 |
Asia-Pacific Forum
on Science Learning and Teaching, Volume 7, Issue 2, Article 11 (Dec.,
2006) Sufen CHEN Views on science and education (VOSE) questionnaire
|
Summary of the Quality of the Instrument
VOSE assesses both the subjects’ conceptions of NOS and attitudes toward teaching NOS, as well as their underlying reasons. The conception and attitude parts consist of ten and five questions, respectively. Each question is followed by several items that represent different philosophical positions. Participants are instructed to read all items of a question before ranking each on the five-point scale.
The items were empirically derived from the learners’ perspectives. On the one hand, VOSE yields reliable results because the items originated from the respondents’ viewpoints instead of experts’ presumptions of reasonable responses. On the other hand, the test-retest reliability is high. The latest version of VOSE was field tested with 302 college students. Twenty-four of the subjects voluntarily filled out the questionnaire again within 1 to 3 months and achieved a high test-retest correlation coefficient, 0.82.
I explained in the original article that the commonly used internal consistency or Cronbach’s alpha is not applicable to empirically-based instruments. Readers may also refer to Aikenhead and Ryan (1992), Rubba, Schoneweg Bradford and Harkness (1996), and Erlandson, Harris, Skipper, and Allen (1993) for more discussion about the reliability of this type of instrument. For VOSE, the Cronbach’s alphas of all issues ranged from 0.34 to 0.81 (see the tables in the original article) and were used to verify the appropriateness of discarding some items in the pilot test, but not a main criterion for reliability. Some issues involve independent subcategories of views and thus have low alpha values.
The content and the interpretation of the items were validated by two panels of experts, each consisting of six experts. The 24 subjects who participated in the retest were interviewed for two hours following the retest to establish validity concerning the ambiguity issue. The results showed that on two of the 85 items, item G of question 1 (1G) and item D of question 3 (3D), five interviewees (20.8%) and three interviewees (12.5%) interpreted the items differently from the researcher and the panel of experts. For the remaining 83 items, over 90% of the interviewees interpreted the items consistently with the researcher and experts. The researcher, experts, and respondents attributed similar interpretations to the items. VOSE therefore measures what it purports to measure.
VOSE is designed for college students and adults, including pre- and inservice teachers. For 10th to 12th graders, VOSE may be appropriate because the language is similar to Views on Science-Technology-Society (VOSTS) (Aikenhead & Ryan, 1992), which was designed for high school students. In addition, a high school student was interviewed to check the clarity of the items in the pilot test. However, for younger students, I would suggest researchers interview some of the students to ensure that the items are comprehensible to them. It should be noted that the NOS and the teaching attitudes parts can be implemented separately. Researchers may use only the NOS part for students below college level.
Moreover, the English version needs to be validated, and, for Chinese users outside Taiwan, the Chinese version may need further revision. Although VOSE was based on multiple sources of empirical data, including Western and Taiwanese learners, only the Chinese version has been field tested, and translation could cause semantic differences. Furthermore, the style of Chinese is based on the tradition in Taiwan. Therefore, the local context and language usage should be taken into consideration for both English and Chinese users.
Finally, respondents who are not familiar with NOS issues may not have a firm position and may be swayed by the first item they read. To reduce the order effect, researchers should inform the respondents to read all items for a question before answering, and may use forms that switch the order of the items.
