Kankam boadu

Reliability of the instruments

The two types of reliability that can be considered to reduce errors or unintended outcomes are internal reliability and external reliability (Cunningham, 1985). Internal reliability refers to the consistency of the instruments. The internal consistency can be obtained by using two methods such as split-half reliability and Cronbach alpha (Cunningham, 1985). The external reliability, on the other hand, can be obtained by other researchers using the same methods and procedures in the same or similar settings in order to discover whether the results will be consistent or not (Wiersma & Jurs, 2005). Its reliability can be arrived at using a coefficient of stability “test-retest”.

Consistency and repeatability are the two main characteristics of reliability. The reliability of the questionnaire can be obtained using Kuder-Richardson “KR” and Cronbach’s “alpha” which means that a high KR reliability or high and both indicate good reliability. Obtaining over 0.7 from alpha test can be regarded as an indicator of good internal consistency of the test (Cunningham, 1985).

In trying to maximize the value of pre-testing the questionnaire, the respondents were provided with a space to write their comments, problems and suggestions for any development. The questionnaire was pre-tested using various social studies tutors and teacher trainees in Holy Child and Akrokeri Colleges of Education. Reliability was computed by using Cronbach alpha and the results 0.732 for teacher trainees and 0.765 for tutors when the alpha level had been set at .05 (Appendix E). While reliability of quantitative data can be achieved by using the statistical procedures, ensuring the reliability in qualitative data is a bit different. The following procedures have been suggested to improve the reliability of quanlitative research;

1. 
   Checking multiple sources of qualitative data to ensure themselves that the data obtained is consistent
   
2. 
   Thinking carefully about the procedures used to obtain the data about trustworthiness of their sources of informants
   
3. 
   Applying internal criticism (e.g. compare what the informant says against what is said by other informants) (Mertler & Charles, 2005).
   

Ethically, it is always prudent to seek the consent of respondents before instruments are administered. Accordingly, a letter of introduction was obtained from the Director, Institute for Development Studies (see Appendix D) spelling out the intent of the research to whom it might concern. After the questionnaire and interview were tested for validity and reliability, the researcher went to the various colleges of educations’ principals for accessibility in November, 2010. Official access was obtained and each of the colleges was separately contacted to arrange an appropriate time for the administration of the instruments. The researcher administered the questionnaire personally in all the colleges. This was done in order to address enquiries by the respondents on the instruments.

One tutor, in each college, upon being given the letter of introduction, was obliged to provide the researcher the required courtesy to administer the questionnaire. The tutors’ questionnaires were distributed before the teacher trainees’ questionnaire. The tutors’ questionnaires were given to them personally where they were around and for those who were not present at the time; theirs were handed over to their Heads of Department.

In administering the trainees’ questionnaires, the following sequences were followed:

Firstly, the subject tutors welcomed the researcher to the classroom, introduced him and mentioned his mission to the trainees and left. Then, the researcher explained and highlighted the reason and benefit of the study and the rights of the trainees during the course of the study. The trainees were allowed to ask questions for clarification. Copies of the questionnaire were administered and collected after filling. A short discussion was granted in order to take additional comments regarding citizenship education and its preparation in the colleges. The researcher recorded names of class leaders and their contact information to undertake follow-up interviews. These proceedings were followed to ensure gathering of credible data and to abide by the ethical consideration in educational research. The return of the questionnaires was quite encouraging. In all, 255 out of 332 questionnaires were returned by the trainees given the return rate of 77%. On the part of the tutors, the return rate was 94.44%. The data collection was done November, 2010 to January, 2011. The data collection took seven weeks to complete.