dynamics studied in single fields (Evans and Kitchin 2018; Buxbaum et al. 2019; Hirsch-Kreinsen et al. 2019), where data depth and granularity make comparisons and considerations of overarching phenomena difficult.

4 Our Mixed-Methods Approach

The fuzziness and shortcomings in both the quantitative as well as qualitative research on digitalisation indicate that there is a need for an integrated approach that overcomes the current state of debate and empirical results on digitalisation (Creswell and Creswell 2022; Greene 2007). We argue that in the current fragmented debate, a mixed-method approach will advance the methodological repertoire for measuring permeation and analysing the impact of digitalisation. Recent years have already seen an expansion in the use of mixed-methods approaches in research on digitalisation (Benitez et al. 2022; Scott et al. 2016). Our mixed-methods approach will follow a pragmatic paradigm (Morgan 2007; Scott et al. 2011). Unlike previous approaches, we do not consider qualitative and quantitative studies to be complementary, and we do not apply them consecutively (Cronholm and Hjalmarsson, 2011). Rather, we use them simultaneously and juxtapose the results of qualitative and quantitative methods for a joint, interrelated interpretation of the state of digital permeation. With this approach, we strive to capture permeation more precisely and more thoroughly, which increases the validity of our empirical assessment on permeation (Völcker 2019).

Qualitative methods allow us to investigate the multidimensional structure of the fuzzy concept of permeation, its underlying mechanisms and the subjective perceptions and consequences for individuals. Qualitative measures may refer to quantitative, subjective, or qualitative dimensions of permeation. Quantitative measures enable us to cover large entities (departments, establishments, sectors) and a large group of employees (and customers), addressing quantitative, subjective, or qualitative dimensions of permeation. The qualitative methods allow us to validate our quantitative findings and supply us with sufficient contextual information to interpret them. Using the qualified quantitative results then equips us with triangulated information which we can use to generalise findings to a certain extent.

We argue that different methodological research methods vary in their ability to measure the different dimensions outlined in our conceptual triangle. While quantitative and qualitative methods have strengths and weaknesses depended on the dimension that is to be measured, we argue that it is sufficient to capture each dimension with a single methodological tool, either a quantitative or qualitative one. In Table 2, we list some of the tools we consider best suited to investigating the different dimensions of permeation separated by methodological research strategy and data source. This table already illustrates the necessity for using a mix of multiple methods.

Table 2 Our proposed research tools to investigate each dimension for different data sources
Full size table

According to this, we propose a multi-method process strategy by combining a mix of qualitative and quantitative methods. On the one hand, this promotes the depth and granularity of the data and, on the other hand, also the broader contextualisation of the results. Hence, the study design facilitates both an explorative perspective and in-depth analysis of specific issues and the possibility to generalise results across sectors, organisations, and occupations. The approach is depicted in Fig. 2. It follows the idea that some research tools can better capture aspects of each dimension at different time points in the research process. However, to gain a complete picture of permeation, the researcher will have to combine all collected information from all steps.

Fig. 2
figure 2

Illustration of our proposed mixed-methods approach

Full size image

To capture the concept of permeation in a given sector, organisation, or occupation, our starting point is an extensive literature review of qualitative and quantitative studies on digitalisation in the investigated unit of interest. Then, this review should be combined with expert interviews on the qualitative methods side and secondary data analysis on the quantitative methods side. This will enable researchers to gain an initial understanding of what permeation means for a given sector, organisation, or occupation, and it enables researchers to identify possible candidates for further empirical investigation.

We then propose conducting expert and employee interviews in the identified units of investigation on the qualitative methods side combined with, and enhanced by, participatory observations. Based on these insights, we propose a quantitative survey among organisations or organisational sub-units and among employees. To enhance the depth and granularity of the data, we propose combining the interviews with observations.

