• No results found

5. Phase One: Discussion

5.4 Strengths and Limitations

5.4.1 Design

This phase of the study used a cross-sectional quantitative design. A pre-existing data set was used that contained various measures of symptomatology, functioning and possible selves at a particular time point.

There are a number of strengths and limitations of undertaking secondary data analysis. Generally, the use of secondary data is an efficient use of resources such as time and money: projects are able to be completed and findings disseminated in a timely manner and, as such, any contribution to new knowledge is accelerated (Johnston, 2014). The efficiency of secondary data analysis was considered particularly advantageous for this project considering the constraints on time and resources of completing a Clinical Psychology Doctoral thesis. This efficiency allowed for the inclusion of a qualitative aspect (phase two) consequently meaning that the study provided a more comprehensive contribution to the understanding of the topic.

Using secondary data has potential limitations. Usually the study population and/or the measures collected are not exactly what the secondary researcher would have chosen (Smith et al., 2011). In the dataset made available here, the main study measures were all well known with established psychometric properties and the

research questions were designed with the measures in mind. Considering the study population, participants were recruited from a specialist mental health team providing a service for a very specific group of individuals, i.e. those experiencing FEP. This was the population of interest here. In addition, the researcher was able to access the same service to conduct phase two of the study, which is a strength of the current research.

A further potential disadvantage of secondary data is that the secondary researcher was not involved in the processes of collecting, coding and inputting data.

A such, they may not be aware if there were any specific problems with any of the aforementioned (Boslaugh, 2007). In order overcome some of these issues, the researcher gathered as much information as possible about the collection of the dataset. This is detailed in the relevant sections below. Time was also spent “getting to know” the dataset. Consideration was given as to whether it was representative of the population it applied to by comparisons of participant demographics and

diagnoses with other data collected in first episode psychosis samples (section 4.3).

The correlational design of this phase only determines if there are significant relationships between variables but does not allow for conclusions regarding the existence or the direction of causal relationships. For example, the study revealed a significant positive correlations between functioning and optimism about future hoped for possible selves but cannot explain whether better functioning causes higher levels of optimism or indeed whether higher levels of optimism lead to better functioning.

As data were collected at one time point it is not possible to observe any changes over time, such as how individual’s possible selves might change over the course of FEP.

A particular strength of this phase is the use of an open-ended measure of possible selves. This provides a more detailed picture of how those who have experienced FEP describe their possible selves compared to the study by Norman et al. (2014) where pre-defined lists of possible selves were used. This allows the study to offer more of an insight into how experiencing FEP may impact an individual’s sense of self and views regarding their future.

5.4.2 Sample

The heterogeneity of this sample in terms of diagnoses is a strength, being representative of the variety of presentations of FEP seen within EI services.

Considering demographic variables, the sample of participants in the dataset is comparable to other EI samples from the region but due to a lack of ethic diversity the sample is not comparable to samples recruited from other areas in the UK. This could be seen as a weakness, placing limits on the generalisability of findings.

Sample size calculations recommended a sample size of 80. When the dataset were screened for missing data, it became apparent that there were varying amounts of missing data across the study variables and this resulted in some analyses being underpowered.

5.4.3 Measures

5.4.3.1 Balance

In this study, an overall measure of “balance” was used due to the relatively small sample size: balance was calculated across all categories of hoped for and feared possible selves. A more specific targeted measure of balance between hoped

for and feared possible selves in one particularly category, e.g. “emotional/physical wellbeing”, or between specific possible selves, e.g. those relating to FEP, may have yielded different results. In research investigating possible selves in academic achievement, researchers commonly examine balance between academic possible selves rather than overall balance. In this area results have consistently found support for the motivational impact of balanced possible selves.

Quinlan, Jaccard and Blanton (2006) have further highlighted that calculating balance by counting the number of hoped for possible selves that are “matched” by a feared possible self in the same domain is problematic. This method does not control for the main effects of hoped for or feared possible selves, nor does it control for the number of possible selves listed. In this study, participants were only asked to generate three hoped for, expected and feared possible selves but it may be that if an open ended measure had been used that the fourth, fifth etc. possible selves listed may have shown participants to have balanced possible selves.