14544 articles - 12260 brèves

2 Methodology

Mise en ligne : 26 July 2005

Texte de l'article :

Methodology

Viewed as a form of mathematical modelling, this study of recruits’ adaptation to correctional officer work required at the outset a thorough knowledge of the field ; a very rigorous methodology allied with theoretical principles validated by 19 scales and codified measuring instruments whose psychometric features, which are presented in each case, referred to earlier, validated studies, including six new scales. The six new scales were pertinent to the research being conducted by the Correctional Service of Canada (CSC) for our specific needs (Tellier & Serin, ibid).

By codification we mean an exhaustive inventory of data, relevance, systemic analyses, interpretations and classifications. The use of rating scales is required to avoid the bias
resulting from self-disclosure and in order to characterize the various underlying variables in each of the themes presented here.

Four questions underlie the general responses or generalizations arising from conceptualization proper to the individual. Each of these responses, including their linguistic nuances, have been textually classified by keyword and in terms of the vocabulary used by the respondent.

To limit duplication and overlap for some questions, we chose questionnaires adapted to Canadian realities. A complete study of new recruits’ attitudinal changes before and after their training and then as they adapted to their institution, that is, with verbs changing from anticipation to distance in time, required an in-depth analysis of the variables underlying their attitudinal and behavioural adaptation. It has also been important to avoid going off track by constructing a group of responses based solely on quantitative results and also not to get into an assembly-line production of data. The validity of a model depends far more on the consistency of the linkages, the soundness of a hypothesis and its specific nature than it does on piling up questionnaires and overloading the intent of the analysis.

Moreover, the sample involved a degree of randomness related to the respondent’s desire or lack of desire to remain with the study over a 15-month period. The randomness become more evident as time passed, but could be felt more definitely beginning with the third month at the institution since the split occurs between the times when the person shares their aspirations (expectation phase) ; what they learns as they begin the training process (learning phase) ; and, lastly, when their classroom learning and the application of that knowledge in the field is reconfigured (findings phase). We will also pay special attention to those who abandon the training for personal reasons or because of failure.

The significance of taking the sample from a pool of all new recruits in the five regions was based on 249 participants ; the kind of relationships the group has maintained from the outset at the staff college and then later with the prison population and on the three shifts ; the impact and the many repercussions of the duality of coming to terms with institutional security and the helping relationship ; the specific nature of the correctional officer’s work and the various aspects of working in a closed environment ; the restrictive interactions between a formal task description and the many unknowns ; environmental and daily visibility ; the confrontation between what the individual has learned and the reality and, lastly, public recognition for a correctional officer’s work compared to other sectors in the criminal justice field, all while observing the cultural and ethnic characteristics of each region. We are aware, however, that the questionnaire, which could be completed in whatever timeframe the individuals chose and where and when they chose, does not reflect certain sensory variables such as non-verbal body language. The words supplement our understanding, which is intended to be as comprehensive as possible as is the case for interviews.

As another point of departure for our methodology and the achievement of optimal objectivation, one of our primary concerns was not to neglect speculation about certain
assumptions and time to distance ourselves from my own experience of the prison environment.
A steering committee consisting of regional representatives, career management representatives and union and labour relations leaders was officially invited to attend a presentation on this research project in November 2001. Each jurisdiction was informed by letter of the operational launch of the research beginning on September 1, 2002. On September 23, the first questionnaires were distributed, starting with the Atlantic region.

In order to avoid distortion, in terms of the process, the questionnaires were presented
by the research at the first CTP class depending on their respective region. Later classes were taught by the instructors on site in accordance with the memorandum of understanding and the procedure limited to the text provided by myself. This was supported by a set of PowerPoint transparencies explaining the six research phases in detail. Each questionnaire was coded with a number corresponding to the region and was distributed in an envelope marked with the recruit’s last name and first name. Only the PRE A Questionnaire had to be signed first in order to make the participant’s acceptance official. After completing the questionnaire, each recruit placed it in a sealed envelope and forwarded to Research Branch at National Headquarters by internal mail.

In closing this introduction to our research methodology, it should be stressed that, although many research papers have been produced in the correctional field in the last twenty years, a number were eliminated from our bibliographical references, some because they date back to a time when programs were still in the experimental stage, others for contextual or legal reasons or because of developmental details specific to each country and thus not in tune with the Canadian reality, or, lastly, because the databases often did not change, causing the repetition of certain biases in addition to the negative effects of the performance measurement method characteristic of researchers, which too often boils down to the number of time an individual is cited in a text.