Participants
This study was a longitudinal analysis of students as they progressed from the beginning of seventh grade in the 2004 – 2005 academic year through their completion of eighth grade in 2005 – 2006. Subjects were selected from a statewide sample of 34 middle schools and school participation in the study was voluntary. Using the U.S. Census Bureau’s criteria for identifying urban, rural, and suburban a total of 35 schools were selected with 16 defined as suburban, 7 as urban, and 12 as rural. The sample of 34 middle schools consisted of approximately 70 7th grade teachers and 3700 7th grade students assessed throughout the 2004-2005 academic years. The 2005 – 2006 academic years included a sample of 34 middle schools with approximately 70 8th grade teachers and 3500 7th grade students. The number of students participating in the 2004 – 2005 and the 2005 – 2006 academic years was approximately 1600.
Table 2 illustrates the demographic information for ethnicity, SES, and English proficiency reported as the mean and standard deviation percentage of subjects in the sample and state. An examination of the ethnicity of the sample population when compared to the statewide school population revealed the percent of Asian-American middle school students in the current sample was statistically comparable to the representation of Asian-American middle school students statewide ( = .19, p > .05,α = .05, d=.02). The percent of American-Indian ( = -5.32, p < .05, α = .05, d=-.23) and Anglo-American ( = -5.39, p < .05, α = .05, d=-.20) middle school students were significantly lower compared to that of the state’s middle school population; whereas the percent of Hispanic-American ( = 3.70, p < .05, α = .05, d=.25) and African-American ( = 4.86, p < .05, α = .05, d=.20) middle school students in the sample were significantly higher than Hispanic-American and African-American middle school students in the state. Additionally, the percentage of sampled middle school students who were economically disadvantaged was also significantly larger than that of the state ( = 2.17, p < .05, α = .05, d=.10). The sample of middle school students was statistically comparable to the state’s middle school students in terms of the percentage who were not proficient in English ( = 1.20, p > .05, α = .05, d=-.11). Though some differences were statistically significant all effect sizes were small (Cohen, 1988), which indicates little difference between the state/population and sample means. Participating schools were randomly assigned to one of three treatment groups: (1) Big6TM, (2) 6+1 Trait Writing®, and (3) control. Of the 34 middle schools, 11 were randomly assigned to the control group, 11 schools to the 6+1 Trait Writing® group, and 12 schools to the Big6TM group.
Teacher Training
In the summer of 2004, teachers from the participating school districts assigned to the 6+1 Trait Writing® or the Big6TM treatment groups were required to attend training sessions on their assigned instructional models. Teachers assigned to the control group attended workshops unrelated to the treatment models. The sessions were conducted by consultants who specialized in training teachers on the respective models. Treatment group teachers, were administered a pre-test questionnaire (Appendix A) at the beginning of the first day of training in an attempt to measure existing knowledge of the treatment models and to ensure congruence of treatment model familiarity between groups. At the conclusion of the training workshop, a posttest teacher questionnaire (Appendix B) was administered to only the teachers assigned to the 6+1 Trait Writing® and the Big6TM groups to determine the extent to which the workshops increased teachers’ knowledge of the treatment models. Both the pre- and posttest questionnaires asked teachers on a five-point Likert scale how comfortable they felt (i.e., not comfortable to comfortable) and their capability incorporating tenets of the 6+1 Trait Writing® and the Big6TM into their daily instructional practice, as well as how comfortable they felt integrating technology within the instruction of the respective model. Thus, the overall purpose of the teacher training session assessments were to determine the efficacy of the training sessions by considering how teachers would: (1) implement the training model within their classroom instruction and (2) integrate technology in the instruction of the training model. Specifically, the goals for administering the pre- and posttests were to examine whether differences occurred between the control, 6+1 Trait Writing® and Big6TM groups in their pre-existing knowledge of the models as well as to measure change that may have occurred in teachers’ ratings of their capability in teaching the learned skills following the training sessions.
In addition to the pre-test assessing teachers’ familiarity with 6+1 Trait Writing® and Big6TM models, it also included questions that asked teachers if they had attended workshops or trainings over the past five years on higher-order/critical thinking instruction. The purpose of this additional questioning was to assess the level of complexity teachers require of their middle school students in classroom work.
The formal training on the two experimental models (Big6 and 6 + 1 Traits Writing) consisted of an initial 40 hours during the summer of 2004.
The focus of the training was:
1. Model design and implementation both to incorporate inquiry based teaching and learning.
a. Big6 schools also had a local “coach” trained to support their techers.
2. Technology integration training
3. Apply model to one instructional curricular unit in preparation for implementation
a. Expectation was that 6 Traits instructors would incorporate 6 + 1 Traits Writing as frequently as they could in all lessons.
b. The Big6 instructors were required to deliver 2 instructional units throughout the school year
There were two additional training dates during the school year, one in October and one in February of each of the two years of intervention. This training was to share and learn with each other based on their experiences as well as to refine some of their implementation skills. An additional 40 hours of training was held for teachers during the summer of 2005. For additional preparation time, teachers were also allowed up to seven days of release and collaboration time throughout each school year.
In addition to the face to face training, each of the project facilitators/trainers, held online support sessions with specific expectations of instructors in both experimental groups. For example, each teacher was required to submit an outline of their instructional objectives, how they were going to implement specific aspects of the model and to provide feedback on the results of the instruction and student learning. This online forum was very helpful for many teachers, however, there were approximately 20% who had limited involvement in the online support.
Scores for both the pre- and posttests were standardized in order to compare across groups, since group sizes were unequal. A multivariate analyses of variance (MANOVA) was conducted, with treatment group as the independent variable (Big6TM, 6+1 Trait Writing®, Control) and four dependent variables (including 6+1 Trait Writing® familiarity, 6+1 Trait Writing® and technology integration, Big6TM familiarity, and Big6TM and technology integration) to determine whether all groups, regardless of treatment assignment, entered the training sessions with equivalent knowledge about the models. Results indicated that all groups were equal in there model familiarity, with and without integrating technology (F =1.23, p=.29, α=.05).
Four paired-samples t-tests were conducted to determine whether the training workshops led to an increase in how teachers’ self-rated their own teaching ability in: (1) the treatment models (Big6TM or 6+1 Trait Writing®), and (2) the integration of technology within the instruction of models (refer to Table 3 for Descriptive Statistics for Workshop Assessments).
Using the treatment models scores, the t-test analyses revealed that teachers in the 6+1 Trait Writing® group rated their abilities significantly higher (t = 6.04, p < 0.01) following the workshop training when compared to their perceived abilities prior to the workshop. Teachers in the 6+1 Trait Writing® group also indicated significantly higher (t = 2.79, p = 0.01) abilities following the training in the integration of technology with the tenets of the 6+1 Trait Writing® model. The teachers who attended the Big6TM training also demonstrated significant differences in pre- versus posttest scores (t = 6.32, p < 0.01) and in rating their capability in integrating technology with the models (t = 5.37, p < 0.01). Overall, the results indicated that teachers from both treatment groups had significantly higher posttest self ratings of their perceived ability to implement the treatment models and to integrate technology within the instruction of those models. Thus, it was concluded that the workshop improved teachers’ perception of their ability to implement tenets of the Big6TM or 6+1 Trait Writing® models and to integrate their respective model with technology.
