Last night, my doctoral students presented their findings from a semester-long research project. Early in the semester, the students developed a Web-based survey that consisted of items from the following six previously validated scales:

1. Perceptions of School Climate (Johnson, Stevens & Zvoch, 2007)
2. Collegial Trust (Hoy & Tschannen-Moran, 2003)
3. Commitment to Teaching (Ware & Kitsantis, 2007)
4. Attitudes toward Professional Development (TAP) (Torff, Sessions & Byrnes, 2005)
5. Teacher Self-Efficacy (Woolfolk & Hoy, 1990)
6. Teacher-Directed Student Use of Computers (Bebell, Russell & O’Dwyer, 2004)

There were also a bunch of demographic items. The students asked their teaching colleagues to complete the survey. Ultimately, they got nearly 700 completed surveys from teachers in five school divisions. Each of the five teams of four students chose two of the scales to examine; they analyzed the degree to which the attitude scales were correlated and looked for differences in each of the two scales by various teacher characteristics.

The results were very interesting. Among the major conclusions I drew from the presentations:

• High school teachers have worse attitudes than their elementary- and middle-school counterparts.
• Male teachers feel slightly less efficacious than their female counterparts.
• Teacher perceptions of school climate are strongly, positively (statistically significantly) correlated with both commitment to teaching and collegial trust.

Most of the findings confirm what we know from the existing literature. But, looking at these different teacher attitudes measures together in one study is pretty unusual. None of the student groups chose to analyze the “teacher-directed student use of comptuers” scale. I now have the data and can see if those scores are related to the teacher attitude measures and/or individual teacher characteristics. That should be interesting. For now, take a look at the following graphic which shows how the scales correlate with each other (all correlations are statistically significant; p<.05); obviously not all of the combinations were tested. Any thoughts?

Tags: survey, teacher attitudes