Observations offer the advantage of collecting in-process and situational data as well as information and contextual interpretations about the dynamics of permeation that actors are unable to explicate in interviews (Girtler 2001). For example, observations can help to understand in more depth how the situational use of certain technologies changes work activities, processes, structures, and the work environment. They can also reveal why certain technologies are not used in specific situations, leading to workarounds and extra work for the respective actors. For systematic comparability and triangulation of data, we propose developing observation guidelines to record data in a structured manner, for example via field notes and survey protocols (Girtler 2001). These observations further enhance the interpretation of the collected quantitative survey data. Finally, combining all collected contextual information and empirical findings allows us to converge on a measurement of permeation in the investigated sectors, organisations, or occupations and possibly beyond. To illustrate the benefits of our proposed methodological approach, we will briefly present and discuss empirical findings from our study on the German healthcare sector.

5 Healthcare and German Hospitals as an Empirical Example

We chose the German healthcare sector as an empirical example for four reasons. First, the healthcare sector in Germany has been undergoing significant changes and major reforms, which have induced rationalisation and restructuring processes in the past (Pfeuffer and Gemperle 2013). Overall, the sector has become more diversified in terms of occupational profiles and economic actors, including private healthcare providers. Rationalisation processes have been the key driver of restructuring since the mid-1990 s, and they have led to reducing staff and economising work organisation, work processes, treatment, the utilisation of material (including medication), and cutting down patients’ hospital days, among many other measures to control and reduce costs. Likewise, this sector experienced a strong increase in the use of (networked digital) technologies in work processes. Nurses, caregivers, and other healthcare staff are confronted with technology-enhanced processes ever more often in their daily work. Furthermore, automation, robotics, and the interaction with networked digital technologies are expected to increase even further in the near future. Second, investigating digitalisation and the digital transformation in healthcare has received growing attention from the disciplinary perspective of sociology during the past two decades (e.g., Bollinger et al. 2005; Kälble and Borgetto 2016; Hertling et al. 2021; Iyanna et al. 2022; Li et al. 2022). Thereby, most studies focus on the major restructuring processes, particularly in terms of the organisation of care, rationalisation, changing work processes, and the development of job profiles of healthcare professionals (e.g., Estermann et al. 2013).

Third, we selected hospitals, as there have been increased governmental efforts to drive digitalisation in hospitals in recent years, especially since the E-Health Act came into force in December 2015 (BMG 2022). The most recent measure was the Digital Care and Modernisation of Care Act (DVPMG), which came into force in June 2021 (BMG 2022). Digitalisation of hospitals is not only a relevant topic in Germany, but also worldwide (cf. Mangiapane and Bender 2020). The Healthcare Information and Management Systems Society (HIMSS) developed a cross-national maturity model for the digitalisation of hospitals (Mangiapane and Bender 2020). This model captures the status quo of the implementation of IT systems and IT solutions in hospitals and has been used in the USA since 2005 and in Germany since 2010 (cf. Mangiapane and Bender 2020). Likewise, an increasing number of measures can also be identified at the hospital level to drive digitalisation forward, such as the establishment of data integration centres (MII 2022), the creation of a “Digital Transformation” department (Charité 2022), and other projects such as the “SmartHospital.NRW” (KI.NRW 2021) or the “Innovation Center Digital Medicine” (IZDM n. d.). Fourth, given that hospitals are high-reliability organisations and particularly affected by digitalisation, they offer a unique and exceptional empirical basis. To provide essential and life-saving medical services, many different professions work together in highly specialised teams under high physical and psychological stress. Likewise, they often work in environments with rapidly changing work intensities, a high standard of technology, large amounts of data and are confronted with critical decision-making situations in which small mistakes can have fatal consequences.

