Methodological basis of the study

The complexity and the dynamic nature of the concept CS, the study findings on the flaws of reliability and validity in the earlier, mainly quantitative studies (see chapter 2.3) and the aims of this study guided the planning of the study design. Since the purpose of this study was to describe effects of team supervision and the effects of this intervention on quality of care the focus of the study was clear, but the challenge was to find the methodological solutions to describe the multifaceted complexity of responses to the intervention in the teams, but also individually within the team members. It was thought that no single method would be adequate to ensure a comprehensive approach in this kind of study, but a set of complementary and confirming methods would be needed to accomplish the study and to ensure the validity and credibility of the findings.

The use of two or more data sources, methods, theories or investigators in a study of a single phenomenon, that is, triangulation, has been utilised in nursing studies since the early 1980s (e.g. Mitchell 1986, Murphy 1989). The strategy of triangulation refers to the use of multiple methods or perspectives to collect and interpret data on a phenomenon, in order to converge an accurate and/or overtake a complete representation of reality (Mitchell 1986, Denzin 1989, Dootson 1995, Begley 1996 a, b, Polit and Hungler 1997). The different types of triangulation (Denzin 1989, Burns and Grove 1997, Polit and Hungler 1997) such as data, method, investigator, theoretical and multiple triangulation have been found useful in nursing studies (e.g. Murphy 1989, Connelly et al. 1997, Shih 1998, Hyrkäs and Paunonen 2000), because the phenomenon under investigation is complex and dynamic in nature, but also cutting across the traditional disciplines (Mitchell 1986). Thus, the strength of triangulation for overcoming the biases of ‘a single method, investigator or theory’ has been evidenced for increasing the confidence in and validity of the results, deepening and broadening understanding of the domain under study, but also for overcoming the biases, and on the other hand, the holistic fallacy of naturalistic research allowing divergent results to enrich explanations and descriptions of reality (Mitchell 1986, Dootson 1995, Begley 1996 a, see also Murphy 1989, Connelly et al. 1997, Shih 1998)

The use of a triangulative approach integrating research methods is based on the assumptions that (1) the world is viewed as a whole, an interactive system with patterns of information change between sub-systems or levels of reality, (2) both subjective and objective data are recognised as legitimate avenues for gaining understanding, (3) atomistic and holistic thinking are used in design and analysis, (4) all those involved in the study are study participants and (5) conflicting views are not ignored but sought with provision for systematic and controlled confrontation, since conflicts are seen to offer a potential for expanding questions and consequent understanding. (Myers and Haase 1989, Burns and Grove 1997). The assumptions were acknowledged by the researcher and guided the course of the study.

Multiple triangulation was seen justified because of the complex and dynamic nature of the concept containing many dimensions. Triangulation was utilised in this study as an attempt to increase the information obtained from the participants in team supervision to provide a more holistic view on the effects

of the intervention (see Foster 1997). The advantages of triangulation were seen to improve validity and credibility of the study, but also to enhance the completeness of the findings by combining various techniques of triangulation (see Knafl and Breitmayer 1991, Dootson 1995). Multiple triangulation was seen to increase the investigator’s possibility to describe and understand in depth and breath and thus more fully and thoroughly the perceptions and experiences of the intervention in the ward/team context of the supervisees studied (see Mitchell 1986, Begley 1996 b). Further, the purpose of multiple methods was to overcome the deficiencies and biases that stem from any single method (see Mitchell 1986). The aim was to achieve findings in which the variance that was obtained reflected the trait being studied rather than the method being used to measure (see Mitchell 1986)

Multiple triangulation was accomplished in this study by means of data and methodological triangulation (see also Mitchell 1986, Denzin 1989, Begley 1996a, Burns and Grove 1997). Data triangulation involved using multiple data sources, in other words patients, supervisees and teams. The patients and participating health-care professionals represented two ‘data source groups’ with a similar focus on provided care, but from their own and divergent perspectives in order to validate the impact of self-monitoring and patient satisfaction feedback on the intervention. The other sources of data were the supervisees and the teams, representing triangulation by person, to obtain the different views about the effects of the intervention on different levels in order to contribute a more complete description and understanding of the topic under investigation. Data triangulation also involved using the five participating wards as ‘data sources’ and applying triangulation by space in order to test the consistency of the findings in multi-sites.

