II. 1 CONTEXTO HISTÓRICO.
II.4.4. TEATRO ITALIANO EN ESPAÑA: LA ÓPERA.
First proclaimed by Auguste Comte in the 19th Century, the positivist approach has developed various mutations and is associated with a number of disparate philosophical schools of thought (Hughes and Sharrock, 1990). The intellectual debate on the authority of positivism, despite its long journey, remains alive today. Hughes and Sharrock (1990) refer to positivism as
orthodoxy as its legitimacy was unquestioned for some time. Endorsed by John Stuart Mill, Herbert Spencer, Emile Durkheim, and Karl Marx, albeit in various versions, there was a belief that society could follow the same logic of enquiry as that employed by the natural sciences (Hughes and Sharrock, 1990). Hence, the social world can be studied according to the same principles, procedures, ethos and laws as the natural sciences.
As social processes are viewed as subject to casual laws, applying objectivity, rationality and rigorous scientific methods of enquiry to establish truth, it is assumed the researcher can identify regularities and causal relationships of social phenomena. The research process starts with a hypothesis or tentative explanation based on the assumption that the investigator is objective and remains detached from phenomenon under study, (Clarke, 1999). Testing the hypothesis to either accept or disprove it involves collecting facts, while the hypothesis remains fixed throughout the research process. To achieve this, survey methods and experimental designs are employed that limit the interaction that occurs between the researcher and the researched (Clarke, 1999). Research techniques include highly structured questionnaires or interview schedules that contain predetermined, standardised categories into which individuals responses are fitted. Systematic sampling techniques are employed to control bias and ensure internal validity (Bryman, 2001). Box 4 details the scientific method in relation to evaluation.
Box 4 Evaluation and the Scientific Method
The evaluator is separate from the practitioners and from the practice supposedly in order to ensure neutrality and objectivity.
Practice is conceptualised as informed by defined inputs and measurable outcomes.
Decisions are made about intended outputs depending on their
susceptibility to measurement, thus simplifying what may otherwise be complex, diffuse and multi-faceted goals and processes.
Different interventions are applied to control and experimental groups so that differences in outcomes can be measured and compared, and these differences are related in causal ways to differences in inputs.
Interventions in the control and experimental groups are controlled for the period of the intervention so that measurements can be made, thus not allowing for practice as a developing and changing process.
Causal relationships are sought between inputs and outputs. (Source: Everrit and Hardiker (1996, p. 46-47).
4.6.1.1 Experimental Designs and Evaluations
Experimental research designs provide a sound method of arriving at causal explanations in evaluations (Rossi and Freeman, 1993) and are “strong in terms of internal validity” (Bryman, 2001, p. 39).
That randomised experimental designs are of proven scientific merit in evaluations is an acceptable view. Thus, logic and rules of scientific method are an indispensable component for establishing the effectiveness of interventions and amounts to identifying relative causality.
In humanitarian evaluations following and during civil conflicts such as Somalia, pre-test, post-test, control group designs can prove problematic considering the nature of interventions. Similarly, it is rather difficult to conduct a pre-test, post-test, control group design in rapid-onset disasters where there is little or no warning. The South East Asian Tsunami in 2004 was a natural
disaster where there was little, if any warning at all. Even, in slow- onset disasters, the treatment in most cases is based on appropriate eligibility criteria, making it impossible to construct a conventional group. Both case studies within this research for example, had clear eligibility criteria as populations without access to clean, safe water and sanitation. Creating a control group would have been problematic but more importantly unethical, as it would involve withholding relief aid to those who met the criteria, which directly conflicts with the purpose of relief aid – to save lives.
Clarke (1999) identifies the following potential limitations of randomised experimental designs:
Ethical considerations such as withholding beneficial resources from clients in the name of research;
Comparability, non-random processes can threaten group equivalence; The potential for creating inequities between groups; and
Generating conflict and resentment between experimental and control groups.
Humanitarian agencies would likely object on ethical grounds to randomisation as a method of assigning individuals to treatment and non- treatment groups (Clarke, 1999). How would a humanitarian agency provide clean water to one group in Somalia, while denying another group access to the same service for the purpose of carrying out an experiment to meet evaluation criteria? Furthermore, this creation of inequities between groups would likely generate conflict as a result of resentment among those not receiving aid. In ‘complex emergencies’ however, the observed effects often occur by chance due to problems of attribution (O’Keefe, Kirkby and Cheetham, 2002 and Buchanan-Smith and Collinson, 2002). Thus, factors operating simultaneously to the programme may cause effects that are not related to it.
Evaluation is about generating evidence of the effectiveness of humanitarian action and the positivist approach provides one of several methods of accomplishing this. Consistent with Everrit and Hardiker (1996), this research
recognised the merit in employing the positivist paradigm in evaluation research. The objective of this research however, was not to establish whether there is a cause-effect relationship but to assess the effects of the programme on adaptive capacity, resilience and vulnerability in the study locations. The limitations of the positivist paradigm form the rationale for the subjectivist paradigm.