5.1 Sampling Strategy

Following our outlined approach, we first conducted desktop research and used exploratory expert interviews to collect differentiated expressions of opinion and context interpretations within the healthcare sector to gain initial insight into the dynamics of permeation and the language of the empirical field (cf. Behnke et al. 2006). It could be shown that, depending on personal interests and a specific institutional context (cf. Mieg and Näf 2006), permeation is mostly understood as the usage of technologies like software and IT tools (I8; I9; I10; I12) and that (university) hospitals are relatively digitalised in comparison to other types of healthcare facilities (I8). A secondary data analysis of different sectors was based on data from the adult cohort of the National Educational Panel Study (NEPS) (Blossfeld and Roßbach 2019; NEPS-Netzwerk 2022). The results implied that permeation and its impact varied on performed tasks and thus on occupation. We therefore wanted to cover as many different occupations in the healthcare sector as possible to be able to survey differences in permeation dynamics in particular. Following this, we planned to conduct both our qualitative as well as our quantitative primary data collection in multiple hospitals, as we could potentially cover a wide range of occupations—from medical to IT or administrative—in this context. Our plan was to conduct surveys in one hospital presumably more permeated by networked digital technologies than the other. Our sampling strategy was thus influenced by our contextual knowledge to identify possible candidates. Our information suggested we focus on two sets of hospitals, a university hospital as an example of a highly digitalised hospital (Sample A), and a foundation that runs multiple smaller hospitals which are presumably less digitalised on average (Sample B). The COVID-19 pandemic affected our data collection efforts immensely. Due to the high workload and the sanitary regulations during the pandemic, we were not able to include observations and group interviews in our qualitative company case studies of the hospitals, and we could not always follow the research plan (Fig. 2) with respect to the time order.

5.2 Qualitative Findings

To capture the three dimensions of permeation, different qualitative methods were used: expert interviews, company case studies including interviews and observations, desk research, document analysis, and study reviews.

Interview and observation guidelines were developed to facilitate a systematic comparability of the primary data (Bogner et al. 2009). Besides this, it was ensured that all three dimensions of permeation were represented in the guidelines. As secondary data the main sources of information were hospitals’ own presentations on their homepages, newspaper articles, and sector reports. The collected information from the secondary data was included, among other things, as contextual information in the collection of the primary data. For example, we ascertained that the university hospital was not only considered to be highly digitalised in its own perception, but the field also perceived it as such, so that other hospitals contact the university hospital to ask for advice on digitalisation (I12).

Furthermore, based on the qualitative data collected, it can be shown that it is not only possible to qualitatively survey all three dimensions of permeation, but also that they mutually influence each other. One interviewee described the permeation of the university hospital with the statement: “I think we have several thousand different IT tools in use here” (I12; own translation from German), which refers to the quantitative dimension. Directly after this, he used the example of a particular tool to explain how the work process had changed (qualitative dimension) and how a specific project had been discontinued because it was not working as expected. He then related the failure to other similar projects and general challenges of hospitals, creating a subjective justification for why the failure of this specific project was not so relevant.

The representative of the foundation’s hospitals stressed the innovative aspects of the hospitals. The foundation was among the forerunners for the implementation of a new digital device, and the hospitals have a co-ordinated baseline standard on digitalisation. The potentials of digitalisation are mainly seen in the interaction with patients, more so than in other areas and processes within the hospitals. According to the representative, the financial potential of the foundation’s hospitals is limited compared to university hospitals.

Overall, we used primary and secondary data to survey the three dimensions of permeation at each hospital. The qualitative methods used to capture the aspects of the qualitative dimension like the contextual use of specific technologies and their influence on tasks, processes, structures, and the work environment, the non-use of existing technologies, possible workarounds and additional work were particularly suitable. Aspects of the subjective dimension like the understanding and relevance of permeation for the empirical field could be captured in an adequate manner. However, we found the quantitative dimension of permeation such as more general statements about the permeation of hospitals in Germany difficult to capture with qualitative methods, since the depth and granularity of the data make comparisons difficult and the contextual and situationally collected data do not allow for generalisations. For example, it is difficult to ascertain whether “several thousand different IT tools” (I12) are a lot or few compared to other hospitals, and which and how many IT tools are used in German hospitals on average.

