2.1 Mentors’ views of factors essential for the success of beginning  (Page 5/11)

Along with the 27 closed-ended questions, respondents were asked four open-ended questions. These questions were designed to evaluate support provided in the teacher-mentoring program, the most difficult duty of the program, and what areas they would have appreciated more support in the teacher-mentoring program. In this qualitative portion of the study, data were collected through open-ended questions from the survey instruments that were distributed to the first-year teachers.

Validity. To ascertain the validity of the Likert-format questionnaire items, the survey was initially reviewed by experts (n = 17). This group consisted of the dissertation chair, (n = 1), dissertation committee members (n = 2), a human resource director (n = 1), secondary school principals (n = 3), and secondary veteran teachers (n = 10). Each expert evaluated the instrument for content, clarity, and appropriateness (Patton, 2001). Amendments were made in wording and arrangement and construction of response options, as recommended by committee members.

Reliability. The most frequent method for improving reliability for surveys is to work towards refining questions, clarity, and instrument design. Good development procedures should result in a reasonably reliable survey instrument (Creswell, 2003). To ensure reliability of responses to the scale items, a reliability analysis was conducted. For the 27 survey questions measured on the factors being viewed as essential or not, the Cronbach’s coefficient alpha was .86, with the range of corrected item-total correlations ranging from a low of .14 to a high of .67. Concerning the six survey items that comprised teacher involvement/support, Cronbach’s coefficient alpha was .66. For the six items that constituted the staff development cluster, Cronbach’s coefficient alpha was .77. The internal consistency of the administrative support factor that was comprised of seven items was .75. Finally, the last factor, resource materials, had a Cronbach’s coefficient alpha of .65. As such, all four factors yielded sufficiently high reliability for research purposes.

Procedures

The list of practicing first-year teachers and their mentors, obtained from the Education Service Center, Region One, Edinburg, Texas, was used to create a database in which every fourth secondary campus was selected as the sample for the study. A self-administered survey instrument was mailed out to the mentors of all first-year middle school and high school teachers identified in the sample, with the permission of the district’s school superintendent. A pre-contact post card was sent to the identified mentors in the districts. A pre-contact involves the researchers identifying themselves, discussing the purpose of the study, and requesting cooperation (Gall, Gall,&Borg, 2003). A follow-up contact was sent out to non-respondents a few days after the deadline. The response time for the survey was a 30-day window. A few days after the time limit specified, non-respondents were contacted by mailing a follow-up letter along with a copy of the questionnaire and another self-addressed envelope (Heberlein&Baumgartner, 1981). All participants were sent notes thanking them for their participation in the study.