• No se han encontrado resultados

4. ANÁLISIS E INTERPRETACIÓN DE LOS RESULTADOS

4.5. Procedimiento: Encuesta dirigida a los trabajadores

4.5.3. Medición de Campos Electromagnéticos de las Subestaciones de la

Multiple-Case Study has been selected as a research strategy to investigate a contemporary developing phenomenon of social media use in organisations and

answer the research questions. This section applies the validity-tests to critically assess the design and address methodological challenges.

3.6.6.1 Construct Validity

Case studies, especially those which are trying to build theory and seek to identify emergent constructs in the empirical data, are generally high in construct validity. The constructs are not created a priori based on theory but emerge from within the data. The process of theory building following the Grounded Theory approach as discussed at the beginning of this chapter starts as close as possible to the “ideal of no theory” (Eisenhardt, 1989; Glaser & Strauss, 1967). Although the literature review and pilot case studies provided an initial framework (Figure 24) and indicated a candidate theory, the constructs and relationships suggested by the framework might not find reflection in the final theory. Multiple data collection methods allow triangulation and so provide another pathway to improve construct validity (Eisenhardt, 1989).

This study employs at least three different sources of information – interviews with HR representatives, interviews with employees and information from the public (and where possible private) social media sites, the findings can be triangulated based on the data source. The interviews are conducted with representatives of different actor- groups (e.g. managers, employees, candidates), and thus the findings are triangulated based on the informant. Comparing data and findings across the cases allows further triangulation across cases. The theoretical constructs, such as the activities on social media and participant groups, are developed from the empirical data and are grounded in the data rather than in theory.

3.6.6.2 Internal Validity

Comparing findings across cases increases internal validity. Internal validity is further improved through the application of data analysis methods which address rival explanation, and generation of propositions based on empirical data and not solely on theoretical constructs (Yin, 2009).

The relationship between actors (actor groups) and activities is to be established as part of the study. The internal validity is increased through replication across multiple cases. Comparison and location of findings within extant HRM and IS literature is a third strategy to improve internal validity.

3.6.6.3 External Validity

The multiple-case study is designed with improved external validity in mind, especially in comparison with single case studies (Yin, 2009). The qualitative rhetoric prefers the term “transferability” over generalisability when evaluating external validity (Denzin & Lincoln, 2008). The definition of study boundaries (e.g. a single industry) initially reduces the settings in which the findings can be transferred, but at the same time sharpens the focus and so improves the transferability of the findings within these boundaries (Gioia & Pitre, 1990). The application of replication-logic during case selection phase, cross-case comparison of findings during the analysis phase, rigorous, transparent and reliable study design and reporting of findings are tools which improve transferability at several stages of the study (Yin, 2009).

This study focuses on a bounded eco-system of large UK firms. The environmental – cultural and legislative structures in the UK frame the population of organisations under investigation. This conscious limitation of potential and actual cases sharpens the theory and helps to specify a population of organisations to which the findings apply.

3.6.6.4 Reliability

The reliability of findings from multiple case studies depends to a great extent on the discipline of the researcher. The qualitative test for “credibility” is a better-chosen term here (Denzin & Lincoln, 2008). What can be (or has been) done to make the findings credible? A detailed case study protocol should allow another researcher to set up a replication study (Yin, 2009). Consistent field notes and a well-maintained database would allow reviewers and colleagues to comprehend the study’s execution and data analysis, and to critically appraise the findings and conclusions (Eisenhardt, 1989). Openness and transparency during planning, execution and analysis of the data reflected in detailed documentation improve credibility.

During the study, several sets of notes are being created and maintained. These notes cover preparation, data collection, over data analysis, theory formulation and reporting. Field notes are taken as handwritten notes and where applicable transcribed for analysis. Interviews are digitally recorded and transcribed, websites are downloaded, and screen-shots are taken. Anonymised (where applicable) data is permanently stored to allow going back and forth between data collection and analysis.

A case study protocol with (a) interview dates, (b) interview scripts and (c) informant’s anonymised details as well as (d) content and dates of other information is kept. The protocol also includes a record of actions such as (e) phone calls or correspondence exchanges so that support can be provided for replication studies.

Methods and techniques employed to analyse the data and derive theoretical constructs and formulate theory are also transparently documented to contribute to reliability and credibility.

3.7 Summary

The chapter introduced a process-view of HRM as a communication system as the underlying framework for this study. The research questions, aimed at investigating whether, when and how social media use in organisations is strategic, are formulated and stated. A presentation and analysis of multiple theoretical paradigms such as subjectivism, interpretivism, and critical realism follows a discussion of different approaches to theory building. Each paradigm is critically assessed as a potential candidate for providing theoretical lenses and methods to answer the research questions.

Selection of qualitative multiple-case study as a research strategy leads to a detailed description of the methodology and design. Each practical step of the study execution – case selection, data collection, qualitative data analysis and interpretation are presented and discussed.

4

Data Collection, Interview Data and Case Description

The research project inevitably requires empirical data to be collected. This chapter focuses on the description of the selection process for case organisation, interview method and the study participants (Figure 27).

Figure 27 - Data Collection Chapter relative to overall Research Design

This chapter provides detailed descriptions of the case selection and data collection process. First, the detailed description of the analysis process improves the replicability of the research. Second, rigorous notes about the work done add to transparency and make the findings more “believable”. Finally, building up to the data analysis and findings, this chapter provides descriptions of who the participants were, which helps us assessing and understanding individual informant’s statements, motivations and viewpoints.

4.1 Introduction

Following the input from Literature Review and the Methodology chapters, this chapter is concerned with the “method”. The Literature review identified the research setting: social media in Human Resource Management, the Methodology chapter

Philosophy and Methodology

Literature Review Research Paradigm

Subject Matter Literature Review

Research Question / Problem Data Collection Methods

Analysis Method

Research Design

Research Method

Design Type Selection

Interview Data Data Collection

Coding First-level and second

level constructs

Theory Analysis

identified the methodological and philosophical basis for this research as a qualitative case study. This chapter is continuing to build upon the previous chapters by providing more detail on how the research was carried out in practical terms. As discussed in the “Research at the paradigm crossroad”-section of the methodology chapter, data quality is an essential feature of good research (King et al., 1994). The content of this chapter is structured around the five rules for data quality: (1) recording and reporting of data generation process, (2) collection of data from multiple sources, (3) recording of observable and not observable/not measurable data and concepts, (4) recording of data collection process, and (5) recording and reporting of data analysis process. This chapter sets off with a discussion on different case selection approaches, providing a report on data collection and data generation process. Next, the case selection methodology which was developed specifically for this research is introduced. The application of the selection methodology within this research is then presented and individual cases (the organisations) are described in detail providing a record of data sources and observations. The detailed description of the case organisation is followed up by description of individual interview participants, focusing on the methods and procedures of the data collection process. The chapter concludes with a description of the interview and transcription techniques applied: providing an overview of data analysis procedures which are further described in Chapter 5.