As a means of methodological triangulation, both qualitative and quantitative methods were used in the study. This occurred at the level of data collection and design to address the same research task (see Kimchi et al 1991, Morse 1991). The selection of the methods was based on the assumption that each would tap a different aspect (i.e. individual as a team member and the team) and dimension (i.e. prerequisites, changes and effects) of the research tasks to produce a rich, comprehensive and complete picture of the phenomenon under study. The multiple procedures for data collection and analysis are illustrated in Table 1 (Table 1). The approach to triangulation was simultaneous. The ultimate purpose of the across methods triangulation was to look for commonalties and thus obtain more confidence and convergent validity in the findings. The unit of analysis triangulation occurred in this study through incorporating two levels (i.e.

individual and team/ward) in the analysis and thus trying to take into account individual team-member’s perceptions, and to describe these across the team members, the extent of a coherent team perspective, whether shared, conflicting or complementary (see Knafl and Breitmayer 1991, Begley 1996 a) for obtaining a more complete description and understanding of the phenomenon under study.

The process of triangulation progressed by conducting first the qualitative and quantitative studies true to the paradigmatic assumptions of each method (Dootson 1995, Foster 1997). Pertinent results within each method were distinguished next, and the confidence and validity of the findings were examined. Integration across the methods occurred after the qualitative and quantitative results were achieved and the integration was guided through conceptual validation of the findings (Mitchell 1986, Foster 1997).

Table 1. Methodological triangulation, unit of analysis and the instruments used in the study Prerequisites for professional

development (quantitative, individual)

Changes in the selected effects of CS (quantitative, individual)

Described effects of CS during the intervention (qualitative, individual)

Effects of team supervision at the end of intervention

(qualitative, team)

TEAM FACTORS Atmosphere (AF1) Team spirit (GF1)

Team’s functionality (GF2)

Commitment to work and organisation (OC2)

Effectiveness on team work Multi-professional collaboration

In relation to team Human relations

Interview theme:

Team supervision and the team

INDIVIDUAL FACTORS Performance motivation (OC3) Growth motivation (OC1) Reflectivity (RF1)

Effects on expertise

Clarified theoretical approach to practice Deepened self-awareness

Personal strengths at work

Work patterns Supervisee

Interview theme:

Team supervision and work

ORGANISATIONAL FACTORS Work’s encouragement value (WF1) Possibility to influence (WF2)

Participatory management style (MF1) Performance oriented management style (MF2)

Task and goal systems (OF2)

Strengthened practical facilities One’s contribution to the ward’s overall functioning

Working on the ward

Quality of care

Interview theme:

Team supervision and quality of care

4.2. Intervention

Finding supervisors for the teams was a challenge. The university hospital database listing the qualified supervisors was extensive (see Hoitotyön työnohjaus 1997, see also Työnohjauksen koordinointi 2001), but many of the contacted supervisors hesitated to commit themselves to an extensive team supervision project.

The reason for hesitation was that although the supervisors had experience of one-to-one and group supervision, many were unfamiliar with multi-professional team supervision in somatic health care. The majority of the contacted supervisors suggested a solution of two supervisors working together, and a list of those accepting this settlement was sent to the participating wards. This procedure enabled the teams to choose their own supervisor pair, and the supervisors were provided with profound information about the arrangements for team supervision and the study.

The chosen ten (10) supervisors, two on each of the five wards, were trained and experienced professionals from a variety of disciplines (i.e. nursing, medicine and psychology) within the organisation, but not belonging to the unit in question. The mean of the supervisors’ work experience in CS was 15 years.

They had used a psychoanalytic, psycho-dynamic and systems theoretic approach or an eclectic combination of these as a frame of reference in their work as supervisor. (Hyrkäs et al. 2002a)

The team supervision intervention started on the participating wards in August/September 1995. The sessions were organised at intervals of about 3 to 4 weeks, with sessions lasting for an hour and a half. The team supervision sessions were arranged in a secluded room or on other premises in the hospital. The criterion emphasised while choosing the place for team supervision was that it was not too far from the ward, but at a distance that the sessions were not interrupted or disturbed by the ward’s activities. The intervention intervals were from January to May and from August/September to December. During the summer months (from June to August) the team supervision was not organised because of the holiday season. The planned duration of the intervention was three years, but the final decision was transferred to the supervisees on the participating wards.