CHAPTER THREE
B. Test of Reliability
4.4 EVALUATION OF THE RESEARCH METHODOLOGY
Virtually every research conducted must be evaluated to determine the extent to which the methods and investigative methodology adopted were apt in analyzing the phenomena for which the research was initiated. The methodology evaluation procedure enables the researcher highlight unexpected changes to and the limitations of the research design, the ethical considerations and the challenges faced in the course of the research (Ngulube, 2005).
As the research progressed, it became imperative to adjust items three and four of the research questions and objectives and subsume them under item two of the research questions and objectives. This was because it was evident that they reflected components of the non- administrative criteria which comprised one of the variables under study. As a result, item two of the research questions and the research objectives were broadened to reflect the contexts which the subsumed items three and four originally portrayed. This reduced the research questions and objectives to three from the five initially proposed, making it more concise. The adjustments neither affected the hypotheses which were the broader issues for investigation and provided clarity on the non-administrative criteria being studied, nor did they in any way affect the subject matter of the research.
As per the primary data collection procedure, the research adopted the unobstrusive methods approach. One major justification for the use of the unobtrusive methods lies in the
162
methodological weaknesses of interviews and questionnaires (Lee, 2000). The addition of the unstructured observation technique to bring the total methods of data collection to three was therefore beneficial in enhancing the quality of data collected. This was done because of the tendency of interviews and questionnaires to, as noted by Lee (2000), create attitudes in part whereby respondents commonly try to manage their personal impressions in order to maintain their standing in the eyes of the researcher. As noted earlier, the use of the unstructured observation technique was however encumbered by the inability of the researcher to observe the employees activities outside the confines of their workplace environments.
At the initial stage, the questionnaire and interview respondents were to be drawn from Grade Levels 6 to 17, divided into three groups, 6-9, 10-13 and 14 to 17. However, the researcher was duly informed that the Grade Level 17 only exists in the federal civil service as the highest level of bureaucrats in the state civil services were those on Grade Level 16.
Consequently, the respondents were derived from across three cadres of staff -the middle (GL 6- 9), the upper-middle (GL 10-13) and the upper (GL 14-16) - within the scope of the study which is the Plateau State Civil Service. The researcher attempted to achieve an even spread by ensuring that all the seventeen local government areas (LGAs) in Plateau State were equally represented by the respondents taking into consideration, the extent of ethno-diversity within the state. It is noteworthy that within these local government areas, there are fifty-two ethno- linguistic groups basically subsumed underabout forty (40) diverse ethnic groups because of the fact that some of the groups spill into other local government areas with which they share boundaries. There were a minimum of ten (10) questionnaire respondents representing each of thelocal government areas of the state. This consciousness was to facilitate the assessment of the socio-cultural environments of the respondents, the work environment as represented by the
163
Plateau State Civil Service and the relationship, if any, between these environments and employee motivation.
Three hundred and ninety (390) questionnaires in all were administered but only three hundred and forty-three (343) copies of the total returned were considered valid enough for the analyses. For the remaining forty-seven (47) copies, few were too damaged to be put to use, others were not filled properly and the rest were not returned at all. The response rate was therefore reasonably high and satisfactory. The guarantee of anonymity and confidentiality encouraged the respondents to give maximum cooperation and to exhibit a degree of honesty in filling the questionnaires. The researcher had to allow periods of two to five-week intervals before retrieving the questionnaires because of its bulk and the fact that the staff still had to accord priority to their official and administrative responsibilities. Also, some of the respondents proceeded on their annual leaves during the period, further delaying the retrieval and collation of the questionnaires by the researcher.
The interviews were conducted after the questionnaires had been administered and retrieved. The focus of the interviews was more on achieving a spread of the respondents across the seventeen local governments than on achieving a spread across the eighteen ministries of the Plateau State Civil Service. They were originally to be conducted with eighteen (18) respondents. However, after the first six interviews were conducted, the interview schedules had to be reworked when it was observed that they did not adequately cover the variables as contained in the research questions and so could be inadequate in effectively achieving the research objectives. The initial six interviews were thus, treated as a pre-test. Consequently, another fifteen (15) interviews were conducted bringing the total number of interviews to twenty-one (21) since the information derived from the first six was not discarded. The gaps on
164
the pre-test interviews were captured under the term ‘missing values’ in respect of the questions and information lacking on them at the time.
It was also observed in the course of the interviews that the higher the grade levels, the better and more tangible the quality of information received because of the level of education, wealth of experience,etc. that appeared to be lacking with the staff on GL 1 to 5 who had been approached by the researcher to serve as respondents. This further informed the decision to exempt the staff on GL 1 to 5 as a result of their inability to fully grasp the contexts of the interview questions on the basis that the technicalities involved were beyond their comprehension and that they lacked the requisite experience for tangible responses. Many staff in this cadre referred the researcher to interview officers on the higher grade levels and notably, no hitches were experienced in the course of the interviews.
As per the methods of data analyses, the Spearman’s rank correlation coefficient had to be adopted to boost the quality of analyses alongside the Pearson’s product moment coefficient formula and Bi-variate two-way frequency table correlation method which were initially proposed. It served as a buffer for the proper measurement of the qualitative aspects of the data (Gupta, 2011).