5.3 Quantitative Findings

Our quantitative findings are based on secondary data analyses with NEPS data and on an online survey that we were able to conduct in both hospital settings among employees. The hospitals were responsible for distributing a link to all their employees via internal communication channels, such as mailing lists, intranet etc. from June to September 2022. In Sample A we received 1143 responses and in Sample B we received N = 333Footnote 1 responses. In the online survey, we addressed all three dimensions of permeation (qualitative, subjective, quantitative, see Table 3).

Table 3 Questions implemented in our survey to measure permeation
Full size table

Looking at demographic variables and occupational distribution, we find that our surveyed sample largely represents our basic population within the hospitals with office occupations being somewhat overrepresented.Footnote 2 In Sample B we knew in advance that employees are deployed in multiple hospitals. About 30 % of our sample work in multiple hospitals and we thus decided to cluster all employees of the foundation as a group for further analyses.

We adapted and built upon several items from different surveys to cover multiple usage and subjective operationalisations to measure digital permeation as can be seen in Table 3. For the usage operationalisations, we used an item from the Understanding Society Panel Study and provided a list of devices (variable A), combining and building upon items from the National Educational Panel Study (Friedrich et al. 2021) and the Linked Personal Panel (Ruf et al. 2022). Respondents were then asked to indicate how often they used ten networked digital technologies with a progressive increase in the complexity of the technology (variables B1 and B2).Footnote 3 Using factor analysis, we then assessed whether we can measure one construct and create an index of summation out of these items. We used all except one item to create two such indices.

For the subjective operationalisation, we again built upon two items used in NEPS and included one from the European Working Conditions Survey (2022). To be able to compare across all variables and indices, we standardised them so that a “0” indicates no (perceived) permeation according to each operationalisation and “1” the maximum amount of (perceived) permeation.

The results for operationalisation A are depicted in Fig. 3 and for operationalisations B to E in Fig. 4. All figures first illustrate that the degree of digitalisation depends heavily on the operationalisation used. Second, they also evince that, contrary to our contextual expectations and qualitative findings, sometimes Sample B is more permeated than Sample A or is equally permeated. This is especially true for the subjective operationalisations and is consistent across occupational groups. We can, however, observe that permeation is dependent on occupational groups.

Fig. 3
figure 3

Use of digital devices by sample

Full size image
Fig. 4
figure 4

Degree of permeation according to each operationalization

Full size image

5.4 Dealing with Contradicting Findings

Looking back at each single step of our empirical approach and how we implemented it for our empirical example, we see that we gained a vast amount of different and unique pieces of a puzzle. These pieces capture different aspects in the multiple layers of permeation in healthcare. After we applied most of our proposed steps, we received a mixed picture related to the permeation of the sampled hospitals. The quantitative studies suggest that the degree of permeation is very similar between the university hospital and the foundation’s hospitals. The use of digital networked technologies is similar with respect to the complexity and the frequency of use between our samples. The similarity in this quantitative dimension is mirrored in the results of the subjective dimension of the quantitative survey. The role of digital networked technologies during work and in the work environment is seen similarly between the university and the foundation’s hospitals. The employee surveys suggest a similar degree of permeation and differences are primarily between occupations, not hospitals.

The information from the qualitative data collection partly points in similar directions. The experts in our interviews emphasise the forerunner status of their respective hospitals. The university hospital stressed that qualitative as well as the quantitative dimensions demonstrate their leading role in digitalisation. The foundation’s hospital pointed to forerunner projects. However, in contrast to the university hospital, the foundation does not approach digital transformation in its (potential) full breadth. Instead, its representatives see specific areas (interaction with patients) as being the most promising. Likewise, they point to the financial limitations compared to large hospitals when it comes to digital transformations. In sum, even though several indicators suggest that the university hospital and foundation’s hospitals are at a similar level, the triangulation of results indicate that the university hospital has a higher degree of permeation and a higher pace of digitalisation.

