• No results found

Step 12: Data analysis.

7: Specification of the structured, self reported questionnaire as measurement instrument discussed in depth the design, compilation and

3.4.2 Research process

The research process followed in this study was linear, systematical and sequential and was also presented in the same way (Gerrish & Lacey 2010:14). A quantitative research design was regarded as appropriate to answer the research question. This choice resulted in data collection by utilising a structured, self reported questionnaire, capturing of the numerical data by a statistician, and a statistical analysis to discuss the research results in order to formulate recommendations.

Polit and Beck (2010:241) state different techniques are available to control a study. Applying these techniques will ensure that the study complies with criteria for internal and external validity (Polit & Beck 2010:246). A definite distinction can, therefore, be made between internal validity and external validity.

3.4.2.1 Internal validity

Internal validity validates whether independent variables have any influence on dependent variables. Polit and Beck (2010:246-247) identify and group the threads to internal validity into different classes.

Selection threats occur when pre-existing differences between the groups under study exist which can affect the dependant variable in a way unrelated to the effect of the independent variable. To bridge selection threats as a potential risk to internal validity, the researcher utilised random sampling to test-retest and pre-test the structured, self reported questionnaire. In this study selection threats was acknowledged but not regarded as a major threat because other control steps were in place to ensure the validity of the study.

Chapter 3: Research process

- 168 -  Rigour in the study

Burns and Grove (2009:34) explain rigour as striving for excellence in research. It is possible to achieve rigour through discipline, adherence to detail and strict accuracy. This study was rigorously conducted by strictly following the specific steps of research as tabulated in Table 3.1 and a logical, deductive reasoning process as indicated in Figure 3.1. Burns and Grove (2009:34) state this ensures detail and logically link the scientific investigation process. Rigour was also ensured by the careful selection of a representative sample for this study, objectivity of the structured, self reported questionnaire, data collection and statistical analysis.

 Precision in the study

Burns and Grove (2009:35) accept precision as an aspect of rigour. Description and details of the research design and methodology followed in this study enhanced accuracy, detail and order, which are regarded as crucial aspects of precision. The conciseness of the research purpose and the detailed development of the research design were favourable to claim precision in this study.

 Control in the study

According to Burns and Grove (2009:35), control in a study involves all the steps taken by the researcher to eliminate mistakes or decrease the possibility of error. The application of control in this study increased the probability of accurate statistical findings and ensured correct reflection of the competencies required by child nurse practitioners. The achievement of control in a study is made possible by the design of the study (Burns & Grove 2009:33). Different methods of control should be in place and evident in a study in order to answer correctly to a quantitative design. The descriptive design, as applied in this study, was an example of control because of the little or no influence the researcher had over the respondents in their natural

Chapter 3: Research process

- 169 -

setting. Polit and Beck (2010:107) validate research control as the elimination of contaminating factors which could influence the variables in study. A positive outcome of the application of control in a study is, according to Burns and Grove (2009:689), the reduction of bias.

 Bias in the study

Bias is defined in the Concise Oxford English Dictionary (2006:130) as “a systematic distortion of a statistical result due to a factor not allowed for in its derivation”.

The distortion of research findings leads to unreliable results, which is not a true reflection of the outcome of the research; it therefore affects the quality of evidence of a study (Polit & Beck 2010:238). Research studies are conducted to discover the real answers to an existing phenomenon. Burns and Grove (2009:220) and Polit and Beck (2010:107) agree that certain methods or strategies should be utilised to prevent bias in a study. According to Burns and Grove (2009:220,353), not all bias in a study is easily recognisable. Some of the examples identified and prevented in order to exclude any influence of bias in this study included the correct selection of a suitable representative sample, the objectivity and composition of the structured, self reported questionnaire, the data collecting process, statistical analysis and the generalisation of the data. Polit and Beck (2010:107) opine that bias can never be avoided in total and refer to research control as a method to prevent bias. A detailed description of control for the purpose of this study was presented in Section 3.4.2.1: Internal validity (in this Chapter see paragraph: Control in this study, page 167).

The researcher is of the opinion that in this study bias was combatted by the application of rigour, precision and control methods.

Chapter 3: Research process

- 170 -

3.4.2.2 External validity

The strategies to enhance the validity of the study are discussed.

 Generalisation

Within a quantitative study external validity relates to the degree of generalisation of the research results to the population (Polit & Beck 2010:248). If the research results are generalised to a target population, it is important to select a representative sample (Polit & Beck 2010:248). In this study the implication of generalisation was the possibility for new research.

 Replication or repeatability

Repeatability refers to future repetition of the research in another context or population (Polit & Beck 2010:248). The authors elaborate on the importance of multi-site studies for generalisation purposes.

3.4.2.3 Limitations of the study

Limitations of a study are regarded by Burns and Grove (2009:41) as restrictions or problems which may reduce the generalisation possibility of the study. These authors identify two types of limitations, namely, theoretical limitations and methodological limitations.

o Theoretical limitations include weaknesses in a study framework and

conceptual operational definitions of variables. These weaknesses lead to restriction in abstract generalisation of the results and findings (Burns & Grove 2009:41). No theoretical limitations were identified in this study. o Methodological limitations include weaknesses in the study design.

These weaknesses can limit the credibility of the findings and restrict generalisation (Burns & Grove 2009:41). In this study limited generalisation occurred due to the non-probability sampling, including convenient and snowball methods used. Generalisation was further limited

Chapter 3: Research process

- 171 -

by the fact that the study was conducted in Gauteng only. Results could therefore not be generalised as representative of the rest of South Africa.

3.5 CONCLUSION

This chapter outlined the research process which was guided by a quantitative design in order to deal with the research question in a scientific manner. The chapter focused on the five phases of the research process. Each phase and how it was applied in this study were discussed in depth.

Chapter 4: Research results and discussion

- 172 -