**3.4 Instrument**

The questionnaire was structured in two parts – (I) demographic data and (II) conceptual questions. The latter part was designed to focus on the hypotheses each of which consisted of four questions.


<sup>1</sup> The information provided here was personally gathered as the official government portal (*see*http://sec urity.gov.vc/security/index.php?option=com\_content&view=article&id=387&Itemid=8) includes only some basic contacts.

*An Analysis of Emotional and Psychological Issues among Males at the Correctional Institution… DOI: http://dx.doi.org/10.5772/intechopen.93526*

yourself/Self-hate", "Avoiding what is happening by drinking, smoking", and "Withdrawing from others"). Feeling comfortable to talk to other males in prison about emotional issues was constructed from the following experiences: "Most males here share the same struggles", "We are all on the same level", "As males we are not as tough and aggressive as we often show", "There is a greater sense of bonding", and "There are less chances of being labelled a homosexual". Inmates were also asked to name whether mental health, relationships, sexual issues, money matters, and/or depression currently affect them the most.


#### **3.5 Data collection**

Permission from the prisons' authorities was necessary. This process took months to complete and initially, it created many doubts that the prison officials would want to grant such access to conduct the research. Having received permission from the Superintendent of Prisons, a pilot study was carried out few months before the actual administration of the questionnaire. This preliminary study served as a means of refining the final instrument to be used for the survey testing its reliability and validity. The piloting stage also helped in identifying some of the problems that loomed on the horizon, namely the lack of cooperation from some prison officials and the full conceptualization of the content of the questionnaires

*Criminology and Post-Mortem Studies - Analyzing Criminal Behaviour and Making Medical…*


#### **Table 1.**

*Reliability statistics.*

by some prisoners. Applying the abovementioned sampling techniques, the questionnaires were distributed and collected personally by the researcher on the same day. Amidst many constraints, the collection process initially began to look fruitful. However, there were days visiting the institution realizing that only ten (or no) questionnaires had been completed amidst excuses of being short of staff or that too many other activities are taking place at the same time. Having completed 150 questionnaires after months visiting the institution it was obvious that completing the entire population was an impossible task; and having started collecting the data with that notion in mind, the data collection was ended.

## **3.6 Data analysis**

The statistical software SPSS (v.25) was used for all data analysis. In terms of methodology, I used multiple correspondence and regression analyses in order to describe the pattern of relationship between (categorical and binary) variables. Such analyses were meaningful given that a prior measurement of internal consistency (*see*<sup>2</sup> **Table 1**) provided an overall high-enough3 reliability coefficient (*α* = .812). Multinomial logistic regressions and tests to assess the associated factors with our dependent variables were also run. These were feasible because, despite the relatively small sample (150), the assumptions about the scale of the variables, their linear – parametric (Pearson's) and non-parametric (Spearman's rho and Kendall's tau-b) – correlations, and homogeneous, normal errors were reasonable. Generally, any related difficulty to discern or recognize possible relevant deviations from assumptions with small samples did not invalidate the test, which remained valid under these assumptions.
