Tilburg University Modular Care Provision

68
4.2  Developing  the  research  constructs  and  research 
framework 
The starting point for the case research is the research framework and questions 
(Eisenhardt, 1989, Voss 2009). The research questions to be answered by means of 
case study research have been stated in Chapter 1. 
Modular Care Provision
Table 4.1  Overview of study design
Research stage
Activity
Our design choices
Research framework
Definition of research 
questions
See Chapter 1
A-priori constructs
We developed a conceptual narrative containing 
constructs, but no specific relations nor 
propositions
Choosing cases
Sampling of cases
Multiple case design: four cases
Case sample is based on purposive sampling, literal 
replication and maximum variation in case history 
and accessibility strategy. 
Within case focus
Total care process
Research protocol and  
instruments
Data collection methods
Case study protocol
Multiple sources of evidence

Semi-structured interviews

Company documents

Observation visits
Conducting the field 
research
Sampling of target 
respondents
For each case, respondents covered different 
organizational levels, various service types and 
different phases of the care process. For each 
case: regional director, team leader home nursing, 
team leader home help, front desk employees, 
start-up nurses, key nurses and representatives 
from complementary services, call centers, and 
marketing.
Actual data collection
Conducting 38 interviews of 1 – 2 hours, collecting 
a variety of company documents and three one-day 
observation visits per case. 
Receiving feedback on interviews and case 
narratives by means of member check and expert 
meetings.
Data analysis
Preparing for data analysis
Interview transcription
Data reduction: coding
Data analysis
Within case: Data display
Cross case: data display, pattern identification, 
comparison with existing literature

---

69
In case research focusing on theory development, it is important to consider a prior 
view of the general constructs and categories we intend to study and their relations, 
no matter how inductive the approach (Voss 2009). This can be done through the 
construction of a conceptual framework that explains, either graphically or in narrative 
form, the main things that are to be studied (the key factors, constructs or variables) 
and the presumed relationships amongst them (Miles and Huberman, 1994). Building 
a conceptual framework will force the researcher to think carefully and selectively 
about the constructs and variables to be included in the study (Eisenhardt, 1989, Voss, 
2009). The a priori framework used as a basis for the empirical part of our research 
was presented in Chapter 2 of this thesis. This chapter included a narrative on the 
operationalization of demand-based care in four dimensions and various aspects and 
practices of modularity that are of potential importance in bringing these dimensions 
into practice in long-term care. As such, starting from literature, we specified the 
constructs that are potentially important (Eisenhardt, 1989). However, our narrative 
did not include detailed relationships among variables, since researchers should avoid, 
as much as possible, thinking about the specific relationships between variables and 
theories, especially at the outset of the research process (Eisenhardt, 1989).
For the phenomenon central to this research, i.e. the concept of modularity in the 
context of long-term care for the elderly, several theoretical insights and concepts 
were already available, however many were missing as yet. On the one hand, there are 
several contributions that address particular elements of modularity in services and 
health care such as the possible complementarity and intertwinement of modularity 
in service product and service process, and the vital role of people in bringing about 
modular services (e.g. Meyer and DeTore, 2001, Bohmer, 2005, Pekkarinen and 
Ulkuniemi, 2008, Voss and Hsuan, 2009). On the other hand, we lacked information on 
how exactly various aspects and practices related to modularity, such as components, 
interfaces, and the achievement of customization through mixing-and-matching, 
come across in and are influenced by the long-term care context. Hence, following 
Zomerdijk (2005), the challenge for this study was to ensure an open mind with a 
minimum amount of bias, while at the same time doing justice to and benefiting from 
theory that was already available. Therefore, we summarized our narrative conceptual 
framework into a relatively simple conceptual model that displays potentially important 
variables regarding modularity in long term care for the elderly (Figure 4.1). The 
exact application of modularity, however, was a black box and the conceptual frame 
did not contain propositions or hypotheses other than that it might be advantageous 
to use modularity in order to advance long-term care towards demand-based care 
provision.
Resear
ch desig
n and methods
Chapt
er 4

---

70
4.3  Choosing cases
When conducting case research, choices to be made include the number of cases to 
be used, case selection, and sampling. In addition, the within-case focus needs to be 
determined (Eisenhardt, 1989, Voss et al., 2002, Yin, 2003).
4.3.1  Sampling and case selection 
The study of a single case provides a great deal of depth of observation, however, the 
generalizability of findings is limited. Multiple cases may reduce the depth of study, but 
can augment external validity (Voss et al., 2002). Furthermore, analytic conclusions of 
multiple cases will be more powerful and the external generalizability of the findings 
is improved through either literal or theoretical replications (Yin, 2003). For these 
reasons, our research included multiple cases. Being able to compare insights on various 
modularity dimensions from different cases in long-term care for the elderly will 
strengthen the precision, validity, and stability of our findings (Miles and Huberman, 
1994), even though some depth of observation may be lost (Zomerdijk, 2005).
A vital choice to be made when using multiple case studies is the case selection. Our 
case selection was bounded by the field in which we intended to do research, being 
the sector that provides long-term care and related services to elderly people living 
independently. Within this research field, we used purposive sampling to select cases. 
Modular Care Provision
Figure 4.1  Conceptual framework
Demand for long-
term care and 
related services

Modularity aspects
Components

Modules

Interfaces

Packages 


Improvement towards 
demand-based provision of 
care, i.e. enlargement of:
Choice options

Variation

Client involvement

Joint delivery


Possible influencing factors
Care
Nature of demand

Position of client

Services
Role of people

Intertwinement of product and 

process
Intangibility, heterogeneity, 

inseparability, and perishability
Other
…


---