Regarding the puzzle that our qualitative and quantitative results on the degree of permeation in each hospital are contradictory, we can also give methodological ideas on how to resolve this. As discussed in our proposed approach, we see the application of observations in qualitative company case studies as a way to consolidate the quantitative and qualitative findings. By applying this research tool, we would expect to gain a greater understanding on the specific usage of networked digital technologies in each hospital. The collected qualitative information paired with the quantitative findings have the potential to explain several questions in more detail, such as which specific tools they are using at their workplace, how complex these tools are and how much time the people actually spent using the tools. Comparing these details across hospitals could then answer the question on whether and why the degree of digitalisation seems to be similar in our quantitative findings.

In doing so, we could have gained an insight into why respondents in Sample B perceive their degree of digitalisation to be as high as those in Sample A even though our other evidence indicates different usage patterns. This further substantiates the benefits of using our mixed-methods approach.

6 Conclusion and Discussion

This article addresses methodological challenges that researchers face when they study the digital transformation of enterprises and societies. Digitalisation is arguably a fuzzy and fragmented concept, leaving researchers somewhat baffled as to which methodological toolkits to apply in order to grasp digitalisation. In the article, we propose a multi-method approach with quantitative as well as qualitative methods to assess digitalisation in organisations or societies. Following the framework of the Priority Programme “Digitalization of Working Worlds”, we argue for breaking down the concept of digitalisation into three motion dynamics: permeation, perpetuation, making available. Using permeation as a representation for the degree of digitalisation in the work context, we further differentiate three dimensions of permeation: a qualitative, a subjective, and a quantitative dimension. For these three dimensions, we argue that single-method approaches cannot fully grasp the degree of permeation. Qualitative methods exert their virtues by extracting highly detailed, highly contextualised information which may leave researchers somewhat uninformed about the scope and the generalisability of these findings. Quantitative methods may provide profound information on a large number of units, be it organisations or employees, opening the potential to generalise these findings to other entities. However, this information usually lacks context, and it may miss important aspects that do not surface in quantitative studies.

Our empirical study on the healthcare sector in Germany offered several contradictory findings from the quantitative and qualitative analyses, and these could have lured strictly quantitative or strictly qualitative methodological approaches to misleading conclusions. We argue both qualitative and quantitative methods are needed to more fully grasp permeation. Using a mixed-methods approach allows for a validation and triangulation of our empirical findings.

For our empirical analyses, we proposed a set of data collection methods, e.g., expert interviews, employee survey, desk research, secondary data analysis, and observations. Our fieldwork got hit by the COVID-19 pandemic, so we could not roll out all data collection methods. Still, we were able to conduct the first four methods, even though we had to adjust our timetable to the availability of the relevant actors within the hospitals. These data collection methods produced different data types (quantitative survey data, text data, qualitative survey data) that we analysed to come to a well-grounded assessment of the degree of permeation of the respective hospitals.

From each method we received different parts of the puzzle on the degree of permeation in hospitals. For instance, our quantitative findings give us information on how often certain networked digital technologies are used, which devices are used, and to which extent they influence tasks. We also learned that variance in usage depends on occupations. Our qualitative findings give us details on specific networked digital technologies and their usage, contextual information on hospitals in general, and in-depth details about our sampled organisations.

However, our mixed-methods approach is not bound to specific types of data. Although not provided by the hospitals in our empirical study, digital trace data could offer valuable insights for our research as well: quantitative and qualitative analyses of digital trace data could inform researchers about qualitative, subjective, or quantitative dimensions of permeation. In this sense, our approach offers a blueprint for systematic research designs in the field of research on digitalisation regardless of the specific data types available. It highlights the suitability of qualitative and quantitative methods and emphasises the benefit of jointly applying these methods.